#Philips CD-i
1 messages · Page 9 of 1
ABCD-iFGHEJKLMNOPQRSTUVWXYZ
Phillips lost a lot of potential customers with their public feud with sesame street
That's interesting Mame is looking at Photo CDs, how are they supporting them? Via the CDi emulator? Funny they have picked the most complex format of Photo CDs to check first
Yes, this is the driver file of mame which defines the parts to build a CD-i. Not only is it responsible for constructing a home usage system, it also handles the quizard arcade cabinet. This is a special modified CD-i board with copy protection hardware.
Fun fact - Sesame Workshop actually co-designed the Roller Controller on CD-i, along with designing two educational games for it as temporary exclusives. (They would later get ported on to 3DO, PC, and Mac, being re-released as late as 2009.)
Roller controller?
OH SHIT I PLAYED IT
omg I had a friend who had that and we played it as kids
I didn’t know that was CD-I lol
A literal big trackball, as opposed to just a big trackball enclosure.
Atari really cared about left handed people
It's a cool trackball
Sometimes I wish the trackball technology has kept up with the mouse. We have optical trackballs today but the DPI is rather limited. Especially with finger trackballs while the thumb trackball in general seems to get more love. Especially by Logitech.
Sometimes I'm chasing a white rabbit and at the end of the rabbit hole there is just some weird software doing its thing. Today it was Burn:Cycle. In front of the psychic roulette there is some animated text rotating. I'm currently busy, checking the CPU speed as our 68k is running way to fast. So I've decreased the speed and the animated text started flickering. "Well shit.... I've created a bug". Then I thought, how is it on the real thing? And It is flickering on the real machine as well! So what is this? Is it a bug on the real machine? An overclocked CPU removes the flicker. Is the flicker a result by choice? Is it caused by some random glitches but intended? Does the flicker belong to the experience of standing in front of the Psychic Roulette terminal? Is the flickering only occurring on the german version of the application? Because german words are longer than english?

The same goes for the Burn:Cycle demo. At some point it returns to the main menu. Is it returning because of an error? A CD reading error? Is it returning because of artistic choice so the main menu is shown again to have the title of the game in demo mode? I was so confused I've checked the real machine as well. And yes, it seems to be random... Is it random because of seeking time? An internal counter?
Another MiSTery!
Don't get me started on Myst. Whenever you're rotating in the game, there are glitched lines at the bottom of the screen. Those are created because the Video IRQ to start the software doing something is too early. With the intended CPU speed you don't see anything. Zenith has the same issue.
in a way its nice to have so many strange cases already found
I actually have started a list. It will be part of the next release. It shall help keeping track of weird stuff which is expected when using the core.
sometimes the game library is so tolerant of hardware variation you struggle to know whats accurate
psx is a bit like that
re: burn cycle, i wonder if any dev ever used the cd-rom seek time as RNG
Thanks for all of your hard work on this core Slamy. At one point I don't think anyone even thought we'd get cd-i on MiSTer let alone one that is capable doing as much as your core does. 👏
I actually owned that model and that controller in the picture back when this was still being sold. Of course I had to mail order any games or software I wanted ;) I always had to be different though. I think I was the only one of two kids who I knew that owned a TG-16 as well.
Ahhh, because it is CD-Valent-i-ne's day
For all my CD-i pals that need that last minute Valentine gift for that special someone.
true art
CD-i can't believe it's not butter.
Since it is a special day, I've decided to make a release of the recent changes that were accumulated over the last days (or even weeks by now).
MCD212: Improved timing of ICA execution
- Fixes hiccups in Burn:Cycle demo mode
- Fixes flickering cutscenes in Flashback
MCD212: Limit CPU speed by reducing memory bandwidth
- Fixes flickering Mario animation at the bottom of the screen in Hotel Mario
- Fixes flickering lines during transition when turning around in Myst
- Fixes screen tearing at the bottom of the screen in Zenith
- Added debug option to make the CPU fast again (about 2x speedup)
- Might fix speed of Plunderball (must be tested by an expert)
- Causes sound glitch regression in Philips Logo of Burn:Cycle
CDIC: Rework of audio player state machine
- Added abort flag for playback
Fixes some audio map related sound glitches in Zelda's Adventure
and also some glitches in the Help cutscene of Wand of Gamelon
- Added mysterious bit reset
Forces Unmute during Audio Channel playback (otherwise it would be muted)
Function not yet understood
- Removed artifacts caused by software emulator inspiration
AD7528: Added two instances for stereo mixing
- Fixes audio panning in Zelda's Adventure and Zenith
- Fixes stereo to mono mixing (SFX + BGM)
- Hotel Mario
- Zelda - Wand of Gamelon
- The Apprentice
CDIC: Added seek wait time
- Fixes sound hiccups in "Felix the Cat"
Edit: I had to remove the uploaded file. I've uploaded the wrong one. 😦 Please check here #1103404843512631357 message
Its a great day. Celebrating a Massacre I like your style
What massacre? O.o I thought it is valentines day
The St Valentine Day Massacre
The Saint Valentine's Day Massacre was the murder of seven members and associates of Chicago's North Side Gang on Saint Valentine's Day 1929. The men were gathered at a Lincoln Park, Chicago, garage on the morning of February 14, 1929. They were lined up against a wall and shot by four unknown assailants, two of whom were disguised as police off...
Oh boy, I'm sorry, I wasn't aware of that.🫠
its what the day is for. Its also for Stalkers to send cards to unrequited loves and then watch them panic as they dont know who sent them
I love the whole day !
Massacre and stalking all in one day
Do we still need bespoke MiSTer main for this, @terse flare? I can’t remember
No! Everything should be integrated by now.
🎉
Did you get my card
Pins updated
Please remove the special MiSTer main. It is no longer required.
Done, thank you for all your hard work
No it may be in the entire pallet of cards and gifts though
Did you get my Pigs Heart ? I sent it to you
I hope it was the pigs one I sent. Its kind of hard to remember with all the body parts in the basement
Is everything now in official MiSTer Main, or just the latest unstable?
You must be super fun at parties.
Thank you for asking this question. It seems I've underestimated the release cycle. The current stable Main is 241227. This means, my changes from 250119 are not part of them. This indeed means that the unstable Main is still required.
Sorry for the silly question but is the date of this build right?
This is so embarrassing. This is actually the second time this happened. Maybe I should delete the older files.
In my defense, 20250214 and 20250124 are really close together >.< And it was a long day...
I hope I haven't made any other stupid mistakes. There is at least one regression with this core already known to me. The Philips Logo intro in Burn:Cycle has some new audio glitches.
Ehhh, it doesn’t matter. If it were publicly released, sure.
I saw that too and just renamed it on my MiSTer lol.
Yep - Plunderball feels like it’s running at the correct speed now.
The recent changes also appear to have fixed the silenced voices in Flintstones-Jetsons Timewarp, as well. (That was a bug dating back to the 20241230 release.)
Master Labyrinth was also having issues previously with crashing, which appear to have been resolved in this new release.
Ohhh, I wonder if Secret Agent is fixed. The intro is completely silent.
No, Secret Mission still has silence during its intro.
The issue with the background music during Wordplay occasionally not playing also appears to have been fixed, though.
Thanks for checking, bro
This build fixed the audio issues with the Felix the Cat game, great job!
- Added mysterious bit reset
this commit message gave me a chuckle thanks
Jan Pienkowski - Haunted House is now Completable. Issues with that title crashing now appear to have been fixed.
@terse flare Do you know when the core might start getting released in #unstable-nightlies ? I have a script from Porkchop Express which downloads all unstable cores and puts them in an /Unstable folder, obviating the need to manually download cores on a pc and shuffle them over.
when a core gets added to unstable nightlies it attracts more attention. Slamy is still experimenting. The extra people would not realise the core was being built and so the number of bug reports and complaints would rise
I do recall something similar was done with the N64 core. There was a custom json path to allow the standard MiSTer downloader to grab the latest unstable build without causing to much commotion. Maybe that would be something to help here. However, I do not know how this was performed. It's also quite some time ago. It would indeed have certain advantages concidering that there would no longer be mistakes on my side of the upload and your side on the download. In general I don't know which would be the best time to start having this core in unstable nightly. I remember that the Saturn core was there much too early as it was still in a rather broken state. I think it was even too early in stable... but that might be caused by special situation.
(that was me)
robert published things as github releases and I had a github action that created a downloader db from the latest release
but you are effectively publishing it in public at that point and it is unavoidable that you will get a bunch of new users/bug reports etc from that
I'm undecided on this topic. The core already has quite a lot attention. It even has an issue on my github repo asking for the DVC >.< The N64 core was also not complete when it appeared on those databases... hmm...
my memory is slightly fuzzy here but I think one of the motivating factors for putting the 'not done' n64 core into a database was the addition of n64-database.txt that also needed to be downloaded alongside the core. I can't remember if we did that before that feature or not
Of course do whatever you want to do. My personal opinion is that the core is in amazing shape and would greatly benefit from the added exposure.
GitHub issues can be ignored until you’re ready and you can rely on the community to corroborate and test for you.
I got a feature request because I am a lazy son of a bitch
Can you auto load a game if it’s the only one in a folder
That would be a nice QoL thing before it went to a nightly, as it'll work just like the other CD cores do
it's part of main
and because it's main, I think that functionality is gated behind some long ass is_psx() || is_pce() conditional
I'll look later
ur a _psx() || is_pce() conditional
you can make it so people can't open issues on your github if you want
My understanding is that to be in unstable nightlies the core needs to be merged into official MiSTer GitHub repo, so it generally happens after official release or far along in the dev process
Which CDi could be, there is no blocker there
CDi is on the WIP download script, yeah?
I only ask because I don’t think it’s grabbed the latest version
I haven't been at my computer today, I can add it in a few
Actually, an excuse to try do this from my phone and was able to easily, so that is in there now
Didn’t knew about this script. Where can I find it?
If you have a read here
https://discord.com/channels/647909397477195803/1096872957030174750
Thank you and congratulations for the script!
Hah, well thanks, but all it is, is me chucking files on a GitHub folder structure, there is no clever wizardry my side 🙂
Am glad some people find it useful though
does the CD-i core have any support for Y/C video so far?
It does? Even I didn't knew that. Is that just part of the MiSTer framework? In the past special YC builds existed. I only test on HDMI and on SCART
Yes it works great. I think it's just part of the framework now. I'm using the latest Mister Addons analog io board with built in yc support and luma trap.
A person on GitHub complained about the game speed of "Escape from CyberCity" without even trying it on a real machine to compare. I can agree on the speed being too fast... even on the real machine. "Escape from CyberCity is too fast" is a statement which can therefore be added to the CD-i discord bingo card. Please don't even try the game. It's just FMV like Dragon's Lair but faster and with guns.
Reminds me of an issue that was outstanding on the Sega CD core for a while. Choppy fmv in Syndicate. Turns out Molyneux was just like "screw it, close enough" and shipped it that way.
Escape from CyberCity is just inherently plagued with all sorts of problems, so it's easy to mistake the flaws as emulation issues.
@terse flare not sure you know this but Escape from CyberCity is running too fast
It also has emulation issues sadly on the core. There are hiccups in the audio which can be barely noticed because your eyes are overstimulated with the animation. The animation itself is also not bad. It reminds me of the technology used by Burn:Cycle since a DVC is not required. Though this time, there are no visible RLE artefacts.
The Flowers of Robert Mapplethorpe seems to have had a regression with this build. Worked just fine in the January 24th build, but the CD-i intro animation gets all choppy and then the main music plays over said unfinished animation in this Valentine's build.
thank you! i got it working; it was just giving me trouble before i had selected a bios (weird vblank). everything's working great now
Escape from Cybercity is actually a port of a Laserdisc arcade game, similar to Dragon’s Lair or the ALG titles.
The original game was titled Freedom Fighter, and it was based on the anime Galaxy Express 999.
It’s a pretty abysmal port, but as of right now, it’s probably the best way to play that title without access to a real arcade cabinet.
The original arcade game is actually just as fast as the CD-i version, but transitions seamlessly with the animation, leading to this frenetic gameplay that’s actually really well-synchronized with the Laserdisc visuals. I’ve only played it once at a now-defunct arcade convention, but it’s certainly worth trying if you ever see an original cabinet.
A lot was cut back on the CD-i version to work on the non-DVC hardware, though, meaning that many of the animated sequences are replaced with stills and more limited animation. It really compromises the feel of the game, which makes it a real shame it hasn’t been implemented under emulation. (Star Rider, which shared the same hardware, is in a similar boat - it’s a racing game by Williams with some really impressive tech behind it.)
There’s a SINGE port, but it’s somehow even worse, with tons of features missing and audio being replaced throughout the game with the Laserdisc version’s intro theme. I’d avoid it, as it’s borderline unplayable.
Thank you for this insight. I wasn't aware that the regression is as big as that.
On a positive note, this release seems to have fixed a lot. The only title in the Non-DVC Games library which doesn’t get ingame now is Earth Command, and several new titles appear to be Completable.
I’ll work on updating the testing sheet throughout today.
The missing audio on Shaolin’s Road has been fixed in the new release.
The speed fix on Plunderball also seems to have fixed a similar issue in Rise of the Robots, as well.
Those games seem to have a bad habit of using the CPU speed to calculate the game logic. This is like ZX Spectrum style of programming. There was a certain time at the beginning of development when even Wand of Gamelon was too fast. But that was a result from the OS time running too fast.
Right now, the CPU is even in this release slightly too fast. I just had the feeling some games had more issues so I didn't went too far. I'm actually in luck that there is no direct way of influencing the peripherals in a "racing the beam" manner. Therefore the CPU speed doesn't need to be exactly correct.
