stable-diffusion / AI - PART TWO

[Johnny Cage]

Been experimenting a bit with in-painting to replace certain elements in the background.

I started out with a prompt asking for a topless girl posing in her dorm room. I liked how the girl and the pose came out looking, but I thought some of the background elements looked a bit boring:



So I decided to first replace the headboard with a nicer looking upholstered one:



I think the headboard itself looks fine but there still seems to be something off about how it merges with the rest of the image.

[Johnny Cage]

Also trying out some img2img using non-SD sources.

I used this still from a 2015 Hannah Ferguson Love Advent video as the basis:


To create this:

[pelham456]

Originally Posted by maxhitman

For using (parenthesis) in your prompts, and other useful information,
here is a guide...

Stable Diffusion prompt: a definitive guide

Updated March 3, 2023 By Andrew
Categorized as Tutorial - beginner, text-to-image

http://***************************/prompt-guide/


......................................................

How to Write an Awesome Stable Diffusion Prompt
by NICK LEWIS

http://howtogeek.com/833169/how-to-write-an-awesome-stable-diffusion-prompt/

......................................................

Stable Diffusion BEST Tutorial for Prompts, Beautiful Results 
- Master Prompts for Stylized Art - by AI ART ALCHEMY
( 1 hour in--depth tutorial video )

http://youtube.com/watch?v=9ErPUGlifQQ

......................................................

ON-LINE BOOK

http://openart.ai/promptbook

no, those were not my questions.

nvm. apparently i am banned in this thread.

even when i ask specific ON-TOPIC questions, the posts just get deleted!

i will ask on some other forum instead. thanks anyways.

[maxhitman]

Originally Posted by pelham456

no, those were not my questions.

nvm. apparently i am banned in this thread.

even when i ask specific ON-TOPIC questions, the posts just get deleted!

i will ask on some other forum instead. thanks anyways.

YOU Sir,
are merely here to de-rail this thread, to get attention,
Because IF you were REALLY interested in using Stable-Diffusion ,
you would have paid close attention to the links I have posted.

LOOK AND READ the posted links I have given you
to research for the answer you have asked, otherwise you will
never learn anything

Stable Diffusion prompt: a definitive guide

Updated March 3, 2023 By Andrew
Categorized as Tutorial - beginner, text-to-image

http://***************************/prompt-guide/

......................................................

SCROOL DOWN to the middle of the webpage... where it says

( ) and [ ] syntax

An equivalent way to adjust keyword strength is to use () and [].
(keyword) increases the strength of the keyword by a factor of 1.1 and is the same as (keyword:1.1).
[keyword] decrease the strength by a factor of 0.9 and is the same as (keyword:0.9).

You can use multiple of them, just like in Algebra… The effect is multiplicative.

(keyword): 1.1
((keyword)): 1.21
(((keyword))): 1.33


Similarly, the effects of using multiple [] are

[keyword]: 0.9
[[keyword]]: 0.81
[[[keyword]]]: 0.73


Keyword blending
(This syntax applies to AUTOMATIC1111 GUI.)

You can mix two keywords. The proper term is prompt scheduling. The syntax is

[keyword1 : keyword2: factor]


factor controls at which step keyword1 is switched to keyword2. It is a number between 0 and 1.

For example, if I use the prompt

Oil painting portrait of [Joe Biden: Donald Trump: 0.5]


for 30 sampling steps.

That means the prompt in steps 1 to 15 is

Oil painting portrait of Joe Biden


And the prompt in steps 16 to 30 becomes

Oil painting portrait of Donald Trump


The factor determines when the keyword is changed. it is after 30 steps x 0.5 = 15 steps.

The effect of changing the factor is blending the two presidents to different degrees.

-- see image example

You may have noticed Trump is in a white suit which is more of a Joe outfit. This is a perfect example of a very important rule for keyword blending: The first keyword dictates the global composition. The early diffusion steps set the overall composition. The later steps refine details.

[Johnny Cage]

We lost around 20+ hours of posts due to forum issues, so I'm not sure if your latest posts were deleted by a member of staff.

[maxhitman]

Here are some recent images created by my A.I. for testing.
I am also doing Famous Male faces and their exact body types.
NO... I am NOT making men naked - I love women too much!

I´m using Realistic-Vision created by SG_161222 as a Main Checkpoint
and doing face+body tests on only 2 batches of 6 pictures for each Actor.
Right from the start, the results are impressive.

I am sure these faces will become more detailed and perfect with a higher number
of repeated batches of pictures.
The more pictures you create of the same person, the better the A.I. will become
"trained" and learn how to make the faces better.
This can also be said for famous female faces too, which I have been doing
many of them - both new celebrities as well as the classic Hollywood Stars.

