AI image generation software
I have seen with great interest a set of images posted to the critique gallery generated by Midjourney AI, and being highly impressed, I thought I would have a go myself.
Well, I had a quick go this morning and my first thoughts about using Discord as an interface were confirmed…I don’t know why particularly but I really dislike Discord, so I didn’t enjoy it much.
I know it’s only in Beta testing format at the moment so I hope they move it onto a decent platform…and lower the price about 10 times whilst they are at it. $30 a month for the standard package seems greedy to say the least.
The image generated as first attempt was not bad, but not particularly inspiring either.
It did make me wonder though if the images generated by AI will end up causing people to find real world photos boring.
I am interested to know what other peoples experiences with this tech have been and what people think about the future of photography as an artistic medium.
Well, I had a quick go this morning and my first thoughts about using Discord as an interface were confirmed…I don’t know why particularly but I really dislike Discord, so I didn’t enjoy it much.
I know it’s only in Beta testing format at the moment so I hope they move it onto a decent platform…and lower the price about 10 times whilst they are at it. $30 a month for the standard package seems greedy to say the least.
The image generated as first attempt was not bad, but not particularly inspiring either.
It did make me wonder though if the images generated by AI will end up causing people to find real world photos boring.
I am interested to know what other peoples experiences with this tech have been and what people think about the future of photography as an artistic medium.
Hi, as it was my post yesterday and I've had another day of experimentation I'll hopefully provide some insights.
The $30 + VAT at the current exchange rate is basically £36 a month, which is a lot, but you can cancel after a month or go down to the basic package. Everything I've read seems to suggest that you are addicted for the first month then can let it go as the enthusiasm fades.
I am a great fan of graphic art, which is why I became interested in trying the package. I took photography seriously when I put my pens, inks and paints away, so I found the idea of creating random images appealing and all of my early images were of that bent. However a request for a 'fantasy' background had me delving into landscapes and I was astonished at some of the images produced.
Yesterday I mentioned photography and I spent some time last night and most of this morning trying out some photographic prompts. The Ai is just a computer, isn't intelligent, and basically looks for ways of putting together an image from your prompt. Like I said yesterday, it has no real understanding of anatomy and is notoriously bad with hands, when you'll find a bunch of sausages, or in one of my images 6 fingers. Occasionally you'll find an extra arm or one arm missing altogether and the same with legs. (I'll post a couple of those). What I have found this morning, is that I have got some really good images and if nothing else they have given me some great ideas for when I next venture into a studio with a model. I'll post some of them on for critique, so you can see what I mean.
I have the feeling that as this gets better you'll see a reaction from the people it's going to affect most, though I'm not sure how that will go. I presume it's similar to the fact that CGI in films now no longer needs the actual actors. Last Saturday I saw ABBA Voyage at the ABBA Arena and the show was spectacular and the theatre was packed, but we were watching hologram projections, which admittedly, were astonishingly good.
https://youtu.be/iEikjzZO2N8
Norm
The $30 + VAT at the current exchange rate is basically £36 a month, which is a lot, but you can cancel after a month or go down to the basic package. Everything I've read seems to suggest that you are addicted for the first month then can let it go as the enthusiasm fades.
I am a great fan of graphic art, which is why I became interested in trying the package. I took photography seriously when I put my pens, inks and paints away, so I found the idea of creating random images appealing and all of my early images were of that bent. However a request for a 'fantasy' background had me delving into landscapes and I was astonished at some of the images produced.
Yesterday I mentioned photography and I spent some time last night and most of this morning trying out some photographic prompts. The Ai is just a computer, isn't intelligent, and basically looks for ways of putting together an image from your prompt. Like I said yesterday, it has no real understanding of anatomy and is notoriously bad with hands, when you'll find a bunch of sausages, or in one of my images 6 fingers. Occasionally you'll find an extra arm or one arm missing altogether and the same with legs. (I'll post a couple of those). What I have found this morning, is that I have got some really good images and if nothing else they have given me some great ideas for when I next venture into a studio with a model. I'll post some of them on for critique, so you can see what I mean.