I've wondered what the issue with Earth Command is. As the hangup occurs in the startup phase I've tried to investigate this problem by simulating the core again using Verilator. It's slow (3 seconds per frame) but precise. For some reason, the game actually loads in the simulation to the game screen with the world map. This is one of the worst cases in FPGA development as this seems to be a deviation between simulation and execution of the synthesized result. Or it might be a stochastic problem and the simulation by accident got the golden route through the software.
The simulation doesn't take the MiSTer main executable into account as I'm faking the HPS via custom C code. One of the things to check here is to create a checksum from CD data and read it out during loading phase to check for differences.
Is anyone else having issues with up registering? I've tried multiple controllers, remapping it to another button, keyboard...but UP seems to only read 1/3 times for me. Holding seems OK, but pressing seems to either have a super high lag or just doesn't trigger the shift register or something. Overclocking input makes no difference.
Jumping in Flashback is nearly impossible, and that's a game I can play the first 2 levels of blindfolded (not really...don't ask).
Down seems fine. Left and right seem fine.
1/2 sometimes feel a little laggy, but nothing like up.
(And I do not know if this is just CDi since I didn't have one back in the day...I've just played Flashback on about 50 different platforms, and Hotel Mario, which relies on up constantly, also barely reads it, so I can't imagine they'd have released like that, but maybe they did and this is perfectly normal)
I sent a wall of text so it was buried, but that didn't fix it.
Overclocking got the other buttons more responsive, but not up.
<RickRollResponses = "false" />
It makes Galaxy Express 999 (or whatever the game that uses it is called) unplayable. Hotel Mario is fine because I can just hammer it. But anything that uses up to jump is a no go for me and anything that requires reticle aiming up and down is a non-start. Arcade Classics is somewhat playable, but up/down are still not responsive...it's just not as much of an issue because holding works after like 300ms, but holding doesn't work for jump, obviously. So here I am... left with all the better ways to play Flashback. :/ (it was for capture, so...)
Sorry I missed that.
Nah, my fault. It was late and I wanted to get everything noted before I passed out. It wasn't well organized.
Then I just hit enter today to send it.
My computer room is a disastger right now. I decided to 'clean' an am now swimming in cables and plastic.
So...just like the Pacific.
This doesn’t help but I have input issues as well. It’s as if the core is consuming inputs sometimes. I do not know if that’s just an issue with my controller, the core, or that’s just how it is.
That's what it feels like, exactly.
is there any input-test rom for cd-i?
I do not know but while unrelated look what I found while looking for it:
ooh neat
@terse flare is that helpful? The CPUTest I posted above?
I first thought yes. A test to verify the accuracy of the CPU is always helpful. I find it slightly scary though that "No CD-i hardware currently passes all tests" is written on the README. So maybe this is something to look at when I have time. This core once had an issue that the multiplication instructions didn't work. This test would have found that. I'll look into that when I have time. The repo links to the CD-i discord. I'll probably should investigate. But not now, I have to go to work in a few minutes.🫣
https://github.com/TwBurn/cdi-debugcontrol/releases
@fierce shoal This one is good for testing controllers. I've used this before to test the core and my real controllers on the real machine.
i wonder if its related to that er, you mentioned the pad is much faster on the core than on real hardware a while back iirc
The refresh rate of the pad is increased with the "Overclock input device" setting. If this is disabled, the speed should be identical. The core having issues with the input is something I haven't heard before this night. This seems to be valuable information but I need to be able to reproduce that. The mentioned debugcontrol test should at least give direct insight. If there is no issue there, I probably can't do much.
It might have been important that we had this talk. The recent core seems to have some problems. The debugcontrol application shows flicker in the input data. If the dpad is constantly pressed, the is some erratic movement going on. It does happen every few seconds.
@fierce shoal The issues you are explaining are totally new to me. The MiSTer comes with a utility core just to test controllers. It's in the menus in Utility->InputTest. If the issues are there too, this is probably just your gamepad.
Nope, input test is perfect.
I'll try the debugcontrol release and see what shows up.
When your standards are so impossible you can't even meet them yourself
@terse flare sorry it took me so long to get back. I just got my capture setup reinstalled. There's definitely an issue. You can see on the DebugCtrl that up and down throw WAY short compared to left and right. And there's TONS of bounce/jitter. On MiSTer Input Tester, same controller, you can see it works just fine. Thank you so much for including those links! That really helps visualize it.
I have additional questions. Do you have a mouse attached to your MiSTer? How does the analog input test look in the input tester. I think you can see it when pressing select. The last question would be, is this issue also present in the cores of January?
No mouse, I'm not using an analog pad (RetroBit 6 button Genesis style, D-Pad only) but I can add an analog controller and test it if you think it would help. And I'll take a quick look at a Jan build and let you know.
Late December is perfectly choked. This response feels great.
It also doesn't throw as long, it seems. But yeah, the controller feels responsive in the late Dec. build.
It's super odd that I'm the only one experiencing it though. @stiff flare - how does your controller test look? Since you mentioned you did feel the controller had some issues.
What do I gotta do? I can help corroborate your findings.
Oh I see
Ok cool I can give that link Slamy shared a shot
I’ll test it, too. Haven’t noticed any movement issues like that, but I’ll record my findings and post them here.
Not seeing it on mine. This is with the D-pad on an 8BitDo Pro 2 over Bluetooth:
It is the NTSC mode somehow. I've noticed you are running the Magnavox version of the menu. This means others cannot see it because of PAL. I'm able to replicate your regressions now locally too. Huh. That's interesting. I usually test mostly the PAL version of the core so this flew over my head.
That is actually some very interesting finding. It can also be slightly improved by using the CPU Turbo in the Debug Options. The erratic movement is then replaced by a screen tear...
One good thing about this issue is that this input debug application happens to be open source software and I have the CD-i SDK installed. This means I can dissect it to check what this is doing. I can also check once how this application behaves if I configure my real 210/05 to a NTSC machine.
Cool, thank you for looking into it! I'll switch to PAL for captures in the meantime.
I've just checked with my PAL 210/05 which I have rewired to NTSC. The issue also occurs there. For yet unknown reasons, the debug application only shows this problem when the input comes from the wired controller. If I use the RC controller, the problem doesn't happen. That's curious. The issue also doesn't depend on the port used. The front port has the same issue as the back port, even so they take different routes in the software. I don't know what this means. The test application might be to blame here. I think I'll create an issue on github so maybe the dev can take a look at that.
Do you experience problems in other applications still with PAL?
It looks like the test application functions as intended on my NTSC 220/37 (apologies about the flickering, my phone camera’s shutter doesn’t really agree with this TV):
Is the input source a wired controller? Or is this the RC?
Wired. I have the Tecnoplus-branded variant of the 3-button controller hooked up at the moment.
Here's a better quality capture (composite via RetroTINK-2x to Elgato HD60S, as I couldn't find a S-Video cable on short notice):
In that case I'm totally lost. It might be possible that a real NTSC machine does behave different compared to a PAL machine which was rewired to 60 Hz mode. I would have thought it was nearly the same. This can also mean that the "NTSC" mode of the core is totally wrong.
Hello, where can I find the latest CDi core?
Hello, please check the pins here at Download Core.
Okay, not seeing it right off on mobile app, will check on desktop. Excited to to test it out.
#1103404843512631357 message
Thanks!
On mobile, click the channel name then click the settings button. You'll see pinned messages there
I think... that with everything that's going on in the world right now... everybody just needs to cool out a bit... and chill with some cd-i tetris on the cd-i core.
Does this core look for a specific folder for the games?
Yes, I beleive it’s //media/fat/ganes/cdi
thank you
Hi slamy, thanks for this core, is awesome!, some thing I notice is that I can´t load games thru a hdd (connected to usb hub). Looking to load the Apprentice, but didn´t work yet for me.
Do you have the recent unstable main?
make sure you have the boot0.rom and boot1.rom in /games/CD-i on the USB drive also
folder name is case-sensitive if the drive is NTFS
Since MiSTer 250216 it is no longer required to have the unstable Main. The CD-i Linux code is currently 100% part of the stable MiSTer.
Edit: I've edited this statement because it was wrong. MiSTer 250216 is still unstable.
Please provide more info on the issue. Does it hang with a cyan screen? Is it just a black screen? Which version of the MiSTer main do you have? Does it only occur with the Apprentice? In this group here there is at least one other person which has managed to use this core with a USB HDD. In general it should work.
Yeah, I've used USB HDD exclusively and not had issues with loading.
Have you managed to play any other games off your HDD? I have a USB hard drive that I run all my CD games off of and the same setup worked fine for me this morning.
The apprentice loaded ok for me from a USB drive. It did sit in the cyan screen for a bit longer than normal before loading
i strongly associate this bastard with edutainment booths in small european museums
Roll the ball and "click" on the images to learn more about them!
Ok I know this has been asked before but my searching is not getting me to an answer. I have just started to dig into the CD-i on MIster. Using the latest unstable MIster main and CHDs from the archive. cdi220 is now boot0.rom and the other one is boot1.rom. All i can see is a blue screen and nothing will load. What, I'm sure, obvious step am I missing?
Just to be sure, do you have checked the checksums of the said boot files? Is the cyan screen occuring without seeing a menu? Or was there at least once a menu?
cyan screen is there as soon as i start the core and with the menu pulled up.
So you never see any Philips CD-i menu with buttons and some artwork?
I have not did the checksum. Where can I find the checksums to use?
I never see the CD-I menu itself.
2969341396aa61e0143dc2351aaa6ef6 cdi200.rom
3d20cf7550f1b723158b42a1fd5bac62 zx405042p__cdi_slave_2.0__b43t__zzmk9213.mc68hc705c8a_withtestrom.7206
```Those are the md5sums. It is vital that these numbers match.
That was it! I am playing Hotel Mario now. Thank ya!
Earth Command might be the first game that actually reads the time and reacts to that. You can fix the game with the current 250214 core by going to settings in the menu and set it to a valid date. As the timekeeper implementation is wrong right now, the MiSTer will store that data on the NvRAM. Afterwards the game should work.
This might be a good time to maybe actually start putting some real time in there.
Ah yes. We shouldn't forget as - at least in Germany - it is not too late: Happy CD-Friday everyone!
Be fun to see some of the cherry on top features like RTC coming in, will give you a bit of respite from the rabbit holes on the audio and colours sides as well.
Might be worth checking .sys, I think there may have been a small update since you last updated although may not have any tangible change on the core.
several cores have a clock to worry about if you want reference, gba, ao486, amiga, probably others
saturn i think?
Here is a list of them all
Although may need updating if Saturn has one. Does PSX?
Does the core work with mister main?
yes
Unless something has changed that I missed them you need unstable main still
For things to work as expected anyway
I thought Slamy just said that everything is in normal main? Or maybe that was unstable and you didn't need the bespoke main anymore
Ahh yep, good catch @spring gate - still need the unstable #1103404843512631357 message
@sharp plume You need the latest unstable, but you don't need the custom mister anymore #1103404843512631357 message
That's my answer....if only I wasn't lazy
Lol
What part of Slamy's post linked above states that you still need unstable MiSTer main?
Since MiSTer 250216 it is no longer required to have the unstable Main. The CD-i Linux code is currently 100% part of the stable MiSTer.
MiSTer 250216 is out?
Oooh... I see now. 😄
@pine stump @sharp plume @gray badge I'm terribly sorry for causing confusion. I could have sworn, that update.sh has downloaded 250216 on the last run I've executed. Maybe I was just seeing things. It is still 241227 which indeed means that unstable Main is still required. At least for the latest features. Especially for The Apprentice you need it.
no need to apologize at all, duder. When I read it I knew 250216 was the unstable, but it was cataloged in my brain as "CDi no longer needs the unstable". Was my mistake, not yours 🙂
I want an apology and a monkey serving me ice cream
we are all monkeys so you’ll need to be more specific
speak for yourself I am evolved from an intelligent shade of blue
Don’t brag on it
ok Monkey boy
Chimpshamer
This core travels through space and time. A freshly booted MiSTer seems to use the date 01 Jan 1970 01:00:00 as reset time. This is actually quite correct, considering we are talking about a unix derivate here. For the CD-i this is a huge issue as it is prone to Y2K because the time keeper only has the last 2 digits of the year. The system is confident that it is the 1 JAN 2070. I need to take care of that with a hack I think. If no one else has a better idea, I will hack the year 70 to 94. It is one year after Earth Command was released so I assume for a real time game, that is ok.
makes sense
This comment has it all!! Earth Command! Y2K! The year 2070! Travels through space time!!
Windows - 1 point
Linux - 0 points

Reverse y2k?
I'm so confused. why does it think it's 2070 and not 1970?
With only 2 digits for the year, the first 2 are based on an assumption. Since this machine is from the early 90s, a date of 1970 would not make any sense to be set. The earliest supported year is 1989 and it ends on 2088.
aha
What do we do when it's 2089?!
I think you mean 1989 
I saw something the other day that another, potentially worse Y2K style situation is on the horizon as 32bit machines can't go beyond some year around 2038. If so the younger generation enjoy the non event that was Y2K.
2038 is like 30 years away, nobody will be using any of that stuff by then
Someone’s never worked for a hospital or government agency
...or a bank.
The number of banks I would visit on service calls still using OS/2 long after it was discontinued was shocking.
Nobody will be using them in the sense that we'll be in a fallout style dystopia at that point
That’s fair. YR24 or whatever will cleanse us all in galactic fire
The fact that it didn’t seem outlandish enough that I said 2038 was 30 years away means I’m obviously among my people.
😆 I didn't even notice that math
Sorry I’m like a week late. My controller test does weird shit. It vibrates as I push in a direction.
Looks like we're NTSC Pals.
Not PAL buds.
Chum.
@terse flare Here’s my controller test with the core:
oh wow you’re right! Weird that it doesn’t move much in a direction though, like in @strange crater’s videos
Oh, setting the M30 to analog mode fixes that
Like a real patriot. 🫡
Oh holy shit games feel 10 times better to play LOL
I feel like I can actually control the ball in Zenith
Yep. It's insane how much worse it is.
