🙋│image-help

[thin tinsel]

ok

this is you leonardo created image?

[rigid folio]

yes

[thin tinsel]

using what method?

[thin tinsel]

ok cool. It's doing a good job seeing the logo and the verticle zip

[rigid folio]

yes, this is not bad. but i some stuff with patches, panel etc. and it is very challanging

[thin tinsel]

yes I think one plain colour is easier for it to replicate

[rigid folio]

that could be it. trying now with something more complicated

[thin tinsel]

Did you get a reply to your question?

[night vault]

not gonna lie, this looks great given limitations

[thin tinsel]

Did you get a reply to your question?

[thin tinsel]

yeah I'm pretty impressed

[night vault]

sadly not good enough for showcasing real commercial use regardless

[rigid folio]

not a single rendering got the shoulders right.and for this use -case it is critical

@thin tinsel do you offer consulting, training or coaching?

close but still to far away.

[thin tinsel]

so easy to edit this one though!

[rigid folio]

you think? cool

[thin tinsel]

That's still really good when this isn't how it's designed to be used 😆

[thin tinsel]

photoshop on the logo and add the red on the pockets. Job done

[rigid folio]

it got the collar really good

[thin tinsel]

YEs, that's great. THat would be fiddly to edit later

Just adding the red will be fine

[thin tinsel]

w have a mentoring service on the discord just type /mentoring

[rigid folio]

awesome, didnt know,. will dive in there

[heavy dome]

Wooow! Did you manage to do this without post editing? So via promting?
This is really impressive!

[steel snow]

how can I find out what images i used for style reference. I don't get this info when I switch machines. I know the content reference I used but there seems to be no option to actually view the source files.

I can't find any such option in the generation info

[eternal plover]

It is looking so beautiful! what is the prompt?

[night vault]

you simply can't, it is non reversible process.

[steel snow]

that's f'ed up....the issue is i re-installed windows and can't remember the style references I used as I was experimenting with different images.

[west pivot]

Thank you, Yes, I used Phoenix/Image2Image at 0.55 with the same settings (prompt, style reference if available). 'Prompt Enhance' must be set to 'Automatic.' After getting the exact image I wanted, I used 'Edit with AI' to achieve the desired expression.

[feral mica]

Hey everyone!

@stable carbon created a quick guide that is useful for both beginners and seasoned users. We are going to post it here and pin it, so if someone looks like they need it we can always come back to it.

Thank you, Thom!

[real lynx]

Hello, you can also describe their clothes so that the full body appears, for example brown boots, medieval clothing, a hood, and you could also add some background, a magical forest.

Another thing that comes to mind is the views

A bird's-eye aerial shot would also show a full body

[real lynx]

Let me look for it, they told me about it in the general and I shared it there and I forgot to bring it here

The prompt is beautiful

I always liked it a lot

And with a few changes you can modify it very easily.

XD

This Grinch came out from the same prompt

She still dances like a ballerina on a flower pot

Under the rain

[manic root]

Hey Leonardians! Checked out today's daily theme yet? Join us in #1108130218260041758 for some creative challenges! 🎨 Show your support by voting with the Leonardo emoji . Awesome prizes up for grabs! Enjoy and good luck!

[thin tinsel]

Yes! Thank you to Thom! We really need this guide. 🔥

[real lynx]

A whimsical miniature figure, crafted from a vibrant medley of green leaves, assumes a lively dancing pose atop a worn, earthy pot, surrounded by a lush forest scene awash in the soft, misty glow of rain-kissed greens and blues, evoking a sense of mythical wonder and serenity, with delicate, intricate leaf veins resembling delicate brushstrokes, as tendrils of foliage swirl around the figure, blurring the lines between dancer and environment, against a backdrop of gentle, glistening raindrops.

[steel snow]

what's the best mode for text?

[raw thicket]

Phoenix

[real lynx]

Whatever you want to be text has to go in quotes and a short description of what style you want it to be.

[thin tinsel]

Realtime canvas maybe another option or part of the process.

[half halo]

so if we all follow this. would you say.. were all. ASPER'gers 😄

im so pleased you got that joke

i will die for you and your cause now

whatever it may be

[thin tinsel]

We got the joke, but I can't believe you said it out loud.

[half halo]

I think everyone should voice their honest puns

and wear their (terrible sense of) humor on their sleave

i think humor is awesome, no matter the way, shape or for it takes. It brings us together and often makes us think in ways we would not otherwise 😄

[rigid folio]

Interesting. How does that work. Color placement is correct

[thin tinsel]

have you opened the mentoring thread?

[rigid folio]

Can't find the chanel😬😬😬😬

[thin tinsel]

you need to type
/mentoring
and fill in the form

[half halo]

good idea

[thin tinsel]

thanks

[half halo]

/mentoring I am in love with @thin tinsel and i would love to go on a date. I request @real lynx as a chaperone. I suggest we find a nice place, out of the way somewhere, maybe an Italian place, order some pasta, maybe one of those big two person plates, where we can eat and have our lips meet like in lady and the tramp.

oh woops

i think it didnt work

[thin tinsel]

https://tenor.com/view/adam-levine-embarrassed-gif-21343983

[half halo]

the bot freaked out!

omg im so embaressed rn..

[thin tinsel]

lmao

[half halo]

❤️

https://tenor.com/bVzgl.gif

This was just friendly memeing 🙂

the love is universal!

my true love is @copper warren

https://tenor.com/bbUBU.gif

I would say Pookie but discord does not recognise this.

[drifting swallow]

Hey gang. Is there a tutorial on the best practice for having an existing image edited? In my case, I have a previously generated image of a man wearing glasses. They are round, but I need the glasses to be more squared-off instead. Taking it into canvas editor, doesn't do well (without erasing his entire face, which creates an entirely different face)... Trying to erase the roundness of the glasses doesn't allow for a square'ish replacement.

[half halo]

You want to use both the character reference, content reference, and probably style reference

you can put each of them on a slider

then mesh it with the prompt

so you have a lot of variation to play with, which tbh is the most fun about leo

[drifting swallow]

That's how I started... and was getting everything modified, except the glasses... which was the odd part.

[karmic pulsar]

Whats the maximum weight that can be used on phrases/words in legacy for prompts? I just tried 5.0 and its working without breaking anything o.0

nvm I see it broke at 10.0 lmfao

[manic charm]

do you guys find it easy to use the Canvas to fix errors in images? I dont know if I use it the wrong way, but I can never fix errors in hands, eyes, etc

the correct way is use the model which the original imagem was made?

[real lynx]

https://youtu.be/hKf2kxbgmhE?si=ZKToDPj-P7t37N5P

[manic charm]

I will watch, thx!!

[hardy sentinel]

yes theoretically it is possible, but I have never try it 😄

[hardy sentinel]

you need to use the model and settings of the original image

[hardy sentinel]

I like to use sketch to image mode, this require a bit drawing skill, I am using mouse

[hardy sentinel]

have you tried sketch to image mode at canvas editor?

[hardy sentinel]

you can use leonardo phoenix and try with short text

[karmic pulsar]

I started getting distortion at weight of 7-8 lmao. But i ended up staying around 3.5. That’s wild compared to SD1.5 and SDXL models from Civit.

[hollow breach]

Did you specify anything about the glasses?

[hollow breach]

Some models are better with hands than others. So sometimes using a different model help

[hollow breach]

I thought it was around 3 but maybe it is 5

[karmic pulsar]

Higher. I got up to 7-8 before it started having artifacts. XD

8-10 just couldn’t see jack lol

[opaque geode]

hey, can someone help me

[wise canyon]

Hi, what would you like help with?

Feel free to post your questions here, we have plenty of members and Mentors available to help you any time 🙂

[opaque geode]

I have a picture, can I change only the background? or put it in motion mode?

[wise canyon]

You could either remove the background with a "Background Removal" tool, if it's a Leo generation it may have that tool option under the image (also depends on Model used),
or change the background using the Canvas Editor, by masking it and prompting for a new background.
Adding Motion to an uploaded image requires a paid Leo subscription, Motion is temporarily unavailable on the Free tier sub.

[opaque geode]

Can I add you as a friend so we can speak instead?

would appreciate it

[opaque geode]

And I have paid for leo

[real lynx]

Hi

[hardy sentinel]

Thank you for testing and sharing. Back in sd 1. 5 it is usually under wright of 2

[real lynx]

You have the option by clicking on the image

[opaque geode]

Yes image creation

[wise canyon]

We can speak here, or one of our Mentors can open a personal thread in the server to assist you one on one.

[hardy sentinel]

You can remove the background, if you have created the image using leonardo, but for changing the background people use canvas editor

[wise canyon]

Background removal can be found in your personal feed images also.

