What an amazing response, thank you for taking the time to do so.
We’re totally on the same method of thinking; I’ll use a kitchen scene as an example again (its also part of the story I’m writing) and within the kitchen is an old granny. What I’m hoping to achieve is a similar effect to classic animations but minus the animated character also. My plan: generate a kitchen image in a specific perspective and then use that image to further generate replica images just with different viewpoints/perspectives. Then generate the granny (say full body side profile) so shes “facing the cooker” in a kitchen backdrop. And simply overlay and position the two together. Doesn’t need to be anymore technical than that. No movements etc. Next scene, same backdrop for example but now generated a granny with her back to us (but exactly the same details from the previous scene i.e. granny looks like the earlier granny) and now overlay the two differently to incorporate granny at the sink. Next scene…close up of granny at sink. So backdrop has change focus and zoomed in, so has granny but all still same features.
This is the first encounter I’ve had of perchance and even coding for such a thing but after discovering several plugins like ‘pose’ , ‘physical description AI image’ and ‘image layer combiner’ I had a good feeling what I’m looking to achieve would actually be possible. I kinda wanna go through the challenge of it a bit too as I’d like to get a deeper understanding so when I want to manipulate stuff later I have more control and greater scope of what can be achieved.
Again, really appreciate the considered response. I was looking through some posts and could see that its a lovely community. :)
p.s. Ive just looked up the site Playground you suggested. Landing page it highlights the very thing, two images integrated. Ive not been on the site so have no idea as to the UI and whats involved…but if Im honest, I quite like the idea of coding it myself and really deep diving in. Also I think having the images seperate still might be handy, I dont know!? But it’ll be quite fun to explore. This whole project has evolved for me…I want to create a portfolio of music compositions for film/TV etc… and was going to use old silent movies. But then pondered the idea of creating my own. Animation would be too labour intensive (I think…well in these preliminary stages anyway) so thought about creating stills. Started exploring AI possibilities, discovered perchance today and was blown away by it! Fits what Im looking for perfectly. Coding it bespoke to my project will provide so much versatility and potential. So it looks like Im back to school!
Also, I wasn’t sure if the image would differ if I altered the prompt itself and so would lose continuity? Was thinking that maybe setting up the generator to have conditions that cover the alterations so would only change say ‘viewpoint’ and use the same seed? again, completely new to this and just piecing bits together so most likely completely arse about face way to do it! lol
Thank you for your very informative reply (although now the cogs are whirring in all manner of directions!).
I totally understand what you’re saying and see that if I were requiring that realm of functionality I.e. in-painting then a different approach would be better. I’m intrigued by the suggestion of my own local Stable Diffusion setup. I didn’t even know that was a possibility and in the long run may possibly be the better option even if what I require is available here. So in my thinking (and please forgive my naivity, I’m green as grass at present!) I would generate one image, say background first. Write the generator to have quite specific conditions I.e. depth of view, positioning, scale, size etc etc (I’m totally clueless as to the limitations here so again please correct if not possible). Result, one image. Backdrop. Then generate the character image. Maybe whole separate generator. Again, very specific with scale or pose, positioning etc and without background. So now both images are created in the manner required. My thoughts were the locked-layer-combination!? To combine the two images 🤷🏻♂️ and voila. The next step would be to then adjust parameters within each generator but revolving around the same seed (might be talking nonsense here as again, I’m just learning) to adjust in the desired way but retaining the features I.e. colours, style etc for continuity.
Am I just wishing on a star do you think and better suited to my own local?
Also, I’m not seeking any prompt advice here and respect that this isnt the place for that too. I just wanted to reach out to experienced minds to question the plausibility of my thinking as with my limited knowledge I could only piece together an idea from what is already available, tailoring to my needs. I mean, even if I create 2 images and use photoshop to layer its not an issue. My main focus is the retention of detail whilst manipulating the image.
Thank you again for a very thought provoking reply. Off I go to see what a ‘local’ setup looks like! :)
Thanks guys, this is greatly appreciated :) I’m going to sink my teeth in and try and get up to speed a bit (well, maybe a hurried walk, or a jog in fact. Not quite speed! lol)
ahhh I see, tags will be involved for sure then. I’ll take a look at your generator too and try and get my head around whats going on. I wonder; you mentioned ‘cropped image’ in your reply…well it just got me thinking about how images are ‘recognised’ and if there is a possibility in the AI to differentiate the character from background (if background is given transparency) and create the crop itself? As in, a lasso plug in type thing that can auto detect? Just an abstract thought so don’t know if its completely ludicrous or not!? 🤷♂️
Yes, GPU…might be my achilies heal with that at present. Im just running off of a laptop, 8GB RAM kind of crap! Apparently could run Stable Diffusion from it but not SDxL without it being slooooow! Would having a local Stable Diffusion and not xL be pointless?