Has anyone finished Link the Faces of Evil on the core yet? I managed to finish it tonight
You conquered each
Ok, it seems that the 250214 core has many regressions. Here the things I know so far from you all:
- NTSC seems to have input problems
- Flowsers of Robert Mapplethorpe is not booting
I do know of some other regressions as well. - Philips Rotating Logo of Burn:Cycle is broken (audio issues)
I'm unable to reproduce the input problems on NTSC on any other software than the debugctl test cd. But at least on that I'm able to reproduce it and there is a workaround for that.
Workaround: It seems my reduction of CPU speed went to far. The core is possibly actually running slower than a real CD-i. Go to debug options and enable Turbo CPU. This reverts the change of 250214 on CPU performance.
The Flowers of Robert Dingelsgedöns is caused by a wrongly timed ICA execution. Maybe I should provide a hot fix for that. It might have hurt every title that uses the ST mode of the videochip as well.
The thing with Burn:Cycle is actually more complex. It can be fixed by enabling Turbo CPU as well. But it seems more to it which I'm not yet 100% sure.
In general maybe an update. The TG68k cpu core which is used by this project has one disadvantage compared to the fx68k that is used for the Minimig. It was never meant to be cycle accurate when it comes to instruction timing. The way I've made the CPU slower is by slowing down memory access. But that was the wrong approach as I assumed my measurements on multiplication time. Even when assuming on memory bandwidth it is still wrong.
As I can't simply use the fx68k because it lacks the features needed for a SCC68070 and is not very maintainable because of microcode I'm still aiming for the TG68k as I understand its source code better. (even when written in VHDL...)
Just to give an insight in speed. The SCC68070 required 76 clock cycles for the MULU instruction. In case of the TG68k it requires 1. Luckily it can be modified to remove the mulitplication unit. It gets much slower with that.
Still it is not right and needs some work.
You might want to add that to the testing sheet.^^
Love these updates, thank you
also lol “Dingelsgedöns”
It means silly things in German
"The flowsers of Robert Dingelsgedöns" I like that title
Secret mission (Germany) doesn't have sound with current core and current unstable main. Anyone can confirm?
THE cd-i is here in the cd-i chat 
I can check when I'm back home (if someone doesn't beat me to it)
Yes correct, no prior build has had sound either.
Might be worth doing a run through the region specific European games tab, those haven't been looked at for a few months
Yes, could look at the German region stuff as I'm German an prefer German versions
Totally random but we have a lot of Germans in this chat. I think that's cool as hell. I would love to visit your country one day. Every German person I've ever met has been such a good person.
It doesn't? Maybe I was hallucinating but I do recall we had sound once. The current core seems to be lacking audio during the intro, that is true. But ingame, audio seems to work.
That's nice to hear, thank you
I think that's because the CD-i was more successful in Germany then in many other countries. Especially in professional applications the CD-i got a good foothold in companies like driving shools
that’s interesting, I didn’t know that
Adde your controller workaround to the Questions Answered post that's pinned - #1103404843512631357 message
Sometimes I wonder which release frequency might be the most healthiest for the community. I release a core and everyone documents its advantages and disadvantages. And then I release another one like a bulldozer and everything starts all over again.
I wouldn’t say it bulldozes it. I’m happy to test games if that’s useful - I think the trouble I run into is figuring out when it’s useful to do a run through all the games
To be blunt, do whatever you want to do. Whatever is easiest, don't worry about the community. We are here to support you and test whatever you need.
Since this isn't the general public, we are all very comfortable with regressions and expect it.
We did according to my testing sheets about three months back, when we were getting the first cores with mouse support.
Not quite sure when the regression occurred. The first I heard about it was about a month ago, when Robby mentioned the game was missing audio.
@terse flare What would be most helpful for us to do for you at this point?
Back in January 2020 I was confronted with the same question. I had a small inner circle of closed beta testers to test my Amiga game. I really didn't want to wear them out. Playing the same game over and over is rather tedious. I can't stop you from going through the list and testing the games. If the core was a game it would still be in very early alpha state considering the amount of bugs and its experimental nature. I don't expect anyone to complete a game right now. It's however sometimes nice to hear that there is a certain amount of stability in the core and things don't just crash or hang very often.
What would be the most helpful? More developers probably. I do realize that is a difficult request considering not even the software emulators have enough developers to handle all their problems. That said, I don't have an answer to your question.
There is however one thing. Please run the core in 50 Hz / PAL mode for now. I don't have enough time to always test both modes and usually I try to get the PAL mode functional as most games are usually designed around the 280 vertical lines. The CD-i is in that regard very similar to the Amiga which also had more PAL than NTSC games. It's sad when recordings for YouTube are failing just because the rarely tested video mode is used for that recording.
Should we turn Turbo on as well?
A hotfix release on this sunday.
CDIC: Adapted audio control register to measurements
- Lowest bit is now set when the audiomap has finalized the playback after processing the 0xff decoding.
- Might make things worse. Checks needed. So far nothing went worse for me.
MCD212: Added debug option for ICA execution line
- The new behavior (based on 210/05 measurements) added in 20250214 is hidden in the debug menu
- Original video timing from before is restored. More measurements needed (ST and NTSC timings might be wrong)
- Fixes boot of "The Flowers of Robert Mapplethorpe"
MCD212: Reduce CPU starve by ICA
- Just an experiment. It doesn't hurt but also seems to be not very relevant
MK48T08B: Added timekeeper (time setting from MiSTer Main)
- Add time related registers
- Fixes/Removes NvRAM change detect on time registers
- Fixes Earth Command bootup (needs a valid date)
SCC68070: Slightly better timing accuracy of multiplication and bit shift
- Removed barrel shifter and multiplication unit
MCD212: Slightly lifted memory bandwidth reduction
- Only a patch! Needs accurate overhaul.
Updated MiSTer sys to fa4c17a.
This release somehow also fixes the "Curator" audio track of Flowers of Robert Mapplethorpe.
It also somehow fixes the audio track of "Secret Mission" during the Philips logo and the intro.
The ICA revert breaks the demo mode of Burn:Cycle. Use Debug Options -> ICA at VBlank -> Yes for Burn:Cycle running on PAL. (Not tested with NTSC)
The core seems to be sometimes unstable with the slowed CPU. Use Debug Options -> CPU Turbo -> Yes to be on the safe side.
-> Disadvantages are glitched graphics at the bottom and game speed.
In general it seems to be more stable actually. I've made a mistake with the way I've implemented that apparently. I still don't feel like making that the default setting though as it fixes some graphical glitches caused by bad programming and also some game speed issues.
Wow that was a fast fix for my secret command pleasure ^^
Slamy hates it when we do this but I am here to make his life harder so, just as a reminder, if you appreciate the work he's doing here, consider supporting him directly https://www.patreon.com/AmigaSlamy
frankly, if you have loaded voyeur (because you were "just curious") your shame should compel you to support him
Nice update slamy 👍 dvc working great with the new hot fix. Like Toasters toasting toast, mah boi
||trying to get bingo||
New to this core. I saw it says it accepts CHD format but when I try that I always get a blank screen
when you load the core, do you get the philips bios and load disc screen?
Nope. Realizing I probably need to download a bios lol
there are a couple of em. check the pins
Just finished Zelda the Wand of Gamelon so that's completable
In today's core
Had one hard crash during the beginning of a cutscene while completing Link Faces of Evil, when i was buying bombs and hit store keeper by accident which triggered the cutscene.
In wand of gamelon i had 2 hard crashes basically same reason, when a cutscene started
Very sweet to see the core progressing! Hopefully there's room for DVC support! Wanna play Lost Eden on the CDI
My first contribution to Redump is out in the wild, for all you Photo CD fans, you can now experience "Cumana Photo Album - Volume 1 (UK)" 🙂
I found a Photo CD that hadn't been dumped before (on bad dump out there that doesn't work) and mailed it to a nice man in the North of Scotland who dumped it redump standard and submitted it to Redump.
Oh nice
I also sent him a copy of Autosport Magazine Presents Grand Prix Special (UK) that had different cover and disc art, but turned out to be identical to the other version dumped, but that is now confirmed to be the case, so that is good to know.
Last time I Cumana photo album I got kicked out
Moondandy does like his PhotoCDs (͡° ͜ʖ ͡°)
you're all disgusting
Let he who is without sin cast the first stone
That is sad to hear. For quite a while I haven't played very deep into the games. This means there is indeed still a stability issue. This reminds me a little bit of my work. We usually do integration tests which are running for multiple hours to ensure the machines are surviving many hours of operation. This is something I did once to test the audio map and it seems it might be required to write an automatic test as well for CD reading. Random errors are a pain to debug. I put this also on my TODO list.
Edit: I assume it is about CD reading... also was it the beginning of the cutscene? Or was it the abort of a cutscene?
You dare bringing light to this lair? You must write into the testing sheet! flash flash
That was a happy accident. I've just tested it with my current work in progress core and it simply worked. Debugging takes usually days or weeks.
The crash might have been when i tried to abort the cutscene by spamming buttons because thats what i was doing when i accidentally activated the shopkeeper cutscene when i was buying bombs and in wand of gamelon one crash was when i entered a boss area and cutscene plays immediately
But was the crash during the loading of the cutscene? Or when it automatically ended?
Was on gibdo cutscene
It loaded and crashed assuming when i pressed a button to skip it
That's pic of crash where it just freezes there
Ok, this is a known issue I haven't seen for quite a while. I've actually added this as an issue for MAME https://mametesters.org/view.php?id=8955 The bug is confirmed but is not yet fixed. In that case the CPU just refuses to move on.... I had hopes high that this issue was not striking my core as well as I haven't seen it for weeks.
At some point in time the problem was just gone.. But I see it now, it is very much reproducible. There is a certain chance that talking to the shopkeeper kills the game. The issue is back. 
It's not such a big deal if make sure to save often but yea still an issue that happened to me three times between across both games. Only with the cutscenes so should be easy to reproduce by spamming the shopkeeper cutscene
One would assume
Not only assume. It is possible. About 10% chance if your are unlucky. I had a sessions once in which I've talked multiple times to the shopkeeper to test if the issue was gone or not. It was tedious but I thought I had it...
This is not related to CD reading. I've actually made a modification to MAME to track it down. The CPU is in spurious interrupt state. The CDIC wants attention but the interrupt status flag is never reset. This also correlates with remaining sound issues as the expected state of the software seems to not represent what the hardware is doing.
How did you get on with the RTC? That proving to be another rabbit hole?
The RTC is already working 😄 The datasheet of that thing is public.
Oh nice, need to check the latest core (added to WIP DB for people who use that). Cool you got that one added in. 🙂
Building hardware that is documented is not that big of an issue. It took me about one day to get that down and tested. The timekeeper is also rather primitive.
If you ever need another break and side quest, if you figure out how to get Photo CD player to save when you set High Res on that would make my Photo CD testing adventures a lot more enjoyable 🙂
Silly question. Apologies. I downloaded the CDI bios. Does it need to be renamed or just dropped in the CDI folder as is?
Everything is in the pins: The name and the location
My first question is probably whether the real machine also has this issue. If yes, I don't think I can actually do something about this. I'm no magician.
Certainly some of the discs explicitly state that the once turned on the option should remain
I think I posted an example before but can look again and find one that has that message
How would that work is there a Battery Backed Ram isnide the CDI where details like that could be stored
I don't think you get what I am asking. Currently:
Load Photo CD image in Low Res
Open OSD in Photo CD player
Select High Res tick box option
Image reloads in high Res
Progress to next image
Image loaded in low resolution (High Res tick box has defaulted back to being unticked)
well you need proof it did that on the real CDI
Curiously though, for images that have been loaded before in high Res, when you come back to those images they load again in high Res, so it has saved that info
And yes, I am trying to get this confirmed. If someone with an actual CDi can check that would be great. It does explicitly state on the help text within some Photo CDs that the option should be saved once turned on
still dont get looking at low rez photos when better stock photos are available
one would argue it already does since you state once turned on it stays on for that photo
It sounds like you haven't explored the format. Many discs are more than just images on a screen, they have music or voiceover and are interactive.
I dont care for interactive cd's . They were rubbish back when first out and now are even worse
Cool, then please go find another chat to post in.
no
this is the CDI chat not the I love Photo CD chat. Please create a I only want to hear nice things about Photo CD channel
Channels are not meant to be worship channels.
You should rename yourself Tim the Disenchanter 🤪
I would advise you to go read the server rules.
explain which rule
These channels are to discuss the Consoles/ Computers. I am stating my view Just because it doesnt agree with yours has no bearing. I am not attacking you in any way.
The channels are not meant to be a worship channel where nothing bad can ever be said about the channel's topic
All consoles have good and bad and in my view photo CDs are one of the worst "features" of the CDI
I like you Tim, and you have said your piece, now take the negativity elsewhere.
No I will continue to discuss the CDI at any time. It not negative to say photo CDs were a bad idea then and still are now. If you want an empty echo chamber start a new channel
Don't derail this thread into petty arguments. Thanks.
Photo cds whether good or bad are fascinating to look back on and explore though right?
I mean photos are fascinating
It's all kind of moot to discuss if we don't even know if the original CD-i did what was mentioned above anyways.
Its a look what we can do with at the time was "infinite" storage
As I stated if you read what Moon said. Its higher likely that it actually means once each photyo is set to high res it will remain hi -rez for that one photo
but yes we need a real CDI
each photo cd software might have a different software-based method of turning that flag on, so it might not be as simple as just an OSD option.
it would need some sort of storage though
and the CDI was a read only medium
and as far as I know there was no battery backed user setup
maybe you could do it with a small ram disk
it has nvram
ah
Each CD-i comes with 8kB of SRAM inside the timekeeper. The SRAM is powered by a cell frustratingly contained inside the same package.
so each CD id could be stored with a flag
If the software does so, yes. An entry like that can be found on the system menu when selecting memory. Keep in mind that the disc must be inserted before as this core, like many others, is reloading the NvRAM based on the inserted image.