I am posting below my first selection of Prompts which I have used to create
these images, BUT keep in mind that for better quality AI-photos, MORE words
and descriptions will have to be written into the prompts
(both positive and negative prompts), which will order the A.I. to create more
diferent hair styles, hair colors, beards, clothing styles and for
background scenes - both indoors and outdoors.
Notice also I did not include in the Positive Prompts a camera Lens type (for
better distance and close-up details ) and no face expressions on the models.
Some face expressions can make any portrait photo look much better.

Guess who they are


-

-

-


images have been reduced in size

---------------------------------------------------------------------

Copy these prompts and give your A.I. a TEST to see if it can create these faces , so
later you can also use them in your Artworks or realistic-looking action scenes
with A.I. create photos.

Play around with the words, adding and removing some words to create a diferent
selection of images for every 2 Batches of 6-pictures. I have discovered
that this is the best way to "teach and train" your A.I. to create better looking
and realistic faces.

---------------------------------------------------------------------

MALE Faces ----------- BEST MALE PROMPT used so far... for Testing

Full body portrait of 25 year old Actor BRAD PITT man with short hair style, Detailed face, (perfect eyes), (highly detailed skin:1.1), (Smile), perfect body, wearing a shirt and Jeans pants , Modelshoot style, Professional Photography, soft lighting, PHOTOREALISTIC, Realistic, standing in a dark studio background, blurred background, RAW, analog, sharp focus, 8k, high resolution, DSLR, high quality, Fujifilm XT3, film grain, award winning

Full body portrait of martial artist Actor BRUCE LEE man with dark hair style, Detailed face, (perfect eyes), (highly detailed skin:1.1), perfect body, wearing a shirt and pants , Modelshoot style, Professional Photography, soft lighting, PHOTOREALISTIC, Realistic, standing in a dark studio background, blurred background, RAW, analog, sharp focus, 8k, high resolution, DSLR, high quality, Fujifilm XT3, film grain, award winning

Full body portrait of martial artist Actor CHUCK NORRIS man with blond hair style, blond beard style, Detailed face, (perfect eyes), (highly detailed skin:1.1), perfect body, wearing a shirt and pants , Modelshoot style, Professional Photography, soft lighting, PHOTOREALISTIC, Realistic, standing in a dark studio background, blurred background, RAW, analog, sharp focus, 8k, high resolution, DSLR, high quality, Fujifilm XT3, film grain, award winning

Full body portrait of 1985 Film Actor CLINT EASTWOOD man with short Dark hair style, Detailed face, (perfect eyes), (highly detailed skin:1.1), perfect body, wearing a shirt and pants , Modelshoot style, Professional Photography, soft lighting, PHOTOREALISTIC, Realistic, standing in a dark studio background, blurred background, RAW, analog, sharp focus, 8k, high resolution, DSLR, high quality, Fujifilm XT3, film grain, award winning

Full body portrait of sportsman CRISTIANO RONALDO man with short Dark hair style, Detailed face, (perfect eyes), (highly detailed skin:1.1), perfect body, wearing a shirt and pants , Modelshoot style, Professional Photography, soft lighting, PHOTOREALISTIC, Realistic, standing in a dark studio background, blurred background, RAW, analog, sharp focus, 8k, high resolution, DSLR, high quality, Fujifilm XT3, film grain, award winning

Full body portrait of Singer JIM MORRISON man with Long dark curly Hair Style, Detailed face, (perfect eyes), (highly detailed skin:1.1), perfect body, wearing a white shirt and leather pants , Modelshoot style, Professional Photography, soft lighting, PHOTOREALISTIC, Realistic, standing in a dark studio background, blurred background, RAW, analog, sharp focus, 8k, high resolution, DSLR, high quality, Fujifilm XT3, film grain, award winning

Full body portrait of Singer JIM MORRISON man with long messy dark Hair Style, (Long dark Beard), Detailed face, (perfect eyes), (highly detailed skin:1.1), (Fat overweight body), ((wearing shirt and pants)) , Modelshoot style, Professional Photography, soft lighting, PHOTOREALISTIC, Realistic, standing in a dark studio background, blurred background, RAW, analog, sharp focus, 8k, high resolution, DSLR, high quality, Fujifilm XT3, film grain, award winning

