Playground Tutorial Thinking Outside Of The Box

Playground Tutorial Thinking Outside Of The Box. Hello everyone! This is a very specific tutorial on how to create a centaur and I realize not everyone would want to create one. However the concept can apply to may different things. Sometimes there might be limitations on what you can prompt and it may be a better idea to get creative and find a solution by editing your images a particular way to get the results you want.
Prompt for the man: muscular greek man standing poised on a mountain top, nature mountains background
Prompt for the horse: horse standing poised on a mountain top, nature mountains background
Final prompt for the centaur: analog photo of a human Centaur, wearing skyrim armor, standing poised, mountains of Thessaly and Arcadia, high details
Negative Prompt: ugly, nsfw, nude, cgi, 3d, 2d, drawing, sketch, painting, blurry, smooth
Music by Artlist
Find us on
Playground AI playground.com
Discord / discord
Twitter / playground_ai

[@ranadeepdebnath1892]

Dear Sir, I sincerely appreciate your guidance. As a dedicated user of the playground for the past two months, I've gained valuable insights from your videos. I've progressed to creating impressive images, yet consistently encounter challenges such as vocabulary limitations due to my less-than-perfect English proficiency. Moreover, the playground often exhibits glitches in its free version, impacting the functionality of its tools. I believe this issue stems from a lack of comprehensive information and images, particularly regarding ancient topics. While the website is commendable, it faces significant drawbacks in meeting specific user needs. Additionally, the platform tends to slow down after prolonged use. Thanks for amazing creative ideas and teaching ❤

[@playgroundai]

Thanks for the encouragement and we appreciate the support. There will be limitations to some subjects like in this video. That has a lot to do with the data sets used for the models. Overtime the models will be able to recognize more subjects more widely. As for the slowdown, that is expected for free users, during peek times priority will go to paid users while those on the free tier will be put on a queue system.

[@ranadeepdebnath1892]

Oh, I see, thanks sir, i didn't know about this. For now, I can't take the paid version, but waiting if someday I can. Until then I am learning the technique....thanks for your kind response ❤❤😊

[@fictionaddiction4706]

@@ranadeepdebnath1892 I think there was another video about how you can just write in your native language and playground understands as it uses chatgpt to translate. I haven't tried it, but you can also just use chatgpt or bardai to transform your language into good commands in English.

[@user-zr2gu2gg6w]

You guys/ladies are just too good. With all your continuous high quality work you need a holiday; enjoy Christmas.

[@playgroundai]

Thanks so much for the support and Happy Holidays to you as well!

[@faisalzishan4616]

Thank you for the tutorial. Very detailed and delivered patiently 😊

[@quentincaldway]

This was absolutely amazing!!

[@hounsfieldunit]

Wow, there is so much to learn! I will definitely try this out. Happy Holidays and Thank you!! 🙏

[@playgroundai]

Yes it can be a deep rabbit hole but take your time and have fun! Happy holidays! 🌲

[@fictionaddiction4706]

How is this not more popular than mid-journey? the user interface alone is great and easier.

[@Dreamseerapp]

Olá. These tutorials are incredibly useful. Thank you for doing that. And keep up the great work! ☀

[@playgroundai]

I'm happy to hear that and thank you! We appreciate the support. 👍

[@billjones3720]

Excellent! Thank you.

[@playgroundai]

You are welcome!

[@captainpike3490]

That is geat, Creative

[@GryphonDes]

What a fun little exercise! Nicely done!

[@playgroundai]

Thank you! It's definitely a fun technique.

[@kagtisleri2418]

Perfect. Thanks for information

[@playgroundai]

My pleasure!

[@user-sf8hi3qw1r]

this is very very useful!

[@playgroundai]

Glad to hear that!

[@abhinandanpandey6513]

EveryTime i see your videos i understand bettere ( thank you sir)

[@playgroundai]

Glad to hear that! You're very welcome!

[@ytv2117]

nice tut. thank you.

[@playgroundai]

You're welcome! Glad it was helpful. 👍

[@staydoogan]

Merry Christmas, my dude!

[@playgroundai]

Happy holidays bud! 🎅

[@RalfKurtsiefer]

Really good Dragons are hart to find

[@playgroundai]

Try with SDXL model and the various filters. I do them all the time with success

[@phil2fr]

I already tried to make a centaur with only one prompt but, as you say, it's too difficult 😄 thanks a lot too bring us the solution.

[@playgroundai]

You're welcome 😊

[@sharmasalilrahul]

Amazing 👏 ❤ It's 🎁 for us.

[@playgroundai]

Glad you enjoyed the vid!

