👋│accademia

[shrewd adder]

dunno, I was just asking if someone already knew something about it ^^

[formal mountain]

In image gen or in 3d textures?

[shrewd adder]

ah okay, now I know what you mean. No, I'm meaning that projection type for large 3D images that get projected on a sphere to create 360° images

Like this:

[ripe wind]

@formal mountainGood morning, any idea how long it will be before we have a collaboration option available on one acc?

[digital walrus]

You mean like collab with your team?

[ripe wind]

@digital walrusyes

[north pike]

Does any body know how to create a basic cover template prompt

[digital walrus]

I'll send you a friend request and dm

[forest reef]

I have heckload of these

right from when control net was first introduced here

this is new one made in phoenix

[shrewd adder]

Nice! Are you into role-playing, too?

[forest reef]

used to be, it can be fun at times

[shrewd adder]

I came to Leonardo because of role-playing, to create assets for it, and all kinds of visualization ^^

[forest reef]

funny enough I came to generate NSFW with AI. That was original intent lol

that was too horrible at that time, so I just stuck with SFW

[ripe wind]

@digital walrusthx

[fickle dock]

Do you dm me that tournament whole story again?

[forest reef]

its hard to find for me now

[formal mountain]

Haha I was just looking back at my old images. All SFW, but man were they weird looking 😂 I completely understand not wanting to generate anything nsfw back then 🤣😂🤣

[forest reef]

when I tried hard to it was pretty bad, but when I didn't try I did get alot of it...

especially with RPG

[formal mountain]

omg yep.

[formal mountain]

I was just about to say!!!

[forest reef]

Dreamshaper was still very tame but RPG was giving out naked people left and right, to the point it annoyed me to put in negatives for it

[formal mountain]

At one point there was a single word that if you typed it in the prompt it created explicit images 75% of the time, with RPG 😂

I reported it to Ayu. Was very surprised 😂

[forest reef]

you type fair skin in deliberate, you get completely naked people. There were multiple issues with Adherance back then

also filter here was extremely strict

[formal mountain]

Yeah they do plan to work on the filter but they’ve had to kind of use an extra aggressive method

[forest reef]

even just relatively good looking people were blocked

[formal mountain]

Haha

[forest reef]

the filter ruined caused me to lose chance to enter a challenge 3 times

I had to try for 3 days to get in

[formal mountain]

Oh no!!

[forest reef]

someone suggest I should crop and post in nobles to check if it works first then put in

that worked

[formal mountain]

Ahh smart

[forest reef]

there wasn't anything NSFW in Image, it was just getting randomly blocked without reason

it was just Yogi meditating with woman on side, fully clothed.

perhaps it had issue with male nudity?

but that couldn't be because I had posted alot of shirtless martial artists before

cropping it got it past the filter

[formal mountain]

Omg yes!!! I got images blocked with yoga too

And it was just older women, well clothed

[forest reef]

oh no, holy church is at it again with propoganda

[formal mountain]

Hahahaha

[forest reef]

persecution against pagans begins anew

[formal mountain]

🤣😂

They’re just jealous to see images of anyone flexible lol

“That can’t be natural!”

[forest reef]

"Your image has been blocked for the following reasons...Its Satanic!"

[formal mountain]

“Can’t tell you why, but it must be!!” 😂

[forest reef]

Well funniest thing I got blocked was actually a landscape picture

[formal mountain]

“Dude that’s just tree pose”

“IT IS EVIL!” 😂

[forest reef]

it was the weirdest thing

[forest reef]

Must've been the Evil tree

[formal mountain]

Omg that’s priceless 🤣😂🥲😂🤣

[formal mountain]

Right! The secret asana 😂

[forest reef]

because Ravana also attained his boon from meditating in posture, must be evil

man went on to become a tyrant ruling over half of the world, his empire was like atlantis...way too ahead and rich for its time

[formal mountain]

That’s just science! 😂

[forest reef]

then he gets into beef with Prince Man in UP, who comes in with monkeys to beat him up and take back his wife he kidnapped

and this is also why we celeberate Dusshera

[formal mountain]

Yeah it’s bad to kidnap wives, especially when the husband is friends with the monkey king lol

This was one of my favourite stories to read when I was young

[forest reef]

well Hanuman isn't monkey god, he's just extremely advanced in terms of evolution despite being primitive. Man could've won the war alone and ended story on the spot

it was just...not directly his mess to solve

[formal mountain]

Yes I realized it after I typed it 😅

[forest reef]

people in Hindi chat often just want to create his pics

[formal mountain]

I saw that!!

[forest reef]

Hanuman, Rama, Krishna...very iconic figures

[formal mountain]

Yes

[forest reef]

all of them are like walking nukes

[formal mountain]

Radha too?

[forest reef]

no she was just ordinary woman but she became extra ordinary in terms of temperment after she understood Krishna. They didn't marry but she became immortalized because of her realization

women dont really wish for power, even if they get it they are very judicious in use generally but their curses can ruin entire civillization

[formal mountain]

Yes I notice she is often mentioned in different ways in movies

[formal mountain]

When your curses are dangerous, you must be wise or else no one is left lol

[forest reef]

Mahabharata happened because of a woman, the Cultural empire that stretched from Iraq all the way to Indonesia was now reduced what you know as current India

Afghanistan is also said to be under curse of Gandhari who lost her 100 sons in war

it never saw stablity again

[formal mountain]

😳

Lesson, don’t anger a mother

[forest reef]

this is why it is always said to never piss off women

[formal mountain]

Very wise 😂

[forest reef]

also Krishna's Kingdom got submerged into water due to her curse too

all Yadavas grew decadent and killed eachother

[formal mountain]

😨

[forest reef]

Bandits took away all the fleeing women, this was start of kaliyuga

dark ages

old powers stopped working due to lack of etheric content, world order changed

this is the time when west slowly started developing upwards

[formal mountain]

Ahhh

Never mess with women lol

[forest reef]

well not all of it was woman's fault, it was a change of yuga...it brings all sort of drastic changes....climate changes...wars...upheavels of all kinds

especially this advent of dark age ruined all morality

you look at the rules of war in Mahabharata you will be shocked...are they doing war or doing sports?

[formal mountain]

Yes I read it a very long time ago. It’s quite long, many stories and happenings

[forest reef]

there were rules such, "No fighting after sunset" "No killing of women or children" "No killing men of lower rank" "Fight must only happen on a dedicated location"

[formal mountain]

Unfortunately of course I only can read a translation.

[forest reef]

of course since the war happened in 13 days with more than million people and after coming near end everyone forgone the rules and went mad with hatred

it all started with night attack

[formal mountain]

Yeah the more people are hurt, the more reasons people have to hate

[forest reef]

war was actually not a place for hatred in tradition, it was to settle enemities and decide hegemony

without cause harm to general populace

[formal mountain]

Yeah

[forest reef]

you are not supposed to hate opponent even if he kills your father in war

he died because he was weak and other guy was stronger

as a warrior you accept this and move on

you are still allowed to take revenge if you think you have capability

this is still somewhat prevalent in China

Mahabharata also talks about China

[formal mountain]

I guess I don’t remember that part, it has been like two decades since I read it lol

[forest reef]

it was very short, it was about the king talking to his sons about how a Tyrant emperor was overthrown by his subjects

that tyrant was suspect to be Xin who ruled a dynasty with iron fist

[formal mountain]

Ahh I see

[forest reef]

eventually got rebelled against

[formal mountain]

As will happen lol

[forest reef]

he was the one probably who was ruler of united China, first emperor

[formal mountain]

Yes I know which one you mean

[forest reef]

so its estimated war was done after his demise

also another big secret, its sort of suspected that Ravana's original Lanka was Atlantis

which was submerged later on

current co-ordinates of Lanka provided by astronomy of Ramayana doesn't match up with today's Sri lanka but rather maldives

maldives is relatively submerged

its suspected that it was a huge landmass over there before being submerged

[formal mountain]

Fascinating!

