So glad to hear it! We'll dive deeper into prompting with another video to come! This may be a regular series I'm thinking of doing, so much to cover. 👍
Just know that negative prompts will only minimize those deformities but it does help. You can also upscale the final image if needed. Some filters you can even get away with sightly bigger sizes like 640x832 anything bigger is prone to duplicates. It’s just the nature of stable diffusion
It's like a hidden cheat code 😁 Now that you know...it can be a deep rabbit hole hahaha! Honestly though, it's not something that is well known and no one really explains it in-depth. I thought I would change that today and there is even more you can do with seeds! 👍
Your video under the Playground AI name is the best one yet. Comparing my images to those at the top of the community feed left me wondering how to bridge the gap. However, with your Prompt tutorial, I feel that I have learned enough to significantly enhance my image creation skills. 👍
great tutorial. I just started using free version. My prompt is simple and no spelling/grammar errors. I wait more than 300seconds and time runs out. I don't get to create anything. Any suggestion for troubleshoot would be greatly appreciated.
Like anything else my friend, it takes time to develop your own style and way of prompting. I've been doing this for awhile so some things come naturally to me. It did take some time to wrap my head around the concepts, what prompts worked, which filters to use etc. If there is anything I can help with please let me know. Be sure to check out the other videos to learn more!
I appreciate it! You can start with this playlist but we have many more other tutorial on the channel. ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-Mc2rNiEljp4.html&ab_channel=PlaygroundAI
So glad to hear this Kryss! Once you understand how this works, the next video I cover on this topic (part 2) will make a lot more sense. I want to get people away from copy and pasting prompts but to really understand what's going on. 👍
The best tutorial so far as you had explained the unknown. Thanks. Please keep these structured step by step videos as they are very helpful. Further, if you don’t mind to start giving a tutorial how to build a complex image and how to change specific items in the image. I am delighted for your kindness to put the effort and time to teach us. Thanks a million
Hi Hisham, yes we are working are way towards that video, hence why I posted this video first. Once this foundation is understood we can move forward to more complex prompting. 👍 Stay tuned!
@@pedroj3432 just write the things you don't want it to generate, like if want a picture of desert and dont want it to generate cacti you write cactus in negative prompt
This tutorial is KILLER! Very well explained, easy to understand (and follow!), quite user friendly. Like others said, anyone who wants to use the tool must see it.
This is the video I have been waiting for; thankyou ! I have been goofing around with Playground and Leonardo for about 6 weeks (having fun) but frustrated that I have been unable to get an outcome that I had in mind. I have been stunned at the beauty of some of the images that AI has produced but more by 'accident' than design. However, after watching your video I was finally able to produce an image that was closest to what I had envisioned and really pleased with the outcome. Many thanks - please keep doing what you are doing. Chris.
Hey Chris, really appreciate the positive feedback! Now that you are getting familiar with the basics, I think you'll really enjoy the next videos to come when we dive deeper into prompting. Stay tuned! 🙌
This video is a must for new or old SD users! I learned lots of new things and I have been using Stable Diffusion for some months now, amazing job and thank you so much!
Why such a surprise? hahaha! But I'm proud to represent! I'm just the content creator and Marketing specialist for Playground. We have an amazing CEO and a very diverse team! 🙌
I love your tutorials❤ Tell me please, do you recommend to generate images via board, and then play around with them in canvas? Or I can start creating directly in canvas? Is there any difference?
Personally I really like the ease of use on board and the simple interface however I almost always generate on Canvas now just because of my personal workflow and I like having an open canvas to work on. A lot of people start on board to generate images then bring it into Canvas to edit. It's really personal preference. We have been a discussing internally on how we can make it more of a seamless experience. 👍
I'm so thrilled! Those videos are exactly what I have been looking for for a long time. Thank you so much! Moreover I'm learning that I won"t need 3-5 different programs for post-processing a picture, I can do all with playground. 😍
Hello, I am struggling to generate images (backpacks,shoes,watches...) that are so close and large in the frame. I've tried using long shots or 24mm lenses, etc., but it did not work. What should I do?
Hi Punita, Yes, this is definitely on my list of video topics to cover. The nature of stable diffusion is very random but there are some workarounds you can do. It can be a bit challenging to be honest but now with the new features in Canvas there are several things we can do to get some level of consistency although it's not 100% perfect. It's still a very new technology but we're not far off from being able to produce those desired results.
The only thing happening different was the generation times and other images created, same prompt was used during those steps, nothing different except changing the filters. Is there something you want to clarity on?
Yeah some people I guess don't want to share their prompts, what they most likely do is bring the image into img2img, regenerate the image at 100% strength and then use the prompt box to put in a short description.
Yes it's 75 tokens. Stable Diffusion models have a token limit, which refers to the maximum number of words or phrases that can be used in a prompt. Tokens are not the same as words, as the model breaks down text into smaller units known as tokens. You can use something like this to check your prompts platform.openai.com/tokenizer
DON'T BUY A PLAYGROUND AI SUBSCRIPTION! The site actually tries to create porn from safe prompts! I actually got sick of fighting it's bias towards nudes and decided I wouldn't renew my PAID SUBSCRIPTION. *OH BUT IT GETS EVEN BETTER* Once my subscription expired, instead of just being downgraded to the limited "free" version, the a--holes actually showed how petty they are by wiping out my account completely!
I think you just have to know how playground works, why don't you start here ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-373yzYmnFa8.htmlsi=_r215F7mHx7JlgZn
75 roughly, however that could include commas too. A helpful rule of thumb is that one token generally corresponds to ~4 characters of text for common English text. This translates to roughly ¾ of a word (so 100 tokens ~= 75 words). This is a tool you can use to check your prompts. platform.openai.com/tokenizer
So I have a question. Can you get consistent characters with playground ai like you can in Mid journey? I tried a lot but have yet to get successful. I like Playground ai but this was the deal breaker that had me look elsewhere.
There are methods to get close but in general 100% consistency isn't possible at the moment and that includes midjourney, there will have to be some post production done afterwards. I will cover this very soon. 👍
Great tutorial as usual, one thing that i am struggling with is to complete the half body potraits or half body images of people, i am not able to complete the image. It always gives me distorted figures. Is there a specific method to do it.
Are you talking about in Canvas? Typically I would just simplify the prompt a bit, if it's legs you are focusing on, then change the main subject to legs, leave the rest. It may take a few tries.
Hello and thanks for awesome tutorial! I am new to text to art, been doing this for only some months and I really love it. What is the best program in Playground AI to use for landscape or medieval village/fantasy scenery pictures? If someone could tell me a bit about this, I would appreciate it a lot. I am planning to subscribe to just one AI art creating engine, and I really like Playground AI so far. I know Midjourney makes fantastic pictures for just about everything, but I really like "underdogs" so if someone could tell me a bit how Playground AI could compete with Midjourney when it comes to nature scenery and medieval villages/fantasy scenery, I would definately be subscribing immediately. I am not interested in creating photography/portrait pictures, not at this stage at least. In advance; thanks! :)
I think they should not use the word engineer, especially since it is generally used for people who study careers that last at least 5 years and what you need to learn in this field is an afternoon or hours.
@@playgroundaiVargas Llosa said that the language should be as clear and concise as possible, this does not give clarity, it is as if from one day to the next we call a cook a doctor.