Playground AI Tutorial - Make AI Art for FREE!

yet another free air platforms at the scene this time it's playground AI this is actually the website here you can see they have their Community Feed or images so you can see the kind of quality you can expect and all you gotta do to get started is click here and log in with your Google account and once you're in you get this feed of images that you can of course explore but we're going to skip right past that and go up the top here to create and you can see we get this interface so uh they do invite you to join their Discord server it's going to X that for now but essentially all we need to do to get started if you don't want to do absolutely anything yet is just simply go here to prompt and I can type in a prompt or something that I'd like to see when a radio hit generate and this is our first image now let's start exploring some of the options here straight away when we create this image we can create variations we can straight up download this image it is currently 512 by 512 pixels I also have these actions here I can edit copy link I can make this private I can in paint there's all of these awesome tools that you can apply to this image so straight away the first thing we're going to look at is we're going to upscale this by four it's fairly fairly decent not as good as I probably like but the resolution is actually pretty good at 2048 by 2048 pixels but let's see if we can get something better I'm going to bend this one and move on you can see there is also a Pro Plan available but we're focusing on the free stuff today and the cool thing is the model here used was stable diffusion 1.5 I can switch to stable to Fusion 2.1 and type something in here again and I hit generate so it's created this image using stable Fusion 2.1 which is another sort of it's even better sort of model a little bit more updated but I can also switch to dial E2 however it does require a paid add-on so you can use Dali 2 here if you're willing to pay for the pro license which is ten dollars per month for 800 images but we'll come back to that later and see what's available well what's cool as I go through and create these images I can also change my columns here to view more as I go I'm going to stick with stable Fusion 2.1 and we're going to keep on playing so I'm going to say here again a and we can now play with a filter so we're going to start improving and adding to this as we go I'm going to click on this filter and I can actually add something in here so I've got these two here such as close up portrait are all timing so it looks like we have just these two for now I'm going to choose the close-up portrait if we have close-up portrait somewhere around the mountains meditating and we can also remove certain things from the image this has been put in automatically so you can add some more things in there if you wanted to the other thing too is so this is the left column over here over the right we have more options for the image Dimensions 768 by 768 is all we get with a stable Fusion 2.1 we can adjust the guidance so we can bring that up a little if we want it to be a bit more closer to what we want we can bring it down if we want to be a little bit more free I'm going to bring it down to see what kind of result we get the quality now this up to 30 is sort of how far it'll actually go with rendering the image you can take it Beyond 30 to increase the quality but it actually tends to only sort of uh change the image past a certain point so you can't just crank up the quality and get these awesome images but you can't play with that and see what kind of results you do get the seed is the starting point so if you want to try and work with the exact same generation and starting point each time you can pop a number in there so I can turn this off and play with this number uh specifically we're going to keep on randomizing and at the moment I've got one image I'm going to say I want to see four images and if I want a private session I can switch to a Pro Plan but for now we've got four images and a few settings for stable Fusion 2.1 let's hit generate and see what we get and now check out these images we've added that filter we've got some really high quality images that have really sort of blown me away the detail in the skin everything is actually really cool so what we want to do is go up to the close-up portrait change it to old timey and see what we get with that and we get something a little bit different again so you can see how we get these cool different sort of options with our prompts the other thing too is if I switch back to standard Fusion 1.5 our options change again I can even change the resolution to something different uh most of the other settings are the same but I can change the resolution and I'm not sure but it does seem to have more filters so let's say we go for a retro anime filter and we go samurai meditating in the mountains so you can check out some of these images they're pretty cool we've got a lot of different options there so that is essentially you've got more filters with with uh steroid Fusion 1.5 so I would highly recommend going and exploring these and having a play because there's definitely a lot of things we can play with here but we've generated some images so let's go down to one of these really cool really high quality 2.1 images and see what options we have again if I hit the plus symbol I can edit this so I'm given instructions so it says here add a mustache to his face maybe I just say open his eyes so let's generate and see what that does now that actually hasn't opened his eyes at all but uh let's take this step further and see what else we can make happen make his mustache green give him an angry expression you see it's kind of tinted the image a little bit making a green that hasn't done a ton but uh it's still that's something we can play with a little bit further I would say this image is probably not the best what I could say even more so is say remove facial hair making clean shaving or something like that now we've had some success but the image is a little bit odd but that's just the what happens with air it takes a lot to figure out exactly what works but uh you sort of get the idea I'm going to hit cancel here and we're going to move on and see what other options we have I can copy a link to send to someone I can make a private by going pro I can download this but we're going to play with in paint now I can actually remove areas here what I can do is I can remove maybe this part of his ears and his eyes and what this does it allows us to generate something within these areas so now I click done and you can see down here under image to image we have our image so I'm going to say samurai samurai face with white eyes no pupils and pointy elf ears seeing see now his eyes are open his ears are a bit pointier and uh overall it's it's been one way we can basically edit exact parts of the image so we're going to come back down to our image hit the plus symbol and play with a few more things so we can create variance or use in in image to image so what happens is if I click this using image to image remove this using image to image and it'll pop that image in there so now all I can say it'll actually reference that image I can keep this same Samurai