[forest reef]

the golden city hanuman burned might've been atlantis

but it didnt submerged due to war...well it might've been triggered by it

since there was alot of use of Brahma-astra

[formal mountain]

It is interesting how many stories there are of a city that was submerged

[forest reef]

also Ravana has had connections with Mexico of that time

[formal mountain]

I am most fascinated by the stories that are shared by many cultures

[forest reef]

yea old world order was way more wilder than you think

at that time, even Mexico was bigger power

[formal mountain]

Yes that region has a very rich history

[forest reef]

during Mahabharata there were alot of places Sanjay who had divine sight to see anything he wants saw many more high quality civillizations other than Bharata to the extent that they'll be put to shame

[formal mountain]

Yes I think that portion of the story is the most famous to those outside it. With Sanjay and his wisdom

[forest reef]

its also suspected wife of Ravana...her father was of mexican descent

its not sure

[formal mountain]

he sees also Arjuna and Krishna, am I correct? I mean Sanjay with his far sight

I might be confused haha

[forest reef]

yes divine sight allows you to see anything you want

almost to the point of omniscience

back in the day, Races were divided as Aryans(Civillized) and Rakshasa...Civillized people were expect to follow law and keep order while Rakhasa did whatever they liked. So they were generally aggressors despite following Dharma

Ravana was born a mix between both, father being a sage and mother being from Rakshas kula

after learning way to obtain power, he went on to dethrone his brother Kubera and took over Golden Lanka

it is possible it was atlantis and much huger landscape than what you see today

to the point even europe knew well about it

perhaps also the origin of name of 'El Dorado'

if the connection to mexico is made more clear in future, there might be more light on this subject

[ripe wind]

Is it possible to get more than one consistent characters in an image, I need 3.......

[tough mist]

depends on level of specificity, definitely try on phoenix

[ripe wind]

@tough mist thx, i did but it seems only 1 character can be used in same prompt

[tough mist]

2 works for me

Ultra detailed, Anime screencap, konosuba. A dimly lit dungeon room with stone walls and floors. On the left, a large treasure chest with sharp teeth (a mimic) has its lid open, with an elf white hair twin tails wearing a white cloak with gold trim, green pants, and brown boots with upper body partially inside, kicking their legs as they are pulled in. On the right, a young woman with long purple hair, dressed in a white dress and a dark green cloak, sits on the floor, watching the scene with an indifferent or slightly amused expression. The background includes stone steps leading up to an unseen area.

Getting eaten by mimic chest is diffult part but beside that character descriptions are pretty much on points

[sacred crow]

Good afternoon, it is translated as mimic chest? is this a term specific to the anime series? which series?

[tribal hornet]

Chest with teeths is quite a complex idea, unless it's trained with trigger word, it may not work, nor describe it(It could give you the result, but very inconsistent at best) usually done by dedicated loRA model

[digital walrus]

tell me more

[forest reef]

well its not really fully confirmed yet but the father in law of his was supposed to have been king of a mexican kingdom, he was treated as Rakshasa kula(clan) "Mayan Race & Civilisation: As earlier said, the Swarna Lanka was built for Sumali by Mayasur. Maya ruled the land around Mexico in North America. That is the reason why we heard about the race and civilisation called “Maaya” in North America. Even this is also the reason for Mexicans even now celebrating RamSita festival.The question arises whether the ancient Hindus of Indonesia had contact with Mayan civilization across the pacific which is evident from the pyramid constructions in Indonesia very similar to that of the Mayans.The “White Island” is a symbolical name. Asuramâya is said to have lived (see the tradition of Jhána-bhaskara) in Romaka-pura in the West: because the name is an allusion to the land and cradle of the “Sweat-born” of the Third Race. That land or continent had disappeared ages before Asuramâya lived, since he was an Atlantean; but he was a direct descendant of the Wise Race, the Race that never dies. Many are the legends concerning this hero, the pupil of Surya (the Sun-God) himself, as the Indian accounts allege … Asuramâya, “as great a magician as he was an Astrologer and an Astronomer.”"

[formal mountain]

Our Accademia channel is full of surprises 😁

[formal mountain]

You’re finding Phoenix isn’t supporting multiple characters?

[near haven]

If you're using the latest Phoenix model, you can go in greater details for more characters.

[lyric zephyr]

I am here 😄

[near haven]

If you're using other models, you may need to consider post-editing for more reliable results.

[lyric zephyr]

if anyone wants a challenge, make a heterochroma Cyclops 😄

[tough mist]

The mimic term popularized by board game dungeon and dragon since its 1st edition in 1977. Since then there are numerous reference to this origin and supposing also captured in most LLM (though there is disambiguation issue to deal with)

[sacred crow]

I see, ok thank you

[lyric zephyr]

https://www.seikens.com/bestiary/mimic/

[formal mountain]

Mimic chests are evil!

[tough mist]

ikr a fictional monster is difficult subject

[formal mountain]

“Ooh look loot!” “rawr”

Yeah lots of games have taken on mimic chests as a thing

[lyric zephyr]

there is also this new Pokemon

[formal mountain]

Because who doesn’t like dashing the hopes of gamers by making them fight their way to the shiny box only for it to attack them

Oh and then there’s the chests in Sea of Thieves like the chest of sorrow 😂

[lyric zephyr]

we got this Pokemon actually

[formal mountain]

They don’t fight you, only sink your ship 😂

[formal mountain]

Okay that’s kind of cute

https://tenor.com/view/neverwinter-winter-festival-mimic-gif-19551491

https://tenor.com/view/sea-of-thieves-sot-chest-of-sorrow-gif-15642213

[sacred crow]

Greetings @formal mountain @lyric zephyr

[lyric zephyr]

Greetings Steven... how is the Weather

[digital walrus]

Now I’m deep into the 🐰 🕳️

[forest reef]

its just one of the theories, its also thought that Atlantis wasn't fully sunk, it shifted drastically into what today is known as Indonesia. Parts of it were broken forming Maldives and Sri-Lanka. This might've been much huger landscape than expected. However the only known fact is that according to Co-ordinates the actual swarna lanka(Atlantis) was not around Sri-Lanka but Maldives. Where Indonesia comes into play is not even discovered. Indian culture entered during Chola dynasty which is why you see alot of Hindu temples and symbols over there. No one is sure how powerful was ancient Mexico was at the time and if this Maaya was different from Mayans or was it a Indian name by which they called them. Another part of mystery is why Europeans knew about Atlantis so much if it wasn't near them.

[sacred crow]

it is cloudy here

[forest reef]

very lucky for you, its almost always 40c+ over here its almost painful

[sacred crow]

good evening, i wish you a cooler weather

[sacred crow]

the knowledge was told by Egyptian to the Greek, so it was pass down probably through the silk road

[forest reef]

ah good its gone now

[forest reef]

that would make sense

[sacred crow]

according to research, knowledge and goods travel 100-150 km through land route and I don't know how long via sea route.

the majority of the volume of the trade of course travel through sea

[forest reef]

it would also make sense that its known as different name in Europe...they call it Atlantis, Mexicans might've called it differently and Indians knew it as Swarna Lanka(Golden Lanka) which is why there was such dissconnect from east

[sacred crow]

yes

[formal mountain]

@dark sage

[swift glen]

@charred nacelle hello, are you able to speak here?

[charred nacelle]

Salutations! Indeed, I am capable of conversing here. How may I assist you on this fine day? 🎨
Our apologies if this message did not answer your question.

[lyric zephyr]

got my graphic tablet working so I can offer really bad sketches again xD

show me content reference 😄

not too bad on MID

seems too simple for MID

[lyric zephyr]

so if anyone needs "something" to use as content reference ... hit me up 😄

[lyric zephyr]

but basically you could use such a bad tracing 😄

or copy and create your own character 😄 LOW - MID - HI)GH

[lyric zephyr]

or you wanna bad detailed sketches 😄 ?

[uncut seal]

I'm running low on ideas what i can do with a selfie. I've done me as sorcerer, cowboy, knight, ninja.. anyone got more ideas? i dont just want myself in cooler clothes than I own. wanna be something.. 🤔

[nimble leaf]

👉 #💡│suggestions message 👀

Style reference idea

[uncut seal]

i think some people just dont want to reveal what their reference image is/was. want to keep their secret sauce to themselves. so unless it would come with a toggle on/off, like the public/private, idk..