That makes sense
but you could read it strictly as saying that its per photo
but without a real CDI it will remain a mystery
Yeah I remember this, there's a special place in hell for whoever designed it like this lol
no worse than the Genius at Commodore who put the battery backed clock on the Amiga 2000 right besides the 68000. Great in the 90's not so good when they leak and corrode all the CPu traces
Please stay on topic folks, I don't want to have to keep removing random gifs and meme comments.
Here we go, this was the message I was talking about:
#1103404843512631357 message
On "God Streams"
That certainly implies that once turned on it should stay turned on to high res for the rest of the disc and not default back to low res
So if anyone has OG hardware and would be willing to check, God Streams (Japan) is the disc to try (it is also a really fun disc with early 90s ambient score)
What if the meme is Hans Gruber saying You asked for a miracle, I give you the C..D...i?
Like this?
pins updated
You should pin that comment.
Sorry to hear that.
Now who's gonna be brave enough to finish Zelda Adventure because that game is just plain bad. The other two zelda games are least entertaining to play with the weird art style and meme worthy cutscenes
Truthfully I expected Zelda Wand of Gamelon and Link Faces of Evil to be much worse, they actually were not the worst games i've played. They're jank af and terrible compared to any other zelda game but on their own, not a bad way to spend 4 hours to finish
compared to the rest of the catalogue of games on CDI from what i've looked at, that don't require the dvc, these are Mozart
probably the only two games I'll finish on the core 😛
Badly recorded footage of end of link faces of evil on the mister
Congratulations?
"I won"
I played a bit of the Gameboy remake of that game lol.
Oh neat, I didn't even know that existed
Is that a recent fan remake/demake?
I think you and Elphive are our champions
@teal grove was in this, who knew
For a game like that I should really add an option to remove seek time💿 as the screen transitions are killing the fun. Right now it is based on measurements from a real machine I've made once. Maybe 180ms is too long for a track change. Games like Flashback are also suffering from frequently changing track. If not done already, the transitions can be shorted by overclocking the CPU in the debug menu.
The overclock does work to make the transitions move faster between screens but it's still an unbearably boring game 😛
Hey all you can do is try to make it accurate. You can't make bad games good unfortunately 😂
Cores like the CD-i make me have insane thoughts like "Somebody should make cool homebrew for this" or "Why hasn't anyone made a hack to fix this game"
Do the tools exist to do either of those?
From what I've seen the 90s era authoring tools for making CD-i games are out there, but I haven't looked into if any of them are still usable.
Every year I like to put a retro-style yule log on my MiSTer and the CD-i core might be the best option yet. There's already a Christmas sing-along disk from back then, and authoring your own CD-i disk is a lot easier than making a FMV disk for the Saturn or PlayStation.
Plunerball is a pretty neat homebrew pinball game. Physics are a little wonky but it’s a fine time.
Last time I've played I got to the Bowser Hotel, Stage 7. When I gave up, selected no on the continue screen and was able to save the progress, eventually I'll try finish it.
I need tips. Once the caterpillars hit in level 10, it becomes stupid frustrating
The tools from back then are still usable today. I know this, as I'm writing my own CD-i software to test my core. Some people on the CD-i discord server are very open minded and publish their source code online for others to learn from. The homebrew title Nobelia is on github: https://github.com/TwBurn/Nobelia The same goes for Frog Feast, which exists somewhere as a zip file.
Keep in mind that the tools are old and while the compiler and linker are 32 bit windows applications, the mastering software is a MS DOS tool.
As I'm a linux user, I'm running the compiler with wine. This works without much issues so far.
It's also nice to know that every CD-i made by Philips has a serial stub loader. This means you plug a RS232 cable into the port on the back and with cdilink you can download an application for execution without burning a disc. Neat.
Nobelia. Contribute to TwBurn/Nobelia development by creating an account on GitHub.
The C compiler of the official CD-i SDK can only be described as archaic. You can't even make // comments, so old is that thing. For my small projects that is ok, but for larger applications there might be better solutions. Something like m68k-elfos9-gcc should exist according to a branch on Nobelia. Haven't tried that yet but I assume that the GCC will deliver a more optimized result and a more modern type of C. For more reading, the guys at https://discord.com/channels/675054061665386544/797170589604511754 will help you out
I know the guy who released Plunderball - I had him for a game history professor, and built his arcade machine.
The game was originally designed to be released during the system’s lifespan, but got shelved due to poor sales of the hardware. He licensed it from the original creators and released it.
Ohhhh, that explains the high production values. Very cool story!
Does any CDi emulator support DVC? I tried the one Batocera uses and it doesn’t.
Also the CDi MiSTer core is a nicer experience, for me
Mostly no. Several have the expansion RAM, and I've heard it's possible to get DVC working in the closed-source CD-i Emulator, but I haven't been able to get it to work myself.
There’s a closed source CDi emulator? Is that available to the public?
I dream for a DVC emu
That’s pretty nice
You don't get to decide that
That's allegedly pretty nice
That is a responsible response. You should avoid possible litigation that way. @stiff flare on the other hand...
I won't disagree with you there, I'm assuming you're jumping on the caterpillar heads to kill them at least? Bad part about them is it takes like 5 jumps to kill them and the jumping mechanics are slightly jank
Just little off center on the landing and mario dies 💀
Our favorite file type 
As one of the few kids in the USA that grew up with a CD-i I can't wait to get this working on my MiSTer. Still have my CD-i but finding working controllers has been rough.
Yep! But the sneaky bastards regen quickly and it becomes a circus act. 🤣
Every time I tried to just jump on its heads directly, I died. Only worked jumping into a head and back to ground, wait a little until new head appear, jump again, repeat until there is no more head. I played until the bowser castle without knowing flower Mario could throw fireballs.
There are multiple projects available which enable you to attach gamepads from other consoles to the CD-i. The protocol is no secret.
I did get an adapter last year that does let me pair Bluetooth controllers. That has definitely made things easier for sure. The he controller hunt was prior to the availability of these projects.
I've not yet checked on a real CD-i player but I think there might be a misunderstanding. The Photo CD is indeed storing some configuration data on the NvRAM. The high res setting is stored for every single photo and not for the whole CD. I've performed a quick test on MAME to check for that. So ... it might be working but maybe not the way you wanted it.
Also some bad news. The recent core has a regression. 😦 The score screen of Hotel Mario has glitches again in the audio. I assume this resulted from my recent changes. It might be possible that this game is indeed capable of crashing during the Score screen, although I haven't seen that yet. That was one of the issues in MAME i've somehow resolved without fully understanding it.
This glitch was actually something I've build a test tool for. Investigation is needed. You have been warned.
The problem is that the cheapest solution - a Bluetooth adapter - doesn’t work correctly on about half of all U.S. consoles, and the Genesis controller adapter is $65, which is up there with a price of an original controller.
I know there are plans to build an adapter from scratch, but I haven’t been able to find anything currently available in terms of a prebuilt solution.
Oh, that's interesting if that is how they work on real hardware. That seems like bad design, and the help text is very misleading. If you check it on real hardware please let us know what you find.
I've wondered what you need to configure to actually get notified whenever someone tries to change something in certain parts of the MAME code base. You seem to get notified very quickly. What did you do?
This Vincent person is currently very active in the CD-i code of MAME. I've already told him, he should join the CD-i discord server to get in touch with the other emulation authors. It's nice to have another one join the CD-i forces.
i'm just checking each day following closely mame project because it is my main software emulator
Same here. MAME is about the only software emulation I use now (thanks to MiSTer), and I've been using it consistently the longest out of all of them (since approximately '99).
I've played Hotel Mario until the 6th hotel so far and god I hate this game but I refuse to give up..
I hope you haven't experienced a crash so far. The audio might be currently in a more unstable state. I'm actually not sure whether I should integrate a debug option to revert that change. But that somehow feels wrong.
No crashes at all! Kinda hard to tell if there was anything with audio but it wasnt something me or my friends noticed while they were watching me stream it.
I've had the same issues with my M30 2.4ghz that was noted earlier with extreme input lag, but that was fixed by swapping to PAL and enabling the input overclock!
So the core itself feels great
Hmm, so when I try to load a game and hit the play button in the BIOS I get a full turquoise screen and then it blanks out and goes back to the BIOS.
Are you on unstable Main?
Also, are you checking games confirmed to work and not ones needing a DVC?
Yeah, I'm using the pinned version of the core and games on the recommended list that don't say (DVC).
And my bios files seem to work on cdiemu.
I've gotten 3rd Degree and Burn Cycle working on cdiemu so I think they're functional dumps.
Are you definitely using unstable version of Main? That is likely the issue there. If you can double check on the MiSTer menu OSD what version it lists you are on we can confirm or rule that out as an issue
The filename is CDi_20250223.rbf
Oh, I understand what you're saying now.
No, I'm not.
Thank you!
Yep, works.
Hi where can i change or find the main file for play cdi?
in the pinned comments
Unstable MiSTer main doesn't look too be pinned, the latest one is here #unstable-nightlies message
Rename the current MiSTer file on the root of the SD to MiSTer.old or something, copy new one over and rename to MiSTer, then reboot
ty so much
NP
I assume the new main that just went out has the CDi stuff
Ahh nice, 20 mins ago
I can confirm in just a sec. Pulled apart this monitor to get pics of the controller board. I’ll put it back together and test CDI
Yeah I'm sure it will, hadn't seen the patreon email 🙂
distribution_mister just updated
The apprentice seems ok with the new stable main 👍
nice
@stuck bane no need for the unstable main any more, run an update and the new stable version that just came out has the CD-i fixes
@south urchin Thanks for catching I didn’t have the latest unstable pinned, just fixed it
Haha, a new stable was released a few mins before I posted :p
Oh lol the latest main has the CDi stuff
RIP
Ok edited it to say to update to the latest Main
Cheers to Jesus for the heads up, n you for keeping current info pinned 🙂
Is save states an option you're working on @terse flare or is that something that would either be difficult to implement or something you'll save until you have a better grasp on how "full" the core will be in the end? The games i've tried so far has fairly robust save systems so it isnt exactly the end of the world without it
I don't yet know how other developers did that. Save state functionality would require a full redesign - I think - as every single register needs to be backed up and restored. I don't have plans for that right now. It also would require to be able to halt the core. My priorities are still within the accuracy.
We had a similar question about cheats around here. That would be easier, though I don't know whether a cheat collection exists for this machine, considering the lack of popularity.
I can imagine that games like Dragon's Lair would make slightly more fun with a working save state system as it is quite unforgiving. As you have already noticed, most CD-i games seem to be designed around short playing sessions with their save system. A save state mechanism would not help that much.
Since I'm already here, I think I can give some updates. Some of the accuracy issues this core still has, are coming from missing understandings of the inner workings of the CDIC. Back when I started this, I just took the implementation of MAME which was based on cdiemu and turned that into Verilog. This implementation only used reverse engineering of the CDIC driver in the boot ROM as base and has issues with certain edge cases.
Therefore a new project is born:
https://github.com/Slamy/CDIC_BlackBoxAnalyzer
This is an application written in very archaic C, running on real Mono I hardware. It stimulates the CDIC and documents the responses. My idea here is, having an automated application to first analyze how the chip really behaves and then make this tool usable as a verification mechanism for emulators to check the accuracy. So far this is very new and I don't know if this gets me anywhere. But evidence shows that this approach was not yet performed.
Vincent-Halver, wo has recently joined MAME development, is more focused on graphics and maybe DVC. I'm curiously observing. He also doesn't have a physical CD-i right now.
Reverse engineering at it's best, ladies and gentlemen! 😮
I think we would much rather have the digital video working 😉
Excited to give the new update a try
I can't let this thread go 2 days without a comment, so desperate measures are called for
Im going to post a joke that occurred to me and was deemed to lame to post:
What is a birds favourite console?
A SEEDY-i
😞
Ah, I'm sorry for having no updates right now. I'm still busy torturing the CDIC.
Why do I feel so guilty for not giving updates every day? Maybe I shouldn't. I'm not an entertainer 😄
You should not feel guilty, at all.
Its your hobby Slamy, we are just along for the free ride 🙂
How you getting on with the CDIC stuff? Feeling positive about your new approach?
I'm most of the time pretty negative but I think it is ok. I've started writing a register map and also started writing a manual. The repository mentioned above has now a collection of tests and some logging of the registers going with them. My goal was to replace the "what we assume" with "what really is going on". It seems most CDIC emulator code was written around how to satisfy the software and just don't touch it too hard and close ones eyes and hope that everything will be alright. The real thing behaves sometimes differently and has additional bits, I've not yet found in any emulated implementation. I don't know in how much detail I can go here without causing mental damage to you all.
I will probably write the deep technical stuff mostly on the CD-i discord in hope to spark some curiosity. Vincent seems to be pretty motivated at the moment as well in improving the situation on the software emulator side. I hope that this is yet another tool which helps all emulators.
Just seen your post on CDi Discord, looks like some great discovers and documentation you have done there
Yes, I thought it made sense to put some caffeine into their thread by posting my findings. The analysis is also still not complete and I'm not yet sure where to apply these findings first. I could try to put them in MAME first to have some faster testing time. In the end I hope to have all audio related issues resolved.
Do you think CDi Fan will try add to his emulator?
I don't have much hope right now. I don't know how I even should phrase that in public. He seems to be occupied otherwise. The emulator seems to stagnate. But maybe I'm just seeing things wrong and in secret he is very busy further refining it
Ah well, maybe trying out in Mame is a good option then
I'm just messing around of course 🤪 but rest assured, if the thread goes dormant for too long, I have a full rewrite of the nursery rhyme "Old McDonald" using the phrase "CD ee ay oh" which I know the community will appreciate.
So never feel any pressure to post
I got your back
Has anyone donated a DVC yet? Just curious.