Yes I do make music. This is a really cool idea to have in your back pocket…once Im a bit further down the line will see if it comes in to fruition! thanks
Ive just opened this one up…thank you. That Artist list of yours is a bloody handy addition! 👍
Thank you for your support. Its really surreal how coding has entered back in to the journey…Its a story for another time. Really excited to see what wonders will be discovered along the way
Thank you too. Its really lovely to see such welcoming people. For me, I’m super keen to explore the possibilities/capabilities of these incredible tools. Also I’m super aware that to do that requires a thorough understanding which I am most definitely without! I find it interesting when first starting something like this. Everything is so alien. Nothing fully makes sense. Slowly the picture starts to becomes clearer and you begin to see the possibilities growing. 👍
Ive been looking in to the image cropping and finding some jscript plugin subreddits so will delve a little deeper to see what people have been coming up with. I’m also going to try SDXL on my laptop to see what happens…if it takes an age to do anything will resort to SD…Thanks for all the pointers 👍
Great stuff, thank you! So I think I’m going to wrap my head around the elements Im looking to utilise and give myself priority learning topics (theres a lot of random generator stuff that I wont be using, lists etc). I built a 3D printer many moons ago and so had to dive in to visual studio and python so I have a very limited understanding of coding, certainly not html specific, but I get the premise. What I’m mindful of is not loosing sight of what Im also looking to achieve. I all to easily follow rabbit holes and lose track of the main objective 😂
Thank you 👍 As I say, Im going to see if I can setup a local and then start learning and developing what Ill specifically need for my project. I don’t need to get absorbed in all facets of SD just yet. and generally as things move along more comes in to light and I’ll swat up on the hoof. I like to know what Im doing or what things do rather than just blind copy and paste, so I’ll use the elements I use to teach me
@Alllo@BluePower@VioneT So I wanted to give little update. Not only to keep you guys in the loop but to pull myself back in to check. There are so many rabbit holes I keep delving in to and the reality is, I think (bar a couple of bits) I have the functionality and tools here anyway. I’m going to develop the structure of it here. Or at least a base version to expand on.
My plan… Although it has been suggested that one generator would be sufficient, my idea is to use multiple gens. and funnel down the outputs. As an example…First phase, scene/background generation. All the variables available to create the initial style/scene. 1 image created and put to the side temporarily (Output = Image A). Continue on to character generation (initial variable being how many characters in the scene. In which more cycles would be added to accommodate character volume. The image created from this run is the base position character (Output = Image B). The overlay/combining images ‘How to’ is still in the works so we’ll just say that’s happening now. What I would like is for that to be a relatively straight forward process (Output = Image C). BUT…That’s not the end! The next generation is the image manipulation phase. So most of the required data will automatically be funnelled down from the previous. Initial selection to ascertain what outputs are changing (which will then automatically populate the variables that are the firm Un-changeable’s. The variants for Output A will be different to Output B (A’s being framing, focus, lighting etc, B’s being poses, focus) and input selections will be limited to suit (as the core data that’s already been funnelled and populated is fundamental to retain continuity so wont be amended in any way ). Output D &/or E generated. To then utilise the Output C generator again, maybe with a further addition of referencing the earlier Output Image C as well. My thinking to further ringfence what the intended style/feel is. (but maybe it would cause a negative effect…not sure). Output = Image F. This cycle can then be further repeated to create additional scenes.
I intend to utilise this system (creating scenes of multiple variations, all generated from the same core data) to be my method for creating ‘Acts’. Meaning all of the relevant data is still live, as well as being utilised that way too by funnelling it throughout. Once I’ve created a base model of this idea it’ll give me further insight in to what works, what doesn’t, what it needs, what it doesn’t etc etc. and if I’m barking up the wrong tree or not!
Would love to hear your feedback/suggestions/concerns/criticisms! In my head it works…but that doesn’t mean it actually works! lol!
Yeah I agree…what I’m trying to do is wrap my head around the components and then further expanding on those. In doing so new approaches/methods come in to the light. Example…earlier in my head I envisioned the ‘stacking’ of generators…that image is now evolving as I’m discovering input output formatting, advanced hierarchical lists, multiple sub listing etc etc.
That’s great man, nice work 👍 We sound very similar in our approaches. The last few hours I’ve pulled up generators that have applicable functions and taking a good look at how they’ve been coded. Just breaking it all down and wrapping my head around it.
Is there a way of viewing a plugins source code at all? Just I really want to look over a few and see how they’re put together. The interesting thing I’m seeing with coding is the multiple approaches for achieving the same result. Some look crisp and smooth running…some look convoluted and hard work for all involved even the AI! lol
Absolutely 👍 I hadn’t really thought that far ahead but as you’ve mentioned it and I’ve pondered I think I would initially provide a condensed version (The background gen, character gen and combination) publicly. Just until I’ve had some time to explore the boundaries of the original. But who knows…as and when the time comes I’ll be sure to be clearer with my explanation though 👍
If i paste it at the end of perchance.org in the address bar I get ‘random pose’ generator…so is {import:pose-generator-simple} using the generator at perchance.org/pose-generator-simple as the plugin? And therefore the html IS the code?
Woah that is minimal indeed! ;) Its cool though man I understand whats happening now. I presumed all plugins were programmed using jscript so didn’t make the connection that the generators were indeed “plugins” themselves. It totally makes sense now. 👍
keeping features the same but adjusting elements within the image
Hi all,...