[nimble leaf]

it's true! I hope that your good judgment is also shared by the developers.

the balance between privacy and new possibilities is ideal

[lyric zephyr]

Cyborgs ... Anthropomorphs .... Movie Heroes/Villains

this is new for me

but if you wanna do Manga-Stuff... its great to use a bad sketch on content reference 😄

@steep shoal

[formal mountain]

Hang glider

[formal mountain]

I’d really like for this to be visible on my own generations, to me. Kind of like how when you use a custom model you can see it but it’s not visible to the public 🤔

[nimble leaf]

it's true... but most likely there will be a price on privacy

[rose forge]

already want to try 😅

[lyric zephyr]

just make a bad sketch... make it mid or high and make Manga Preset with your normal Prompt 😄

[rose forge]

I loved it

I gave him a prompt that I saw on home

Image content

And then I put character image to this image

Super magical thank you

[lyric zephyr]

❤️

[lyric zephyr]

we had one time that "Centaur" Theme here 😄

here you go :KEKW:

[lyric zephyr]

simple sketch translation 😄

[rose forge]

😅

@digital walrus

It reminded me of the first time I met you when we made an elf from a sketch

Using the same technique but with advanced tools :3

[lyric zephyr]

@rose forge u want a specific sketch to try ? I can draw you something up 🙂

[steep shoal]

Very cool friend, these characters were really good e.e

[lyric zephyr]

bad sketch + content reference + manga preset ... hit me up if you need something you wanna turn into something great 😄

[digital walrus]

it was very fun that day

[rose forge]

I drew master that I transformed him into a fairy xD

[rose forge]

@lyric zephyr

Thanks for the sketch 💚

[lyric zephyr]

❤️ glad it worked out quite nice for you

[rose forge]

I add these elements to the sketch

[lyric zephyr]

🙂 nice

[rose forge]

"High-definition miniature of Tinkerbell, with intricate details,Blonde Hair, vibrant colors, and a magical glow. Captured in a close-up shot, highlighting her delicate wings, expressive eyes, and iconic green dress. Set against a softly blurred forest background with sparkling fairy dust surrounding her."

Prompt

When the first stage of the sketch is ready I upload it to content reference to the new image and in character reference I use my photo and that's how I got the last photo

[lyric zephyr]

❤️

[uncut seal]

any of you have experience using character reference to turn yourself, or any human I guess, into a robot/cyborg..?

[lyric zephyr]

not tried that so far

[lyric zephyr]

for some reason it made me black :KEKW:

[uncut seal]

thats a guy in armor. not a robo

[lyric zephyr]

Cyborg and Robo is not the same

[formal mountain]

Cyborg maybe

[lyric zephyr]

Technology-enhanced Human 🙂

Char on HIGH turned me into Willam Defoe LUL

and yes on high it seems to miss the cyborg features

????? Char Reference on MID .... became black again hahaha

[formal mountain]

what is your prompt?

[lyric zephyr]

Movie Poster, Cinematic Photography of a Cyborg-Soldier wearing a futuristic Body Suit with robotic Elements holding a sniper rifle, Robotic Face-Components, LED-Eyes, standing in a futuristic Glass-City, 8K

[steep shoal]

It was really good, I'll ask you for some tips there. e.e

[gentle light]

super⚡ and what model kino or... and preset... please like this type of image

[lyric zephyr]

Cinematic Kino yes

[gentle light]

and preset?

[lyric zephyr]

Cinematic

[gentle light]

thanks⚡ 🙏 ⚡

[fallow tree]

Cyborg + Movie probably makes the AI reference Cyborg from DC

and he's a black man

(that'd be my guess anyway :D)

[lyric zephyr]

or its the model 😄

Lightning doenst know what a cyborg is

[formal mountain]

Well technically that could be an entirely robotic chassis

[formal mountain]

Oh smart!

[lyric zephyr]

well I have Cyborg + Movie ... but Lightning makes this

depends entirely on the model I suppose

[formal mountain]

Yeah that looks like it could be a cyborg. Robot abs

You need to add more descriptors if you want a specific style of cyborg

[lyric zephyr]

I added a lot more

[formal mountain]

Are you using the Edit With AI feature to hone your prompt?

[lyric zephyr]

best shot so far 😄

[formal mountain]

Like, what is missing? What makes it not look like your idea of a cyborg?

[lyric zephyr]

there is no "Edit with AI" on my screen 😄

just "reuse prompt"

[formal mountain]

have you tried using Phoenix to see what it does?

[lyric zephyr]

ow I saw its only on Phoenix 😄

would be cool if you can Edit with less Tokens

Problem is if I wanna make Fantasy Figurines, Phoenix is not good with that 😄

same Prompt.... Game Concept 3D (Albedo) VS Phoenix

[near haven]

I suspect that Phoenix deliberately does that to avoid replicating any copyrighted materials.

[formal mountain]

the edit is just the cost of the next gen, it's not adding to your gen cost

[formal mountain]

yeah, they worked really hard to be selective when sourcing the dataset.

it's super powerful, and with the right prompts you can make almost anything. and it's only going to get better with future updates and finetunes etc. but for sure there aren't the same shortcuts to mimicking copyrighted styles. if we want something in particular, we have to work for it, and make it our own lol

[near haven]

That explains a lot.

So Leo is indeed taking a very different path from MJ.

[lyric zephyr]

I know that @formal mountain but "editing" is basically just an edited prompt 😄

[formal mountain]

no that was a reply to siloa lol

but in phoenix, edit with AI is really intelligent about doing what you ask.

[lyric zephyr]

just saying if you Edit your prompt ... it makes a new gen ... thats why it should be cheaper 😄

if its just minor changes, it still costs a lot

[formal mountain]

you'd like repeat gens to cost less?

the cost is based on how much is being asked of the servers

[lyric zephyr]

but like some minor stuff can also be done with canvas.... like changing a color

[formal mountain]

then do that instead of edit 😂

but if you're asking the AI to do a whole new image, it's going to cost tokens.

[lyric zephyr]

I dont want "a whole new image" ... i just want my current image to be changed slightly 😄

[formal mountain]

try turning off quality or reduce the number of images, if you want to save tokens

[formal mountain]

that's a whole new image, to the AI hahaha

it doesn't go in with a pen and edit

that's canvas

if you want to do a canvas edit, go for it!

but when you do edit with AI, it's modifying the prompt, using the same seed, and running a new gen

[lyric zephyr]

its just my opinion ... still curious whats better than Leo Veteran XL ... Rob says somethings big cookin

[formal mountain]

there's tons of big things cooking hahaha

[lyric zephyr]

I just wanted a Model based on Images of us Veterans 😦

[formal mountain]

yes, I know, you told me

[lyric zephyr]

sadly he told me somethings better comin 😦

[formal mountain]

sadly better?

haha

[lyric zephyr]

I was hoping for that Model 😄

[formal mountain]

if you pin your hopes on only one thing, there's going to be a high chance of disappointment 😂 meanwhile, there's things being worked on that we don't even know to anticipate. better to enjoy the ride 😂

[lyric zephyr]

the "ride" is like riding an escalator

[formal mountain]

https://tenor.com/view/deadpool-elevated-webspin-gif-12196624

[tough mist]

https://tenor.com/view/chores-cleaning-housework-tom-and-jerry-housewife-gif-20706096

[formal mountain]

it's these little pep talks that are the golden nuggets that I hope for from our mentori 🤣

[formal mountain]

exactly hahahaha

I'm pretty sure that's how the ml team looks at all times 😂

[tough mist]

usually each ml person own one feature. imagine a team

[lyric zephyr]

whats ml

[tough mist]

https://tenor.com/view/machine-learning-baby-crying-deep-learning-sad-math-gif-17208484

[lyric zephyr]

ah you mean Proxima

[formal mountain]

well actually it's a team 😂 if it were just proxima it would be especially unrealistic to expect a special model just for the veterans of the discord server 😂

[lyric zephyr]

then she is the Lead I suppose 😄

[formal mountain]

https://tenor.com/view/dog-cat-pat-고양이-강아지-gif-10280691

[formal mountain]

yes, I believe so. there is more curation involved in the building of the dataset for phoenix, which takes more work but results in a superior model, without the baggage of SD

[lyric zephyr]

https://tenor.com/view/help-wasabi-gif-12718090

[tough mist]

It is blessed and cursed at the same time. As it would be difficult to get new stuff online with backward compatibility, like img2img controlnet lora.. (Sorry @lyric zephyr my educated guesses that your control net will be at best next year in Phoenix 😂 )

[formal mountain]

I think something that is such high priority will arrive sooner, but to manage danny's expectations, lets go with that 😂

[lyric zephyr]

there is even HyperXL out 😄 ..wondering when that arrives here

[formal mountain]

I honestly don't know, but I think once the Leonardo team developed their own foundational model in Phoenix, it got a lot less likely that they would prioritize outside models

Phoenix is only in its first generation and it's already fantastic. That seems a better place to direct energy

[lyric zephyr]

Phoenix just dont know people and has too many ringlights xD

[tough mist]

My honest feedback, while Phoenix is a great milestone, this launch alone is a catch up to competitor since they shipped model with similar capability 9 months ago. I like leo a bit more. But if the tools I mentioned above are well integrated. I think that would be a step function for novelty in this space.

Or maybe I dont understand customer segment Leo want to target..