[opaque geode]

that would be much appreciated

[opaque geode]

I Might be stupied but I have tried so many times but cant get it right

[manic root]

Hey Leonardians! Checked out today's theme in #1108130218260041758 yet? Jump into the challenge and don’t forget to cheer on others with a vote on your favorite entries. There are awesome prizes up for grabs! 🎉

[wise canyon]

You can ask Cami, an excellent Mentor, or go to: https://discord.com/channels/989166677390426132/1230345629776085023 and type "/ mentoring" (no space) to arrange one on one mentoring.
#🙋│image-help message
(fyi the Mentors have yellow names) 🙂

[karmic pulsar]

Yes. I remember if hit 2 it would break. But wow can’t believe it can now go so high.

[opaque geode]

thank you

[wise canyon]

Successful background removal can sometimes depend on the subject outline, some members prefer manually removing backgrounds in Canvas.

[opaque geode]

I just want to remove the background simple, or put the picture in a motion mode

[wise canyon]

Try the option Cami is showing here: #🙋│image-help message
Just under the tab Cami has circled, you will see the add Motion option tab.

[hardy sentinel]

That means they changed something in the programming

[opaque geode]

Are you here Cami?

[real lynx]

Hi

[opaque geode]

Is it possible to be guided step by step in voice?

[real lynx]

I'm working but you can ask questions when I have time I'll answer you

[real lynx]

Mmm I speak Spanish if you understand Spanish I have no problem xD

[opaque geode]

we can take it when you are free then, because I cant fix it

[opaque geode]

haha no only English, arabic, swedish and Danish

[real lynx]

share your image so I can try it?

[real lynx]

Hi, so you want to remove the background from this image?

[opaque geode]

yes

or this one

[night vault]

you cant. better use other tools like macos has remove background built into os

[placid python]

How to get full-body portrait image?

[opaque geode]

are u sure?

[hollow breach]

It really depends on the type of background. Sometimes you need to finish the job manually

[hardy sentinel]

Set your aspect ratio to 9:16 if possible

Or 1:2

[hardy sentinel]

If possible set the aspect ratio to somewhat vertical.

[hollow breach]

Add hair and shoes in the prompt

[thin tinsel]

correct!

[steel token]

Witam, czy ktos z was orientuje sie moze jak moge zmienic stosunek postaci do tła? Tzn mam wilka i sarne, ale one sa duze, a lasu jest malo, nie zrobie z tego animacji. Unzumowanie grafiki nic nie daje, a na komendy tez nie reaguje.

[hardy sentinel]

please post your image too

[hardy sentinel]

which model do you use ?

aspect ratio ?

[real lynx]

Dzień dobry

[hardy sentinel]

Hola Cami 🙂

[steel token]

I don't even know where to look for such options

[hardy sentinel]

as you have two subjects here, maybe try to use leonardo phoenix

[steel token]

yes, i try it with leonardo phoenix

When I use Leonardo phoenix from this photo and in the edit option I enter that I am asking for distant perspectives, the generator creates completely different images

[leaden jacinth]

My apologies. Are you still wanting assistance? Maybe we can help.

[leaden jacinth]

Awesome resolution 😍 looks like Pixar aesthetic 😻

[leaden jacinth]

The Generator will produce slightly different images with the use of each model and preset.

[steel token]

Yes, I still need help, I would like to know how to change the perspective so that there is more of the background in the image and less of the characters themselves.

[steel token]

Thank you, wolf model created from a photo of my dog

[leaden jacinth]

Magnificent.

[leaden jacinth]

One second. BRB.

[steel token]

Thank you, is there any possibility for me to get the effect I need? The other generated photos also show mainly characters and I don't have enough space to create animations. I suspect that the built-in AI robot for generating animations will also not be able to cope with such a small background.

[leaden jacinth]

Are you seeing this option @steel token when you attempt to change the aspect ratio?

[steel token]

Yes, i have it

[leaden jacinth]

You can press "More" for additional AR based on your current plan.

[steel token]

Yes, I can, but will it change the proportion of the forest to the heroes? These settings look like changing the image resolution so that the image is displayed vertically on the smartphone, or horizontally on the TV.

[leaden jacinth]

Changing the AR shouldn't affect your image based prompt results. You may have to tweak it or generate more AR until you get closer to your desired outcome.

[steel token]

I understand, I've already generated a few images from this seed, but the hero is always in the foreground, maybe I'll try the command to put the heroes in the background. because requests to increase the volume of the forest in relation to the size of the characters do not produce results.

[leaden jacinth]

There are a wealth of mentors here to help. Use the "/" command to summon one of them for 1:1 assistance if you require a deeper experience.

[steel token]

Always like this

you mean discord?

[leaden jacinth]

Beautiful images ✨ Perhaps you require help beyond my purview.

[leaden jacinth]

Yes. Here in this channel.

[leaden jacinth]

The words foreground and background are interpreted differently. So when you use both words in the prompt a contrast of the two can appear.

[steel token]

Thank you, for now I will use the traditional technique, maybe in the evening I will come back to further attempts with AI

[leaden jacinth]

You are welcome 🤗 Happy Creating ✨

[steel token]

I'll try to figure something out. I was hoping that there was some option that I hadn't noticed that could control this. By the way, it would be useful to have more slider options to issue the appropriate command.

It's a great tool for fun, but if we want to achieve a specific effect, it becomes difficult. It's possible that this is also the case, more generating the appropriate image consumes more resources. 😉

[marble cape]

Forgot having hard time adding this again

[manic root]

Hey Leonardians!🌟 Have you checked out today's theme in #1108130218260041758? Jump into the challenge, share your creations, and don’t forget to cheer on others with a emoji! Big prizes await! Enjoy!

[hollow breach]

Start the prompt by describing the background.

[thin tinsel]

Use the OUTPAINT feature in Canvas Editor to build around the image.

@steel token , if you don't know how to do this, please ask.

[thin tinsel]

I drew around the image with a black brush in Canvas Editor. Made one generated image like that.
I used image-to-image on full strength to bring it into Leonardo as new image. Prompt doesn't matter.
Then used the remove background option to remove the background

Also the remove background tool in Realtime canvas did a great job, but you'd end up with quite a small image at the end

[thin tinsel]

Hi Yuri. At this time, you are not advised to use the model you created the image with.
Please use Dreamshaper or Absolute reality for best results.
If you require step-by-step guidance, then please request a mentoring thread, by typing /mentoring.

[thin tinsel]

If you can share the image then we can have a go and let you know what works best

[thin tinsel]

You can do it but it looks weird because there is no drop shadow and lighting inside and outside is different, so it depends what you wanted. But just to show you it is possible to some degree

[real lynx]

It is very beautiful

[nimble bay]

Hey Willow.

I just upscaled an image but need to fix a hand

it messed it up

do I go back to canvas?

[thin tinsel]

Hi WinterOak.

[thin tinsel]

Oh did the upscale not fix the hand, that's a shame!

[thin tinsel]

That's unlucky. 😬

[nimble bay]

Nah it messed it up strangely

[thin tinsel]

Yes it'll need to be fixed in canvas again. Do you want me to do it?

[nimble bay]

Im gonna have a go first

if I need more help I will reach out

Image is looking dope

[real lynx]

A captivating cinematic watercolor masterpiece , featuring a graceful ballerina suspended in a dreamlike, ethereal sky. The ballerina, adorned in a flowing, elegant gown, is mid-dance, her movements delicately captured and frozen in time. She floats amidst a backdrop of a starry night sky, teeming with celestial bodies that twinkle brightly, mirroring the magic of the night. The scene exudes tranquility and serenity, effortlessly fusing the worlds of fashion and art into a mesmerizing, harmonious visual symphony that transcends reality.

[pure geyser]

@real lynx #🙋│image-help message

[real lynx]

Hi

How can I help you?

[pure geyser]

#🙋│image-help message

[real lynx]

The image was not created in Leo you just want something similar

[pure geyser]

Yes, in leo Which preset and elements I should use get this

[real lynx]

Can I use any tools or only free tools?

[pure geyser]

Any

[pure geyser]

It's okay if you are busy.
Thank you!

[manic root]

Hey Leonardians! Ready to join today's daily theme in #1108130218260041758? Jump into the challenge and express your creativity! Don't forget to cheer on your peers by voting with the Leonardo emoji . There are exciting prize tokens up for grabs! Enjoy and good luck!

[real lynx]

Hi

Sorry if I was working

Only with image Style

It already takes the colors of your image

[real lynx]

Use the image to have the AI describe me and give me a prompt

Manga style anime

[hardy sentinel]

Beautiful anime

[eternal plover]

I hate anime, it dosen't make any sense to me but looks good

[hard sand]

So im using anime realisim and Im trying to make the entrance to a cave with native american totem poles carved out of stone for the entrance, I want it to be realisitic with blurred, and depth of view like in these images but it seems to just make them cartoonish is there a prompt to make it like these images, I type depth of view,blurred background or (sometimes have the other parts blurred) extreme realisim and it doesnt wanna do it.

