Key moments:
0:05 Tour of the Dream Studio Beta
1:06 Overview
3:08 Setting
5:49 Redream
9:30 Cfg Scale
19:34 Sampler
You can sign up for the DreamStudio Beta here: https://beta.dreamstudio.ai
And join the discord for tips and community: discord.gg/stablediffusion
0:12 like what you see? Follow these featured twitter accounts to get your #stablediffusion AiArt fix:
@illustrata_ai
@KyrickYoung
@iminescent
@delta_sauce
@KarolineGeorges
@nin_artificial
@TomLikesRobots
@pharmapsychotic
@esmeromebooks
@zippy731
@RiversHaveWings
@d_horizons_ai
@devdef
@nvnot_
@CoffeeVectors
#stablediffusion #aiart
so great, very helpful as i'm playing with the DS beta too!
he's back!
I highly recommend checking out what RJ palmer said about this tool.
much like all other AIs it is stealing art from thousands if not millions of artists in order to train it's dataset, it's immoral and it infringes so so so many copyrights.
they even advertised this particular one as being able to copy artstyles from still living artists which is a huuuuge no no within the art community, much like tracing. do not endorse AI art, it is trying to make a monopoly of art by stealing from individuals
Wait… this isn't music
Wait how do I download this
Although I'm not personally too excited about the possible implications of AI generated images, I'm really excited to see technology advancing so quickly. You seem like you really know what you're talking abot when it comes to both AI and oscilliscope music and i'm excited for whats in this channel's future.
I'm especially interested in the applications of AI as a tool in assisting artists in creation and am wondering if there could be a way to somehow cmbine AI and drawing things on vector scopes.
thats enough from me, have a great day!
Have you requested it to generate "listening to the voice of energy" there? jus askign.
don't tell me it's credit based , that you have to buy some credits instead of a subscription like midjourney ?
thanks zip!
I have been testing DreamStudio Beta since Monday and find it very user friendly and lightning fast. I have tested all the other text to image generators and this is my favorite so far. I have created three art videos featuring my Stable Diffusion results and they are posted on my channel now, to give an idea of the quality I have been getting😍
How do they actually feel about NSFW stuff? Like I know they don't endorse it but they give you the ability to turn off the filter. So will they ban people.
This is amazing. I would want to use those for my music.
Thanks for the great guide! You managed to explain some of the pre existing settings better than I had kind of known already. Great job!
Well that’s odd
I always imagine his voice to be similar to that of Kevin MacLeod, even upon hearing that it’s not even close.
I guess underrated producers of good things have an unfinished spot in my brain or smth
How can I gain access to it ?
I am the creator of Dream One, the beta test event's last !dream command to run before the beta test closed at 05:33 AM, PDT. This day on 08/20/2022, Humanity truly saw its dreams come to life for the first time. Today we woke up,, and in a few days, all of humanity gets to experience what we did. It was an absolute pleasure to be a part of this and live every moment of our night dreaming as a group of close to 16,000 beta testers. This technology changed us, and making it accessible to all of humanity will change the world. We love every one of you. I made life long friends out of today, and I seriously hope you did, too. Go forth, and give humanity this new dawn. This wasn't just ours. In a few very short days, it will be yours.
This is now a public beta at http://beta.dreamstudio.ai – new accounts get 200 free generation credits.
Watch my other video https://youtu.be/UN1YVU6AkrQ to get the most from your Image Gen Credits!
Speaking of NSFW, is this locked down MidJourney's "big brother" garbage where you can't even make artistic nudes?
And get banned if you try to prompt around it?
What I mean is, do they watch what you do and judge you like MidJourney does with their flawed logic of what art is.
🙌
amazing. great demo
Played with it for a bit. Wasn't really happy with most of the results I got in the open beta test. Also a good 1/3 of my images came out blurred even though I wasn't generating anything NSFW, and the site wouldn't allow me to turn off the filter from my profile (the switch didn't respond to my mouse clicks) so these were just wasted prompts.
I'll check it out later if it improves. After my testing recently with Dall-E 2, Midjourney, and this I think I might put my money into a month of Midjourney.
I tried on my Android device and it works. Thanks buddy.
This is so cool! Once this matures, it can help artists work on their personal projects.
So with the weights being released on Mon does that mean we'll be able to use it with disco diffusion in a couple of days? I don't have a gpu with 10gb 😭
Nice was looking for a Dale type generator that like this
It keeps telling me check my membership
Awesome!!!!
i suppose uploading your own images to dream coming
Thank you guys for you great work.
DREAM BUTTON????? #nice
Why am I costing double credit? 18 credit points when I have generated 9 default images,and 150 steps for 9 image just costs my 54 credit! it should be 27credit for 3×9, it doesn't seem to be a display error because when I ran out of credit I was banned from generating
That is great!
How does something like this compare to the discord journey ai?
@11:39 looks like the CFg the higher it goes its paying more attention to the Artists reference as much as the creative direction in the prompt. So I am sure if you looked up the artists work you would see more of the painterly landscapes. Because in the prompt there is not a mention of a valley or planet etc.. it just says a distant galaxy.. if you told me to paint a distant galaxy I would paint your a milky way or andromeda or even the sombrero galaxy…..not a landscape painting.. so this tells me the AI is trying to conform the artists work into the concept of the galaxy, daytime, a dream…the artist must have images of people contemplating..like the famous -Wanderer above the Sea of Fog…
Pity their pricing model is whacked.
Thanks for guide !! especially how to use configuration to enhance the initial results
can you change the width and height to 1920 x 1080
Hu why some results are blurred or censored???
The same prompt that worked the 1st time, after some seed modification the result is blurred/ censored.
How can i solve this because those images costed me money
On my account there is no NSFW switch. Was this option removed in a later release?
Is this not re-patching of someoneelse's work? I wonder if that can be considered arttheft in itself? Re-imagining someone's work is still not your work right? Whether you re-composition it or not… shouldn't you use it with your own artwork as base line? I am curious as to what constitutes art here then and how do you compensate the ones you borrowed the base line from?
ρяσмσѕм
Thanks for the information, but I have a question. At 21:52 you show the option "Safe Mode restrict access to certain features". This option is not shown in my membership account settings. Is there a way to enable it. I got so many blury pictures, just because I want to create something with human bodies. Do you know, if I have to pay for all the blury pictures?
I’m a bit peeved off, there is no safe mode toggle switch in v1.5, it’s a shame one can’t experiment freely after paying for my credits…a great engine thought none the less… superb clarity overview, much thanks.
This was so informative and helpful. As someone who finds it really hard to focus, your tutorial was easy to understand and to follow. I’m brand new to AI art and have only been playing with it for a few days. I tried Midjourney, Stable Diffusion and StarryAI and I like Stable Diffusion best so far. I find a lot of software lingo impossible to wrap my head around and I really need tutorials like this. I hope you’ll do more.
I don't have a Features Safe Mode in account settings. Why?
Hey Chris , is it possible to feed the dream studio 3-4 actual images to use as reference for certain concepts?