[@MrX-nh1xm]

cool stuff, definitely looking forward to crerate such " composite" styled images, definitely good for creating complicated subjects and ideas that the usual text prompt that would find difficult in one image generation

[@playgroundai]

Yes that's the idea. You can only get so far with prompting at the moment. Much easier to edit the image you want sometimes.

[@MrX-nh1xm]

@@playgroundai absolutely !

[@umutcankilic2974]

​@@playgroundai Hello, I would like to ask you these two questions: Will there be a phone version of your application? Are we able to publish the images we make from the application on social media applications and Pinterest-type applications? I would be very happy if you answer.

[@Tieungoandong89]

Great video! Thank you ! And Merry Christmas ! By the way, I have a problem that I don't know how to handle. That is, I have a photo of an outfit, and I have a model character (wearing a different outfit). How to create a photo of the character wearing that outfit? And how to change the character's pose while keeping the outfit the same?

[@TiDiRSHORTS]

*Happy New Year) It turned out beautifully and most importantly a good example) There is no way to pay for paid versions of the site, but for a free opportunity to create a separate respect)*

[@playgroundai]

Happy holidays my friend!

[@CarlosRodriguez-cj8oo]

There is much "food for thought" here! Thanks. I'm new to Playground, just a few weeks, but I discovered early that I must "show" playground what I need rather than "telling it." And with that shift in perspective I now have an easier time producing the images I want! Do you plan to give the eraser tool different shapes? There are times when a square or an elliptical would have made erasing and editing much easier. Thanks as always and happy holidays!!!

[@playgroundai]

Yes, you also need to think like AI and be very specific. I'll give the feedback to the ream about shaped brushes. Happy Holidays as well! 🎅

[@nn-db4fw]

Hi and thanks for another fantastic video! I have one complex picture I really struggle with and that is to create a classic 3 headed Norwegian Troll. It would be amazing if you could do this and show us (me at least), the process how to achieve this. In any ways; keep up the great work. I am always around and completely in love with PlaygroundAI!

[@playgroundai]

Hhhmmmm I can see how that would be difficult! 😊

[@nn-db4fw]

@@playgroundai Yeah, I would love to be able to create it as I am writing some childrens books & fairytales, and will illustrate them with the pictures I create myself. As Norwegian I grew up with our famous fairytales, and who doesn't like a gigantic 3 headed troll? I bet the kids would love it for sure! ;)

[@AlchemistAli]

Bravoooo,

[@playgroundai]

Thank you!

[@chandrasekar9990]

wonderful explanations, super sir. I ask you one simple question. how to make a movie or cartoon video? if you had a solutions, could you please explain it in a video.

[@TMtheScratcher]

Great video! Now where you have a good representation of a centaur: Can you use that one OR even multiple ones are "Subject-reference" - so the generated background etc depends on prompt and parameters as usual, but the foreground object uses some traits from your reference - is this possible or planned?

[@playgroundai]

Not at the moment, I mean you can get variations via img2img but it will still be similar. We are looking at other tools like more controlnet features to bring in the new year. I just don't have a timeline at the moment but personally I'd love to have them!

[@TMtheScratcher]

@@playgroundai That sounds great! img2img is my current way to go, but the results look either too similar or too distinct. My use case is a card game and I need to depict many characters of the same fantasy race. While I got one great result of such a character, I struggle with getting a similar character into new environments, i.e. new backgrounds, poses, variations in clothing and holded objects, etc. The differences in face, details etc work fine, but the image composition is basically the same with img2img. Cutting out and erasing will solve the background, but not the pose etc. Nevertheless, playground AI got me already faaaar beyond what I could have managed to draw or imagine by myself. And while it often fails in what I would think of (for humans) simple prompts, playground generated much better results than other AIs like Bluewillow or even Midjourney. Love how much control you have with this superb interface!

[@DritzzHistoryFacts-pi8sm]

I really like your videos. have an image that i really like and don't want to change its looks. But what I want to do is change posses, expressions etc. How can i do this and do you have a video on this?

[@Lokpui]

can you create a video teaching how to create consistent character in playground?

[@HolisticVibeStudioG]

Hola Y Aloha Learning practicing Gracias Grattitude Happy Holydays 🌲❄️🎵

[@playgroundai]

Happy holidays! Appreciate the support!