face of white eyes no people's pointy elf ears and say lightning coming out of his [Music] and let's generate so you see how it's actually gone and made our image a bit more closer to what we said here with the white eyes or not so much the white eyes but the pointy is the lightning and it's reference that same sort of layout to the image it's not the exact same person but it has referenced the image with the colors the way that's framed we can see that it's actually used that image as a bit of a guide to creating this one but if I come down and the image strength is bumped up to 60 something a lot stronger and hit generate again it's gone closer to that image and still sort of tried to find somewhere in between the two so we can actually get really good refinements on our images and play around with those pretty successfully now I'm going to come back down again hit plus and this time I'm going to create variants and you can see it's produced this second image up here so it's pretty cool if it's create a variation on the same prompt if I click on this again and yeah the option is not there at the moment but that's a pretty cool feature now the next thing I want to look at if I go back here I can do a face restoration but the face actually looks pretty good so what if we come down to this image down here and perform a face restoration it's actually smoothed out the face just a little bit and I can go ahead and download that or compare I'm going to hit close but uh we can actually work by improving the faces on our images if we want to we can do the same with this image where the face might look a little strange again we go Plus and face restoration and the face is looking a little bit more normal although it's actually blurred a little bit but uh it's pretty cool moving back to our original image again I can view this full screen if I want to preview it which is pretty cool but uh considering the amount of detail I wouldn't mind upscaling this to see what kind of result we get from it also so hit plus upscale by four and this is our upscale image and the quality is actually really amazing if I zoom in you can see it's a very large image the detail is pretty awesome and it's one of those things where when you compare this to something like midgen which does probably the best imagery out there it doesn't produce these high resolution images so the image resolution of this picture is 3072 pixels by 3072 pixels which is pretty insane now you notice before we had some options disappear from here but if I switch from stable to Fusion 2.1 to 1.5 some of those options come back but another thing that we can explore with this we can import images to edit so if I go up the top here where it says import image to edit I click that I'm gonna bring in this skull I made about 20 years ago on Photoshop literally about 20 years ago and I have the original here which also means I can go in I can say make changes here so add a mustache to his face what I can do is give the skull ice pupils and make him purple and I can also edit my instruction strength so it goes a bit better add the quality and details a little bit and I can choose to remove things from before so I have a negative prompt so I'm going to leave that for now I'm going to hit generate so you see how it's edited my image using the AI so you can actually even edit this image further if you want to but you can actually go in and make edits to your images and also if I go to cancel now use image to image here I actually don't even have to edit the image I can just use my I can just basically Bend this and I can go here to image to image and simply upload the same picture again say glow in the dark skull demon and hit generate and you can see it's created a skull based very closely off the layout of my image now once again if I go into filter although I'm going to come out I'm going to use stabbed with Fusion 2.1 although we don't have images to image with stable diffusion 2.1 so we'll go back to 1.5 add in our demon again choose a filter we'll choose a Dark comic this time or geometry Style add a prompt and generate and again it's taken the layout of my image and worked off that there's a very cool setup and a very simple interface if we use something like mid Journey where you want a few more options and Tools around to play with this is definitely the way to go and if you've used Leonardo Ai and you think it's just a bit too complex this is much simpler and has plenty of features and you simply log in with your Google account and you can start using it straight away well except for the dolly and some of the privacy but everything else here is uh is pretty cool and I highly recommend if you want to get better results definitely play around with some of these Styles because that seems to be the strength is these are all styles that have been models uh been trained around a certain style as you can see here this skull image is looking pretty cool if I upscale it another thing we can do is head back to hit the playground AI logo at the top left here we can go back to our community feed if I want to go to say I've got my feed here it shows images of people I follow I don't follow anyone yet but if I go to say animals let's say I want to recreate or play with this I can click on this image here and I can see the prompt I can actually copy that prompt if I want to and try it I can copy the negative prompt and try that you see here copy prompt I can edit the image so if I click edit I can use the same edit function as before I can remix the image so if I click remix I've got all the settings here they've used so I can essentially do what they've been doing but I can change some of the details change some of the settings and play with that image further as well so that's one of the cool things about playground AI you can really just have a lot of fun with it you can explore other people's images and try to sort of improve on those or work on those for yourself so it's a great platform to play with and like I said I sign up with Google and it's what I've been doing here today is totally free I haven't been paid a cent only if you want to use dolly or the Privacy maybe a few other features we can check out the plans by going up the top here and going to pricing free for everyone you get a thousand images A Day images can be used commercially so straight away you can get started making images uh you've got the 15 a month plan 2 000 images a day with double diffusion no weighting No Limits uh classrooms generation or you've got ten dollars a month which is 800 Dolly images so uh it's kind of I think there's one or the other it doesn't look like you can actually so this is an add-on I should say an add-on to the Pro Plan so you can pay up to 25 a month if you want to really get the most out of playground AI so check out playground AI the link is in the description otherwise hope you enjoyed the video if you did please consider giving it a like or check out my channel for more AI art videos otherwise have a great day and hope to see you again soon

Channel: Wade McMaster - Creator Impact

Views: 19,574

Keywords: playground ai, playground ai tutorial, playgroundai, stable diffusion, ai art, getting started with playground ai

Id: KmYvQV7sFs0

Length: 14min 27sec (867 seconds)

Published: Fri Mar 10 2023