ivop

No, it was a typo. But when I noticed, an hour had already passed Indeed, although process is a bit like unbaking a bread Probably. I think especially modern mod formats can handle all the effects. On the other hand, it has no notion of linked channels. The main problem with RMT is not its editing capabilities (there are some though), but mostly the player is lacking features.

Yes, exectly.

So many forgotten SID sounds in that soundtrack! Sometimes it suddenly "opens up" and ascends the muffled/synth/filter sound that a lot of SID songs are known for. There's so much more possible with the SID chip. Three 16-bit channels.... Edit: Just listened to all the tracks/stages and it is like listening to a new 70's progrock album. Very nice!

On YouTube DerpDerp3001 says: And a theremin Didn't know this song. It entered my top-10 SID tuned immediately. The amount of different timbres is staggering.

Yes, that's exactly what it is. A raw 50Hz register dump. And then try to reverse engineer that As proven, it can be done by hand, but automation is something else all together. Emkay and I have talked about that, but there are just not enough goattracker files available to justify the effort. Or can you point us to an archive of goattracker files? Hadn't seen it yet. Looks nice! But the three colors on the keyboards are just the three SID channels. There's no detection of instruments or anything more sophisticated. But it's a nice visual effect!

As long as the schematic is free, it's all good IMHO. Here's how I have redrawn the 1088XEL schematic in KiCad: https://github.com/ivop/1088xel-kicad Gerbers are a work of art. Specifically when they are routed manually. Whether to share them or not is a personal choice. But if you do, you can be surprised one day, years later, by somebody that sends your gerbers to a Chinese PCB manufacturer

During the first 48 seconds, a full stroke did not occur After that, it's not too bad. I like the break at 1:06! 1:20 and onwards, there's a bunch of patterns that need better "sound design". 1:46 it gets better again, 1:55 the band kicks in. 2:04 sounds pretty nice, although at 2:17 it's a bit out of tune Rumble, rumble 2:43 wake up. From there on, it's a lot better, but hit or miss sometimes.

It turns out to be a lot harder than I expected. One reason for doing this manual conversion was to get more familiar with siddump's output. There are a lot of problems to automate this. 1. Find out how many bytes frames to skip at the begining (0 for noisy pillars, but 4 for freeze). 2. Do a test dump and try to figure out the pattern length (96 for both, although XLent's transcription used only 64 because there were no triplets). 3. Dump with siddump -f4 -n6 -p16 -t117 Freeze.sid , which means skip 4 frames, one (16th) note is 6 frames, 16 notes per frame (1 bar), 117 seconds. So far so good. Now you can clean up the start and end manually (exact pattern lines). After that I want write a script that cuts that file up in pieces of 96 lines. That are all the song lines. Then cut up all the 96 pieces in three files. While creating a new file, check if one similar has already been created. If so, symlink. If not, write to disk. Afterwards, you end up with a bunch of files: songline00_pat00, songline00_pat01, songline_00_pat02 songline01_pat03--->symlink to songline00 etc... Now comes the hard part and that's extracting the right notes. Siddump puts non-notes between (), arpeggios for example, but fails to notice notes that have a percussive hi-hat in front. Best would be to check for each note that has the gate bit set, no matter whether siddump thinks it's a note or not. Arpeggios and glissandos must be done manually during instrument creation. Speaking of instruments, determining to which instrument a note belongs is not that easy, too. Perhaps compare the WF/ADSR/Pul columns between setting and clearing the gate bit?

Sorry, there's not yet a way, other than to type in all the notes, and patterns, and song data manually. That's what I did, and recognize the different instruments and those that start with a hat. You have to know how to read the register dump. Some knowledge about the SID chip's working is necessary. 81 is noise with gate bit, 41 or 21 is note with gate bit, i.e. the start of a new note. And recognize 0370, 0380, 0470, 0490, 0580, and 0590 arpeggios at the drum channel. They are prepended with a hi-hat too. And the zero note of the arp (the 0xx0 arp) is not always necessarily the basis of the chord name, either major or minor. For exampe playing G-2 with the 0580 instrument is a Cm chord (G C D# G ). But you typed in a song with the help of a YouTube video, didn't you? This is the same, but more accurate. It's monkswork But IMHO, in the end, it's pretty satisfying when it all works out.