[lyric zephyr]

I am missing the Switch Ratio from Legacy

If I want 3:4 I always have to go to "More" and use the Slider

[tough mist]

I can write to a script if you know how to run the stuff

[rose forge]

Hola estás disponible

[rose forge]

It reminded me of my anti mufa letter

This card protects the selected comment from being a bad omen for speaking too soon.

The other day in Argentina's game you couldn't beat Chile xD and I said I think the mufe (mufar is to bring bad luck)

[rose forge]

I didn't know you could make letters xD I edited it with an external program

[lyric zephyr]

That's what u use Phoenix for

[rose forge]

I understand but I can't upload my image there

I sent you a pm

I want to know if I'm right

[formal mountain]

Yeah so the point is that Phoenix is a first generation of its own thing. Instead of spending more time and resources building stable diffusion fine tunes, they have spent a ton of time creating something that in v1 is on par with other tools’ v6. The foundation is the hardest part, and from there there’s a lot of cool new opportunities for features not limited by the stable diffusion ecosystem.

It’s not to say that there isn’t merit to other approaches, and again I do not know what’s planned. But it just seems unlikely to me that they’d spend months and months creating Phoenix, only to then stay shackled to stable diffusion

[lyric zephyr]

lets hope we get ICLight soon

[formal mountain]

No you’re only allowed to hope for one thing at a time 😂

[lyric zephyr]

no its my bday today... I have more wishes 😛

[formal mountain]

Whoa what, it’s my bday too!

By some fraction 😂

[lyric zephyr]

and its too hot here -.-

[formal mountain]

Ugh yeah here too

[tough mist]

They did right thing. Reading SD3 license agreement make me bruh

[formal mountain]

So cute!

[lyric zephyr]

@formal mountain when style ref stream 😄 ?

[formal mountain]

What are the specifics of the stream you’d like? 😁

[lyric zephyr]

make a Stream about the Texture Thread ... using it with the imagination of the chat 😄

[fallow tree]

Indeed! And now Stability is… not… doing great…

[near haven]

They just want to ruin their reputation once and for all with that license agreement lol.

[fallow tree]

gooooood morning Macadamias!

[formal mountain]

Good morning fellow nut! 😂

[formal mountain]

Omg beat me to it!

Aaaaand now I’m hungry 😂

[fallow tree]

🤣 can't come in hot with a pun like that and not follow up with an image

I'll get booted from the server

[formal mountain]

I’d be the one to fire the ban cannon! 😂😂😂

[fallow tree]

I believe it

[formal mountain]

Kind of want to change the channel emoji to 🌰 now 🤣😂

[fallow tree]

😈

myesssssss

[wheat ridge]

I love cookies

[lyric zephyr]

@graceful badger

[lyric zephyr]

I can make you custom figurine sketches for content ref if you want 😄

[rose forge]

I'm going to have my killer Tinkerbell

[lyric zephyr]

hehe okay 😄

sadly I am out of tokens ... but I make you something better 😉

@rose forge if you still have tokens

as for the funsies

shall I create a thread here for "abominations" XD ?

[rose forge]

Later it generated it :3

[lyric zephyr]

this I know 😄

I hope the "good" tinkerbell sketch makes you great figures 🙂

[lyric zephyr]

Misfigures (Discounter Figurines)

anyone wanna help out xD ?

[rose forge]

Hello

I had fallen asleep

Do you want me to put my sketches?

[lyric zephyr]

sure

[lyric zephyr]

hello... cant DM you for some reason 😦

[rose forge]

It's because you didn't accept my friendship

I was receiving a lot of messages...that had nothing to do with the AI

So I block all unknown people

[lyric zephyr]

❤️

[formal mountain]

Omg who would refuse your friendship?! 😱😭

🤣

[rose forge]

He likes to be begged, he still doesn't accept my friendship, I had to change my settings to talk to him again xD

[gentle light]

This is Great @lyric zephyr can I know prompt preset ...

[lyric zephyr]

its basic Phoenix

[gentle light]

Basic Phoenix? I tried to made the trunk with golden money but nothing( like pirates) and the prompt (is it possible?)

[lyric zephyr]

just use Leo Phoenix and be very descriptive 🙂

[gentle light]

very descriptive how?

[lyric zephyr]

just describe everything

[near haven]

He means you just have to describe everything naturally

Phoenix takes natural language quite well.

[lyric zephyr]

instead of "trunk with golden money" you go like "You can see a rare and old treasure chest filled with many golden coins"

[lyric zephyr]

the more you write into phoenix the better the result 🙂

[gentle light]

or can I use the pheonix prompt?

[lyric zephyr]

I can make you one if you want ... just tell me what u need

[gentle light]

I wish a trunk under water with golden coins. I cannot create coins...Sent my attempt

[lyric zephyr]

what model are you using ?

[gentle light]

This is an old attempt, now I finished token, here I used rg. However I used also kino with dark element I made several photos with pirates because I like them. But I cannot made a simple trunk under water with treasure

[formal mountain]

With phoenix, our newest model/preset, there are several options:

• use a short-ish prompt, and it can expand on it and add detail
• use a long prompt, clearly describing the content you want, and it does an amazing job of adhering to your prompt
• click on the pencil ✏️ icon to the top left of images generated with Phoenix to use Edit With AI to continue to refine your prompt

So I would try “A chest underwater, open and filled with golden coins”

Specifically with Phoenix

[gentle light]

But I prefer kino because phoenix makes images too standard I wish to make more personalizeted, "my" coins , "my" pirates

[near haven]

You can always use Phoenix to create a base image first. Then, use content reference/ img2img/ or other means to convert to other models.

[gentle light]

Yes This is true, but I want also learn to make an image one attempt too, is it too hard?

[near haven]

One thing you can consider is not to use the dark art element when generating your chest.

It contradicts your wish of creating a chest full of shiny treasure.

[gentle light]

why? I love the mood of dark art, the light - I use dark art almost with all , change weight but I like the mood

[formal mountain]

Also do try using edit with AI to refine your image and prompt. I’ve found when I do this I get more interesting variety and it is not too stylized

[gentle light]

Ah my god!!! edit with AI is my TROUBLE . I finished token and delete images

[near haven]

Is this something close to what you want?

[gentle light]

you and you all are masters, but what if I want learn because I want to be able to make this

[near haven]

It’s all trial and error. I also started as a newbie here just a bit more than a year.

[near haven]

I’ve learnt quite a lot from the others as the community here is quite willing to share and learn.

I have a typo there in the prompt, but that actually doesn’t matter much 🤣

[near haven]

The dark art element makes the image darker and more eerie.
If you still want your treasure to shine, it has to be set low, and you probably need some other elements to help counter that “dark” effect as well.

[gentle light]

thanks 🌟

[near haven]

Hmm… The prompt works way better than I thought. The treasure still shines regardless of the dark art strength 🤔

However, if you simply use dark art alone with Kino, it doesn’t look as realistic.

Dark art 0.1 alone

Dark art 0.3

Dark art 0.5

[gentle light]

like the light

[near haven]

At 0.8, the chest closes

“Aura” and “glow” help create that effect

[gentle light]

don't smile this is my pirate, very dark, so i need a dark but shinning treasure

[near haven]

You can combine the two images with Canvas Editor

And merge with inpaint to enhance consistency

If you ever want them both to appear in the same image, that is.

[formal mountain]

If you’d like, we can create a thread for you to continue to work with mentors on building your skills

[gentle light]

Canvas is my big problem. I used up 700 tokens on an image, and then it had to be deleted.-

[near haven]

It’s almost 1am here.
Do let pookah know if you need more help with your project.
I may demonstrate a bit of how it can be done next time.

[gentle light]

yes please first CANVAS EDITOR 🌟

[near haven]

Good night 🫡

[formal mountain]

Good night, rest well!

[outer plaza]

Hola, alguien podría enviarme el enlace de los términos para el concurso, por favor

[near haven]

https://discord.com/channels/989166677390426132/1108130218260041758

[near haven]

https://discord.com/channels/989166677390426132/1193949717952987228

[nimble leaf]

https://discord.com/channels/989166677390426132/1124543230306955436

[uncut seal]

hi @near haven would you mind sharing part of your prompt that gave you the 'First person shooter view' in the current daily? lookin for ideas..

