AI generated art

[Mon0Zer0]

Curious, which AI programs are you all using? I'm making a list of AIs I want to play with at some point. There are a ton of them out there and I experimented with a few of them before taking a (hopefully) temporary step back from them, and some are definitely better generators than others .

All of the above are using Disco Diffusion v5.3 - it's the earlier project of the coders behind Midjourney. It's very techy with hundreds of parameters to tweak and slower, but it's free and you can get as good results if you know what you're doing. DD also means you can start with an image (such as a roughly photobashed landscape) and DD will render it in an artist's style.

The berkey painting above started life like this:

 

[Mon0Zer0]

Recently, I've been using them as starting points for paintovers. With DD you see a snapshot every 50 iterations that are just fantastic for generating compositions.

that started off as one of the earlier iterations that I added to with a mix of overpainting and photobashing.

 

[Laura R Hepworth]

@Mon0Zer0 Thanks! I hadn't heard of that one so I'll check it out and add it to my list . Oh yeah, I think their main contribution to artists will be as a source of inspiration. Those I see on DeviantArt fretting and complaining about AI 'taking over' and replacing 'real art/artist' are really whining about nothing.

 

[Mon0Zer0]

 

[Mon0Zer0]

 

[Mon0Zer0]

 

[Mon0Zer0]

 

[Laura R Hepworth]

Nice! I've yet to figure out how on earth to install/set up Disco Diffusion .

 

[Laura R Hepworth]

I have, however, successfully applied to the beta of Midjourney . Applied a week or so ago and got the invite today .

 

[Mon0Zer0]

I have, however, successfully applied to the beta of Midjourney . Applied a week or so ago and got the invite today .

use your credits wisely! I used mine up in about twenty minutes!

 

[Mon0Zer0]

Nice! I've yet to figure out how on earth to install/set up Disco Diffusion .

Try this:

 

[Laura R Hepworth]

use your credits wisely! I used mine up in about twenty minutes!

Yeah, they went pretty quickly . Personally, I think it would have been better for them not to count upscaling and variations against your credits, but just unique prompts. Especially since you sorta waste the first few allowed jobs just figuring out to use the thing .

Thanks for the video! Will definitely take a look as I'd like to figure the thing out. Even if it turns out I shouldn't use it for commercial work, it would still just be fun to goof around with.

 

[Laura R Hepworth]

These were my favorites of my Midjourney experiments. Just wish the limit was higher for the free trial so I could have played with it a bit more .

The prompt for these two were the same 'magic portal in a forest of glowing mushrooms, semirealism, digital painting, colorful, mysterious.'

And the prompt for this one was 'steampunk fairy.'

Following the tech with interest and hoping all the issues that have me being cautious with it right now get sorted out soon, because these generators are such good fun.

 

[Laura R Hepworth]

Try this:

Unfortunately, it's not working . Followed what the tutorial said, but keep getting error messages no matter what I do.

 

[Mon0Zer0]

Unfortunately, it's not working . Followed what the tutorial said, but keep getting error messages no matter what I do.

Can you cut and paste the error message here?

 

[Laura R Hepworth]

Can you cut and paste the error message here?

This is the main one I keep getting when I click on the 'Do the Run' at near the bottom of the page:

NameError Traceback (most recent call last)

<ipython-input-1-59aa9c154b7a> in <module>()
4 n_batches = 50 #@param{type: 'number'}
5
----> 6 if animation_mode == 'Video Input':
7 steps = video_init_steps
8


NameError: name 'animation_mode' is not defined

Even when I do try to set the animation mode to something, I still get the same error message. In the big long string of code above this error message I found three other ones under the animation mode section:

if animation_mode == 'Video Input':
frames = sorted(glob(in_path+'/*.*'));
if len(frames)==0:
sys.exit("ERROR: 0 frames found.\nPlease check your video input path and rerun the video settings cell.")
flows = glob(flo_folder+'/*.*')
if (len(flows)==0) and video_init_flow_warp:
sys.exit("ERROR: 0 flow files found.\nPlease rerun the flow generation cell.")
if steps <= calc_frames_skip_steps:
sys.exit("ERROR: You can't skip more steps than your total steps")

Was there anything that you had to download in order to make the program run? The tutorial I watched didn't say anything about it, but other places have though with conflicting information as some said those things were no longer needed. I followed what the video tutorial said to do, but something isn't working right so no idea what I'm doing wrong. Though, the version I'm trying to use is different from the version in the tutorial. This was all on Disco Diffusion 5.4 and he was using 4.1

 

[Mon0Zer0]

Right - did you start with "Run All" in the dropdown menu?

 

[Laura R Hepworth]

No, I hadn't. Was looking for it, but looked in the wrong place. Thank you! I'll give it another try and see if I can actually get it to work this time.

 

[Mon0Zer0]

Queen of Hell - DD5.4 + Photoshop

 

[Laura R Hepworth]

@Mon0Zer0 Thanks for the help! After disconnecting/reconnecting Google Drive, using 'run all' worked. Have a test batch going right now to experiment and compare with the AIs I've tried before.