looking for critique | Astrobiscuit | Page 1

gloomy carbon Apr 18, 2024, 12:06 AM

#

Another edit of my unfinished M101 target (because I have 50 hours and I wanna mess with the data, shoot me)

Looking for some critique here, I’m still probably going to gather another 20-30 broadband hits as well as that much Ha.

swift inlet Apr 18, 2024, 12:20 AM

#

@gloomy carbon was arcsinh stretch used

half barn Apr 18, 2024, 12:35 AM

#

Looks pretty solid for the most part. The black point is pushed a bit further than I think is necessary for this. Also the overall background and the solo stars are too red.
I think you need to spend a bit more time on the noise reduction too.
Hope that helps and, of course, feel free to ignore me.

gloomy carbon Apr 18, 2024, 12:39 AM

#

swift inlet <@1014637321460654101> was arcsinh stretch used

No

swift inlet Apr 18, 2024, 12:40 AM

#

Black point is fine imo

gloomy carbon Apr 18, 2024, 12:40 AM

#

half barn Looks pretty solid for the most part. The black point is pushed a bit further th...

I also have mixed feelings about the black point but it’s not clipped. I don’t think more NR is the answer, but more integration as I used deepSNR on 0.7 and nxt on 0.4

swift inlet Apr 18, 2024, 12:40 AM

#

Why deepsnr and noisex?

half barn Apr 18, 2024, 12:41 AM

#

I didn't say it was clipped, it's just further than I would have gone

gloomy carbon Apr 18, 2024, 12:41 AM

#

DeepSNR pre stretch nxt post stretch

swift inlet Apr 18, 2024, 12:41 AM

#

Ah

gloomy carbon Apr 18, 2024, 12:41 AM

#

half barn I didn't say it was clipped, it's just further than I would have gone

Yea I know

swift inlet Apr 18, 2024, 12:41 AM

#

I would recommend trying deepsnr 1.0, I've never had to do any noise reduction past that after linear

gloomy carbon Apr 18, 2024, 12:41 AM

#

I have found that after 0.75ish I get nasty artifacts

#

I’ll give it a shot after the gym

swift inlet Apr 18, 2024, 12:42 AM

#

Artifacts?

#

Could you give an example?

gloomy carbon Apr 18, 2024, 12:42 AM

#

Uh like blotches and fake stars

#

I tried that method of yours to remove fake stars and it didn’t work

swift inlet Apr 18, 2024, 12:43 AM

#

Can try lowering drop shrink next time you stack

gloomy carbon Apr 18, 2024, 12:43 AM

#

This time was 0.8

half barn Apr 18, 2024, 12:47 AM

#

Also earlier, I didn't say to do more noise reduction, but to spend more time on that step. There was just something odd in the background noise that I found weird.

I don't really find the best results through AI based NR. I would probably go for a MureDenoise and maybe touch it up with MLT and compare that against the NXT. I've never had any luck with deepsnr so I don't know much about getting the most out of it.
good luck

gloomy carbon Apr 18, 2024, 1:00 AM

#

half barn Also earlier, I didn't say to do more noise reduction, but to spend more time on...

Hmm I’ve never heard of MureDenoise, and you’re the 2nd person today, and ever, to tell me about MLT. I’ll look into em

#

The background is very funky in this data I blame the 294

half barn Apr 18, 2024, 1:01 AM

#

MureDenoise is hiding in the NoiseReduction tab of the scripts dropdown in PI.

#

I think both VisibleDark and Adam Block have some tutorials for MLT if you wanna play around with that process. Also it's covered extensively in Inside Pixinsight by Warren A Keller, but at this point the second edition is wildly out of date and I probably wouldn't recommend paying full price for it.

gloomy carbon Apr 18, 2024, 1:06 AM

#

Oh haha I can’t be paying for any processing tutorials right now anyways, lol

half barn Apr 18, 2024, 1:07 AM

#

I meant on youtube

vocal void Apr 18, 2024, 2:13 AM

#

gloomy carbon DeepSNR pre stretch nxt post stretch

Try mixing like 10% of the original noise back in via pixelmath

#

Instead of just reducing the dsnr slider

#

Just to retain as much micro detail as possible.

#

Ideally you should have to use noiseX

#

Contrast is good and colors are great

#

Maybe slightly desat background just to get that satisfying neutral color.

#

Just a tad bit of yellow tint from what I see.

gloomy carbon Apr 18, 2024, 2:39 AM

#

vocal void Try mixing like 10% of the original noise back in via pixelmath

How

vocal void Apr 18, 2024, 2:40 AM

#

"Original * (0.1)+deepsnr * (0.9)"

#

Dammit discord

#

Ignore the spaces

#

That's the pixelmath expression I use

#

The image names can be whatever of course

gloomy carbon Apr 18, 2024, 2:42 AM

#

Okay

#

Ty Zac

vocal void Apr 18, 2024, 2:43 AM

#

Np

vocal void Apr 18, 2024, 2:43 AM

#

vocal void Ideally you should have to use noiseX

Shouldn't*

gloomy carbon Apr 18, 2024, 2:45 AM

#

Figured, lol

gloomy carbon Apr 18, 2024, 3:01 AM

#

swift inlet Could you give an example?