This is a alittle to cartoony

[real lynx]

Muy bellas texturas me gusta el anime de este estilo mesclado con texturas reales

[hard sand]

Thanks it's a bit hard to get leonardo to keep making the style all the time

[hard sand]

Im loving the realistic parts of the first image I would also need the person to be semi realisitic

And I would like to so that its realisitic for nature scenes without people

[copper warren]

Hi there. You have to know that this is not something we'd permit on this server. Self-promotion of any kind is off-limits. You've also chosen to issue your challenge in the channel that is intended for users who are seeking help with using Leonardo, which is disruptive. I understand being proud of what you've managed to accomplish so far and wanting to share your progress, but this server is about Leonardo.

[eternal plover]

your rules are wrong but I respect it very much! I will not share again! but what is the thing that I did self-promotion? you mean that there is not allowed images that are been generated from other AI huh?

[copper warren]

promoting your own AI model would be considered self-promotion, yes. we don't allow any form of self-promotion, meaning anything from people who are looking for work, sharing their youtube, sharing links to other tools.

[eternal plover]

so did I sent a link or something?

ok fine

[eternal plover]

but it makes me disappointed

[copper warren]

we should move this to the general channel to stop disrupting image help

[marble cape]

Hey pooka

[candid perch]

Hello, i want to make hyper realistic images about 1960 : this one looks good (2 people) but the second looks like unreal :

wich presset must I use ? I'm now with Lifelike vision

i want all picture with same render (picture one)

[thin tinsel]

What about Portrait Perfect with Portrait cinematic?

[candid perch]

I try

[thin tinsel]

What to do you mean the same 'render'? I thought you wanted like a photo, Is that correct?

[candid perch]

yes

real

real scene in 1960

[thin tinsel]

So what do you mean by render?

[candid perch]

sorry my english is bad

quality of picture

i want pictures like real life in 1960

[thin tinsel]

No it's fine. I just want to understand so I don't send you off on the wrong path and waste your tokens

[thin tinsel]

There is an element that will help with that. Let me just check the name

[candid perch]

better with Portrait

[thin tinsel]

Ok, you've got Modern Analogue Photography, and Vintage photography.
Try the Vintage Photography element first on about 0.4 and see what you think.

[candid perch]

peopel doesn't looks real

Try the Vintage Photography element first on about 0.4 and see what you think. where can i find that please ?

i want real picture like this one

[night vault]

can you try photoreal, realistic vision model?

[thin tinsel]

https://discordapp.com/channels/989166677390426132/1170046881183834192/1306593400971202612

[thin tinsel]

Sometimes multiple people can be an issue, plus you still need to have a detailed prompt.

[candid perch]

but wich presset must I use ? portrait ?

it looks real ?

[real lynx]

looks like an old cereal advertisement

[candid perch]

it not my goal i want ultra realist pictures

[real lynx]

Do you have premium tools?

[candid perch]

yes

[real lynx]

So tell me clearly, what does an old family want?

[candid perch]

don't know haha

[real lynx]

I mean whatever comes out But old and realistic style

[candid perch]

can you please check your dm ?

[real lynx]

If you need it to be private you can use the /mentoring command

This opens a private room where only you and all the mentors in the communities can come in to help you.

[candid perch]

thanks

looks

i write in mentora

[marble cape]

Still trying to implant this background but with different staff

[manic root]

Hey Leonardians! Checked out today’s theme in #1108130218260041758 yet? Jump into the action, try your hand at the challenge, and don’t forget to cheer on your peers by voting with the Leonardo emoji . There are some awesome prizes up for grabs! Enjoy!

[marble cape]

Yeah it will give edges if I do it on my pad

[steel snow]

what model is this?

[thin tinsel]

it's in legacy mode. It's called 3D animation style

[lament bane]

Hello, I am getting conflciting info about the character reference feature. I understand it became available in May but it says "Coming Soon" for me

[night vault]

it is available to all but Phoenix model

[lament bane]

Okay, thank you/

I tried it in the other models and it doesn't seem to work very well. Any tips?

[real lynx]

Hello good evening, what image are you trying to create with character image?

Is it a human face or an anime?

[manic charm]

hey, what settings in canvas do you guys recommend to fix the poor eyes in this image?

I tried this settings in screen, but didnt work

[opaque geode]

hey

[night vault]

try increase rendering density. it can help with small area.

[winged ferry]

HI Fam, I created a image in another AI Platform which has now closed, I want to know which models and fiters would be best to achieve this image in Leo as am I only joined today? I do have my prompt for this

[thin tinsel]

basically you can do anything with Leonardo Phoenix model and chose the appropriate preset.
You'd use "describe with AI" to get the prompt.
This style of image would also be possible with Albedo. but you'll only know by testing and experimenting.

[winged ferry]

Where can I upload the image to generate a prompt from it?

[thin tinsel]

Describe with AI is located in the Sparkles ✨ next to the generate button

[winged ferry]

NOpe it renders this type ofimages

[thin tinsel]

And what do you want to see that you don't see

[winged ferry]

the same image that I shared above, cant get this styl in Leonardo

[thin tinsel]

If you cannot explain it to me, then how will you add to the prompt?
Again, what do you want to see that you are not seeing in the leonardo images?

[winged ferry]

The same output style, I tried using the Original prompt I used from the old AI Platform + also let Leo create a prompt by uploading the image but now sure which models, filtersetc to chose

[thin tinsel]

@winged ferry
How do you describe the output style?
Is it a photo? Is it a drawing? Is it vector illustration? Is it abstract?

[winged ferry]

So here is the original Creation

[thin tinsel]

Using WORDS, please describe to me WHAT about this image, is not the same as the images that Leonardo has given you.
I am trying to help you. Again, if you cannot explain to me what you want then you will not be able to explain to the AI.

Is it the style you want Or is the hair colour, is it the side profile, is it the paint drips at the bottom, is it that her hair looks like ribbon, is it the curl on the side of her face?

[thin tinsel]

If the other platform you used was based on Stable Diffusion then try Albedo, with both prompts.
Phoenix is not based on Stable Diffusion.

[open echo]

seeing from the image info, if you want to start by replicating the settings as close as possible, it’d by going into the Legacy mode and manually choosing the « SDXL 1.0 » model.

Make sure to input exactly the same positive AND negative prompt and use the same seed. Then again, I see they had a « Zavy Chroma » filter added, and Leonardo doesn’t use the Euler sampler (called a Scheduler here).

Every AI image generation system has their own « secret sauce » behind the scene, so it’s near impossible to directly take a prompt from one website and have the same results on another.

From what I see, I’d presume the best guess would be to use « AlbedoBaseXL » in Legacy Mode, or the « illustrative albedo » preset in the modern UI as it’s more of a digital painting type of image. But you will have to change your prompt and see what works and what doesn’t

Just tested, and definitely « AlbedoBaseXL » in Legacy or « Illustrative Albedo » preset in modern will get you closer to what you want.

But they definitely had their secret sauce to make the image like this (Just as Leonardo has some of its own 😉). So moral of the story is: taking a prompt from another service and bringing it to Leo, or taking a prompt from Leo and taking it to another service, will never output the same images. 🫂

[thin tinsel]

Content reference and Style reference if you just basically want the same image.

[winged ferry]

Wow great stuff thanks Guys Stunning Wow - Still new at Leonardo

[thin tinsel]

You're welcome. 🤗

[open echo]

@small garnet
Hey there 😄

If I understood correctly, you have a map with a blank look and you want to generate over it using a specific style ?

[small garnet]

I have an abstract map without terrain. It has two rivers converging and place markers for structures I want to be located in the map area. I can not get Leonardo Ai to draw an image that looks real based on the map I installed in the datasheets and trained with it. How do I get Leonardo Ai to use that image and turn it into a real-life look based on that ?map

[open echo]

The way I would go about it:
Instead of loading it in your model, you should instead use the « Content Reference » feature to use it as a base plate. Then with the prompt and your model, it will generate around the lines and feature of your image

[small garnet]

hope you are here for a bit because I may need pointers along the way

how do i get the images produced to stop saying they might be explicit and greyed out the images every time? they are pictures of rivers

[open echo]

Make sure the prompt doesn’t contain words that could be double-entendre

Also since using a content reference and a trained model, the prompt can be SUPER short. Like ridiculously so

I went from the blank map to the other with the prompt « fantasy map, island »

[small garnet]

u mean i must have toggled something because all iimages i create are doing that

[manic root]

Hey Leonardians! Checked out today's theme in #1108130218260041758 yet? Join in, tackle the challenge, and cast your vote with the Leonardo emoji to support your favorites. Big prizes are up for grabs! Have a blast!