[tough mist]

I think it is his sleep time now. Maybe try any fps references like doom, overwatch, fps ui

[uncut seal]

i dont mind waiting.
i dont wanna name games or artists and such if i dont absolutely need to. havent tried using the element yet, maybe its hella easy. idk 🤷‍♂️

[near haven]

Prompt: First person shooting game still, point of view low angle photo taken by a Gopro, aiming a gun at the terrorist in a war game, phi ratio, rule of thirds, joy, 35mm, kodak portra 400, photograph by hideaki hamada, backlit, light bloom, lens flare, exposure for the shadows, overblown, overexposure highlights, natural Light, portrait photography, iso400, film grain, vignette, depth of field, bokeh, fujifilm

[uncut seal]

Heyy thx. Was about what I expected then. Some of those words in there definitely wont work for other subjects. Guns, shooting, etc.. 😅

Do you think it would understand the abbreviation FPS view/perspective? Or is FPS more associated with 'frames per second'? 🤔🤔

I didnt test a lot yesterday. Just did some style and strength tests for something to put in the daily

[near haven]

I tend to avoid ambiguity in prompts because it can go either way.

By the way, I really like how your Pokemon Game UI entry turned out to be. 🫡

[uncut seal]

You mean my 'generic creature capsule capture' image 😉
Dont wanna get on a certain company's bad side 🤪😂

Thats also why i didnt post one of the better ones in the contest.

[tough mist]

that is the only pokeball, ai knows it is the same thing/concept no matter which you try to steer away 😆

[uncut seal]

Idk man, it gave me a premier ball by accident in the image i put in the daily. 🤷‍♂️
But yea, ik what you mean

[lyric zephyr]

Phoenix creates a perfect version of my GameUI Style ... but cant do 1st Person View

ok seems you have to describe YOUR Hand

ALMOST :KEKW:

well ALMOST perfect

[lyric zephyr]

seems you cant use motion on your own images.... still generating for 5 minutes xD

[lyric zephyr]

besides the Piggy Bank having weird feet.... Phoenix did a great job

[formal mountain]

Did you report the bug on the website?

[lyric zephyr]

well it took some time but it was later generated

[valid lark]

Hello masters

🫡

[lyric zephyr]

xD

hey Elpelon

[valid lark]

what's up dude

I'm new to generating

https://tenor.com/view/cat-milk-ear-twitch-twitch-cat-drink-gif-8647170092448092909

[formal mountain]

heya, welcome! 🤣

[valid lark]

oh damn friend he's the boss

[formal mountain]

hahahahaha

how are you doing today? 😁

[valid lark]

Well here quiet on the beach

[formal mountain]

ooooh, well that sounds nice lol!

[valid lark]

Yes, I am calm working in an accommodation that belongs to my family.

And you?

[formal mountain]

Sleepy today 😂

[valid lark]

That makes two of us

I'll stay a little longer and go to sleep

https://tenor.com/view/waking-up-cat-reverse-mimir-gif-14394846257419819571

[formal mountain]

https://tenor.com/view/bedtime-boss-baby-gif-89862483612162691

[valid lark]

Good morning friend

[torpid scroll]

WOW, First Person, amazing!!

[torpid scroll]

Sensational!

[uncut seal]

Thanks. It was surprisingly easy. Getting a pokemon in the encounter as well is a tad harder. 😂

[rose forge]

Hello good morning

I need help creating an image of this style

Yes it is possible with these features But I can't get it to be behind the fabric

The face is always ahead

I need it to be similar to this

[lyric zephyr]

thats what happens if you are under the blanket and your BF starts to fart hehe

[formal mountain]

and you have fangs.

do you have fangs, Danny?

[lyric zephyr]

yes

anyone tested "Describe with AI" ?

[formal mountain]

yep! how do you like it so far?

I've been combining it with style ref for images that I've found difficult to create, in the past

[lyric zephyr]

Might be hard that the AI grasps the concepts

[lyric zephyr]

yeah it KINDA gets it

[ruby wraith]

Got to get your pre-sets right I'm thinking 🤔

[lyric zephyr]

true

[rose forge]

My profile picture

[formal mountain]

Wow lovely!

[lyric zephyr]

some more textures for pookah ❤️

if anyone has time and wants to do another Style Ref Stream .... I will add a lot textures today 😄

[forest dragon]

@limpid forum

[lyric zephyr]

hehe

[limpid forum]

Truer words have never been spoken. 💯😆

[formal mountain]

There can be no doubt 😁

[formal mountain]

Omg I love it! 😁

[formal mountain]

Oo gorgeous

[valid lark]

Hello máster, I need help

Well, more of a recommendation

[formal mountain]

greetings! how can we help you?

[valid lark]

What model do you recommend to make this type of images, (keep in mind that the gif image is also from ia).

https://tenor.com/view/dark-fantasy-gif-5836843924882828129

This

@formal mountain

[rose forge]

Tipo realista a eso te referis

[valid lark]

Si y no, sino que me salga estilo a televisión vieja

[rose forge]

Kino-albedo

[formal mountain]

https://tenor.com/view/dark-crystal-skesis-nod-age-of-resistance-netflix-gif-15044254

[valid lark]

any recommendation from you

[valid lark]

La primera es kino?

[rose forge]

Si

Pero también use la cara del señor

Con character referencia

[valid lark]

Lo quiero hacer sin necesidad de usar image to image

[valid lark]

Pero también estoy probando el modelo

[rose forge]

Kino

[valid lark]

Hmm

[valid lark]

Bueno mejor intento cuando tenga mi pc en el celu se me es incómodo

[plush bison]

aqui se puede hablar en castellano?

hola, por cierto

tengo un problema con la app, y me han dicho qude pregunte aqui

I have to speak in english here right?

Hi...I need some help

[nimble leaf]

https://discord.com/channels/989166677390426132/989166678053113901

[plush bison]

guay. y se supone que es para resolver dudas?

[nimble leaf]

maybe

in the best of cases, better said

[rose forge]

Que duda tenías buen dia

Los que hablamos español dormimos hasta tarde :3 menos adrean

[plush bison]

😂

tranqui no pasa nada

creo que es problema de la pagina...no me deja subir de plan

[rose forge]

Bueno

Si querés podés mandarme scremshoot y reporto tu problema

Al privado

[rose forge]

https://discord.com/channels/989166677390426132/989166678053113901

Cualquier cosa que necesites este es nuestro canal en español

Te invito a que pases a saludarnos 👋

[plush bison]

gracias!!

[formal mountain]

sure, you can speak whatever language works for you. We have mentori who are able to help out in a variety of languages. All are welcome 😁

[rose forge]

Si podés hablar en tu idioma :3

[true parrot]

Hello everyone

[fallow tree]

Heya Winteroak! I have responded to your request for the Capybara flag! I think we can get it to work, no problem!

[formal mountain]

hey, is anyone having trouble with elements in legacy mode?

[true parrot]

I havent had any issues today even that downloading thing that seemed to affect others

[sacred crow]

Hello, that is awesome, sometimes you are lucky, i was not so lucky

[formal mountain]

Oh no! But better now yes?

[sacred crow]

now is fine

[loud mica]

Hola, soy nuevo y espero me gustaría empezar a trastear en este mundillo, espero aprender con vuestra ayuda.

[formal mountain]

¡Bienvenido! Estamos felices de responder cualquier pregunta que tengas. Si te gustaría trabajar uno a uno con un mentor, solo escribe /mentoring

[formal mountain]

oh good, glad to hear it!

[loud mica]

algún tutorial, recomendación, para saber por donde empezar? Tengo varias ideas en la cabeza... desde generar imagenes de una muñeco/persona haciendo algún tipo de ejercicio (posición), hasta pasar fotos de ánimales/personas a 2 o 4 colores en modo ilustración para poder imprimirlo en 3D como en estas imagenes

[charred nacelle]

Congratulations, @loud mica! You are now a 🍃 Novice! You can view your rank and level by typing !rank in #🤖│bot-commands.

[rose forge]

Hola mucho gusto mi nombre es camila

Vengo a darte una mano

Empecemos por el principio ya creastes una imagen ?

Me gustaría saber si dispones de herramientas premiums

Para poder ayudarte con las herramientas que dispongas

Muñeco persona

Se parece a mi :3

Quería ver cómo me quedaba el pelo rojo

[rose forge]

Te invito a pasar por el canal español también

https://discord.com/channels/989166677390426132/989166678053113901

[loud mica]

Muchas gracias. Ya he estado en el canal #español-es y por ahora no tengo herramientas premium.

Hace algún tiempo ya hice alguna imagen, pero fue al principio y ya estoy olvidado…

[rocky heart]

I wonder how to change the background in a image?

[true parrot]

Using mask in canvas maybr?

[cyan dust]

Hey there, welcome to the cozy corner of Accademia! 📚 Feel free to hang out and mingle. And hey, if you're aiming to level up your talents, dive into some challenging projects, or hit some learning targets, our mentors got your back. Need a bit of one-on-one support? Type /mentoring and we'll set you up!

