Polaris

Three Souls, Each with Natures Divided

A Grand and Ghastly Meal in Tartarus (Version II): The Glad Dirge of the Three:

Presenting "A Grand Meal in Tartarus." The lyrics are mine and taken, once again, from my novel.

A dramatis personae for The Last Marriage:

As the thread title says. So far, I have generated 7 suitable pieces using the AI Suno, which I have subscribed to because I am very pleased with the results. All the pieces I've generated so far use my lyrics. The music is inspired by my novel The Last Marriage. More will be posted in due time.

To quote myself with regard to the program I've been working on, in a conversation with Google Gemini: "...while the processing is very slow even on a fast computer, the results do not any longer sound like an interesting special effect; they sound smooth, clear, rich, and unobtrusive, which is exactly what I was trying to achieve. The program does what I consider a superior job at automated remixing. If one doesn't want to spend hours, days even, remixing a sound file by hand, this is, I think, the way to go, particularly because you're never going to get the mathematical precision of this program if you do remixing the normal way. Given that the processing is now beginning to get very slow, the results have achieved a more than adequate degree of excellence, adding more features would risk breaking the program's workings, and I am low on realistic ideas on how to improve the program, I think at this point, we can consider the project more or less done for now." All the music links in this thread have been updated with the new sound, and here are a few more links to music:

I made one more improvement to the program, and I've also learned to perfect the settings more or less. The pieces in the above post now reflect this improvement, having been replaced with updated sound files, as do the ones in this post.

I am now on version 23 of my psychoacoustic remixer. Here is a screenshot plus three links to music it has modified.

I was wrong. There was room to improve, plenty of room, and I have done just that. Give this remix a listen:

Version 17 has arrived. I completely changed the logic for how panning constellations decide to orient themselves, using a highly complex and original idea that I came up with, and the result is very lucid and pleasing. I think this is the final stage of the program, actually, as I am finally quite happy with the sound. Below is a short sample, I will be adding more to this topic as I continue the rather slow and tedious process of reuploading all my music.

Here's a screenshot of the current version of the program, with which the music in the post directly above this has been remixed.

Now that I have more or less perfected the processing of adjusting panning, I have turned my attention to adjusting the loudnesses of sound partials. This second aspect of the program is already finished, I think. Here are some results.

A vast improvement. Formerly, below a certain frequency, partials were suddenly restricted to being panned hard center. Now there is a smooth transition rather than a hard cutoff, leading to a much smoother effect.

One of the last problems was that, while each stucture of panning was more or less optimal, the transitions between them were sometimes not. Sometimes, instruments seemed to jump about at random. I have therefore changed the logic behind this. Formerly, the program would flip optimally constructed "panning constellations" (partials with their orientations in stereo space) so that transitions would be maximally smooth. Instead now, they're connected to each other based on structural similarity, with the result that clusters of partials move around in a much more logical and coherent fashion.

Further improved the program by taking into account loudness masking and the fact that the panning of low frequencies cannot be distinguished (these are now panned dead center). Some results may be heard below.