Proof Of Concept. Siddump conversion of one of my favorite SID tunes. It took six days with varying hours per day, and about 20-25 hours in total, including creating all the instruments. Perhaps I should start my own Pokey music thread, like fragmare, vinscool, emkay, etc...., instead of burying this at page 31 of Nothing Special Freeze.siddump.txt

Perhaps you can implement a priority system. Rectangle redraws of the currently open application get priority over background applications that are visible that want to redraw a rectangle. I believe Windows 95 had such system, that's why it felt way more snappy then X11 or MacOS at the time. It had other issues though

That's what RMT 1.27 patch 6 by AnalMux is for. Thread: https://atariage.com/forums/topic/175878-rmt-patch-6/ Link to RMT 1.27 patch 6 thanks to emkay: https://atariage.com/forums/applications/core/interface/file/attachment.php?id=248525 I used it for my last Noisy Pillars version: https://www.youtube.com/watch?v=ClO2XifvKiA

Altirra. Its pokey emulation is (almost?) spot on.

I suppose it doesn't do custom 1541 loaders as it doesn't have a loopback IEC cable? That might be the reason. Not really a C64 guy

I have seen a demonstration in Maarssen once, and I found it a marvelous device. But ethernet might be a bit out of date now. I suppose mixing AVG cart with SIO cable and FujiNet brings you the same experience.

Actually, the main problem isn't the tracker per sé, but the RMT player. It can't to vibrato or arpeggio's in 16-bit. I can turn on the vibrato on a 16-bit instruments in the tracker, but it's not played. And using a different tracker won't magically transpose the 16-bit not table I think he means the snare drum sound. It's also what ruins it for me. The rest sounds great!

Okay, it can't be done without patching the 16-bit note table a bit further. It only has bass sounds, so you can't do a lead with it. There are not enough octaves, too. We need a better tracker LOL

I did not have the motivation yet to work more on the automation of generating tables, and replaying the tables only, and run that through aubiopitch. Lately I did some RMT music in emkay's thread and not much more. But I have not forgotten about it.

Definitely Edit: although part of the blame lies with the SID musicians. They almost always go for that heavily filtered and modulated sound. I'm sure you can make SID sound more open, less muffled, like the AY/YM and like this pokey version. Thanks I believe I can also do 15kHz bass and drums, and 16-bit lead, with patch 6, but then you can't have vibrato on the lead

Here's a patch 6 version And as usual, an xex-file for playing on real hardware noisy-pillars-t1-ivo-edit-rmt127p6.xex

Does it have a link back to the IEC socket? It seems it doesn't have that, so I doubt an Atari version will have a SIO connector. An adapter for the Ultimate II+, now that would be cool

PIA is pretty good protected I asked for editting rights and have updated the first post, including your image, and a link to your post and the 3D model. Thank you, too! It's always nice to see people pick up projects which I shared with the public. Perhaps you can sell your surplus of boards on AtariAge?

Yeah, those were the days. My first Fender amp (they make amps, too) in the early 90's had a full schematic in the manual. Nowadays manufacturers can't be arsed to even send you a schematic if you ask them for it. Just a buy a new one.

I removed most of the vibrato, but sometimes it is needed to mask being even more out of tune. 8-bit 15kHz is really coarse in the higher registers. But no vibrato at all makes it very bland and the out of tuneness more obvious. I like my previous 64kHz edit more, but it misses the low bass. This edit is better in the low end. I could even double the bass +$01 frequency to get a fatter sound, but I wanted both edits to stick to 3 channels for now. The idea is to eventually combine the best parts of both. That's what the spare channel is for

Okay, here's my 15kHz edit. Still three channels, no fancy instruments I like the bass and drums and did my best on the lead instruments. 8-bit 15kHz is hard to work with, but it shows promise imho. Best would be a combination of both. The 64kHz version with 16-bit 1.77MHz bass. Edit: included xex for playing on real hardware. noisy-pillars-t1-ivo-edit-15khz.xex Edit2: LOL within 8 minutes there was a spam comment on YT