I have the feeling that as this gets better you'll see a reaction from the people it's going to affect most, though I'm not sure how that will go. I presume it's similar to the fact that CGI in films now no longer needs the actual actors. Last Saturday I saw ABBA Voyage at the ABBA Arena and the show was spectacular and the theatre was packed, but we were watching hologram projections, which admittedly, were astonishingly good.
https://youtu.be/iEikjzZO2N8
Norm
“
Quote:Last Saturday I saw ABBA Voyage at the ABBA Arena and the show was spectacular and the theatre was packed, but we were watching hologram projections, which admittedly, were astonishingly good.”
Hi Norm…thanks for the further insights. I think those holographic projections of the members of ABBA are actually better looking than the originals ever were 🤣
Quote:Last Saturday I saw ABBA Voyage at the ABBA Arena and the show was spectacular and the theatre was packed, but we were watching hologram projections, which admittedly, were astonishingly good.”
Hi Norm…thanks for the further insights. I think those holographic projections of the members of ABBA are actually better looking than the originals ever were 🤣
Like Norm I feel this type of software will find a place for it's self.. I feel like a few others it's main use will be for studio type backdrops and costumes that would be too costly to produce but can be produced cheaply this way. Also the backdrops can be designed in the wording to mix with the subject.
We will see discussions about if this is correct as we have with everything from colour/black and white, film/digital and straight out of camera and the use of photoshop. I think this one has a long way to run yet...
We will see discussions about if this is correct as we have with everything from colour/black and white, film/digital and straight out of camera and the use of photoshop. I think this one has a long way to run yet...
I wonder where Midjourney AI sources it’s images from? Are the artists and photographers who produced the ‘doner’ images receiving royalties?
Its an interesting concept, bit expensive for my tastes but I've just been looking at some of the results, and the conversations surrounding them! There's long been a category for digital art on epz, is this just a progression of the category I wonder? The dark winter nights will be rolling in and its a time when I've generally dabbled in photoshop etc producing various bits that dont resemble photos anymore, rarely post, but its good fun so I can see the attraction of AI.
This looks like a useful thread to have going, though I don't visit the forums often.
For what it's worth, here's an edited version of the comment that I posted on Norm's first set of Midjourney images.
There's a big discussion to have here, and I hope we won't get derailed by whether the technology is good enough now. I'm sure it will be before too long.
The key question is honesty (as suggested by acancarter). It's a form of digital art, and if it's not declared as such, there'd be a big problem for me.
Obviously, photographers can't provide critique on the technical side of this sort of thing, but I see no reason why we can't provide worthwhile comments on the aesthetics.
There's just one big but, almost unnoticed because the overall scene is dramatic. And that's the lack of the primary interest. This is like a stage set, or the opening shot of a film: we're waiting for something to happen, for the characters to appear and the tale to begin.
In short, the lead has atmosphere, but not story.
And I'm sure the program can address that, or photographers can add such points of interest in editing..
Plus one big reservation, and the same one that I have about digitally enhanced portraits. Will a surfeit of AI images mean that we find reality boring? Alternatively, will we find a diet of utterly spectacular scenery that defies geography and meteorology too rich, so that we hanker after a Martin Parr style 'boring image'?
For what it's worth, here's an edited version of the comment that I posted on Norm's first set of Midjourney images.
There's a big discussion to have here, and I hope we won't get derailed by whether the technology is good enough now. I'm sure it will be before too long.
The key question is honesty (as suggested by acancarter). It's a form of digital art, and if it's not declared as such, there'd be a big problem for me.
Obviously, photographers can't provide critique on the technical side of this sort of thing, but I see no reason why we can't provide worthwhile comments on the aesthetics.
There's just one big but, almost unnoticed because the overall scene is dramatic. And that's the lack of the primary interest. This is like a stage set, or the opening shot of a film: we're waiting for something to happen, for the characters to appear and the tale to begin.
