Kylearan

Ha, nice! Both undo/redo and especially more playback accuracy had been on my to-do list (along with porting the improved sound emulation from the newer Stella versions), but I found it hard to motivate myself to actually do it... was/am kinda burned out on this project. So I'm really glad somebody did this even though you probably had to endure some bad code of mine!

In TIM1T, I used the ARPM bankswitching macros found here and found them very helpful - you might take a look at them for inspiration or comparison. Just note that the original macros have a small bug - see my post in that thread for details.

The problem you described appears when the current working directory is not the same as the TIATracker executable when starting the program. This can happen for example if you have linked the .ttt file extension to TIATracker in Windows and start the program by double-clicking a .ttt file (which right now TIATracker doesn't support anyway). This is a known (and admittedly quite silly) problem which I haven't found the time to fix yet, sorry! As I wouldn't know how to do that, probably not. I believe there should already be tools to capture WAV from your soundcard.

I tried Stellerator with some of my demos, and I'm quite impressed. I don't have the latest version of Stella installed at the moment so the comparison is unfair, but right now Stellerator looks and sounds more like a VCS connected to a real TV. This is great!

Thanks! Will try that out. Google gave me https://github.com/dpiponi/Stellarator(note the "a" instead of the "e" in the name), where it says "There is no audio yet.." so I was confused.

Excuse my ignorance, but what is Stellerator? It popped up several times in the past weeks, but all I can find is Stellarator, an emulator that according to its GitHub page has no audio...?

For quick sound design, you could always use a TIATracker percussion instrument. The percussion editor lets you define all three register values per frame by easily dragging around with the mouse or entering them manually, with almost no practical length limit, and you can immediately hear what it sounds like by just pressing a key. This allows you to quickly experiment with very fast turnaround times. Downside: It uses Stella's old sound emulation which is not perfect, but for initial experiments and a lot of normal cases, it should be good enough.

As Thomas has said, you'd have to do it multiple times per frame, the more the better. So for a pure music cartridge that would be possible, but for a game or a demo, where you need the CPU cycles for your display kernel, this would be difficult (and more importantly, very specific to the display kernel). And since TIATracker is aimed mainly for music in games or demos, I decided not to support this. Apart from that, a musician told me that changing parameters multiple times per frame makes a big difference for sound quality only if you'd have more parameters like filters anyway, which we don't have on the VCS. I have no idea about POKEY, but a quick Google search tells me there are trackers for that already.

Just thinking out loud here without testing it: The division into three areas can be done by manual comparisons (2 cmp instructions). If you do that first, then you could re-use the result for the division by 3, i.e. you only need a look-up table of size 12 (0, 3, ..., 33) and then add a base value depending in your previous cmp results (+0, +36, +72). This should be very fast and might be light on ROM. EDIT: Scratch that, this is nonsense - I should refrain from "thinking out loud" and think things through before posting instead. EDIT 2: Or maybe not. (I'm tired today...) After the manual comparisons for the div 36, you could subtract 36 or 72 from the sprite position if it's in area 2 or 3, and then use the smaller, 12 byte look-up table for the div 3 and add 12 or 24 again to the result.

Note that PAL vs. NTSC music is not only about speed but also about pitch. TIA Notes have slightly different frequencies and thus sound differently when played with 50Hz vs. 60Hz. It's a small difference, but it might mess with what the composer (or listener) considers to be an acceptable level of "out of tune" for a given song.

I'll PM you his contact info.

Or you could use deep learning. https://deepmind.com/research/publications/playing-atari-deep-reinforcement-learning/

Yes. Glafouk (the musician) really likes it and by now has made over 40 tracks with it.

Obviously it's not impossible in the literal sense. What I meant was that considering the limitations and the sound quality of the TIA, most people would consider it "imposible" to make a music collection in the sense that it woulnd't be worth listening to anyway. I should have used a different word; unlikeley perhaps? Anyway, I hope you get what I mean.

Hi, the demo group Flush did the "impossible" thing and released a 32k music cartridge featuring 32 tracks (about an hour of music!) for the Atari 2600: http://www.pouet.net/prod.php?which=71561 (Joystich left/right changes the track, up/down the presentation) Who would have thought someone would actually do something like this for the TIA?!? <3