#

haha

#

swift inlet Apr 18, 2024, 3:18 AM

#

gloomy carbon

Wtf that's a lot kekw

#

You are using the varshape 1.5?

spark wedge Apr 18, 2024, 3:19 AM

#

Not drizzled 💀

swift inlet Apr 18, 2024, 3:19 AM

#

spark wedge Not drizzled 💀

No it's drizzled

spark wedge Apr 18, 2024, 3:19 AM

#

Varshape 1.5 has to be done with a Drop-shrink of 0.5-0.7

#

otherwise the drizzle overshoots, since it's using a oversampled drop size.

#

Varshape allows for enhanced detail and SNR at lower drop-shrinks (Something square, gaussian, and circular cannot do)

#

More focused drizzle samples allow for higher resolution in a drizzled image- but at a cost, DeepSNR is actually tuned to the likes of ~.8+ dropshrinks, so it will actually hit a wall where the noise floor cannot be refined further without producing a blur like effect.

#

Which is something you see in my images, I use DS of 0.7 with Varshape 1.5- its a tradeoff, sure, but a worthy tradeoff in my book.

gloomy carbon Apr 18, 2024, 3:31 AM

#

swift inlet You are using the varshape 1.5?

No

spark wedge Apr 18, 2024, 3:32 AM

#

gloomy carbon No

Use it

#

Helps reduce hallucinations

gloomy carbon Apr 18, 2024, 3:32 AM

#

Last time I used it it was gross

spark wedge Apr 18, 2024, 3:32 AM

#

Probably used DS .9

gloomy carbon Apr 18, 2024, 3:32 AM

#

Possibly

#

Idr

spark wedge Apr 18, 2024, 3:33 AM

#

From my testing, a DS of more than .76 causes overshoots

gloomy carbon Apr 18, 2024, 3:33 AM

#

Alright

#

When I stack the rest of the data I’ll do that

spark wedge Apr 18, 2024, 3:33 AM

#

ye

gloomy carbon Apr 18, 2024, 3:33 AM

#

I gotta buy a new drive to stack it all bruh

spark wedge Apr 18, 2024, 3:34 AM

#

You'll lose out in DeepSNR's NR power, but your details will increase.

gloomy carbon Apr 18, 2024, 3:34 AM

#

Hopefully 20-30 more hours will solve the noise issues

#

I don’t think so tbh

#

30-50

swift inlet Apr 18, 2024, 4:12 AM

#

I use varshape 1.5 and a dropshrink of 0.8

#

I've tried going lower but didnt notice much difference tbh

#

so i just left it at that

spark wedge Apr 18, 2024, 4:16 AM

#

The difference is really miniscule but you can see it. Try Varshapes over 2 and go to 0.3 or 0.2, and you'll see a world of difference.

livid veldt Apr 18, 2024, 1:12 PM

#

vocal void "Original * (0.1)+deepsnr * (0.9)"

You can use back ticks to show code in discord (character `)

#

@gloomy carbon liking this a lot. Personally I’d probably use a bit more saturation on everything but that’s mainly a taste issue.

livid veldt Apr 18, 2024, 1:15 PM

#

spark wedge The difference is really miniscule but you can see it. Try Varshapes over 2 and ...

Wonder if there’s a drizzle comparison script that drizzles a preview with various settings for blinking

gloomy carbon Apr 18, 2024, 4:07 PM

#

livid veldt <@1014637321460654101> liking this a lot. Personally I’d probably use a bit more...

Thank you I spent a good amount of time going for these colors

#

I can’t wait to add a bunch more integration and the Ha

spark wedge Apr 19, 2024, 1:42 AM

#

livid veldt Wonder if there’s a drizzle comparison script that drizzles a preview with vario...

I wish.

spark wedge Apr 19, 2024, 3:49 AM

#

howo do I change varshape in deepsnrv

vocal void Apr 19, 2024, 3:56 AM

#

Vershape is altered right before commencing with stacking + cfa drizzle

#

In WBPP it can be altered under drizzle parameters

gloomy carbon Apr 19, 2024, 6:16 PM

#

I’ve still got data to gather. I do like those colors

#

A bit dull on the saturation for my taste but the actual hue is justright

swift inlet Apr 19, 2024, 6:44 PM

#

His color is fine now imo, except for the core

#

I meant the color in his image

vocal void Apr 20, 2024, 12:29 AM

#

I gotta grab this mf before its out of season

#

Any noticeable OIII in there?

#

Idk if it's worth it with 400mm fl

gloomy carbon Apr 20, 2024, 1:11 AM

#

vocal void Any noticeable OIII in there?

I can let you know once I take a look at my data

#

My core do be pretty yellow

#

Ha will come soon

vocal void Apr 20, 2024, 1:12 AM

#

I'll likely stick to Ha

gloomy carbon Apr 20, 2024, 1:12 AM

#

Also imo you should shoot this

#

Leo’s triplet looked really good at ur FL so I think m101 will look solid

vocal void Apr 20, 2024, 1:13 AM

#

My scope wouldn't resolve anything due to just how little OIII is present compared to Ha.

vocal void Apr 20, 2024, 1:13 AM

#

gloomy carbon Leo’s triplet looked really good at ur FL so I think m101 will look solid

Ye def

#

May have clear skies next week

gloomy carbon Apr 20, 2024, 1:14 AM

#

Tn might be clear

#

I need to start testing that svbony scope bruh

#

They hit me up yesterday asking how testing is going

#

I have not replied 💀

vocal void Apr 20, 2024, 1:16 AM

#

Lmao

#looking for critique