In short, the lead has atmosphere, but not story.
And I'm sure the program can address that, or photographers can add such points of interest in editing..
Plus one big reservation, and the same one that I have about digitally enhanced portraits. Will a surfeit of AI images mean that we find reality boring? Alternatively, will we find a diet of utterly spectacular scenery that defies geography and meteorology too rich, so that we hanker after a Martin Parr style 'boring image'?
I don't think I'd mind the images on here so long as they were clearly marked.
I have just spent an hour or so this evening doing some images on Ai for a friend who is an artist. She was asking for abstract images that she could adapt and use in her own paintings. They're not the sort of thing I'd normally do or even think of doing, but the resulting images are really interesting. She's going to use them as a base for her artwork. Likewise the images I produced have given me ideas I'll use next time I'm in a studio.
It's a long way from the days of getting a sheet of tracing paper
I have just spent an hour or so this evening doing some images on Ai for a friend who is an artist. She was asking for abstract images that she could adapt and use in her own paintings. They're not the sort of thing I'd normally do or even think of doing, but the resulting images are really interesting. She's going to use them as a base for her artwork. Likewise the images I produced have given me ideas I'll use next time I'm in a studio.
It's a long way from the days of getting a sheet of tracing paper
Very attractive images Norm. They look a bit Art Deco. Nice complimentary colours …not usually the kind of thing I like but I like these.
Simon
Simon
There are of course free AI things to play with. These are all links our club chair sent round after we played with some AI tools at the club a couple of weeks ago.
https://ommer-lab.com/research/latent-diffusion-models/
https://stability.ai/blog/stable-diffusion-announcement
https://stability.ai/blog/stable-diffusion-public-release
An article about an artist who couldn't find his own work because all the searches were bringing up AI generated work with his style, it also goes into some of the moral and ethical concerns and legality of use.
https://www.technologyreview.com/2022/09/16/1059598/this-artist-is-dominating-ai-generated-art-and-hes-not-happy-about-it/
An article about how some artists are adapting and building this kind of technology into their workflow
https://www.reddit.com/r/StableDiffusion/comments/wyduk1/show_rstablediffusion_integrating_sd_in_photoshop/
Apparently if you want to play with this yourself without any of the technical mumbo jumbo required to get it up and running locally you can use this website;
https://huggingface.co/spaces/stabilityai/stable-diffusion
https://ommer-lab.com/research/latent-diffusion-models/
https://stability.ai/blog/stable-diffusion-announcement
https://stability.ai/blog/stable-diffusion-public-release
An article about an artist who couldn't find his own work because all the searches were bringing up AI generated work with his style, it also goes into some of the moral and ethical concerns and legality of use.
https://www.technologyreview.com/2022/09/16/1059598/this-artist-is-dominating-ai-generated-art-and-hes-not-happy-about-it/
An article about how some artists are adapting and building this kind of technology into their workflow
https://www.reddit.com/r/StableDiffusion/comments/wyduk1/show_rstablediffusion_integrating_sd_in_photoshop/
Apparently if you want to play with this yourself without any of the technical mumbo jumbo required to get it up and running locally you can use this website;
https://huggingface.co/spaces/stabilityai/stable-diffusion
After a little more practice with Midjourney AI I got these images this evening
Hi Willie,
I found in my brief trial of the AI bot that simple seemed better. The prompt text I used was something along the lines of ‘Fantasy seascape, castle, under attack from ship’
It’s quite good fun and the results can be quite a surprise. If you haven’t yet, why not have a go?
I used Midjourney via the Discord app.
Best wishes
Simon
I found in my brief trial of the AI bot that simple seemed better. The prompt text I used was something along the lines of ‘Fantasy seascape, castle, under attack from ship’
It’s quite good fun and the results can be quite a surprise. If you haven’t yet, why not have a go?
I used Midjourney via the Discord app.
Best wishes
Simon