[open echo]

Granted for this I had the b/W map loaded as « content reference », and since I don’t have a trained model I loaded this other image as « Style reference » 😁 (cheeky)

[open echo]

Are you always using really long prompts or it does it with very short ones as well?

[small garnet]

um, no one time I put in the river and the image says click to view all of my messages are doing that no matter the prompt text

[open echo]

Hmmm… I remember there was a « NSFW » toggle to hide risqué image…. But looks like yours went into overdrive 😅

I’m trying to find it again as they changed it recently.

Worse come to worse, at the bottom corner of the website there is a small purple button ( ) that lets you speak directly with support members (after answering a couple of questions from the chat bot 😜 )

I’m looking around where they put the toggle now

[small garnet]

i wondered what that was

[open echo]

Aaaah!! Go into your Settings, bottom left, and there’s the toggle

[small garnet]

that worked however i tried to talk to that little comic book talking icon and it just vanishes after clicking it

But i can see the images now I have a set of dice where I should be clicking to get to settings to change the map to life-like

im guessing it will give me a random prompt

[open echo]

The dice gives a random prompt yeah 🤣

Good to goof around, but may not be the best tool for your task at the moment

[small garnet]

well i did not put it there how do i get the picture icon back?

i turned off legacy mode and it showed back up soed that mean i cannot use legacy mode for changing a map to lifelike?

[open echo]

In the top right, you probably have the « Legacy Mode » turned on

[open echo]

With Legacy Mode on, you can use Image Guidance to load your map as a « Depth to Image »

So it’d look something like that:

Also for the real-life effect, I’d stay away from the word « map » in the prompt. It would guide the generation too strongly towards drawn content. So « landscape » or « photography » would be better choices 🤔

[small garnet]

Thanks so much for taking the time to help me. I've been so frustrated. Why did the dice appear while in legacy mode when it wasn't there before? Or did I not notice it?

[open echo]

Legacy mode has the dice on the left side 😉

When it’s off, the modern UI has the dice on the right side with the other « prompt generation » features.

[small garnet]

Which mode is better for what I'm trying to do?

[open echo]

If you absolutely need to use your trained model, it’ll be in Legacy mode only.

If you can use the Content and Style Référence feature instead, I find they offer more flexibility

[small garnet]

ok thanks for that and the rest ill go try to make a picture hopefully without blowing through the rest of my tokens to get it if so ill owe you a drink

[marble cape]

Hey

[real lynx]

Hey

[marble cape]

Hard with the edges when I do it on my iPad

[leaden jacinth]

What are you trying to achieve?

[marble cape]

Again to put in background but this time it’s different staff

[leaden jacinth]

Ok are you making any strides after all the guidance received?

[marble cape]

Yeah iPad I use give edges

[leaden jacinth]

Nice. Post your finished and approved images in https://discord.com/channels/989166677390426132/1052363298370236527

[marble cape]

No what I’m saying when I add the background using my iPad it gives the edges, it doesn’t completely cover the whole thing

[leaden jacinth]

Keep practicing until you get it. Tell AI in your prompt or use Canvas on a device you can edit correctly on, such as a desktop. Thank you and happy creating 🎨

[marble cape]

Was hoping someone would do this for me but I’ll keep pushing through to do it myself

💪

[leaden jacinth]

Sorry pal everyone is busy creating as well. We've all hit a hiccup where something was difficult to interpret through AI. You must remember it's a tool made by humans, for human use, and is capable of mistakes.

[leaden jacinth]

https://cdn.leonardo.ai/users/46484726-47e0-47c2-8d2b-6ecb58d855f2/generations/2532f146-3416-4f66-9b9b-8bca237d4120/Leonardo_Anime_XL_Mysterious_creature_all_dark_purple_and_shad_1.jpg

[manic root]

Hey everyone! Have you jumped into today's theme over on #1108130218260041758 yet? It's a blast! 🎉 Participate, cheer on others with a on favorite entries, and who knows? You might just snag some major prize tokens! Let's have some fun!

[marble cape]

Nice photo

[leaden jacinth]

https://cdn.leonardo.ai/users/46484726-47e0-47c2-8d2b-6ecb58d855f2/generations/2532f146-3416-4f66-9b9b-8bca237d4120/Leonardo_Anime_XL_Mysterious_creature_all_dark_purple_and_shad_2.jpg

Anime Flat - Consider further exploring Leo presets and models to find what you need. Or you might want to get a Leo/mentor for one:one guidance @marble cape

https://cdn.leonardo.ai/users/46484726-47e0-47c2-8d2b-6ecb58d855f2/generations/2532f146-3416-4f66-9b9b-8bca237d4120/Leonardo_Anime_XL_Mysterious_creature_all_dark_purple_and_shad_0.jpg

[marble cape]

Really magnificent

[leaden jacinth]

Thank you. Download and use any of the images "AI Describe" to create a new prompt for you to work with.

[leaden jacinth]

https://cdn.leonardo.ai/users/46484726-47e0-47c2-8d2b-6ecb58d855f2/generations/2532f146-3416-4f66-9b9b-8bca237d4120/Leonardo_Anime_XL_Mysterious_creature_all_dark_purple_and_shad_3.jpg

[leaden jacinth]

🆒

[marble cape]

Originally cane out of this

[leaden jacinth]

I remember from the Serenity Prayer. I was helping you.

[marble cape]

Looks different but when I get subscription I’ll make it content reference

[leaden jacinth]

🆗

[leaden jacinth]

nice

[real lynx]

Hi

There are many applications that can erase the background of your image, my friend, so don't get stuck, look for alternatives to reach your goal.

[real lynx]

Use the /mentoring command too

[hardy sentinel]

You can post your questions here @nocturne tartan

[hardy sentinel]

Beautiful image

[nocturne tartan]

Thank you! I will remember that for next time. I figured out it was the word mutated that was blocking me

[hardy sentinel]

That is unfortunate. Is it in the positive prompt or negative prompt?

[nocturne tartan]

Positive prompt. I removed the word though and works fine now.

[hardy sentinel]

Ok

[deft charm]

Hello guys. by any chance does anyone know if i can extend the duration if i motion a photo on leonardo?

[stable carbon]

hi. not on leonardo, no. maybe they're working on it, but at this time you need to take the clips from Leo and attach them in external software for that

[marble cape]

yeah dont know why it does that in my pad

[hollow breach]

What is your best tip to avoid trademark, signature, logo that appears sometimes on the generated image especially with Phoenix. I tried to put in the negative trademark, signature, logo. But it doesn't seem to have any effect. I always removed them with canvas but it's annoying. Any suggestions

[hardy sentinel]

which preset do you use? how long is your negative prompt? In my case I have a relatively short negative prompt and it works 🙂

[hardy sentinel]

if you want to, you can use /mentoring command to have a private session

[hollow breach]

I usually use bad hands, logo, signature, trademark. So it's quite short.

[hollow breach]

Thanks for the offer but I'm sure there are other people with higher needs. And I know how to deal with it afterwards

[hardy sentinel]

do you use automatic prompt enhancer ?

[hardy sentinel]

ok make sense

I turned my prompt enhancer off

do you need text to be in the image ?

maybe remove "logo" and "trademark"

usually I use text only

if I want to make very sure, then I use text, signature at the end of my negative prompt

and my negative prompt is slightly longer than yours 🙂

[manic root]

Hey Leonardians! Have you checked out today's daily theme in #1108130218260041758? Jump into the action, tackle the challenge, and make sure to cheer on your peers by voting with the Leonardo emoji . There are significant prize tokens up for grabs! Enjoy!

[hollow breach]

I haven't turned it off, maybe I should. My prompt are usually quite long in Phoenix, so the auto prompt doesn't change it.
And I can add text when I don't need text. Thanks that's helpful

[hardy sentinel]

does it work?

[hollow breach]

I haven't tried yet. I would let you know when I do.