[formal mountain]

can you tell us a bit more about the image? what settings were used to create it?

[keen copper]

So question, found this dataset traning page. Anyone use this and explain a bit on me how it works? I like to try training my own sets for a spcific sequence. I have a paid account and was told to share my info/questions here.

[rose forge]

Hi

[keen copper]

@rose forge Hi

You experinced in this dataset training page?

[weary isle]

Hello RChap, we have this wonderful Discord Community Stream recording about the subject of datasets and model training, hosted by our most experience team members.
It should hopefully answer a lot of questions for you 🙂
https://youtu.be/YvHn4Kw4beg

[rose forge]

Yes

In my experience I think that image character is a newer and better prepared tool.

Data set is missing an update

Here I am as a mini elf

It's a photo of me

Fits very well

[keen copper]

I want to try this to see if I can get the thing to understand mid-morphing sequence, anyway that you for the video. So like training it on a person face in the various stages between human and full wolf muzzel, so the AI does not do that weird wolf-man images

[rose forge]

I never tried anything like that

[rose forge]

I use these two tools

For example, in character reference I put my photo, which is going to be the face, and in reference content, the body.

I am going to use this image to transform you and use your body

In your case I would create the wolf transformations

And then I would show his face

[keen copper]

Hn, that's intresting-not heard that idea. May try. The idea is again to get the idea to, at best, do comic-like panels that show the TF in sequences or if you chopped up the American Werewolf In London TF into about 2 or three dozen frames but still showed progression. That being said also I want to use it on real people, but getting the first part is key.

[weary isle]

You can certainly create the images for the morph sequence with Leonardo,
If using an existing photo of a real person, then definitely give Character Reference a try as Cami is suggesting,
You may also want to consider using a fixed seed from the first image of the sequence, changing the prompt only for following generations with the same seed.

(fixed seed is under the advanced tab)

[rose forge]

It depends on the strength you give to the photo of the elf or your photo, it will be modified

The elf is shorter than me that's why I'm always going to be sitting :3

But in all of them they maintain my facial features very well.

In addition to the pose imitation control that you can use

[keen copper]

@weary isle Ya, my main goal is to stop AI form doing stuff like this where it less of a TF and more randamly fusing of stuff.

And this was the prompt "A young, handsome man undergoes a dramatic transformation into a gigantic oozaru with muscles bulging and his clothing torn to shreds. The scene is depicted in a photo-realistic style, capturing every detail with precision and depth. His expression reflects a mix of primal power and youthful innocence, as his features contort in the midst of his monstrous growth. The intense lighting and intricate rendering of fur and textures make this image a stunning showcase of talent and skill in visual storytelling."

So, hopping if I train my own set, then I can make it understand stuff like this

[weary isle]

Haha, the AI thrives on being random 😁
But seriously, the AI doesn't understand context well, so "transformation" context is ignored, and it tries to create a mix of everything else.
It would be better to describe visually each morph stage, like a story board, i.e-directing the AI.

[keen copper]

So then I would have basically discribe only the head, then the hand, etc, and not the rest of the body, or use character and seetting ref for the body?

Also seed, explain those little more.

[weary isle]

You can do all that in one prompt, just keep it simple and visual

[keen copper]

@rose forge Also when you say character pluse content, do you mean just a image of the face and then content with a full body image?

[weary isle]

A seed is bound to each generation run, e.g- 4 images.
If you like a particular generation you can re-run it using the same seed (found in generation info panel), making a slight change to the prompt, or even changing the model used with the seed.

[keen copper]

Ok, and you said do it in all on prompt, but part of the issue is even if I story board it, I've yet to find a AI that can do panelling well-or has Lenardio solved that, Keep in mind this be multiple pages if I got this to work.

[weary isle]

For paneling of a single image, yes Leo can do character sheets and comic book layouts well, There's a comic book Element (LoRa) available also.

[keen copper]

What the LOA comic element labled and is all the ref tools compatable

[weary isle]

If you mean layer the tools, yes you can, and adjust their strengths too, (with the exception of Phoenix XL model)

[keen copper]

Not seeing a layer tool

[weary isle]

As in you can use more than one ref or Element at once

You will see some variation in the available tools (in the image guidance tab) depending on the chosen model

[keen copper]

Really, only able to do one of each so far

[weary isle]

Are you on paid or free acc?

[keen copper]

Not renewing, but for now on paid

[weary isle]

Yes, only one image for character ref input

On paid you can use multiple images for the other reference types

[keen copper]

I'm on the apprantice for now, so should that limite not come in until my month is up?

[weary isle]

As far as I know, you're correct

It's best to use the purple intercom tab in the webapp for account questions

[keen copper]

Ok, I think there was a miss understanding. I know at the moment I can do up to four images, one image in each category. You were not saying I should be able to do more than that, correct?

[keen copper]

Ok, will play. One ome question, when it comes to a ref photos character expression, is there a way for the AI to take the general shap but not try to recreat the expression entirly?

Been runing into that

[weary isle]

On the non XL models we would use "edge to image" or sometimes "depth to image",
With XL models try just the content ref on low

[keen copper]

Those words to put in the prompt?

@weary isle Also jumping back, when you say story board, means something like "five panel page. Panel One: Man standing beside his apartment window with both hands placed on the bottom window frame. He is looking up upwards into the night sky, head-titled at the slight angle. He has a dazed expression on his face, eyes focusing on the full moon up in the sky [negative prompt: no visible full moon in panel one]. Second Panel shows close up of face with sweat trailing down his body."

[loud mica]

Muchas gracias. Ya he estado en el canal #español-es y por ahora no tengo herramientas premium.

[loud mica]

Hace algún tiempo ya hice alguna imagen, pero fue al principio y ya estoy olvidado…

[fallow tree]

hey @dark sage would you want to join a thread, got a user that wants to generate elementals - figured the creature aspectr is up your alley

[dark sage]

Sure, count me in

[fallow tree]

cool, come click the Winter0ak button in #1260333130724802691

[dark sage]

I joined the wrong one. 🤦‍♂️

[rose forge]

Que imagen querés hacer ?

[cyan dust]

Hey there, and a big welcome to Accademia! 🎓 Feel free to mingle and share in the conversations. Want to hone your skills, dive into big projects, or hit those learning targets? Our mentors are on standby to lend a hand. Whenever you need a bit of one-on-one guidance, just type /mentoring and we'll be right with you! 🌟

[loud mica]

Me gustaría utilizar una imagen como base y sacar esa imagen como una ilustración, pero con solo 4 colores. Por ejemplo, la imagen del perro sacarla como una ilustración de 4 colores...

Como se ve que la cara de la ilustración parece que solo tiene 2 colores morados (oscuro y claro), pues algo así con 3 marrones y el negro.

[rose forge]

Mira conla segunda imagen

Funcionando la a esta vas a lograr solo los colores de la s gunda imagen

Tendrías que conseguir la textura que vos queres

Ponerle a ña primera imagen

Por ejemplo aquí use la textura de la primera foto

Y se la puse a la segunda

[loud mica]

Y como puedo hacer eso?

Ya siento, pero es que estoy muy perdido…

[rose forge]

Tomo más que los colores las formas geometricas

Ahí que buscar texturas o algún cuadro que te guste

[loud mica]

Pero esta bastante bien, me puedes recomendar algún tutorial o curso de como usar Leonardo.ai?

creo que necesito empezar de nuevo por aprender la base de su uso

y luego ir profundizando

pero primero conocer las herramientas y como utilizarlas

[rose forge]

Perdón seguía intentando :3

[loud mica]

Tranquila...

[loud mica]

Por ejemplo, yo hice esta sin Leonardo.ai, pase una foto de mi perra a un dibujo a dos colores

Se ve que es la misma perrita, pero en este caso pasandolo a solo dos colores, blanco y negro...

[rose forge]

Claro si entiendo

[keen copper]

@weary isle So just learned that I guess when using character ref, the sensors are much stricture when trying to do any American Werewolf in London poses. 😦 Still, jumping back to last night on my question on storyboarding, am I on the right idea?

[tough mist]

yes, intentionally they are preventing deepfake scenario. Link to CEO’s answer https://www.reddit.com/r/leonardoai/s/UWVNHl47DX

[keen copper]

@tough mist I personally think this entire industry-assuming governments don't try whole put the genei back in the box idea, which good luck with an ever advancing world-is going to in the end come down to not what is create, but how you use it. Create fifty deepfakes of a famous person doing above pg-13? Does not matter. You start publishing it and claiming it as true to the wider world? You, and only you, are in shit. But that just my thoughts on the whole future of AI Art/Video.