[@im_sorry_dave5117]

Hi really enjoying the tutorials and the platform itself... learning a lot 😊 Could you please explain the 50 generation limit per day for Quality and Details settings.... what should the slider be set at to not trigger the count... hope that makes sense... thanks 👍

[@playgroundai]

That just means you have to keep your quality and details at 50 or less to generate images. Which is more than enough for most use cases. 👍

[@im_sorry_dave5117]

@@playgroundai Ah that makes sense many thanks for the answer... looking forward to the next video 😁

[@johnleighdesigns]

I loved this tutorial and your wisdom on these approaches and very topical for me right now! Im struggling to create the fantasy christmas card I have in mind using an image of my wife and I - watched twice your video "Make AI Magic with Just One Photo" with both approaches of image to image and also cropping an image and outpainting, have tried in playground and Leonardo with the goal to create a movie still christmas action scene as if my wife and I are dressed up in santa clause coats of the impressive kind in the movie The Christmas Chronicles - Any ideas on best approach would be most welcome as im running out of time!

[@playgroundai]

There are a couple ways you can approach it, you can generate an image of Santa and Mrs Claus, then mimic the composition with your wife and take a selfie and just use the faces from the selfie merge them with inpainting. Or like use the selfie or existing image, inpaint the clothing and prompt for santa outfit.

[@johnleighdesigns]

@@playgroundai appreciate the feedback! seems i need to do a lot more testing and iterations rolling the dice to get something in the right direction!

[@vishwarupdev4055]

Please teach us to generate a family images, in different circumstances. I mean creating a storyboard (like movies, where we see more than 1 or 2 character in a single frame) also, consistent face. The real challenge lies in having a consistent face of more than 2 characters in different frames. Please help.❤

[@forreadingoutloud]

Does copy and paste with two images (like your man and centaur) only work with the paid for version of Playground?

[@IbrahemAb1]

Is there a way to get a burst image of a specific scene? Meaning, if I have a story and I want to turn it into pictures, for this I will need a set of pictures in the same format for each scene.

[@chandrasekar9990]

what are the differences between playground ai and Stable diffusion? please explain.

[@zaherdaw9068]

Thank you... just a question away from the subject .. with respect to PlayGround V2 is it better to use it with the filters or no for realistic images...?

[@playgroundai]

At the moment there are no filters for PGv2. For realistic images you want to put some negative prompts like 3D, 2D, cgi, sketch, drawing, hyperrealism to counteract the default smooth look. It also helps to use words like photo of, analog, raw photo etc.

[@zaherdaw9068]

Thank you@@playgroundai

[@fikrihaikal3598]

Sorry, is playground undermaintanance now?

[@livingchina2667]

When it will be available for ipad? Actually all these things for PC are available anywhere else for a long time, is a basic composition, the only thing interesting is the option for ipad as you are kinda chipper than photoshop, but only 5 dollars. I became to like playground text to image software (fabulous one, I proved before plenty of others) but it is annoying that I cannot do half of the things as I am ipad user, and my digital tablet is somewhere catching the dust as it was not so accurate as ipad is. So, will you be doing the version of canvas for ipad? Or better I start to think about becoming the bad part of the world, the Adobe user?

[@playgroundai]

We don't have a definite timeline but it has been something we've been considering for some time. We're very hyper focused on making PG a power image editor and we'd like to get it right on desktop/laptop first before venturing into mobile devices. Time will tell.

[@livingchina2667]

thank you for the reply. It is something to consider as ipad comunity is growing very quickly, consequence of ipad becoming better and better than digital tablets for the general public. One more question. Will be there some improvement for the storage website? Something like pagination, better search system?Thank you one more time for your work and for your answer. Happy Holidays!!!@@playgroundai

[@tianyuasiandrama]

Hey is anyone know which is the best app or program for draw amine with Ai

[@justjules5831]

How to make seamless tiled generations? I’ve used “seamless” “tiled” as prompts but they don’t work properly- the generations are always off. How does Midjourney manage to get it right most of the time?

[@marynapastukhova8119]

Hi! Thank you for your tutorials, they are really helpful! I have one question: I copied an uploaded photo into a generated one. Removed background, dragged, copy-pasted, and repeated all your steps but images won't merge, they remained separated and could only be downloaded as two separate images. Where can the issue be?

[@playgroundai]

Are you using the shortcut key CTRL+C and CTRL+V to copy and paste? Also make sure you drag over both images to select them.

[@marynapastukhova8119]

@@playgroundai , yes I used CTRL+C and CTRL+V, also tried to hold Alt and drag. Will try again. Thank you for your response)

[@re-doppelpunkt5054]

@@playgroundai I thought, I am the only one with this problem. I can not Copy & Paste anything. I selected both images, not working. Even with one image, not possible. If I use ALT + Drag&Drop I can copy both images, but they are still separated. My workaround is to use both images a Image to Image and generate with 100%. Than I have one image. I use Firefox Browser in the lates version. CTRL + V is working if I had something in my clipboard. But CTRL + C never works. Could you give me a hint, please?