I just found one for less than $100
just wanted to say i love reading the detailed updates in here. fascinating stuff everytime. thanks!
Oh don't worry. When I bought my 210/05 it came together with a DVC. The device you have there. It might not be compatible with Mono I. There are actually multiple DVCs with a different pinout.
Edit: Just as I thought. That one is for Mono IV and Roboco. It is the 9956
It'd be pretty wild to be able to load vcd images onto the mister and use this core to play them back!
Which also begs a goofy question...
I wonder if anyone ever dumped laser disc images
A core for a Pioneer laserdisc player would be wild lol
There are LaserDisc images out there - I believe one's already been done for Pyramid Patrol that's circulating in private hands.
I tried to archive a few LaserActive discs, myself, but I ran into an issue where the off-the-shelf player I have refused to read the LD-ROM discs and forced a disc stop. (That, and one of the rarer titles I ordered - an English copy of Manhattan Requiem - cracked in transit, discouraging me from looking any more into it until I figured out a way to play them.)
I think the key issues with LaserActive are going to be a mixture of decoding and bandwidth.
Unlike the LaserDisc arcade games (which just used standard discs controlled by an external program, meaning you could read all you needed to get off the discs by capturing the video & audio alone), LD-ROM discs use a subcode format to store the data internally on the disc, essentially hiding Genesis and PC Engine game code within the disc's signal - you can overcapture, and recreate that decoding in realtime, but that creates a file so unbelievably massive that many devices (including the MiSTer) can't currently handle it in realtime. (And, when you're talking about 100+ GB files for standard definition, no one would want to store that much data for a single game to begin with).
I think the eventual solution needs to be decoding the game data out of a direct RF capture (with something like ld_decode), then multiplexing the decoded file back into the compressed video & audio in a method similar to the Matroska container format.
But, given the rarity of the games and the complex & infamously delicate hardware, I'm not that surprised people haven't put an emulator for the platform together yet. There's a lot of details that need to be hammered out just on archiving it, let alone getting the software in a playable state.
I mean laserdisc movies too. The untouched original Star wars trilogy is still best on laserdisc
'95 is a bit better than the GOUT, but I think that the fan 4K77/D+77 restorations blow most of them out of the water.
Granted, the existing quality of the archival work done with those discs is insane - not every Laserdisc gets that kind of care & attention with digitization work by fans.
I have watched the 4K77 yet. I need. I bet those are fantastic
Heh just repaired a couple of these, and did a full recap on one (it’s intense) and a partial on the other but they both work at least. I’m sure there are lots of failure modes but mostly it’s stuff you’d expect (belts, mechanical issues, lubricant) and then some less expected stuff (ICP transistors meant to protect actually causing a power imbalance that fries the laser pickup when only one fails of the +/- voltage rails). Then output quality issues that were down to donor compatible laser assemblies being themselves quite tired. Good times. That and the pacs all self destruct if you don’t recap them so I’m sure some of the fragility is just people being neglectful
Wow, I never ever even want to try to repair one of those after reading that
Eh, between the console5 wiki and consolemods wiki it’s all well documented, just time consuming
And zaxour is a nice dude who will answer questions
I never seen one in my life so I'd be extremely unlikely to come across one anyways, couple people I've mentioned the CDI to IRL had no idea it even existed and they're 10 years older than me, well within the age to have been there when it was relevant
I mean i wasn't even alive when the CDI was released
I knew about CDI but somehow we already knew it wasn’t good for gaming, I didn’t know anyone who owned one. I vaguely remember a store display, possibly in the now defunct Service Merchandise
Guess gaming mags must’ve been mostly how we knew
First time I heard of the CDI was from AVGN back in like 2007, never seen one ever IRL
Not like we had much other news than talking to other gaming nerds and magazines
I had a Nintendo power Subscription back in 2003 that came with a Conker's Bad Fur Day walkthrough as a sign up bonus, That's as close as I got to the classic magazine scene
Wasn't long after that magazine culture was dead so
Didn’t even know it ran that long
last publication was in 2012 so thats a few more years than you would think
I still have my nintendo power magazines from the first volume through a number of years of subscriptions and the strat guides etc you would get
I bought an xbox magazine magazine once that talked about saints row pre release and i only bought it because it came with a demo disc
I still got a bunch of my 2003 Nintendo power magazines but they're destroyed ish because it was a 7 year at the time handling them along with other 7 year old friends
Yeah mine are a mix of good shape to fairly ratty
Greetings! I tried out a few CD-i games and the wacky worlds mario prototype today and everything seemed to work fine. Can I or someone update the spreadsheet?
That depends on what you tried and if it's any different from the previous reports
I only ever add anything to the spreadsheets in the past if I completed something so i'm super bias to what I report on them
Oh lol, I would say it's "playable", I mean, the protype booted and I was able to load up a couple of different levels. Controls were a little strange, but I think that might also be a problem with the prototype itself as well
Try this to improve controller responsiveness:
If your controller feels sluggish; try either setting the console to PAL, using the analog stick instead of a d-pad, or access the
Debug Optionsin the core menu and enableTurbo CPU.
I’ve got more tips in the pins - #1103404843512631357 message
Oh awesome, specifically the issues we were seeing in the Wacky Worlds game (I was playing with my daughter) was that we had to hold UP and right in order to go forward. I was thinking it may have been a bug in the software rather than the core.
Before we realized about holding up and right, we were able to go forward for a bit, then backwards then forward and repeat
Yeah that prototype is wacky
That's what I thought also, but I did want to report it since I tried it and it wasn't on the list.
ANd it was Mar10 day 🙂
I knew about it from watching the infomercials that constantly played at night. https://www.youtube.com/watch?v=mW8PVEe4qTw
Stitched together from a few videos found on Youtube. Can Phil survive The Wall's deadly game? Or will he spend an eternity of suffering digitized in the CD-i?
Audio is quite low due to some issues with the source videos. Just a heads up.
I bought it a few years later in a mall but had to buy the software and controller through mail order.
Hilarious
Late night cable got weird as shit sometimes ;)
I always had to be different. I was also only 1 of 2 people in my area I think that owned a TG-16.
Of course I picked up a Genesis and SNES later..but then I went different again and had to have a Virtual Boy. Though to be fair those things were only $50 brand new.
Even SMS & Genesis were less common in my area. CD-i & 3DO were mags and kiosks and nowhere else in my world
Zero people with NeoGeo as well, though I knew someone who rented one and even that was expensive
Yeah I didn't know anyone with expensive systems at all, those thing were vapor hardware as far as I as concerned ;)
CD-i was pretty spendy
CD-I was expensive, but that was like the one expensive thing I think I got my whole young adult life save for like a go-cart. I also think it was on sale or something maybe as it was on it's way out I'm sure.
I think I was hoping it was going to be better than my friends Sega CD. 🙄
Turned out neither was good
Yeah..quite terrible for sure. I spent months teaching myself how to use that POS wireless remote until I finally got a controller in the mail. Probably the only saving grace was the sexy scenes in Voyeur.
I only ever had Keith Courage for tg-16. Which wow...it was so bad. Of course everyone knows that now with emulation.
Why they didn't include Bonk is fucking beyond me.
They eventually included Bonk.
Pretty sure Bonk wasn't available yet at release... but yeah, poor choice of pack-in
I bought long after release ;)
It was probably still sitting on the shelf. 😉
likely I'm sure >.< it was also a rural area which tends to get everything last heh
But that keith courage title screen music…
@gray badge With the latest update, it seems like the full Europe set of Games is at least Playable or higher. Probably a good point to go back through testing the other sections, if you’re up to it.
I can do some today. Tracking down a PSX issue too
Godspeed. You do a great service to the community with all your testing, it is much appreciated.
I'll run through the regional exclusives. Excited to see how things look since I last tested in november
I won't know if things are completable though because I was educated in the US and therefore only barely speak english
Not sure if this is present on real hardware but, whenever there is a cursor (mouse or whatever), the horizontal movement speed seems 2-3x the vertical movement speed
Halfway through the EU regional exlcusives and it's incredible to see how far this core has come in the 5 months since I last did this list.
Every game is playable so far, and the only graphical bug I found was that there is a line of corrupted pixels on the edge of the screen in alice in wonderland - but I'd bet that's in the overscan.
Really excellent work, Slamy
Is the line of corrupted pixels on the left or the top edge? I'm asking, because I've seen something similar in Laser Lords which seems to use the same engine.
That would be totally new for me? In every game?
Yeah. It’s really noticeable in earth command because the targets are small.
I’m using a m30 Bluetooth controller if that helps
On the left border
If you are on a roll, be worth checking the stragglers on the Asian and Homebrew tabs, probably only a handful to test there that didn't work previously
Hmmm.. I'm unable to reproduce this problem in Earth command. It feels like a 45° angle. Let's get more specific. There is the controller debug test cd. Are you aware of it? I'm curious whether you have the same issue on that one.
To make matters even worse, Earth command just hanged up during a cutscene. >.<
This is interesting. It is actually slightly different compared to the glitched line in Laser Lords.
I saw you guys playing with a controller test, but I'll admit I wasn't paying too close attention
Is this phenomenon reproducible with the MiSTer input test when viewing the analog signals?
walk me through how to test that and I will
Select Cores -> Utility -> InputTest.
Hold Select -> Select Analog
ok, got it open. what should I be looking for? it's a digital pad so it just jumps to +/-127 on all axis
For both axis it is from -127 to 127?
yep
https://github.com/TwBurn/cdi-debugcontrol/releases/tag/v1.1
Please download the Zip, put it on the MiSTer. It is a very similar test but runs on the CD-i
i'm glad we have all these testsuites to narrow these things down
also, left and right are one step, the start in the middle and then end up on the left or right. up and down are two step. first a little up or down, and then as far as you see in the screenshot
Up and down are digital and left and right are analog.... I don't know why though
is that true on real hardware?
Ok, let me explain. You might have noticed that the CD-i has multiple different controllers and also a mouse as official peripherals. The way the core behaves right now is this order Analog > Mouse > Digital. For some reason the decision gets confused. With a PS3 controller I use, the dpad is digital and the analog stick is analog.
There is something you might want to try.... it's erasing the button mapping. The MiSTer is really bad when it comes to artifacts on the drive.
Please log into your MiSTer via SSH or other means and look for CD-i files inside /media/fat/config/inputs. Delete them or move them away
For some very weird reason, the real analog joystick has a wider range than the digital ones. This ... might be a problem, considering that your DPad is rather aggressive. Seeing this, I would assume that a "Disable Analog" option might be required to avoid issues like these.
@gray badge This is an experimental build for you. It adds a debug option to disable analog input to experiment around with. I would assume that this fixes your issue. I'm not sure how a final solution might look like.
In this regard, the CD-i is more like the Vectrex. It is Analog first and digital second because there is no digital movement in the operating system. The digital is only an interpretation by the program running in digital style peripherals. This is why there are two speeds. The paddle controller is the one in charge for that.
Maybe... if problems like these keep ramping up, the Analog should be disabled per default and only enabled for people who are knowing what they are doing.
Edit: Please don't use this core for anything other than joystick tests. The CDIC is broken in this build. 
yup, if I disable analog, I now I have same sensitivity on all axis
Back to testing: i noticed in Kether (Germany), the bottom....8th? of the screen will get messed up if you pause during stage(?) select. Might be true on all version.
Also Kether (germany) (v3.08) just boots back to the bios. Not sure what the story is with that one as there wasnt a netherlands version of 3.08
Dentnz was working on a dragon's lair core but that was long ago does that count? I assume he gave up
That sucks. I'd love one just for laserdisc image movie playback
Daphne that's the name
Yeah! I have Daphne on my PC
Ok, Europe Regional titles all tested. Only one didn't boot (kether 3.08). I put a couple of notes for things I noticed, but it's come a very long way in 5 months.
I'll leave US and Asia for others to test
For the 7th Guest, the game requires a DVC (Digital Video Cartridge).
I saw the question on the 3DO channel.
Hold Minus and up until the red led blinks once, it will make you stop using the left analog stick.
Even when remapping without the M30 analog mode in Main, Mister still makes the controller move the X axis when the controller is set to left analog stick (Minus and Left until the led blinks).
You will need to remap inside every core you mapped the controller to be able to use the dpad again.
what mode does that put the m30 in (I’d like to keep in xinput if possible)
Doesn’t change mode, changes dpad to work as a dpad
Holding minus and left for 5 seconds changes it to left analog mode
Minus and right is right analog stick
I’ll check this later today
I have seen that image 4 or 5 times now, and each time I’m like “I should save that”
I should save that
It's in the pins in #controllers btw
8bitdo really need it as a sticker on the back. Too obscure.
Yeah I have one of those and all I know is that I occasionally accidentally turn on autofire and have to flail around for a while trying to figure out how tot turn it off.
Confirmed that this works with the 20250223 unstable. Disabling analog doesn't appear to be needed @terse flare
I was able to complete a game of 3rd Degree with friends on the core last night - as such, I upgraded it to Completable.
Wow, looking at the sheet looks like nearly every game is confirmed at least playable now
yeah, absolutely great stuff
It was fun to remember how stuff was when I tested it in november (october?) versus how it is now
I would be surprised if the majority of titles weren't completeable
There are still some issues with accuracy and stability.😬 Hopefully I'm able to iron them out as well.
its in amazing shape, Slamy. really impressed with what you have accomplished thus far. we are lucky to have someone as passionate as you are about the CDI writing this core 🙂
Yeah definitely. Might be a good time to start digging into the non games and see how the software is looking. Not sure if we have packs for any of those yet
You call it passion. My colleagues at work call it insanity.🤣
I found it hysterical reading this. I saw that image n was like " wow I'm saving this to my phone!" 😂
Is there a beta/alpha of the CD-i core available for people to test out? And if so is there a place to support the programmers creating it?
