Horrors of the Pinwheel Galaxy 

I have been refining my process. "The Android Uprising", even though not a popular video, was the video that caused a huge shift in how I approach creating these kinds of videos. The earlier videos I did, had no real narrative. I did try to order the images so they had good pacing, but they were more like a tour of the art I created. The real focus was on the music. Now I try to add at least some sort of narrative thread to the videos to add another layer to them. I like my older videos for what they are, but I think having at least some sort of narrative thread or progression makes for a more compelling experience. Like I have stated before, I kind of fell into doing videos like this largely by accident. :) It's definitely interesting making them though.

I have been considering doing narration. I don't want to make a guaranties or anything on that at this point though. I consider myself more of a musician, so hence the "silent film" style. ;) But I could definitely have narrations and music. I like the challenge of creating a narrative with just images and sound, but it's also a bit limiting too.

@Cheeses_K_Riced I am fine if you don't like the end product, but it takes quite a bit of work and skill to make something like this. It takes quite a bit of time to develop the art styles I use for my videos. With this video I created the ideas and concepts before generating any of the art. I created 4592 images in the process of making this video. I needed to create that many to get the 50 images I used in this video. I used targeted prompts to get the images I needed to fill each sequence. And that's not even counting the music I made for it. I have been doing AI art for about a year now. If you check the art on my other channel Xinferis, it will be readily apparent that creating AI art is a skill. The art for my earlier videos was horrible. There is a definitely progression and evolution of my style. Every day I learn something new about making art with AI. Describing what I do as "describe a situation vaguely to a machine image generator and just posting the outcome in a video", is like describing an artist drawing as just someone just scribbling on paper. Or it's like describing a musician making music as someone just pressing some keys on a keyboard. There is a lot more to it than that. If you don't like the end product though, I am fine with that. I don't expect everyone, or even most people to like my videos. I just have certain creative visions, and I try to bring them to life the best I can in my videos.

@@Cheeses_K_Riced Which is what a movie director does with their human production designers, cinematographers, etc...

Thank you. I enjoy trying to create an alternate reality like this.

Thank you! I won’t ordinarily be releasing multiple videos a week though. :)

If you look at the comments on my various videos, I give a lot of tips and tricks and outline my process. My process does change and refine over time, but that should give you a starting point. If you look in the comments, I don't give my exact prompts, but that isn't really about not giving away my secret sauce. I try to give people tips so they can use any model and get good results. Plus everyone has different taste, so learning how to develop your own styles with AI is much more useful than just getting some canned prompts that only work with one model and only create the specific style used in a couple of my videos. I really should create a separate channel explaining my process, tips and tricks for making AI art. I already have two RU-vid channels though, and I don't want to go too overboard. If someday I can quit my day job and just do RU-vid, I would likely do it though. :) By the way I read a lot of those same authors, and that's where a lot of my inspiration comes from. :) I would say at this moment Philip K. Dick is my favorite author. The first Horus Hersey book by Dan Abnett blew my mind with how good it was. Not sure if everything in the story would make much sense to someone that isn't a 40K fan though. Thank you for watching!

@redpillcommando One other thing to note though. While it's important to first make sure the prompts you are using come as close as possible to creating the results you want, a lot of images need to be created to make a video like this one. I created 4592 images solely for this video, and only kept around 50 to use for this video. My first pass going through the images got them down to 642 images, before I assembled the video using the final appoximately 50 best and most fitting images. The specific prompts I used got very close to what I was trying to achieve. A lot of images just ended up having something too wonky in them that made them unusable. Other images weren't bad, but they didn't fit the scene in the video I created them for. Ultimately I was looking for the best images that had the most "X factor" and fit well with the video.

@@XinferisTV Cool, the Horus Hersey is already on my "to be read" list. Thanks for the tip. I think I want to try something like this for the Boötes Void. Perhaps, "The Gibbering Horrors that lurk in the Boötes Void".

​@@redpillcommando Sounds like a fun project. :)

Thank you for watching! :)

The video below has a storyline, but it didn't get that many views. I don't think number of views necessarily reflects the quality of a video though. ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-Rl1TdriCMbQ.html

@@XinferisTV Thanks for the link! I've really been enjoying your videos.

Thank you. I am definitely inspired by a lot of classic sci-fi art and movies.

Thank you for watching. I appreciate it!

Thank you for watching!

Thank you!

Thank you!

As always, thank you for watching! :)

Thank you! It was kind of a pain to get this one finished. I was editing it, and Final Cut Pro erased my work right before rendering three times! :)

It is freely available, but youtube won't let me provide the url to it. It is available as an nxtbook. Google that to get it.