[@anto6085]

How to setting ratio ?

[@KI-yd3ji]

Could you also make a video for how to make a same character with just different facial expression or same character, same cloth with different situation?

[@playgroundai]

Will be covering this soon, it's a bit tricky but I have some tips 👍

[@ranadeepdebnath1892]

Yes,,,I also want to know this...@playground, we are waiting .....😊

[@virtualwhispers]

I have no filter matching - juggernaut XL - or any of the filters that I see in your video -- I see them, I had to change the model :)

[@playgroundai]

You're probably on Playground v2, change the model to SDXL. Top right you'll see a selection for "model".

[@untossablesalad4423]

How did you prompt to get that wide shot of the man and the horse? Everytime I try I get only the top half (waist up). I'm generating in 768x1024 resolution, not sure if that affects it.

[@playgroundai]

Prompt is in the description 👍I believe I was using 1024x768 or 1024x704.

[@ArtBuddy-bf1is]

is this done with board or the other option >canvas ????

[@playgroundai]

This was done on Canvas

[@DidierJunk]

for me it is impossible to generate Wheelchairfencers in competition. In wheelchairfencing both wheelchairs are fixed and the fencers are sitting and fight. I know its not common but i am wheelchairfencer and I tried thousands of possibilities and cant get throug.

[@playgroundai]

I can see a limitation on the models data set for something like that. I did a quick simple prompt for it and see what you mean. I'll have to think about this one, hopefully I can come up with a solution for you. 👍

[@OfficialLove143]

In future, is there any hope for text to video option???

[@playgroundai]

Currently we are very focused on graphics but who knows in the future?

[@OfficialLove143]

@@playgroundai okay thanks, and in paid version is there possibility to create high resolution hyper realistic image creation like Mid journey or Dall-e 3??? Because I am interested for that so???

[@sharmasalilrahul]

Sir, Can you please make complete detailed video on FILTERS. It's help us a lot.

[@playgroundai]

We have a few videos on the channel but I do have some that I've been procrastinating on editing them hahaha! I will soon my friend. For now you can start with this one. There are a few more, just look on the main page. kzbin.info/www/bejne/nmqTkJ6Yp756m68&ab_channel=PlaygroundAI

[@sharmasalilrahul]

@@playgroundai Thank you so much 💓 for quick reply. A lots of Love ❤️ from India 🇮🇳

[@hk10218]

I am trying to compose ships on water, but don't know how to make the ship face different direction, such as bow, or stern facing the camera, or port facing the camera. Is there a simple prompt to do it ?

[@playgroundai]

Don't overthink it, try front view, side view, rear view, aerial, cockpit view etc. Start with simple prompts like "cockpit view of a ship on the ocean" then build from there. 👍

[@hk10218]

thanks, I will try it out@@playgroundai

[@livingchina2667]

try on the same configuration and THE SAME SEED, more likely to produce similar kinds of ship/boat@@hk10218

[@jnm2148]

Can someone tell me what words I have to include in the prompt to get the object in the centre of the image. Though the object is in the centre when using an aspect ratio of 1:1, when generating an image in 4:3, 3:2 etc. the object is pushed to a side and part of it get cropped. How can we get the object in the centre?

[@playgroundai]

Is it possible you are using the Playground v2 model? I ask because currently that model isn't optimized for aspect ratios other than 1:1. Try using SDXL model which you will find on the top right corner where you can select the base model.

[@jnm2148]

@@playgroundai yes it worked. Thanks a lot. There's another issue regarding editing a generated image. When we add a mask (highlight the particular area) and give the prompt to remove it, still the generated image is the same without the part being removed. For eg. a teacup is generated with two handles. I added the mask to one handle and gave the prompt to remove it. But generated image is same. This happened to other images which I wanted to make changes like adding a (missing) finger, correct the shape of a crooked heart etc. But nothing changes. Any way to solve this? Thanks.

[@playgroundai]

Mask is more for adding details, to remove use the object eraser. Shortcut Shift+E 👍

[@jnm2148]

@@playgroundai thanks👍

[@LouisGedo]

👋

[@playgroundai]

What's good my man! 👊👋

[@rp7438]

i try this, but the copy and paste : not working

[@rp7438]

i select two images and go > image to Image > give prompt and its working GOOD, Thank You!

[@Keimeri803]

I think its pretty Shiet, everytime " Your queue is too high." Dosent really matter when or how you try to make a art.