Check the pins, Shane
Haha, thanks again. 😄
The pins hold all of human knowledge
They now have fancy title fonts and the external links are now just called LINKS
Robby is the kingpin
Thank you for this bug report of those graphical issues. I'll put them on my list. This is a curious issue, haven't seen something similar in other games.
About the boot back to BIOS... that might have been an actual true crash. The operating system of the CD-i will reset the whole machine when something unexpected happens. That is actually not that good. If the crash is reproducible I would like to know the method to reproduce it. Random errors are the worst.
the crash with Kether 3.08 is consistant - it doesn't boot (but I also don't know if the disc image is good - germany is the only region with the 3.08 variant of the game)
Programmers? I wish there were more than one but I don't have much help so far in development. 
Oh that's even more interesting. I'll put that on my list. It might be necessary to also burn the image to test on a real machine... does this version of the game work in other emulators?
@gray badge? More like theGOATfish! 
I haven't checked but I will 🙂
Interestingly, I don't get that 1px width column of corrupted pixels when playing alice in wonderland in DV
Ok, US list done.
- Name that tune freezes when you try to start a game.
- Wacky mini golf freezes after a couple of strokes.
- Couple of games with minor graphical issues.
Made notes on the spreadsheet 🙂
Wacky mini golf freezes after a couple of strokes.
Happens when you get older I'm afraid
It’s just a little weird with Eugene levy watching
You get used to it
That’s a relief
Maybe I'm living again under a rock. What is DV? Digital Video?
Thanks for doing all these tests. I'm usually testing only the things that are knowingly broken to me as fixing them sometimes fixes other issues as well. The testing sheet gives a good overview of the cores state.
DV = DirectVideo in this context
usually using cheap hdmi to vga adapters to get analog output without an IO board or for direct connection to external scalers
yeah, as Freddo mentioned, it's Direct Video. Mister going into a retrotink 4k
This is a weird finding. I wonder if the left edge is always missing with DirectVideo. I don't test DirectVideo very often. For core development I usually have the AnalogIO v6.1 with a CRT and a HDMI monitor attached.
It could be that the crop information in the DV infoframe crops that band of pixels out. I will try turning off autocrop later today and see if the pixels return
do they still put ™ behind every other name like that nowadays? it was/is such a strange practice, like anyone's going to care
@terse flare
We were robbed of a CDI Donkey Kong game.
Yes! Would have been great to have a bizarro DK game in the arcade game style from Philips. I’d love to have some weird Hotel Mario like DK game
I celebrate obscure and weird stuff and don’t shun it. I think we’re way better off with the CD-I Nintendo games than w/out. It’s like a view into an alternate dimension seeing 3rd party devs make Nintendo games w/out Nintendo’s oversight
The Nintendo games are definitely the killer apps of the CD-I and the main reason behind the allure of the system imo
I can only agree. Without those games, the AVGN might not have reviewed the system. Those 3 episodes in 2008 have sparked a certain interest in this machine for me and certain friends of mine as well. It even shaped the way we talked as lots of cutscenes are as quotable as other popular franchises.
I googled goatfish in the hopes of finding some sort of funny meme image. Only to be disappointed to discover that goatfish are a real thing. And they are boring!
fukkin goatfish
Man, they don't even have goat style eyes
I am going to have a T-Shirt made that just says “We know about goat eyes”
There are MANY goats around where I live, and I can say the eyes and the horns are totally wrong there
You live near a veritable goat-sea?
Well, kinda. It's a Natural Park called "Las Batuecas" 😄
It's full of wild goats, gotta be careful when driving at dawn because they come out of nowhere, and in great number
I don’t get it. Can you walk me through this joke?
i feel the same way. i actually played the cd i nintendo games a lot when they came out and always loved them just for being different
bellhop and elvis goombas are cool and nintendo is cowardly for not embracing them as part of mario's history
After starting Alice for the first time, I had the impression that the voice actor of the narrator is the same as Robot from Invincible. It sounds similar. They are different people it seems though.
Oh isn't that Zachary Quinto (spock from the new star trek movies)? That would have been a hell of a credit list :p
Exactly. I looked through the Credits of Alice. The name is not on the list. 😄 Sadly, only Alice has her actor listed next to her. For the others it is the same issue as with Faces of Evil. I think today it is still not known who voiced Morshu.
I've looked into the graphical issue of this game. It is caused by a misalignment of the region flags by one high resolution pixel. The borders contains garbage data which usually is not shown as this area is hidden using the regions. Maybe some insights in the graphical layers: The first layer contains the ingame footage. It is a CLUT8 frame which means it has 256 colors from a palette. The second layer is DYUV which is fine tuned for photographs. After the game is started, this person is hidden in the background.
Now, why even talk about this? It turns out, the game is slightly glitched on the right border.
If you look closely, the man in the back is shining through at the right edge. I first thought this has to be a bug in the core. But it seems that is not the case. This very small glitch is also visible on a 210/05
oh wow
Hmm, maybe a similar issue here:
Space Ranger (USA) (Unl) (DVC Enhanced)
Menu seems to look fine, but game is unplayable. The lower half of the gameplay area is not visible.
Hasn't been checked for a couple of months that one, maybe it is OK now
On the Homebrew tab
Space Ranger... Laser Lords... With all those names I would assume quite some action. I was surprised how slow a game with such a powerful name could be. Speaking of Laser Lords, it has the same issue as Alice. I think, the engine is pretty much the same. The developers recycled that.
MCD212: OFF coding is considered black instead of transparent
- Fixes red bars and single blue line at the top in Hotel Mario
CDIC: Rewritten audio state machine
- Changes based on findings of the CDIC_BlackBoxAnalyzer project
- Removed audio coding sidechannel and audio/cd synchronization
- Fixed behavior of Audio Control Register
- Fixes audio regression during "Hotel Mario" score screen
- Fixed spurious IRQ caused by sector data interrupt after reading stopped
- Should fix hang on shopkeeper cutscene in "Zelda - Wand of Gamelon"
(Cannot be reproduced after some tries or is very unlikely now)
CDIC: Seek time now 19 sectors (14 before)
- Fixes audio glitch in "Help cutscene" in "Zelda - Wand of Gamelon"
"Remember, tools can only be used..."
MCD212: Continue fetching framebuffer data with OFF coding
- Fixes graphical glitches on pause screen of Kether
- Fixes graphical glitches during intro of Zenith
MCD212: Fixed region flag misalignment on the left edge
- Fixes glitched vertical line in "Laser Lords" and "Alice in Wonderland"
MCD212: Fixed Image Control Area timing
- ICA starts immediately at vblank
- Fixes hiccups during Demo mode of Burn:Cycle
- Also reduced CPU starvation caused by ICA memory access.
(Might not have an effect, only more compliant to datasheet)
The last release was nearly one month ago. As some fixes have been accumulated during the last days, I've thought to make a release again. So far I'm not aware of any regressions caused by this. With this release the core might have entered beta phase since quite some issues might be now gone.
uh oh. are we headed to mister_devel and unstable nightlies?
Ah, beta phase only in my mind. I don't know who even decides when a core is considered stable enough to be there... Some have been added to that much too early. Saturn for example. (Iirc Nights into Dreams was not yet working when it was added to stable ^^)
if I understand correctly, it's totally up to the dev. it's whether or not you want it to have wider exposure. If you'd rather keep it "in house" for now that's totally fine.
Saturn being added to the mister devel repos may have been driven by external factors
(well the timing of it being added)
Great stuff Slamy! Might these audio fixes impact red book bugs or just green book (i.e. should I recheck photo CD audio issues?)
Once something goes into MiSTer devel it is opening it up to many more eyes on it and issues being raised in GitHub, so your call when you want to move to there Slamy
I've just performed a quick check on Dark Chaser. The short wrong audio at the start seems to be gone. Though I have to say I lack the patience to play this game. It is very slow and the seek time doesn't improve the situation.
That's great to know!
I wonder if it may also fix that weird issue with repeating the last part of the previous audio track you hear on some discs
New build added to WIP DB
I think that's the problem I wanted to point out. I recall it was present in the last build.
Ah, got you, can recheck the discs I noted that on
Is there more to be done on the CDIC side Slamy, or do you think that is nailed down now?
Good question. There is at least one title remaining with audio issues. That would be "Golf Tips". The menu has some broken audio. (Headphone users have been warned.)
Then there is still an issue with the media player when playing Audio CDs. Sometimes, the software just hangs. I haven't yet seen this behavior with games that use red book audio. As the CDIC commands for CDDA are different to Green Book, there might still be something wrong.
I did see some weird issues when I looked at the enhanced music CDs, things seemed broken in some of the menus, but was hard to tell without having a reference
The wrongly played audio in both Zelda's Adventure and Zelda - Wand of Gamelon was one of the last issues I knew of. Zelda's Adventure is just broken by design. I was able to prove that as the issue can be replicated on the real machine. And Wand of Gamelon just needed some seek time.
I ultimately parked digging in anymore on the previous build
I think Enhanced music CDs... what are they? CD+G? That would still not be supported and requires changes in MiSTer Main because of the subchannel data. Or is it something different? If you can point that out it can be checked on the real machine as well.
No, enhanced music CDs are different, they are like multimedia discs where you can play albums through them. So they have their own menus and can have bonus videos in there. I guess like a DVD menu with content in there
There is a pack of them on archive from the same hero who brought us the non DVC games pack
If you check the relevant tab on the sheet way to the right you can see what discs there are, it isn't many that were released in the format and some require DVC so don't work
We need "Rick Astley - Never Gonna Give You Up" as a CD-i disc. It is nearly April Fools. xD That would be the perfect enhanced music CD.
The Cranberries one is quite involved, there is a house you can go and explore point and click style and unlock music to play by clicking on things
Others are just an inbuilt music player pretty much
The Ennio Morricone one lets you play the album but also has bonus content like a DVD, with an interview video with him
I already got issues on my GitHub. One complained about the missing DVC, the other one complained about the difficulty of "Escape from Cybercity" even so they never played it on real hardware. Can it get worse?
Only one way to find out!
Sorry I haven't checked in some time but the core saves and loads just fine, right?
In general yes. I had a certain case though today when Burn:Cycle didn't wanted to boot in 60Hz/NTSC mode. 😓
Saving? I had no complains yet about broken save games. The default behavior is still a system reset when the CD is changed.
However, it turns out that might not be required actually. I operate the core usually without the extra reset. No issues so far.
Feeling in a pretty good place at the moment?
It felt really great last release
That to-do list must be getting smaller, the core is in a really great place now. 🙂
I feel a little more in control, especially about the audio stuff. And yes, the TODO list shrank a lot with this release. Having the real 210/05 unit around helped.
Alright, bed time. It is near midnight around here.
Pin, pin, pin, all day long. Pin, pin, pin, while I sing this song. Gonna update links, gonna make it shine. Gonna download the core, while I’m online.
Funny timing, I woke up wanting to have another go at the Zelda games and a new core update was up!
Finished Faces of Evil, no audio glitches, no crashes, no issues at all beyond saves not staying between changing cores but I knew that ahead of time.
Amazing work on the core! I never would have dreamed of playing CDi on the MiSTer ages ago.
The saves should stay between core changes from a few builds back
Rank the Nintendo CDi games
Hotel Mario
Faces of Evil
Wand of Gamelon
Zeldas Adventure
Probably
how can you rank your favorite children?
Which has the worst animation?
I repeat myself
The 7th Guest and Mystic Midway: Rest in Pieces is the sole reason i want a cdi core
It’s playable on the PC and someone did a 0Mhz version of it if you want to play it now.
even has mt32pi support
yeah the pc version sucks compared to cdi though
you don’t get to decide that
goes back to @Robbys Mom
Pretty spot on. The only ones I’d flip is Faces of Evil with Wand of Gamelon, but otherwise that’s the list in terms of quality
Yeah I was battling with those 2 as well 🙂
@terse flare wow, your last update fixed all the visualizers and audio for the European multimedia discs Antistatic and Head Crash!
eScape is much improved but it still has this weird row of pixels on the right-hand side.
very happy to see escape from cybercity seeming to not crash beyond the second sequence of the game now lol
As someone who’s played through both - the CD-i version is very slow, and some places lack music present in the PC version. It has better picture quality due to the MPEG card, however, and the added silence definitely changes the tension and atmosphere of certain scenes.
Zelda's Adventure is mostly playable
Crashed in the intro cutscene but that's skippable
More sadly, it crashed in the cutscene where Zelda falls into Ganon's Trial at the end.
Should have saved at some point. Almost had the game beat
That's strange, I'm using the latest one posted here. I haven't tested a lot of CDi games but Link FoE and Zelda WoG don't have their save persist between core changes for me. They'll stay if I reset the core but they're gone when the core is changed.
Have you updated to latest main as well
Not in awhile - that's probably it.
This is one of those games I'm not entirely sure whether it feels accurate. On my core, the game feels like it has hiccups. But I'm not sure yet.
In general, the save games are persistent between core changes. You find them at /media/fat/saves/CD-i and there is no connection between core versions. This is also the first time I've heard of a problem like this. It would be interesting to know where the save files went. Did it create another save file for the same game? The code for NvRAM handling wasn't touched in a while now as it pretty much does what needs to be done.
They work for me but when I get home from work I can try faces of evil and see if it works for me
Carefully documented crashes, hangs and hiccups are valuable for the development. Please provide more info. Does it consistently crash? Or is it random? What does crash mean? Cyan screen and back to system shell? The core is still a little bit experimental. Please save often. If you have a save file which is able to reproduce a problem, please provide it to me for analysis.
Discord might not be the best place to track problems though... GitHub Issues would be another place to document problems.
Just to be aligned on the problem.
You mean this? Without knowing how it actually should look like it's difficult to distinguish between the intended psychedelic effects and the non intended ones.
Yup! Also LOL
maybe these lines should be objects ?
Here’s a throwback to 1993, Hex’s interactive rave AV Art piece for the Philips' CD-i platform.