Full body portrait of Singer JIMI HENDRIX with dark skin and afro curly Hair Style, Detailed face, (perfect eyes), (highly detailed skin:1.1), perfect body, wearing hippie shirt and jeans pants , holding a Fender guitar, Modelshoot style, Professional Photography, soft lighting, PHOTOREALISTIC, Realistic, standing in a dark studio background, blurred background, RAW, analog, sharp focus, 8k, high resolution, DSLR, high quality, Fujifilm XT3, film grain, award winning

Full body portrait of Singer Musician JIMI HENDRIX with dark skin and messy Frizzy Hair Style, Detailed face, (perfect eyes), (highly detailed skin:1.1), perfect body, using head bandana and wearing shirt and pants , Modelshoot style, Fashion Photography, soft lighting, PHOTOREALISTIC, Realistic, standing in a dark studio background, blurred background, RAW, analog, sharp focus, 8k, high resolution, DSLR, high quality, Fujifilm XT3, film grain, award winning

Full body portrait of younger Singer DAVID BOWIE man with short blond Hair Style, Detailed face, (perfect eyes), (highly detailed skin:1.1), perfect slim body, (wearing a chrome suit) , Modelshoot style, Professional Photography, soft lighting, PHOTOREALISTIC, Realistic, standing in a dark studio background, blurred background, RAW, analog, sharp focus, 8k, high resolution, DSLR, high quality, Fujifilm XT3, film grain, award winning

Full body portrait of younger Singer DAVID BOWIE man with Red Ziggy Stardust Mullet Hair Style, Detailed face, (perfect eyes), (Eye makeup:1.1), (highly detailed skin:1.1), perfect slim body, ((wearing a light blue suit)) , Modelshoot style, Fashion Photography, soft lighting, PHOTOREALISTIC, Realistic, standing in a dark studio background, blurred background, RAW, analog, sharp focus, 8k, high resolution, DSLR, high quality, Fujifilm XT3, film grain, award winning

Full body portrait of Singer KURT COBAIN man with blue eyes and messy blond Hair Style, Detailed face, (perfect eyes), (highly detailed skin:1.1), perfect slim body, wearing old torn sweater and faded old jeans pants , Modelshoot style, Professional Photography, soft lighting, PHOTOREALISTIC, Realistic, standing in a dark studio background, blurred background, RAW, analog, sharp focus, 8k, high resolution, DSLR, high quality, Fujifilm XT3, film grain, award winning

Full body portrait of younger Singer ELVIS PRESLEY man with blue eyes and black Rockabilly Hair Style, Detailed face, (perfect eyes), (highly detailed skin:1.1), perfect body, wearing gold suit , Modelshoot style, Professional Photography, soft lighting, PHOTOREALISTIC, Realistic, standing in a dark studio background, blurred background, RAW, analog, sharp focus, 8k, high resolution, DSLR, high quality, Fujifilm XT3, film grain, award winning

[Johnny Cage]

This was the best one I was able to produce with your Brad Pitt prompt:



Most of the time it looks more like a poor man's Brad Pitt. I guess mine isn't yet sufficiently trained to consistently produce good Brad Pitts.

[maxhitman]

Originally Posted by Johnny Cage

This was the best one I was able to produce with your Brad Pitt prompt:

...

Most of the time it looks more like a poor man's Brad Pitt. I guess mine isn't yet sufficiently trained to consistently produce good Brad Pitts.

That looks very cool
I grew up watching his films, great actor, so most people already know
how he looks, both in younger and older faces.
I will say that image is of Brad Pitt !
It´s great that you got it working right from the start ! Amazing !

Keep doing a few more and you will notice the face details and hair will
become even more realistic looking.
It usually takes my A.I. to create at least 40+ images until they become
so Life-like and Real, that I cannot tell the diference if it´s a
real photograph or if its an A.I. picture.
The only way you can tell sometimes that if its A.I., is by looking at the "hand fingers". This A.I. still struggles a little to create realistic looking Fingers and toes.
Maybe in 6 months from now, all pictures will look much better, this AI
keeps evolving every day !


Last night,
my Stable-Diffusion AI just downloaded a bunch of new data and
Information updates. It updates everything with new goodies, once
every 4 days or so.
Now it has been UPGRADED with some new things and selections on
the MAIN Menu !
-- Last week it added over 800+ art styles and Embeddings and
around 200+ new Loras ! I have noticed my images looking
much better too.

I think this new thing has to do with adding extra LoRA´s or Embeddings
into the Checkpoints already working, without having to open a
diferente Menu section..... I think that is what it does. (?)
I am still testing it to see what it does exactly... or if it will launch
a nuclear missile !

I have to go research for some new tutorials and Updates as to
what exactly this new Menu Selection does. I will tell you later.

[Johnny Cage]

Messing around a bit with nighttime lighting in the background.