[proud summit]

who sniped lol

[stable carbon]

🤨

[proud summit]

Someone posted something sus and it got deleted quick, figuired someone was here but I think botti did it

[hollow breach]

Phoenix, dynamic style, i got some scribbles on one out of two.
Phoenix, Cinematic, none.
prompt enhance is off.

the prompt was
In the foreground, an ash-covered rabbit nibbles on a small bundle of carrots, its once-pristine white fur now a dusty gray. Behind it, in the middle ground, a 25 yo male wildlife volunteer stands amidst the charred remains of a burnt forest, his back to the viewer, his clothes smudged with soot and ash the volunteer is carrying a big bag of carrots. The scene is captured in a poignant and evocative digital art piece inspired by the styles of Studio Ghibli and Simon Stålenhag. The color palette is muted and somber, with the cool, desaturated tones of the ash and charred wood contrasting against the warm, vibrant orange of the rabbit's carrots and the volunteer's bright, hopeful yellow safety vest. The composition is layered and thoughtful, with the rabbit serving as a symbol of resilience and hope amidst the devastation. The digital art style employs soft, delicate line work, intricate textures, and atmospheric lighting, bringing the emotional scene to life. The rabbit's fur is rendered with gentle, fluffy textures, while the burnt forest is depicted with harsh, jagged lines and deep, dark shadows. The perspective is from a low angle, looking up at the volunteer and the stark, haunting landscape. The atmosphere is heavy and reflective, inviting the viewer to contemplate the cycle of destruction and regrowth in nature, and the power of compassion and hope in the face of adversity. Prompt by JDS.

[hardy sentinel]

ok, let me try to edit it

[hardy sentinel]

i got zero scribbles out of 4 images

i have shortened your prompt, and used illustration preset

[hollow breach]

Maybe it's the dynamic style. Thanks. I'll use a different one from now on. Thanks for your time

[hardy sentinel]

try to shorten your prompt as well, it is too long 🙂

@hollow breach this is my result

not optimal

[hardy sentinel]

how is it now ?

[hollow breach]

It's fine thanks

[hardy sentinel]

awesome

[hardy sentinel]

you can remove this part to make it shorter: The atmosphere is heavy and reflective, inviting the viewer to contemplate the cycle of destruction and regrowth in nature, and the power of compassion and hope in the face of adversity. Prompt by JDS.

[unborn mauve]

Okay so I need help with prompting, I just don’t get the result with full body, like it’s close up of the face or half the body and face, can anyone help out ?

[wise canyon]

Hi Spartan, this is a common occurrence and can depend also on Model and Presets used, but a most effective way to stop this is to add "footwear" details to the prompt. i.e "wearing boots, high heels, runners, etc"

[unborn mauve]

Thanks, I’ll try doing that next time.

[hollow breach]

or saying things like standing on the grass

[unborn mauve]

Thanks I’ll definitely give it a try.

[hollow breach]

👍

[unborn mauve]

I’ve another question, which model is compatible with alchemy V2 and generates images that stand out?

[leaden jacinth]

Perhaps Phoenix 🐦‍🔥 especially if you want typography in your image design.

[unborn mauve]

Idk but I can’t spot it in the models

[leaden jacinth]

These are in the new models

[unborn mauve]

I was in legacy mode lol

[leaden jacinth]

No problem. Switch the button over lol 😁

[unborn mauve]

and what setting would you suggest for preset style and contrast ?

[leaden jacinth]

Settings are personal to your liking. I generally try various styles. In Phoenix 3D, moody and Pop Art are cool

[leaden jacinth]

[unborn mauve]

I think the model is decent but my preference is something different, I can drop in the image prolly.

[leaden jacinth]

No problem. You may have to try a few. This looks similar to Albedo Model.

[unborn mauve]

Let me give it a try

[leaden jacinth]

You can also try using None as a preset.

[manic root]

Hey Leonardians! Checked out today's theme in #1108130218260041758 yet? Jump into the challenge, and make sure to vote for your top picks with the Leonardo emoji . There are awesome prizes up for grabs! Enjoy!

[unborn mauve]

Yeah this kinda looks similar to what I actually want.

[leaden jacinth]

You may not get the typography (text) on other Models

[leaden jacinth]

Ok cool. You can probably edit in Canvas to add your own text.

[unborn mauve]

Yes, I don’t need text, I just need images to what I sent or they might relate to it.

[leaden jacinth]

Ok perfect. Test out some of the other models to see if they measure up. Happy Creating ✨

[nimble bay]

Looking sharp 🙂

[marble cape]

Thanks!

Next sometime when I have enough money I’ll buy subscription and use content reference

For my other projects

[subtle gulch]

With the right elements you can achieve different styles and touches that match your likings.

[unborn mauve]

Right, I’m still experimenting to be honest.

[subtle gulch]

Leo may not be able to change the color of the Superman logo that easily, but you can get that style quite close to what you want.

[unborn mauve]

Yes this works as well, how did you do it ?

[hardy sentinel]

do you use legacy user interface or the new user interface ?

[unborn mauve]

Legacy user because it’s pretty decent

[hardy sentinel]

ok, next question: which style of art do you like ?

[hardy sentinel]

the models have their own strength

[unborn mauve]

I think acrylic mixed with realism sounds good for me

[unborn mauve]

Definitely something like this

[hardy sentinel]

oh ok, if painting, I like to use Albedo Basel XL

[hardy sentinel]

alternatively you can try Leonardo Vision XL

[unborn mauve]

But is there any elements that would get me similar results ?

[hardy sentinel]

I think the text is added using GIMP or PhotoShop

[unborn mauve]

I’m talking about the image, I can add the text with different app

Basically the image is generate through mid journey but I’d be glad to get the same results through leonardo

[hardy sentinel]

have you tried dreamy acrylics element

[hardy sentinel]

both albedo and leonardo vision are models

[unborn mauve]

Yeah but I don’t know why it’s making my characters look like dwarf lol

[hardy sentinel]

then it is not for your purpose

[hardy sentinel]

for acrylic painting I have used albedo base xl mostly

[unborn mauve]

And what style would you choose ? Like I’m choosing cinematic mostly

[hardy sentinel]

illustration if you want to get acrylic painting

[unborn mauve]

Oh alright let me try that out

[hardy sentinel]

good luck

[real lynx]

I like dwarves, I can see your image 😸

[unborn mauve]

I don’t have it now lol but it kinda looked like a tiny green lantern 💀

[unborn mauve]

Here’s another example of what I kinda want

[hardy sentinel]

awesome

[unborn mauve]

I think cinematic works with albedo base xl

[unborn mauve]

How did you do this ?

This is what I actually want 👀

[real lynx]

The face then does not matter

[unborn mauve]

Yeah I agree w that lol

[real lynx]

From what I see, it's hard for him to get over the Superman suit.

The white suit comes out sometimes

They even mix together

[unborn mauve]

So it’s different all the time ?

[hardy sentinel]

yes, ai assisted image generator work with randomizer

[unborn mauve]

Yeah but what if face swap 💀

[real lynx]

Image is a digital artwork featuring a muscular male character with a strong, chiseled jawline and intense blue eyes. He has short, styled white hair and a rugged, weathered appearance with visible scars on his face. His skin is a light tone, and he wears a metallic, white-colored suit with a prominent red emblem on the chest, resembling a stylized 'S'. The suit has a textured, armor-like surface, and a red cape is draped over his shoulders, evoking a sense of Superman-inspired attire. The background is blurred, suggesting a high-altitude setting, possibly in flight. The overall composition is dynamic, focusing on the character's determined expression and powerful stance. Zdzislaw Beksinski and Wes Anderson styled PVC showpiece Masterful color graded, Highly detailed surfaces, reflective and absorptive, contour and relief, subsurface scattering, specular highlights, ambient occlusion, dynamic range, (symmetrical framing), microdetails, physically-based rendering (PBR), ray tracing, realistic depth of field, displacement and normal maps, (vintage lighting), Shot on (Super 35mm), Depth of view, (F/11)

This prompt uses

Semi real anime preset

[hardy sentinel]

character reference works with real person photo

[real lynx]

plinth, title, text, three-legged, anatomical distortions, anatomical inaccuracies, proportional inaccuracies, blurry, low quality, unclear features, incorrect features, mutation, deformed, disfigured, cloned, cropped, out of frame, out of picture, double head, multiple figures, signature

[unborn mauve]

Yeah but I know a website which literally changes the face

[real lynx]

negartive prompt

plinth, title, text, three-legged, anatomical distortions, anatomical inaccuracies, proportional inaccuracies, blurry, low quality, unclear features, incorrect features, mutation, deformed, disfigured, cloned, cropped, out of frame, out of picture, double head, multiple figures, signature

[unborn mauve]

But like what model and settings did you use for this ?

[hardy sentinel]

leonardo anime xl using the new user interface

[unborn mauve]

And what is the style ?

[unborn mauve]

thanks

[ruby sable]

so is leonardo diffusion not a thing anymore? last time I was here, I was making stuff in pheonix, and then switching to ddiffusion to flavor it and add style. Is there a new way to do that? Or where might I find it.

[stable carbon]

you can swap over to legacy mode at the top right of your screen. in legacy mode you'll find models like Leo Diffusion and Leo Diff XL.

or are you talking about something else?