The sounds of UK techno pioneers B12, Eon, Coldcut and Irresistible Force x computer visuals that the user can intact with using the CD-i remote controller. Check out the 45 minute mash up. Respect to Evan Grant of seeper for excavating this, he tol...
That’s what I think too. Those lines appeared in the other multimedia discs I mentioned above but with the latest core they turned into recognizable shapes.
Not that I particularly care about eScape working perfectly but it seemed to be an easy issue to reproduce and from my ignorant perspective a similar issue that was fixed in other multimedia music discs.
isnt the pc version missing puzzles? that the cdi version has
No, the CD-i version actually had two less than the PC version by removing the soup can puzzle & heart maze puzzle.
SHY GYPSY SLYLY SPRYLY TRYST BY MY CRYPT
Apparently, the designers of the CD-i version cut it intentionally because none of the playtesters had any idea that “tryst” was even a word
The PC version is impossible to 100% now because the one of the puzzle logic is tied to CPU speed. I have always wanted to try the CDI version.
What game, 7th Guest?
yes
If the 7th Guest is so good then where’s 8th Guest?
11th Guest and 13th Doll?
I was fascinated in games like 7th Guest in the 90s. Exploring creepy old mansions. More so when virtual reality became available.
Oh shit that’s right there was an 11th guest
Ah my trolling failed!!
Yeah 7th Guest appealed to me way more than Myst
It had a legitimate mystery I could follow and the introduction was fascinating
I love Myst too! Myst VR sold me the meta quest 2
Myst is playable on the core
I think you mean the 11th Hour
i liked riven but myst always looked a bit too static to me
7th guest soup can puzzle i skipped somehow, because i remember the maze that came afterwards and how much i got freaked out every time by "FEELING..... ... LONELY??"
there is no way my 8 year old non native speaking ass would have figured out that mf
Sorry for the late response, went to bed shortly after that. The cutscene crash is consistent, using the last core posted here. If you watch the intro without skipping it the Wizard reads this letter and it just hangs with the cutscene is no longer being skippable so I'm presuming that's a crash. If you press button 1 at any point prior the cutscene skips.
While cutscenes after collecting the signs started/ended just fine the crash at the end was a similar situation to the intro crash when Zelda falls into to Gannon's trial and it just hangs, no longer being able to be skipped.
ROM-wise I'm just using a standard EU .bin / .cue and I don't believe the game was released in the US.
I'll check out the save stuff in a moment. Did you want me to open this as an issue on the Github?
Trying it real quick (I was testing the other reported bugs) this is wand of gamewhatever, yeah?
Zelda's Adventure, the top-down one.
aha, gotcha. I have never been in the europe folder because elph usually tests that. Loading it now
I don't think I've updated my Mister main in ages either way - does the CDi core still need the alternate main? I've still got the
[CD-i]
main=MiSTer_CDi
in my .ini
nah, it's all baked into stable main now
Probably why saves aren't working lmfao
lol
Yeah, it doesn't freeze for me
update main and try again
(and pull that out of your ini)
Yeah, gonna test now.
Strange, it's still crashing on my end. Wonder if it's a bad ROM dump..
Saves are working fine though which I'm more happy about thanks for your help!
Why are you sorry for the late response? Don't be, it's totally fine.
Concerning the issue, I'm slightly confused. I'm also unable to reproduce the problem on my side. In the past we had some issues around here with USB hard drives as they come with some latency. I wonder if this is really a bad dump... Hmm... Do you have access to other emulators as well? Maybe MAME? If the issue also exists there with your image, then something is probably wrong with the dump.
I sent them a DM for absolutely no reason whatsoever but hopefully it fixes the issue for them
by the power of DMing alone
Oh, it's working!
I didn't change anything but it's working.
The screen transition felt a little longer than it should have been so I'm not sure if that's the game's issue or a me issue. I'm just using a standard sandisk micro SD. It's working and saves are working too now so I'm going to try to power through the game again.
Thanks to both of you for the help!
If you say the screen transition takes longer, do you mean the transitions during gameplay when it scrolls? Or during the cutscene? In general, this game is a "heavy track seeker" when it comes to CD access. During development of this core, the seeking times got slower and slower to fix some issues with some titles. I haven't yet added an option to remove seeking times. It might make sense to add that though. Especially for a game like this.
I find it fascinating that you can replay this game again even so most consider it the worst Zelda game on the CD-i 🙂 I personally still have to try the full experience. So far I've only completed the sidescrolling Zelda games on a real CD-i.
"pull that out of your ini" 😆
(͡° ͜ʖ ͡°)
Just that one cutscene, it just felt longer between the shot change from the letter to the wizard compared to other cutscenes.
To be honest so am I lmfao
But I've been a huge Zelda fan all my life and being able to say I've beaten them all - even the CDi feels strangely nice?
It's an interesting game if nothing else. I'm not going to pretend it's good but I get what they were going for. Giving up when I was close to the end seems a shame, too.
Of course it does. You see the word "Gannon" (with 2 Ns) and seconds turn into minutes as you wonder what went wrong during development. Jokes aside, please provide some video footage showing the delay if you think that the behavior is not accurate. I've just made an experiment and made a CDI 210/05 race against the core during the intro cutscene and it is pretty much spot on. (Works on my machine
)
Oh yeah, the script is all over the place too 
there's a few areas that NPCs aren't consistent on naming, plus the game's not even set in Hyrule and one woman talks about being in Hyrule too.
It's done, though - played through it to completion with no bugs or crashes and saw the ending this time!
I'll check on emu later if that pause isn't accurate, to be honest I've never even seen a CDi in real life so I'm not sure if it's accurate to real hardware or not. Given there was text on screen maybe it's a little longer to let the player read? It's the only odd delay in the game.
Here’s whu happun https://youtu.be/H8QzxSJH0-c?si=gsBGBzUJOFYOKEhE
➤ BECOME A VIP MEMBER (and see early videos): https://www.patreon.com/mattmcmuscles
➤ SUBSCRIBE TO THE FLOPHOUSE: https://www.youtube.com/mattmcmuscles
➤ TWITTER: https://twitter.com/MattMcMuscles
➤ MERCH: https://store.screenwavemedia.com/collections/matt-mcmuscles
Edited By: https://twitter.com/Guarascia
Intro By: https://twitter.com/2Snacks ...
It’s the usual stuff: no time or money, mandates from the cd-i execs who didn’t know what a video game was
Ah, I see now. This timestamp shows the scene: https://youtu.be/IqFZ1iVmhIo?t=89 You thought that during the reading of the letter, there would be actual footage of the wizard. This is not the case. It actually shows a still image during the reading of it. Maybe they wanted to save some space on the disc or reduce efforts by not recording another scene with actors.
Now that you have finished the game (thx for your efforts) you might want to put that entry in the pinned testing sheet to document it.
It's sad, but there were a bunch of people invovled in making consoles back then who didn't play games or know what made games interesting at all. Not limited to CDi, it was also the American side of NEC, and others. Just botching every opportunity,
Even publishers and some developers.
Not names that you'd recognize today though (obviously)
I think this is also where the term Euroshmup came from. Quite some games from here were technically impressive but they weren't as funny and addictive as games coming from Japan. The biggest exception was probably Rare with some very good games for the N64 even surpassing Japanese titles.
Somehow I always have to think about Project X which is a euroshmup for the Amiga. It has a very awesome title music but is probably the most boring shmup I've ever played on a 16 bit machine. The waves of enemies are boring and there is no epic music during the levels.
Back to CD-i. Today I've decided to continue my savegame of Hotel Mario. Not only do I think that the level designer is an "Avengers Level" threat to society but the core actually has some problems in the audio department of volume. As you already might know, the BGM is on one audio channel and the SFX is on the other. Afterwards these get mixed together. Hotel Mario is far louder than both Zelda games I've tried. This is causing some audible clipping. I need to think how to solve this. One way would be to divide the volume of both channels by 2 but this feels wrong. In case of CDDA we would loose 1 bit of resolution. I'm not yet sure. It might also be possible to dynamically detect the loudness and reduce the volume if clipping occurs. Hmm.
Any two signals added together has the possibility of exceeding the maximum, or partially cancelling each other out. In the case of clipping, then both sounds are already simultaneously near the limit (peaking at over -3dB), and clipping would not be surprising. You can either sense for it and deliberately clip, or deliberately divide by some factor, but division is not really easy in digital logic (except by exponents of 2).
Fine tuned this problem for the CD-i, it's not simply adding two signals together as there is a matrix applied here.
left_out = a * left_in + b * right_in
right_out = c * left_in + d * right_in
The typical Audio CD would max out a and d leaving the others at 0. This would mean I don't need to do anything at all and things are fine. But some games are having every channel fully amped up to mix stereo to mono with maximum volume.
In software I could think about summing both factors up and dividing down to ensure that the sum is never exceeding the maximum.
That's interesting, but in order to determine those actual factors (and factors for mixing any other sources), you'd need to examine the ampilifer/mixer stages of the original devices, and perhaps try running some tests like MDFourier to determine frequency response and relative strenghts of those sounds
I'd be very surprised if CDs in the 90s were mixed to maximum levels - the extra headroom was intended to be reserved for things like drum kicks. Actual regular volume levels should be something like 0.25 of the maximum range. It also wasn't until after 2000 that audio mix engineers were given the instruction to mix things "loud" - as in dynamic range compression and trying to maximize the dynamic range on the CD
Do you need 3db clearance for modern DACs?
Sorry, that should have been in the dev channel really
What if Street Fighter II had been on the Philips CD-i? 🤔
Cursed timeline? A kid's desperate wish in 1992? 88 mph?
Either way, here it is! 😋
More info on my Patreon: https://www.patreon.com/Frogbull
#StreetFighter2 #Philips #CDi
I hate how these videos are playing inside of a tiny ass window
woah the sprites sure are tiny...
This is only a theory but these might actually be the sprites of the SNES version of Street Fighter II: The World Warrior
I can see that by the number of toes on Kens feet. The SNES has non square pixels which are wider than usual. The CD-i has nearly square pixels.
Here the sprite sheet https://www.spriters-resource.com/fullview/5572/
Full view of Street Fighter II: The World Warrior / Street Fighter II Turbo: Hyper Fighting - Ken.
I always think it's interesting to find exceptions to the europlatformer/shmup rule, like i think Mr Nutz is a good game for instance because it was made by 2 French young guys
but its certainly true for most games
Living in America, and never really playing any European games till recently, hell I just played my first Amiga game in the Mister last month, I do think European games have a unique style all their own
And I really enjoy it. Gives me something new to experience all these years later
I think the Turrican games are also good.
Apidya is also one cool case for euroshmup. It is like Gradius but with a bee in the main role. You even fight underwater against a Pike with snails attached to it shooting plasma. These are typical things happening in German ponds. The music is also by Chris Hülsbeck. I like his style.
i can never decide whether jazz jackrabbit was a good game
it has good music but the movement and level design was definitely off
Earthworm Jim i'd always assumed to be a europlatformer, but it turns out that was an american game, it just feels so euro though
https://youtu.be/2aCxJBbqzS0?si=5cZPiNSHuHxncCVf
anyway, this pretty much sums up how european developers felt about games at the time
My feelings exactly. Could never get into Earthworm Jim due to the controls.
i had it for snes as a kid, loved it honestly just because i was easily sold on any game that lets you jump into toilets
the hit detection was the biggest problem though
growing up with that sort of game you eventually got used to it, like for me Lion King for snes is a fairly easy game, maybe because i was used to not trusting hitboxes and weird ledge jumping. The cd i zelda and mario games are also just sort of playable to me. Not "good", but interesting and worth putting up with
oh man, Name That Tune boots!
yes you're right
arcade sprites for SF2 are 64px wide, SNES ones 32 pixels
to be fair all 16-bit versions of SF2 seem to use 32px wide sprites. maybe the issue is they're using a bigger screen resolution
Went through and retested anything that I had a note on. Lots of fixes! Only a couple problematic games remain
So is the list in the pin updated? Alot of them say "playable" but only a few say "completable"
Playable usually means they can be player controlled and completable means someone has taken the time to beat the game
gotcha
yeah
I would suspect everything is completeable
but unless we play through it, we don't mark it
Some games are also a little bit too painful to complete. I'm trying my luck with Hotel Mario every now and then and the Hotel of Wendy is much more difficult compared to the ones before. Every level in this hotel consists not of a single screen but two. This means if you fail at the second screen, it is back to the first. >.<
Fear not, Slamy. We will find someone foolish brave enough to finish these games
Well, I don't have plans to complete them all. 😄 There are also games that are not only painful but boring as well. I think some of them haven't aged very well. Hotel Mario is at least a little bit addictive.
yeah...I also don't have plans for them all. But I may try to tackle one of the awful ones. Depends on how much time I have
I think it would be optimal if someone plays them who also has experienced them on the original hardware. To actually be able to see the fine differences.
Also I am not sure how many of them are "games". LIke, is Earth Command a game?
yeah that's a good callout
This is where I stopped when I posted earlier aswell.. I mean I already wanted to throw my controller a few times during the Ice stages but Wendy hotel is next level hell. But i've gotten past it without any audio / graphical issues that I could see
Havent completed the last hotel though
However I have no experience of original hardware, i've only compared to youtube videos that i've scrolled through to compare
Next level hell is certainly the right description for this. I fail to see which target audience it was aiming for. Children? Never! Korean eSport Masters? Maybe? Casual Players? Never! Mario Fans? No...
Starting at Wendy's this game becomes a drinking game. The one loosing a life takes a shot and the next one can play. I've played frustrating games before. Even "Getting over it with Bennet Foddy" is easier than Wendy's Hotel.
The funniest thing is that a few small tweaks and the game wouldnt necessarily be "good" but it would atleast be a mid game. But I wholeheartedly agree that I have no idea who the difficulty is tuned for.