Anyway lonesampler, I still like to hear your response to my storyboarding post.

[tough mist]

I dont have opinion about what should policy should be. You can talk to CEO via that post.

Note that if you have paid subscription, they relax that down and can request to get exception on specific prompts.

[keen copper]

Oh, I was not asking you nor am I arguing the site's policy-though feel what I would like be a no-was just my though overall on how I think this AI industry may go. Bit of a historian, so I think about to other game-changing tech and 'What do we do?" that came with it.

[weary isle]

Hi RChap, I'm back, I just clocked-on ⏰,
No, these are older image guidance options for non Alchemy/XL models, and are found when in the Legacy Mode, with Alchemy off.
(I just wanted to show that there are multiple options in Leo's guidance section, that show up depending on which Model you choose to generate with.)
Leonardo is like a swiss army knife, its wide range of tools allows users to find various workflows to suit their projects.

[weary isle]

Short answer: yes generate each panel (frame) individually, then bring them together how you want.
To create your werewolf transformation, my approach would be to do it the same way they did in the movie, with stop-motion.
Story board, like making a movie out of a story, I mean breaking the scene/story up into a series of images, then creating them individually,
generating each frame image, one at a time, making changes to each one with more werewolf appearance per frame. This is were one of the controlnet options would be needed to keep the images similar in composition.
After creating a set of images that progress well into each other (this may take quite a few generation runs to gather a good set, e.g 12 out of 40 images), you could then build/join the morph scene using a video editor software.
This is just one method, Leonardo has many options that could be implemented here: Realtime Gen and Realtime Canvas, adding Elements for styling, etc. 🙂

[keen copper]

@weary isle ok. Also learning using the AI inhance is next to useless as this is the result of me asking it to do a prompt with a character ref image and then the image of David holding his hand as it changes as the setting. As you can see, AI liked to be random.

this was the AI 'enhacment' "A man's hand, frozen in astonishment, morphs into a wolf paw before his eyes. Captured in a hyper-realistic painting, the transformation is depicted with intricate detail: coarse fur sprouts from his skin, sharp claws emerge from his fingertips, and his knuckles contort into a lupine shape. The artist's skill is evident in the lifelike texture of the paw, the vivid colors that seem to leap off the canvas, and the way the man's expression conveys a mix of fear and wonder. This stunning image seamlessly blends the boundary between man and beast, inviting viewers to witness a supernatural metamorphosis in exquisite detail."

[weary isle]

A relatively new feature, I don't use it much, and have had good and bad results from it, yes its more random.
A fun feature if you're stuck for a prompt to play with, though.

Try a different Model too

btw, AI has always struggled with hands 😩

[weary isle]

This enhanced prompt would need further editing to improve the resulting gens,
Here I put it through AlbedoBase XL, not great but it shows how a change of Model interprets it differently.

[tough mist]

My hunch is that character reference has tendency to produce human character (base on its limitation), and it is likely to impact on a hybrid like a werewolf.

[weary isle]

Yes, a good hunch, and then at that point the AI starts working against you, which almost always leads to frustrating results.

[cyan dust]

Hey there, welcome to the Accademia! 😊 Stick around and mingle with us! Should you be keen on levelling up your talents, diving into challenging projects, or hitting those learning objectives, our mentors have got your back. Need a one-on-one with a mentor? Just hit up /mentoring anytime!

[weary isle]

@keen copper Another note: words closest to the beginning of a prompt carry more weight,
In this one (also Albedo) the AI picked up on the words "frozen" and "wolf" but ignored the hand.

And with Lightning XL it focused on the "mans Hand" (missing a finger, haha) 😁

[weary isle]

And the always beautifully creative DreamShaper7:
(I'm just sharing this for fun) 🙂

[lyric zephyr]

I made this ... so you can make a ballpoint sketch thingie with style reference 😛

[lyric zephyr]

sketch style 🙂

do you like it Cami ?

[rose forge]

Yes I like it

[lyric zephyr]

@rose forge tried something differnet for you 😄

a bit more colors 😄

maybe works better for you with character reference 🙂

[lyric zephyr]

Better ?

[rose forge]

Yes much better

[lyric zephyr]

I make more for you ❤️

do you wish a certain color palette ?

[lyric zephyr]

@rose forge 😄

[lyric zephyr]

@rose forge or try to combine multiple Styles in Legacy 😄

[lyric zephyr]

really nice 🙂 .... Avísame si necesitas colores específicos como plantilla

[rose forge]

I like to improvise

I prefer not to choose

[lyric zephyr]

whatever u can do with this 😄

I like dis 😄

[lyric zephyr]

doesnt look like you haha

[lyric zephyr]

@rose forge here you go 😉

[rose forge]

What is that xD

[lyric zephyr]

use it as Style Reference xD

[rose forge]

XD

[cyan dust]

Hey there, welcome to the vibrant Accademia hub! 📚 Feel free to mingle and exchange ideas here. Whatever your aspiration – mastering a new talent, taking on challenging ventures, or smashing those learning objectives – our dedicated mentori have got your back. 🌟 Need one-on-one support from them? Just type /mentoring and you're all set!

[formal mountain]

Danny, is there a chance one of my Mentori is posting non-Leo images here? 😂

[lyric zephyr]

is there a chance its a Realtime Canvas Screenshot 😉

[formal mountain]

Pretty low chance 😂

[lyric zephyr]

well my sketches for Style Reference are also Non-Leo .... so ....

[formal mountain]

Soooooo

[lyric zephyr]

then make Leo have better tools ....

[formal mountain]

Recommends improving your Leo sketch creation, my mentor friend

[lyric zephyr]

add Layers

[formal mountain]

I think you need to request mentoring 😂

[formal mountain]

We have layers!

[lyric zephyr]

so how can I paint BENEATH 😉

[formal mountain]

Paint, select, type [

[lyric zephyr]

Leo has no pen pressure 😉

so I can paint my Sketch with Leo

but not sketch 🙂

[formal mountain]

And yet it has so many other things

I’m just saying, this is the Leonardo server, where we share Leonardo images, as has always been the rule 😂

I feel silly telling some poor Novice that they can’t share non-Leo images when you’re here being a troublemaker 😂

[lyric zephyr]

on that matter, Leo should delete "Your Upload" and only keep "Your Generations" for Image Reference

since Images would be created by using external sources 😉

[formal mountain]

I think you mean to say that here: #🫂│amici-di-leonardo

[lyric zephyr]

ok let me take another shot ....

[formal mountain]

Soooo are you going to delete the images above or am I?

[lyric zephyr]

which :Kappa:

[formal mountain]

The ones that aren’t made with Leo, because you are not at all a troublemaker 😂

[lyric zephyr]

ok let me take another shot ... if people use their own uploads for images ... how should we discuss this as "Mentore" if asked if people are not allowed to share their source image

[formal mountain]

Something to discuss in the mentore forum 😂

[lyric zephyr]

If someone asks you "I uploaded a photo of my garden" and the user asks for help because some details should be in there.... we cant do anything because we dont know what the original Reference is

since not allowed to share

[formal mountain]

If someone is requesting mentoring from you in a thread here, they can upload an image if they’d like you to see it

[lyric zephyr]

but they are not allowed to share non-leo images

[formal mountain]

I’m saying that I make a distinction when they’re in a private thread for mentoring

But you are not

And you’re not seeking advice, you’re just being trouble lol

[lyric zephyr]

I deleted them .....

wont share anything anymore

btw. there is nothing in the rules that states users cant share Non-Leo Images

so if users accept the rules, they cant be charged for breaking a rule thats not written on the discord

[formal mountain]

Hahahahaha oh my word 😂

Re-read the rules

No one is charging you, I just said they need to be removed

[lyric zephyr]

where in the discord rules is the part about sharing non-Leo Images ?

[green spindle]

HI. I need some help to use Leonardo.AI. Are there some tutorial or information about it? Thanks

[misty cosmos]

Help about what specifically?

[lyric zephyr]

there are tutorials on the Leonardo YouTube Channel

[hexed mantle]

Hi all, I just read about the api possibilies, can I use the api within a ComfyUI workflow?
Is there a custom node for that?

[cyan dust]

Hey there, welcome to the bustling Accademia! 🎓 Stick around to mingle and exchange ideas. Aiming to up your skillset, dive into hefty projects, or hit those learning targets? Our awesome mentori have got your back. Need a personal guide? Just type /mentoring to connect with one. 👥✨

[formal mountain]

We have not created a custom node for this. I believe it’s possible to create your own custom node, using python, and you can connect to Leonardo’s api 🤔

[hexed mantle]

If I have an api key, what is the url?