[ruby sable]

U da best Ty

[charred light]

Hi everyone.

Hope someone can help me. I am trying to create art that simulates AAA games. So far I have made the following with Dall E. But looking to create stuff like the images attached with Leonardo, which models and settings would be best please?

Just to re-iterate, not asking for help for prompts but rather which models and settings would be able to genetate stuff like this with the same look and feel, with the same render effects if that makes sense.

[real lynx]

I made this with semi-real anime

[charred light]

Awesome will check that out thanks 👍

[keen hatch]

@real lynx hi

[real lynx]

hi

[keen hatch]

im trying to generate an image of a character, and i need him to be pointing at the screen but idk what prompt to put in

like this

because when i put in "pointing" he just points up or somewhere random

[real lynx]

Do you want a prompt of that image or an equal image?

[manic root]

Hey Leonardians! Checked out today's theme in #1108130218260041758 yet? Jump into the action, participate, and don’t forget to cheer on your peers by voting with the Leonardo emoji . Awesome prizes await! Enjoy!

[keen hatch]

ye this image

but mostly wondering about how to make him point at me

like i showed

do yk? @real lynx

[ruby sable]

How do you write a negative prompt in pheonix? Legacy just has a little toggle

[real lynx]

Sorry I didn't tell you I had gone to create

To give you the best indication

I was creating elves 😅

Would something like this be okay?

[keen hatch]

yeah but how did u make it point at me

like what did u input

[real lynx]

Yea

[keen hatch]

what do you mean

what did you write as the prompt

[real lynx]

Step by step

First find your anime style

I found it in legacy mode

They are two different systems

For this preset I turned it on

The preset I used is 3d animation style

[keen hatch]

ok and what did u type in

[real lynx]

Then use a guide image with your image

With the Depth image tool

You can also have AI describe your image

This gives you a prompt

(((pointing with an index finger at the camera))) of Image is a digital illustration estilo pixel art in a cartoon style, depicting a classroom setting. The main subject is a male character with exaggerated features, including a large mustache, thick eyebrows, and a prominent nose. He has a light skin tone and is wearing a white shirt with rolled-up sleeves, a brown vest, and a striped tie. The character is pointing forward with an intense expression, mouth open as if shouting. In the background, there is a green chalkboard and several students with varying hairstyles seated at desks, though they are not the focus. The classroom has a bright, well-lit atmosphere with fluorescent lights on the ceiling.

This prompt gave me that tool

All this is using free tools

[keen hatch]

ty

[real lynx]

If you have any other questions, come back here and I'll be on the lookout.

[real lynx]

As long as your progress is

[nimble bay]

When need to make that spear a symbol of this channel somehow lol

[real lynx]

https://tenor.com/es-US/view/spear-gif-15891508659345924813

I have a gif with his first spear

[nimble bay]

🙂 Good grab

[nimble bay]

So cute

[hollow breach]

[marble cape]

Next going to generate this to

[marble cape]

Anything I can do you can do

[hardy sentinel]

Hi, do you mean the first image as content reference and the second as style reference or vice versa ?

[marble cape]

I think when I buy subscription there going to be content reference

Either that or in depth

[hardy sentinel]

ok

[hardy sentinel]

which image should be the reference of the style ?

[hardy sentinel]

these images are very minimalistic, vector art ?

[marble cape]

For content reference I think it’s going to be the one with the candles and also the other one with the ghost but like I said I feel it’s going to be either content reference or in depth image

Now for the prompt I may play around with it a little

Yeah I did on my pad years ago

[hardy sentinel]

ok, sure, do you want the style to be simplistic or I have more creative freedom?

[hardy sentinel]

that is awesome

[marble cape]

Like the same

Maybe for the candles leading maybe reality or anime

But for the other one it would be more like maybe animation

[hardy sentinel]

ok realism or anime

[marble cape]

For candles

[hardy sentinel]

ok, later i will try with the candles first

[marble cape]

3D I don’t know

Oh you want to help me?

[hardy sentinel]

later, it is almost 2 o clock in the morning 🙂

[hardy sentinel]

3D? but your images are 2D?

[marble cape]

No one has ever said that to me 😓😅 it would be an honor

[hardy sentinel]

ok, i will go to sleep first, see you later, i have noted your wishes

[marble cape]

Not sure I was just testing

[hardy sentinel]

Good night

[marble cape]

Ok thank you sir have a good night..

And trust me there be more projects to come

Cause anything I can do you can do

[unborn mauve]

I hope I’m not bothering but can you mind doing some batman images as well ? I wanna see how that looks, I think I’m making some mistakes with prompting.

[real lynx]

hi

You don't bother me Batman is very recognized by the AI just by putting the name I modify my entire prompt

[real lynx]

I was creating mechanical monsters

[manic root]

Hey Leonardians! Checked out today's daily theme in #1108130218260041758 yet? Jump into the action, challenge yourself, and don't miss the chance to cheer on others with the Leonardo emoji . Big prizes await! Enjoy!

[unborn mauve]

Damn that really looks like batman

[unborn mauve]

What model did you use for this ?

[real lynx]

anime semi realista

[open echo]

This one was with Pheonix.

Most of the time, issues with the results are due to a prompt that’s way too long.
The system just gets confused trying to integrate everything asked, And you are left with a muddy result

[real lynx]

I liked the hat xD

[unborn mauve]

Gotcha, honestly I’m using albedo at the moment, I’m getting better results, but still what I want isn’t being accomplished yet

[unborn mauve]

That’s one hell of a hat

[real lynx]

This is very funny

[unborn mauve]

That looks like Lego bat lol

[open echo]

AlbedoBaseXL is great a comic book characters.

Sometimes for complex poses or styles, the best is to leverage the « References » features.

Like this is AlbedoBaseXL with a scribble as style reference

[unborn mauve]

Personally I don’t want it to look like comic, I want it to look like a blend of slight acrylics paint and realism, but I don’t know how to achieve those results

[open echo]

Elements maybe?
We just got a « Dreamy acrylics » one.. paired with Vision or Lightning for the realism. That might do the trick.

I’ll try it out for funsies

[unborn mauve]

Should I try it out with lightning ? Because I tried with albedo and it looks a little whimsical tbh, the green lantern I was generating became a dwarf lol

[open echo]

AlbedoBaseXL is best for drawings, paintings, etc..

Like this is Lightning with Dreamy Acrylics at 0.5

Same prompt but the Element at 0.3

[unborn mauve]

Idk but why does it turn the characters small ?

[open echo]

Sometimes the prompt is too long and somewhere within there’s a word or a turn of phrase that gets interpreted as « short people » or an Element is too prominent or there’s a reference image that was left on that we forget… can be due to a lot of things really 😅

Sometimes when all seems to be failing, best course of action is to go into the left toolbar, all at the bottom, and click on « restore default settings » hehe

[unborn mauve]

I agree it’s really hard to understand where you’re making mistake when using leonardo lol

[open echo]

Lots of different tools, levers and buttons… so it’s easy to leave something as « on » when we work on something else 😜

[unborn mauve]

That’s true lol but gotta keep on researching everyday I guess to get your results

[unborn mauve]

Do you think results alike mid journey can be achieved from leonardo ?

[open echo]

It’s a tool with a learning curve like any other 😜

Some people make it more complicated than it needs to be, but really, it’s way more powerful than Midjourney 😉

[unborn mauve]

Kinda agree with that, at least it’s free, you can’t even use midjourney lol

[open echo]

@unborn mauve
So the prompt I was using for my Batman was:

Batman, standing solemnly atop a skyscraper, night sky, gothic architecture, masterwork

Nothing really special: simple, to the point

[unborn mauve]

Gotcha but like what if you want to change the outfit of batman

For example if I want it to have a Brazilian flag themed outfit or something like that

and also prolly have Christ the redeemer statue in the background ?

[open echo]

Many things are possible: Leonardo Pheonix is great at more complex layout with multiple details

Or! You can also use Leonardo Canvas to generate additional details and change clothes Colors after the fact (it’s like a localized image generation)

(I didn’t have time to really go in-depth with the exemple… but just to show what’s possible with a minute of time 😅)

[unborn mauve]

Gotcha, means a lot lol I mean I didn’t knew about the canvas, I’ll try checking it out tomorrow, but I guess multiple generations does the work and prolly I must switch my prompt for better results.

[open echo]

Some prompt pointers:

• keep it short and to the point
• first subjects will be more prominent
• keep relevant information together (describing a character THEN describing the background. Avoiding back and forths..)

Just some quick tips

And yes, sometimes just rolling the dice and seeing what comes out 😜

[unborn mauve]

Alright, so if I’m to generate a prompt from chatgpt do I have to tell it to write the description for both ?