I sincerely believe that they just thought "You have infinite continues and a save function so its probably fine"
I'm not sure if this game was tuned at all. It feels almost like they ran out of testing time. I also consider it a rather bad idea to have a one-hit-death platformer like this to stretch over more than one screen. If you get the fire flower, things become a little bit easier. But assuming one has that right now, is like Gradius syndrome. I'm not sure if I'm able to finish this. I tried to play it as it might have been relaxing. This core might need cheats so this game becomes playable. I'm not even sure if cheat codes for this game exist.
Ye fully agreed, honestly I think i've gotten a single fire flower on all the maps i've done so far, I have no idea what even triggers those items to appear at all.
I'd be shocked if the game has any cheat codes tbh
The biggest issue I had with wendys is the timer, cause as far as I can tell it doesnt really matter how fast you run through it, cause goombas keep respawning on the platforms that you just left as soon as you go into the elevator.
So you go back down - kill them - run back to elevator.. oh wait the elevator changed direction - waste a bunch of time - go upstairs again - another wave is scripted to spawn - rinse and repeat a somewhat random amount of times.
Oh we also added unkillable things that can literally block doors and there is zero you can do about it if you arent there in time.
There is zero words that can describe just how frustrating the game actually is cause if only do the first 2 hotels the game seems fine difficulty wise, you just keep descending into hell the further you play this is my final warning to anyone that is gonna play this god forsaken game.
If you enter the doors, eventually an animation plays and turns the coins from monsters into actual coins for the rest of the Hotel.
Oh I didnt know that
It is random I guess. I suffered through some stages until I started entering doors to avoid getting hit.
"A hotel where all of the Pennies are disguised as coins, tasking Mario to be careful when collecting them. Mario must enter a room with a piggy bank (denoted by a cash register sound cue from an unopened door on the same floor as a recently defeated penny) to unveil the real coins."
https://www.mariowiki.com/Hotel_Mario"
This page is full of tips. I don't even know how one is supposed to know this.
MCD212: Added Mosaic display file type
- Fixes visualizations of eScape
MCD212: Added RGB555 coding
- Fixes "DYUV decoder" test on the "Validation Disc (Europe)"
AD7528: Reduces volume of input to 56%
- Approximates the clipping behavior of a CDI 210/05
- Required to fit the theoretical 17 bit output
into the 16 bit PCM signal of the MiSTer
- Fixes sound distortion in Hotel Mario
SERVO: Simulate tray operation
- Can be opened and closed with the system shell now
- Avoids reading of an empty tray
I don't have 100% confidence with the audio change but I assume it can't be avoided. A real CD-i 210/05 can mix the 16 bit output of both audio channels to one signal with an analog circuit. This means in theory a 17 bit DAC would be required to be 100% accurate. The MiSTer is slightly to blame here since everything is based on 16 bit audio.
I've measured the clipping behavior with a custom application to execute on a real machine. I've used the old authoring tools (thx to the CD-i discord team❤️ ) to generate a sine wave signal with maximum volume to play on the real deal. I've recorded that and checked how much the sine wave was altered.
It turned out that Hotel Mario is also clipping on the real machine but it is barely noticable. In general it is the loudest game I've experienced so far. To fix the non-linearity during mixing, the core has now a slightly larger headroom for mixing compared to the real machine. It will still clip during Hotel Mario but is barely noticeable.
The tray operation may not be fully experienceable on the standard settings of the core since it currently resets the machine whenever a CD image is loaded. I myself use the core with Debug Options -> No reset on NvRAM change but I'm still not sure if this is 100% safe in every regard. The CD-i is like a Saturn with the cell battery enclosed. The Saturn core currently performs a reset with every disc change. My experience so far is that this is not necessary as it seems that the OS of the CD-i is reading the file system on the NvRAM with every operation. Maybe, in future releases the core will not perform a reset when the first disc is inserted.
Great work slamy!
Really exciting stuff, dude!
I hope that your LSD trip music video thing is now accurate. I can't confirm that really since it is just too random.
It was the first title I've yet found which uses the Mosaic video format. The pictures are actually half resolution compared to the generated backgrounds.
Maybe some simulations of the video signal to show the differences.
Oh damn awesome!
I have an automatic integration test for the video side of the core. As the videochip reads its own configuration from RAM, it is just like the soundchip of the SNES. You have a dump of the RAM and nearly everything is there. So I keep some RAM dumps from MAME around. It's also very good to check for regressions.
I wonder if the digital optical audio out of the MiSTer could support 24-bit audio if Sorg was willing to implement?
or the hdmi?
i'm not sure what it would "cost" to do
quite a few cores have this sort of mixing/clipping issue and could potentially benefit
I had the chance to discuss a little bit with Frogbull earlier today.
About the a recent port of SF II on the CD-i system.
- From our discussions, he could reveal the sprites were imported from the Asia MD version of the Championship Edition & the backgrounds are from the SNES version.
- The choice of the MD version was done because of the directly integrated debug mode from it.
- He is working under CLUT7 colour format instead of CLUT4 as 'it is easier to work like that' for him.
And I have a quick question on the CD-i system from him.
"Does the CD-i allows us to play around with the MPEG-1 layer without the MPEG module ?"
His idea was to create a parallax effect for the backgrounds if it was possible.
A human being will not be able to hear the difference between 16 and 17-bit audio.
no but it gives you headroom so you can mix lots of audio and still produce a reasonably normal/loud output
Hex’s videos are all like that lol - they had a whole trilogy of those releases on CD-i, including Escape, Head Crash, and Antistatic.
They had a variety of different releases on just about all of the multimedia systems around that time, though - they also had a 3DO game called Digital Dreamware, and a CDTV title called Global Chaos.
It was an early multimedia project by Ninja Tune Records, who are still active today. The label’s founder, Coldcut, did most of the project’s music.
They also made Top Banana, one of the most wildly cursed-looking platformers I’ve ever seen:
Top Banana is an environmentally-themed platform game produced by Hex and Psycore for the Acorn Archimedes in 1991 and ported to the Commodore Amiga and Atari ST in 1992.
Its graphics makes my eyes giddy because of background colors.
additional information : https://en.wikipedia.org/wiki/Top_Banana_(video_game)
True, if you need to mix multiple layers together, extra bits are needed.
The "external video" which is generated by the MPEG cards own video hardware replaces the single color background layer of the CD-i base system video chip. Sadly, the CD-i base only has 2 layers to work with. This might also be the reason for the absence of any parallax. This is at least how I get it.
He should discuss it with the actual software developing experts on the CD-i discord. Right now TwBurn is working on another game and this time he also involves the MPEG decoder. It seems you are not even able to manipulate the shown picture when the MPEG stream is stopped. But maybe he is only using the official software interfaces for that...
Gameplay demo of SkyWays for Philips CD-i
Implemented level interaction (walls, tunnels, void), loading animation and multiple levels.
Preview build coming soon! (DVC required)
Yes, he explained and showed me that video earlier today.
I will ask him to see the others on the CD-i Discord.
Thanks.
The CD-i is lacking in the sprites department 😦
Anyway, thanks again for the CD-i core.
My family lost everything in a fire last summer.
And they used the CD-i 205 (910 in the US) a lot, even recently.
Their CD-i player and record collection were destroyed.
I showed them the core a few weeks ago, and it brought back so many memories.
That's horrible. Sorry to hear that. A fire can be relentless.
one delightfully strange and specific regression i can point out: the apprentic makes my composite out signal black and white on my NTSC crt now with this new core. the previous core didn't do that!
no complaints just adding it to the pile of notes in case it ends up being helpful!
Howdy. I sadly don't have a YC adapter to test this so I can only make educated guesses. In the menu you have the TV Mode selection. I assume you have set it to PAL which puts the machine into 50 Hz mode. For NTSC it is 60 Hz. I was already called out for that naming on the CD-i discord as it is a lie. There is no PAL and NTSC for a MiSTer core. From what I've learned, the PAL/NTSC Encoder switches the standard based on the refresh rate. Maybe you have set it to PAL and forgot about that.
Or does going back to the old core really fix this? What was the version of that specific core?
CD-i titles should not have a limit to a specific region. It is against the green book standard. The titles which are optimized for NTSC use a special PAL mode that crops 20 lines from top and bottom.
it is possible somethings slightly shifted timings wise somewhere and broken the composite colorburst, its really sensitive
Story time: In the last days I was trying to wrap my head around a possibility to auto start the application. On a real CD-i 210/05 you can press the Play button on the front panel. Even so, a CD was not yet read the system remembers this and starts the application or the audio CD playback as soon as the medium is detected. Technically, the SLAVE is polling the front panel controller U3090MG via I2C to read the button states. When the play button is pressed, the word provided by the front panel controller is not 87 00 but instead 07 20. This in turn causes the SLAVE to give the main CPU an IRQ and gives A1 87 20 on one of the channels back to the OS. In case of the current state of this core, only 87 20 is returned. This is wrong. I wondered why it is like that and it turns out that SLAVE 2.0 is the wrong version for this core. A real machine uses SLAVE 3.2. There is one issue though. I only have access to 4.1 from cdimono2.zip and 2.0 from cdimono1.zip from the MAME dumps. Its interesting to note that 4.1 from Mono II doesn't have this problem. But this revision doesn't help us as the audio mixing circuit is never controlled by it.
I don't know why someone had put the revision 2.0 into the rom dump for Mono I. I was told that this revision is only used in the even older hardware revisions.
Sadly the SLAVE ROM cannot be dumped that easily. There is a person here https://seanriddle.com/mc6805p2.html which explains a way but it requires removal of the SLAVE from the PCB and some trickery to dump the content. As my machine has it soldered to the board and I have two left hands with the soldering iron I need to think how I can progress with this topic. I know some soldering experts who actually can pull that off. I still have hopes high this can be avoided and a dump somewhere exists.
It might be worth looking into test clip dumping. Some people managed to read ROMs incircuit using Pomona clips or similar, depending on the chip’s design. Could avoid the need to desolder entirely.
The situation became even more confusing. The mentioned method requires a certain pin (NUM for non-user-mode) to be manipulated. This pin doesn't exist on the package of the 68HC05 used here. In the meantime I think I found the person which dumped the ROM for MAME so I have hopes high to gather info how to proceed.
@terse flare fwiw color is still working correctly for me. I'm using
ntsc composite tv
mister addons io analog pro (so not direct video)
CDi 25.03.26
Latest unstable mister main v250331
oh crap I missed it was game specific one sec
update: The Apprentice working fine for me
feel free to dm anyone who wants me to check composite support with this specific setup (latest analog out board)
Please take a look at the CUE file. It might point to a non-existing BIN file. I wanted to investigate why this one version of Kether was not working and there is not really a feedback from MiSTer when something like this happens.
CATALOG 8712581301002
FILE "Kether (Germany) (v3.08).bin" BINARY
TRACK 01 MODE2/2352
INDEX 01 00:00:00
The real filename is probably Kether (Germany) (En,De) (v3.08).bin
CHDMan is good feedback for when something like that happens, haha
I will take a look as soon as the morning meeting rush is done
I don't know if it would help here, but the MSX1 core is able to load MSX computer ROMs direct from the OSD, rather than needing to then do a command line load like other computer cores. Not sure how it achieved this, but may be interesting to check.
So you were right about the cue file. However, even after editing it, the same thing happens. Maybe I did something wrong here: ```CATALOG 8712581301002
FILE "Kether (Germany) (En,De) (v3.08).bin" BINARY
TRACK 01 MODE2/2352
INDEX 01 00:00:00
This is bizzare... hmm.... just out of curiosity. If the core sends you back to the shell/bios and you press play again... what happens?
And just to be sure about the checksum.
/media/fat/games/CD-i# md5sum "Kether (Germany) (En,De) (v3.08).bin"
efec1072e1a891083d0145837c14ca99 Kether (Germany) (En,De) (v3.08).bin
This was user error. I don't know if the old cue was cached, or I hadn't saved it, but it's working fine now
I'll update the sheet
Can i just say, the Tetris music goes hard on CD-i
I'll permit it.
https://soundcloud.com/user-391016128/sets/phillips-cdi
Jim Andron has released the original uncompressed audio as well
Back in '92 (30 years ago!)I did the music for a version of Tetris for Philips CD-i. This music has since taken on a life of its own on YouTube. I've have numerous requests to upload my original track
A weird note from today. I wanted to look a little bit into the NTSC mode. Then I've noticed about 6 missing lines at the top. It turns out, my Commodore 1084 is to blame. It's just not awesome enough to handle the extreme overscan the CD-i does produce with 240 active lines and 262 lines total. You have to keep that in mind when also operating the machine with NTSC. I just wanted to share this finding. (Of course, this issue is not present when using HDMI)
Quite some titles are broken. Tetris for example is missing a number of lines at the bottom.
1084 not awesome enough?!
Oh but don't get me wrong. It is still my favorite CRT for retro stuff. I have 3 of them around in case one breaks. Compared to PC CRTs I've found, there are very bright despite their age. I was just very confused yesterday as I've never expected that it cuts off image data. I have to assume that 20 lines are just not enough to move the raster to the top.
"Having three of them around in case one breaks" is so much like me 😄 I always have to get three of everything...
I'm curious. I'm unable to recall whether someone of the testers have discovered any wrongly positioned video when using the NTSC mode of the core. For some reason, the behavior differs between simulation and how the MiSTer displays it. This is especially noticeable with Tetris. The buttons are cut off at the bottom.
To give some reference picture. This is how it looks like on my machine.
I don’t see it in my notes
Holy moly! It was not a difference between synthesis and simulation. If CPU Turbo is active, the problem doesn't exist.
To speedup the simulation (1 frame every 3 seconds on my machine) I usually have the CPU faster in that case. Curious ...
This seems to be some sort of Heisenbug. For some reason, the issue is no longer reproducible. I don't get it. There seems to be no correlation with the CPU Turbo at all.🧐