I think that AI at some point will just be considered another tool in the toolbox. Just like how we use CG to do effects in movies rather than building models by hand. Or we use video editing software rather than cutting and editing film the old way. At this point AI, although amazing, is extremely primitive. I think people also overstate the effects it will have. It will at some point less people are required to make a certain end product. But it still requires skill and artistic ability to use the tools. Instead of an artist drawing pictures, you might have an artist using AI to assist in creating images. This will apply to many other many other disciplines too. It might make us more efficient, but that will just open to door for us to create more content and experiences. I believe that any tools that make movie making more accessible are a good thing. I am into both video and audio, and for me it's almost like heaven now. Inside of my computer, I have more musical gear by far than even multimillion dollar music studios did back in the day. I also have movie making tools that are powerful enough to make even the highest budget movies (although I definitely don't have the budget to make the content in the movies though :) ). We also have video cameras that are far more affordable than ever that can give world class results. And there are platforms like RU-vid that allow anyone to distribute their videos. Between the tools and the accessibility things have dramatically changed, and it has changed the lives of many creators. More and more people are able to realize their creative visions. I still think that no matter how popular AI and CG becomes though, people will still do things the more traditional way. And I really like that we have that too. Christopher Nolan is a good example, where his movies are still shot on film, and he avoids digital effects as much as possible. I also really enjoy movies with real stunts, whether done by the actor or a stunt double. I have a very large appreciation for movies that are pure like that, where they really do what you see on the screen. I am extremely appreciative of their talent, and love to see it. I do not want to see that go away. This is all my opinion, even though I might state it all like it's fact in the way I write. Everyone has their own opinions, and I respect them. I am definitely passionate about AI technology though, because it allows me to create things I wouldn't be able to create otherwise. I will see if somehow I can get a copy of that magazine and have a read.

@@XinferisTV just google "American Cinematographer nxtbook" open the first one and edit the url to get the proper date that is ac1023

I have been considering doing some videos like this, but narrated. That would be closer to being a movie, but not quite a movie. :)

@@XinferisTV Interesting idea! Or it might be interesting just to provide some tantalising "lore", like occasional names of characters or places that could be woven into a more cohesive story later as things progress. I do enjoy your videos as they are though. The combination of music and images is so evocative.

​@@ian_b I do actually like the restrictions and challenges of only having music, images and a subtitle here and there to tell a story. I did something similar on my other channel Xinferis, where a title, music and a single image is all I had to tell a story, and that was very enjoyable. With that and this form of video, as an artist I can decide what is left up to interpretation and what is revealed to the watcher / listener. WIth narration just by the very nature, it has to be much more heavy handed. For an example of what I did with just a title, image and music, although it wasn't very popular is one of my favorites from my other channel. I leaned heavily on the music to tell the story I had in mind. The instruments selected, how they were played and the progression of the song all tell the story I envisioned. ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-h8eaDsZsE94.html What is kind of interesting is that even though I have been doing my other channel Xinferis for over a year, I had come up with a similar story telling concept before. I had written some unreleased music where I would have 6 tracks, that between the music and their titles would tell a story. I was planning on creating a pretty massive album, a Halloween album with a few of these stories. It had a similar format to the movie Creepshow, or the EC Comics Creepshow was based on. It had intermissions designed in, with little intros and punchlines for the stories. :) I worked on this many, many years ago but never ended up finishing it. I did actually get pretty far through the project though and still have the project files on my hard drive. I think what I am doing here is a more fully realize version of what I worked on back then. So I am contemplating doing something narrative based at some point. If I did, it would still have music though. I do really like the unique experience that can be created with music and images though, so I might just stick with that. :)

Thank you for watching. I really appreciate it.

Hah good one!

Thank you!

Thank you for watching! :)

The Pinwheel Galaxy isn't ALL bad. I personally think they have pretty cool music. ;)

I would argue that we have been at the point where there is more digitally created art than art created by traditional means. I don't do a ton of editing to my images, and mostly focus on creating prompts that generate what I want, and selecting the best images that reflect what I am trying to do. So I am mostly focused on the style, ideas and scenes. Generative AI is evolving quickly to become a more generic tool though. For example Photoshop has generative fill. I have also seen artists take sketches and use Stable Diffusion to create fully realized illustrations from them. So longer term I see it as just another tool in an artist's toolbox. I don't actually think training AI on art that is created with the assistance of AI will actually pose a problem. In the short term, training AI with AI generated data isn't idea, because the results are usually a bit wonky. The technology is advancing rapidly though. We are already at the point where you can create completely new art styles and concepts with it. It would actually make sense to train the AI with that new data. This is of course just my opinion.

Styles meaning what?

Thank you for watching.

Thank you for watching!