[unborn mauve]

That’s what I do, I only choose the realistic ones

[open echo]

ChatGPT is not very good at generating image prompts 😅

Tell it to create a list of details separated by commas.

If you see it going like « Immerse yourself in a mesmerizing picture of …. » you can save yourself the tokens and not try it 😜

[unborn mauve]

Yeah I remove that part mostly from the prompt

Keep the details which generator is going to catch

[open echo]

(Oh! And That’s speaking about achieving a precise results. Often times it’s fun to just throw a huge novel-lenght prompt and see what comes out. But when going for « consistency », it will give too much variations)

[unborn mauve]

Yeah, so I think it’s better to keep it short and to the point

Also which model do you think is on par with mid-journey images ?

[unborn mauve]

This is not what I meant 💀

[real lynx]

xD

[unborn mauve]

That’s Mexican themed batman

[open echo]

Depends on the style

Artistic or photorealistic?

[unborn mauve]

Bit of both

[open echo]

If you scroll the main page of MJ and Leo, you will the results are pretty similar

Artistic, I’d go with AlbedoBaseXL or Anime

Photorealistic, Vision and Kino
Or, Lightning for REALLY grounded photography.. like a picture that looks taken from an iPhone

Leo Lightning 😄

[unborn mauve]

Lightning looks cool

But is it good for generating superheroes ?

[open echo]

It can work

I’ve seen people use a more advanced workflow of creating a complex image in Pheonix, edit it in Canvas, then bring it to another model by using either Image to Image or Content Reference

Leo is a toolbox 😉 So many possibilities !

[unborn mauve]

Sounds good, i think I still need some time to experiment with it.

[open echo]

Of course! 😄

expect a lot of burned tokens and experimentations at first

That’s how we all learned 😉

[hollow breach]

Part of the fun, is to try different options and tools, a prompt can be 100% of the image or it can be part of it. Style reference is a also a good tool, if you know what look you want.

[unborn mauve]

Right, I just can’t seem to find my style, so far albedo and diffusion works for me, I haven’t tried fusion or lightning yet.

[leaden jacinth]

Are you getting the results you're seeking?

[unborn mauve]

Close enough but idk, can I show you the edit of what I kinda want ?

[leaden jacinth]

Sure. Post your image.

[unborn mauve]

It’s not an image, it’s a video: https://youtube.com/shorts/X2g9Fyfl6bg?si=O2SDah1qHKMRLQpZ

But yeah I want images like that.

[leaden jacinth]

Did you enter the prompt to generate an image?

[leaden jacinth]

There's lots of trial and error. Try uploading a still image.

[unborn mauve]

Yes I have but it was from albedo

[leaden jacinth]

ok?

[leaden jacinth]

What seems to be the issue? Image or video creation?

[leaden jacinth]

You've chosen a reputable model. Anime General is a nice preset.

[unborn mauve]

It’s realistic but the outfit is like silicon or acrylic paint

[unborn mauve]

Image, video I can do it easily

[leaden jacinth]

Is this what you generated with Leo?

[unborn mauve]

No it’s someone else’s work

But I want something like this

[manic root]

Hey Leonardians! Have you checked out today's daily theme in #1108130218260041758 yet? Jump into the challenge and remember to cheer on your peers by voting with the Leonardo emoji ! Awesome prizes are up for grabs! Enjoy!

[leaden jacinth]

How are you describing the outfit in the prompt?

[leaden jacinth]

Try describing the outfit how you envision it.

[unborn mauve]

I’m just stating with intricate design and country’s flag colour

[leaden jacinth]

If you want the outfit to be like a video example then you would need better descriptors/keywords.

[unborn mauve]

Alright but what model would you suggest that is like the accurate one to my ideal type ?

[leaden jacinth]

I would test in Phoenix once the prompt is to my liking.

[unborn mauve]

Gotcha and style ?

I’m gonna try Phoenix today but I’m gonna show you my prompt first if you’re available

[leaden jacinth]

Try 3D.

[unborn mauve]

Thanks I’ll try that

[leaden jacinth]

Sure. TTYS. Going to sleep now.

[unborn mauve]

Talk later, night.

[hardy sentinel]

have you tried realtime canvas previously ?

[hardy sentinel]

this is with prompt only

[hollow breach]

You can look at the feed and pick an image whose style you like and use it as style reference

[hardy sentinel]

using your candles image as reference

[unborn mauve]

I can try that too

[hollow breach]

or an internet image, or a photo of some art you have

[marble cape]

Well I was looking more into like a mist comming out of the candles as you see

White mist

[hardy sentinel]

Ok you can try that using various tools. First the area of the white mist should be white

[marble cape]

Yeah

When I generated it for the prompt it gaved me this

Image is a digital illustration featuring three white candles with visible wax drips, positioned centrally against a solid black background. Each candle has a small, bright flame with a yellow and orange gradient, creating a warm glow. Surrounding the candles are stylized, white smoke trails that curve and swirl, adding a sense of movement and atmosphere to the composition. The overall style is minimalist and monochromatic, with a focus on contrast between the white elements and the dark background. The illustration conveys a sense of calm and simplicity, with the candles and smoke forming the primary visual elements.

Does that add right I wonder?

[hardy sentinel]

You can try it

[marble cape]

Does that sound right though?

[hardy sentinel]

I don't know

[hardy sentinel]

You have to try

[marble cape]

Instead of smoke trials that says it I’ll add instead white mist

[unborn mauve]

Should I generate prompt using chatgpt or the one leonardo has ?

[stuck otter]

Does anyone know how to get character reference to work consistently. In the past it was fine, but now when I use it generated character looks plastic not a real person. Both in lifelike and portrait options.

[real lynx]

I sometimes use gpt chat sometimes it generates nice things

[real lynx]

Hello, do you have premium tools?

I would like to see how you generate your images

[stuck otter]

Yes. I have premium

[real lynx]

Sometimes quality is disabled

You have to activate it again.

[stuck otter]

I tried fixed seed as well.

[real lynx]

What preset do you use?

I like these

This is the best for me

[unborn mauve]

Can you suggest which one might work best for me ?

[stuck otter]

Lifelike vision, portrait perfect, tried Leonardo Lighting now.

So original

with reference. Same prompt

[real lynx]

I don't know, that's just a matter of what you're looking for.

[unborn mauve]

Which one is this ?

[real lynx]

What are you looking to create?

[stuck otter]

Leonardo da vinci. Problem is faces become like plastic like smoothed out. While original creation is more life like

[real lynx]

You are looking for an image, not a statue.

[stuck otter]

Yes

[stuck otter]

Ok if that character reference it's much better. Could you share settings? Thanks

[real lynx]

Real image Leonardo da Vinci Waving with one hand raised

[stuck otter]

is this with character reference or just prompt

Ok, I got it. Thanks

[real lynx]

miniatures

[real lynx]

Here are more tools that can help

[unborn mauve]

This @real lynx

[stuck otter]

Thanks a lot

[real lynx]

Too much real came out

[stuck otter]

Lol

[unborn mauve]

Yes lol

I think vision and albedo work best

[unborn mauve]

Relatable 👀

[unborn mauve]

Which one was this ?

[unborn mauve]

Yeah this looks like what I might want

[real lynx]

He looks like an Argentine actor xD

[unborn mauve]

Yeah literally lol

[real lynx]

The image is a digitally created piece of art featuring a muscular superhero character, Superman with a strong, chiseled jaw and intense blue eyes. The character has light skin and striking white hair styled voluminously and tied back. He wears a metallic-looking suit with a prominent red and silver emblem on his chest resembling the letter "S", and a red cape draped over his shoulders. The background is a blurry aerial view of a cityscape. The overall tone of the image is dramatic and heroic, focusing on the character's powerful presence.

Prompt

In this case use the phoen preset

And also use your picture in picture Style

[unborn mauve]

Is it cinematic or dynamic ?

[unborn mauve]

Thanks let me try it out

[unborn mauve]

Can I use cinematic instead of portrait ?

[real lynx]

Yea

[unborn mauve]

I think maybe I’m not getting good results because I don’t have subscription

[real lynx]

I changed the view

[unborn mauve]

Is there any way to do this with alchemy ?

[real lynx]

bird's eye view aerial view

I just added that to the prompt

[unborn mauve]

I can make this image to vid and it’d look bomb

Can you do it with a green lantern ?

[real lynx]

I think so

I never make superheroes 😅

[unborn mauve]

Like have the same look just not the cape and S symbol

[real lynx]

My superheroes are little dwarfs with long ears

[unborn mauve]

I do make superheroes lol

I turn images into videos

That’s why I need decent images

[real lynx]

Because you can't generate the images yourself

If I taught you how to generate it

[unborn mauve]

i can generate them but i don't have subscription for leonardo

maybe that's why quality gets butchered lol

this is what i got for green lantern following your prompt lol

[real lynx]

Images with free tools

[unborn mauve]

smh

so i can;t generate it with free version

[unborn mauve]

okay this is period and it's exactly what i want

[unborn mauve]

mind giving me jpg version of this, i would like to make a video of this

[real lynx]

I am here to help you learn how to use the tools and use them.