I think I know how to do this

I want to try this

If you can provide me with this information I would be very thankful

[tough mist]

I’m afraid you have to have subscription to use API. There is no try.

[tough mist]

You can check what are features here https://docs.leonardo.ai/reference/getuserself

[hexed mantle]

Ofc, I try this after i got a subscription, that is not what I ment, hahaha I want to give it a try that is what I ment

I asked chatgpt and this is the answer: https://cloud.leonardo.ai/api/rest/v1/generations

Is that correct?

[tough mist]

Great! Some users are insisting to try before buy, so I just want to made it clear

[hexed mantle]

I can not paste any links here

[tough mist]

Sound right

[hexed mantle]

I mean

The warnings are flying around my ears here

[tough mist]

But I send full ApI doc above did you check?

It listed all end point, server name, parameter, etc

[hexed mantle]

I did check it, but that only give me a general search page

Oh there it is

its working now

I see the user info about the api

[tough mist]

right, fiddle around there should be generation api

[hexed mantle]

chatgpt was almost correct

[tough mist]

yes

[hexed mantle]

I will just try, and if I am having some troubles I can come back yes?

[tough mist]

I’m surprised it is not hallucinated

[hexed mantle]

Do you know the custom_node Plush?

For ComfyUI

[tough mist]

I have not played with comfy much myself

[hexed mantle]

A okay

[tough mist]

But maybe you can fork Dall-e node, perhaps it is famous enough to have one

[hexed mantle]

DO you think this will work?

[tough mist]

Did you added custom code?

[hexed mantle]

I will try later, I am off to bed, It's 4 am now lol

Thanks for your help so far!

[cyan dust]

Hello and a warm welcome to Accademia! 🎓 Settle in and join the conversation. Looking to level up your skills, dive into big projects, or hit certain learning milestones? Our dedicated mentors are on hand to help! For personalized support from our mentors, just type /mentoring when you're ready.

[formal mountain]

I use the Leonardo API with python, and I do have experience with comfy, although I prefer to use Leo. But I am happy to look into what would be required to create a node like this, to use the Leonardo API for comfy.

[tough mist]

@zenith comet Here 👋

[zenith comet]

Hi!

[tough mist]

I'm leaving the link to start of our discussion in case other Mentore want to join in #❓│platform-questions message

Let's continue, since we cannot combine your logo and tile together
so I think that we need to tile it ourself instead.

For now pick (1) of the image that you like to create a tile from your previous generation

[zenith comet]

Ok got it

[tough mist]

Open it in your Feed, click on image it will have "edit in canvas"

[zenith comet]

got it

[tough mist]

did you manage to get that image to canvas?

[zenith comet]

yep!

[tough mist]

Check on top, if you have focus mode click to change to infinite mode

In infinite mode, you should have select tool

Use it to select the image you just import in, use copy short cut to make duplicate of it

Our goal in this step is we will use design from initial image but will have AI fill in seam between them

By place them like a tile but with some space apart

Like so

[zenith comet]

It doesn't let me change the selection size. and Copy isn't working either!

[tough mist]

THis could work too

worse than copy but should work

[tough mist]

I know canvas has some quirk control when you're using the first time. Let me know which step you stuck.

[zenith comet]

Ok i did your second option there

seems like this should be easier to create a pattern though...how is it that i can't just use a prompt to do this?

[tough mist]

it is easy to create pattern, but it will not have the shape you like, like your logo

or option B, you create some background that is tilable, the merge logo into that titable pattern some how

[tough mist]

Now you do know how to use canvas?

[zenith comet]

hmm ok I just need to play around with it more. i have to go now, but thank you for your help!

[tough mist]

ok, I'll leave you the link to some canvas tutorial then just come here again if you need some help

[zenith comet]

thank you!!

[tough mist]

(1) https://www.youtube.com/watch?v=aw10ykp5n5A
(2) https://www.youtube.com/watch?v=wLk5Pkxnu-g

[hexed mantle]

Hi, wow, that would be awesome! Thanks in advance!

[misty cosmos]

unbelievable..nutz!

[formal mountain]

haha is this something you've worked on, yourself?

[lyric zephyr]

@formal mountain so how would we go on about Mentore Request ? Like me need help with one of my best images 😄

[pallid gorge]

maybe good to put these in community guides ?

[tough mist]

Think so, wdty @flat ruin

[formal mountain]

Do you mean how to request mentoring?

[lyric zephyr]

No I mean between Mentore

[weary mesa]

I'll help 😆, with time and experience you'll figure this out and we will guide you on your art journey🙃😂

[lyric zephyr]

The Question is where 😄 and how to organize

[misty cosmos]

:--D not yet

[weary mesa]

Wel our almighty overlord chatgpt may help us 😂

[swift glen]

Verrocchio doesn't even have access to this channel lol @formal mountain

or specifically, not to send embeds

!setautomodchannel

[swift glen]

there we go

[delicate cairn]

https://tenor.com/view/helldivers-automaton-invasion-bots-dropship-gif-17359944560955706193

[uncut seal]

what 🤨
😂

[verbal haven]

Can anyone suggest me Leonardo preset and preset-style to achieve this - a furry art piece featuring two characters:

1. Zebra:
   • Name: Zoe
   • Gender: Female
   • Personality: Playful, curious, and adventurous
   • Appearance: Bright blue eyes, flowing black mane, and a sparkly horn on her forehead
2. G Cat:
   • Name: Whiskers
   • Gender: Male
   • Personality: Sassy, confident, and loyal
   • Appearance: Vibrant green eyes, sleek black fur with white paws and chest, and a mischievous grin

[fallow tree]

You're going to have difficulty achieving both in one generation with so much specificity

your best bet is going to be Phoenix most likely, it has better aptitude for keeping characters separate

@verbal haven I've got you here bud

how familiar are you with prompting?

[verbal haven]

Want both characters in a furry art piece- this is the client requirement, I am struggling with this.

[verbal haven]

Not much now using chatgpt for suggestions

But results not coming

[fallow tree]

ok, well what style of art should it be?

like a realistic photo of the furry characters?

or a furry cartoon?

[verbal haven]

General

[fallow tree]

alright, and what are these two characters doing

just standing side by side, having a picnic, or what?

[verbal haven]

Aah, this part is creative what looks good

[fallow tree]

ok let me see if I can create a prompt for you to start from

[verbal haven]

Ok

[fallow tree]

I think my current prompt went too literal

[cyan dust]

Hello and a warm welcome to Accademia, where great minds meet! 🌟 Feel free to mingle and exchange ideas. If you're on a quest for knowledge or keen on mastering new skills, our mentori are at your service. Need a focused helping hand? Just type /mentoring and support will be on its way!

[fallow tree]

looks like kids toys, but at least Phoenix got the right details attached to the right characters

[verbal haven]

I also messed up

[fallow tree]

you want Zoe on hind legs?

this is the current prompt I'm working with "cel-shaded classic cartoon, Zoe (an anthropomorphic female zebra with blue eyes and a sparkly horn) meets G-cat (a confident male anthropomorphic cat with black fur and green eyes), forest setting, motion blur, octane rendering, dynamic range, subsurface scattering, ambient occlusion, reflective, absorptive, joyful and playful, smiling"

[verbal haven]

This is far better 💖

[fallow tree]

please take the prompt as a starting point. I'm using 16x9 aspect ratio to try and avoid the characters merging

here is where I took it way more literally

[verbal haven]

Share prompt and r u using pheonix

[fallow tree]

say please

yes this is in phoenix still... it handles multiple characters the best out of current models

"cosplay convention photography, Zoe (an anthropomorphic furry female zebra with blue eyes and a sparkly horn stand bipedally) meets G-cat (a confident male furry anthropomorphic cat with black fur and green eyes), candid photography, captured in professional DSLR, cinematic lighting, diffuse spotlighting, HDR, joyful and playful, smiling"

[verbal haven]

Thank for your time and helping me

[fallow tree]

no problem

[verbal haven]

@fallow tree forget to ask
Pheonix
Prompt enhance -?
Preset style -?
Contrast -?

[fallow tree]

sure, since you didn't have a focused look to go for I left that all default

Auto prompt enhance (although it's long enough that it didn't enhance much)

cinematic preset

and low or medium contrast - I forget, but it was default

[fallow tree]

slap "quality" on there if you have access to Alchemy, but Fast works as well

[chilly jackal]

Hi here, can someone pls help I don’t know what to do

[nimble leaf]

it sounds a little aggressive... let's see that prompt

[fallow tree]

try another adjective other than "blood-red" sky

crimson sky