[maxhitman]

Originally Posted by Johnny Cage

...//...

Messing around a bit with nighttime lighting in the background.

That is a beautiful and very sexy picture
Congratulations! I love those light effects.

I have read a few comments on other websites about the "strange"
behaviour of this A.I. to create too much light on Ai-Photos (not ART)
and in particular some Ai-photo Portrait scenes, in which there is just
too much light focused on the subject face and/or body.
For ART type of Ai-creations this will not have much of an effect
on the overall image, although some times it can. It all depends on what
type of "art" it is.

Fortunate for us, this does not happen too often and usully happens in
1 or 2 pictures seen in a batch of 6 images created. So, there is no
cause for an alarm.
However, there is always a solution to every problem.
One way I got around to fixing this little problem and to make every image
have a more stable and softer tone of light in the Ai-Portrait photos, is to
write in your Prompt "Soft Lighting".

Here is an example where to place these words
(from one of my previous posts)...

Full body portrait of Singer JIM MORRISON man with long messy dark Hair Style, (Long dark Beard), Detailed face, (perfect eyes), (highly detailed skin:1.1), (Fat overweight body), ((wearing shirt and pants)) , Modelshoot style, Professional Photography, soft lighting, PHOTOREALISTIC, Realistic, standing in a dark studio background, blurred background, RAW, analog, sharp focus, 8k, high resolution, DSLR, high quality, Fujifilm XT3, film grain, award winning

You can also write "Dramatic lighting" to have a stronger contrast between
hard lights and dark shadows.

By writing "volumetric fog", or "volumetric haze" right after the
prompt word script "blurred background", will make the Ai create
this fuzzy-soft light shimmer around your subject, and the bright lights
will not be so strong on your main subject if its an indoor photoshoot.
With these same words "volumetric fog", used for an outdoor photoshoot
will create a nice foggy haze behind and around your
main subject. If the image is outdoors and in a Sunset scenario, this
can also create a nice shimmering Bokeh effect and a dreamy-look to
your overall Photoshoot pictures.

If you want to go further in playing around with words for your Prompts,
you can try all sorts of diferent Light tecniques used by professional photographers


Download these images I am posting and save them to your files,
for later use and writing your prompts.

-

-

-

-

Types of lighting used in FILM CINEMATOGRAPHY


13 Film Lighting Techniques Every Filmmaker Should Know
http://nofilmschool.com/film-lighting-techniques-and-examples

14 Film Lighting Techniques for Cinematic Shots
http://topsheet.io/blog/lighting-techniques-film

By writing some of these Professional lighting tecniques for Studio portraits
into your Positive Prompts, your A.I. can create some really nice photos
of your favorite Film or Music stars.

You can also research some Lighting styles used for Hollywood Films and use
those names and details into your Prompts. The A.I. will create these film
light effects too.

Here are some picture examples I did while testing some Light styles,
for studio portraits and Film Cinematography such as the 1982 "Blade Runner".

Alison Goldfrapp
(singer from band Goldfrapp)

-
images have been reduced in size
-
Sean Young
(actress - "Rachel" from film Blade Runner-1982)

...................................WORTH WATCHING...........................................

While searching for some tutorials I stumbled upon an interesting video review
of something which I already downloaded from Civitai a few weeks back.

Offset Noise for Stable Diffusion: Model, LORA and Embeddings
video by Olivio Sarikas

http://youtube.com/watch?v=EbPUQv2c5gA

---------------------------------------------
download them for free at Civitai.com

Illuminati Diffusion v1.1 - by IlluminatiAI (Checkpoint)
http://civitai.com/models/11193/illuminati-diffusion-v11

epi_noiseoffset - by epinikion (LORA)
http://civitai.com/models/13941/epinoiseoffset


Personally I like the epi_noiseoffset because its a smaller file BUT can 
create some amazing light and shadow effects.

These are 2 diferent add-ons for S.D. which can Improve your
light+shadows in your A.I. created photos.
They give you a higher definition and contrast between lights and
darks on your images without the need for specially written prompts.

Of course you CAN also create the same light+Shadow effects just using
your S.D. with NO Loras or extras, and writting a very good positive
prompt, but for those not knowing how to, you can always try out
these 2 add-ons and place them in your SD menu.

Give it a try. Its free !

-------------------------------------------------------------------
::::::::::::::::::::: MUST SEE A.I. NEWS ::::::::::::::::::::

This A.I. is evolving Very Fast !

AI NEWS - This changes everything!
- Nvidia Omniverse, MetaHuman, GPT Plugins, Canva AI

http://youtube.com/watch?v=1Y_35_BhJ6Q