I'm not going to do all the work

If you have questions I can help you with the tools

[unborn mauve]

is there any other way to generate images like you did with other models ?

[real lynx]

With what I taught you, you could try to generate an image

Yeah

You just need to change the model and keep the rest.

To see the new changes

[unborn mauve]

gothca

[real lynx]

This would be with free tools

I think it's pretty good

Power up in the image of your original Superman

[manic root]

Hey Leonardians! Are you in on today's theme challenge in #1108130218260041758? Jump into the action, take part in the excitement, and make sure to cheer on others by voting with the Leonardo emoji . Fabulous prizes await! Enjoy!

[unborn mauve]

whats that gonna do ?

[real lynx]

Generate your Superman texture and a similar image

[unborn mauve]

but what if i don't have a reference image ?

[real lynx]

If you have it you just don't know where it is.

I want you to take a picture of your screen to guide you.

[unborn mauve]

this is what i got so far but with green lantern @real lynx

[real lynx]

Very good

[unborn mauve]

but its with leonardo vision.

[nimble bay]

Dope

[unborn mauve]

Thanks 🗿

[marble cape]

Yeah maybe the candles I’ll anime and this one looks like it can be in anime as well

I’m also working on this design

[real lynx]

And how is that project going?

I would try to modify them with some elements

[real lynx]

Too scary for me 😵‍💫

horror

the demon has my face

[marble cape]

More like anime maybe or 3D animation

[real lynx]

For me it was too scary 😅

[marble cape]

I got more projects

[marble cape]

My next project is implementing my original design style model into this image;

[real lynx]

I thought that was over

[real lynx]

It was the first project you presented.

[unborn mauve]

@real lynx hey what should i do if i want to implement a country's flag colour in a suit

[real lynx]

Mmm for example I am from Argentina

I just described the Argentina soccer team

Blanca y celeste similar a la de Leonel Messi con el número 10

I was with a Minecraft prompt 😅

[unborn mauve]

Oh gotcha so I have to state Argentina flag themed suit right ?

[unborn mauve]

Damn this looks nice

[real lynx]

If the one that best understands everything right away is the Phoenix preset

[real lynx]

This is with premium tools

[unborn mauve]

Yeah but unfortunately only if I had subscription it would work in my favour

[unborn mauve]

I can’t see the picture

[marble cape]

For the background on this image would it look better in that same color or much darker?

[real lynx]

I don't like it as it is.

But don't listen to me, I'm into images with lots of colors 😅

[real lynx]

Oi

Você está aí?

[marble cape]

Ok?

[thick pebble]

to

[real lynx]

I want to see your modifications

[hardy sentinel]

using realtime canvas

[hardy sentinel]

using realtime canvas

[hardy sentinel]

if you want the ghost to be dark then the background has to be bright and vice versa

[marble cape]

Well I visioned only three canceled as you see

And the mist swirling around, like it’s comming at you or something

Maybe in anime to

[marble cape]

So the background is fine I mean there’s that edge color as you see

[leaden jacinth]

Nice. I want to learn RC.

[hardy sentinel]

you can remove that in GIMP or Photoshop

[leaden jacinth]

Very Nice Candles. Can you share the steps?

[hardy sentinel]

ok

[marble cape]

How do I do it

[hardy sentinel]

I have used obliviandark image as reference

this image

[marble cape]

lol

[hardy sentinel]

yes this image

[hardy sentinel]

you are faster, I will delete my post

[marble cape]

Have you tried content reference?

[hardy sentinel]

no, because you are a free user, I try only features that are free 🙂

[leaden jacinth]

[marble cape]

Oh

[leaden jacinth]

[marble cape]

I’m ok free version and it gaved me this

[leaden jacinth]

Pretty cool!

[hardy sentinel]

ok, you need to lower the creative strength

[marble cape]

I explained the prompt and everything but it’s just like different still

Looks different still

[marble cape]

Thanks

[hardy sentinel]

which model do you use ?

[manic root]

Hey Leonardians! Have you checked out today’s theme in #1108130218260041758 yet? Jump into the challenge and make sure to cheer on your peers by voting with the Leonardo emoji . There are some great prizes up for grabs! Enjoy!

[marble cape]

Hold up

[hardy sentinel]

this is key as if you use only prompt would look like this

[marble cape]

On free Version I use that thing to describe the image that generates the prompt, that one

[hardy sentinel]

ok, then, which feature or model do you use?

[leaden jacinth]

Cool.

[leaden jacinth]

Sure

[marble cape]

Mostly anime then phoenix

[hardy sentinel]

so you need to combine the image reference and prompt

[leaden jacinth]

Are you referring to style Ref?

[marble cape]

I don’t know I just been playing around with it before to make exact sake position style ohh

[hardy sentinel]

no, I am referring to realtime canvas, as you have asked for the steps

[marble cape]

Oh you know something?

Msyne I should of added a table or stand on it

[hardy sentinel]

oh ok, just try realtime canvas, it is free as long you do not use alchemy refiner

[leaden jacinth]

I need more practice. The generation is more technical than RG. I will take these tips and try it. Thx.

[hardy sentinel]

yes, it is. I started on 1st of December 2023 🙂

[hardy sentinel]

you are welcome, I need to go to bed soon 🙂

[leaden jacinth]

Ok thx. I will see what I can come up when I get a moment.

[leaden jacinth]

Nite.

[marble cape]

Maybe or just the black background itself

[hardy sentinel]

Good night

[hardy sentinel]

Good night

[marble cape]

I don’t know canvas editor is funny when I use it

[marble cape]

😭

[leaden jacinth]

me too lol

[marble cape]

I’m all alone once more

Yeah I’m doing the best I can

Maybe like a glass table or without it

Image is a digital illustration featuring three white candles with visible wax drips, positioned centrally against a solid black background. Each candle has a small, bright flame with a yellow and orange gradient, creating a warm glow. Surrounding the candles are stylized with white mist that curves and swirls, adding a sense of movement and atmosphere to the composition. The overall style is minimalist and monochromatic, with a focus on contrast between the white elements and the dark background. The illustration conveys a sense of calm and simplicity, with the candles and smoke forming the primary visual elements.

Image is a digital illustration featuring three white candles with visible wax drips, positioned centrally against a solid black background. Each candle has a small, bright flame with a yellow and orange gradient, creating a warm glow. Surrounding the candles are stylized with white mist that curves and swirls, adding a sense of movement and atmosphere to the composition. The overall style is minimalist and monochromatic, with a focus on contrast between the white elements and the dark background. The illustration conveys a sense of calm and simplicity, with the candles and smoke forming the primary visual elements. Yugioh trading card game artwork

[hardy sentinel]

Thank you

[real lynx]

Don't be discouraged

You are never alone

Sometimes we just watch him from afar so he can walk alone too 🫂

[unborn mauve]

The hand lol 💀

[marble cape]

Nice photo

Oh yeah I noticed the hand

[real lynx]

I already fixed my hand and I better leave it drinking a coffee

[real lynx]

What a nice, super descriptive prompt

[real lynx]

What image did this generate and can I see?

[marble cape]

I’ll use content reference for my image when I buy subscription

Well for the free version it gaved me this

[real lynx]

I like candles

[marble cape]

You’ll never go cold I guess

Found this on Leonardo

[marble cape]

There a reason I like mist fog or rain for reasons

[real lynx]

I was creating fire and water girls today

[marble cape]

Wait is realtime canvas is in free version? Where is it at?

?

[marble cape]

Oh live canvis

[leaden jacinth]

Here are your prompts in action. Hope you accomplish what you're looking for.🔼 🔽

[marble cape]

Nice

Thanks right now I’m learning about real time canvis quick

[hardy sentinel]

I was sleeping. Yes, it is free

Here @grizzled swift

[grizzled swift]

I'm trying to use this image as a reference to generate more pictures. What tool should I use if I want to import it then use prompts to change the look?

The image-to-image thing isnt really suiting my needs. Im looking for something where I can type "Change colors to bla bla bla" or "add a cowboy hat"

[hardy sentinel]

To change to what style?

[prisma plaza]

You have hit your fair use limits for this month. If you think this is an error, please contact our support using the chat function.
i have 60$ plan Unlimited credit

[hardy sentinel]

How about canvas editor?

[hardy sentinel]

Adding cowboy hat, you could use inpainting mode of canvas editor