Huzak - Yet another music driver

[Arkhan]

Were not really getting alot of musicians in the first place.

And, like ive said a few times now...

Mml is still being used for Mabinogi, often by like, 15 year olds.

¯(°_o)/¯

Sent from my D6708 using Tapatalk

[elmer]

Were not really getting alot of musicians in the first place.

Yep, and that's a bit of a problem for those (few) folks here that want to create PCE homebrew.

And, like ive said a few times now...

Mml is still being used for Mabinogi, often by like, 15 year olds.

Absolutely ... in those countries where Mabinogi is actually still popular.

And are those 15 year olds then turning around and creating lots of MML music for the MSX or the PCE? 


In the meantime, we have the appearance of deflemask, which despite its flaws, is being used by people like Michirin9801 and JIR-O and others to create PCE music right now, and we don't currently have a way to harness their creativity and passion to help PCE homebrew developers.

From my POV, it's just a question of how to solve the perceived problem of lack of musicians.

I can't force people to learn MML.
I can try to write something that allows them to use the tools that they know to create music that folks here can use.

BTW, I'm not trying to start a religious war ... I think that I've said this before ... there is very little difference between the data format that my driver is processing now, and the compiled-MML format that the System Card Player uses.

My driver currently uses 1-byte note (including octave), and 1-byte length in 1/60s increments, and it allows transposing segments of music data (what I called a "sequence", and what deflemask calls a "pattern").

The only difference with the System Card Player's data format (when it is used in "Direct" mode), is that it stores the note without an octave, and keeps that infomation in a separate place.

Not a huge difference in the basic data structure.

While I find it ugly, I'm not really against MML. I'm just not convinced that continuing down that path moves us forwards at all.

Folks can already use MML with Squirrel ... and if you desired to do so, then you should be able to modify Squirrel to spit out something that my driver could be changed to read, quite easily.

[Arkhan]

Protip:  Mabinogi is popular in USA, and people like my high school dropout cousin dick around with MML in it, lol.

The downside is, most of the Mabinogi players don't really do any retro gaming at all.

You might not be trying to start a religious war.  It's just that you sort of constantly make remarks implying that MML is inadequate or obsolete, and it errs on the side of ignorance.   Doing this is precisely what people have been doing for a decade or more with regards to MML, and all it's done is perpetuate the ignorance, and cause people to run away screaming.     

While you and some others may not prefer it, this certainly doesn't mean it's obsolete or inadequate. 

One great perk to using MML is that it's sort of more portable.   I can move a song from MSX to PCE or back again pretty easily.   Or I can turn it back into a MIDI and make it sound like real instruments were involved.  This is a nice side effect.

[Gredler]

As bad ass as MML and Squirrel can be, it really has been difficult finding someone who is a musician that is not only interested in retro games but is also technically proficient enough to create a specific midi that converts cleanly to mml that is then cleaned up for export through squirrel. I have found multiple friends who do deflemask, for whatever reason.

If only the rest of us were on the same level as you Arkhan, but most of us are not as much as I would like to be or have friends that could be. We're dumber than 15 year olds, sorry

[esteban]

Protip:  Mabinogi is popular in USA, and people like my high school dropout cousin dick around with MML in it, lol.

The downside is, most of the Mabinogi players don't really do any retro gaming at all.

Perfect. I look forward to hearing an album's worth of PSG songs from your cousin. I don't care* what genre of music it is, just guide him (so it sounds halfway decent on PCE). Tell him that constraints (PSG PCE) are actually an engine for creativity. He is not shackled. He will benefit from this. So will I.

Thank you, in advance, for agreeing to convert all of his music to PCE.

*Disclaimer: Naturally, there is no need to convert Rush/Pink Floyd covers. It'll be our secret.

[Arkhan]

As bad ass as MML and Squirrel can be, it really has been difficult finding someone who is a musician that is not only interested in retro games but is also technically proficient enough to create a specific midi that converts cleanly to mml that is then cleaned up for export through squirrel. I have found multiple friends who do deflemask, for whatever reason.

If only the rest of us were on the same level as you Arkhan, but most of us are not as much as I would like to be or have friends that could be. We're dumber than 15 year olds, sorry

whoever you have that has issues with Squirrel, have them get in touch with me.   They might just be making one tiny stupid error

[elmer]

You might not be trying to start a religious war.  It's just that you sort of constantly make remarks implying that MML is inadequate or obsolete, and it errs on the side of ignorance.   Doing this is precisely what people have been doing for a decade or more with regards to MML, and all it's done is perpetuate the ignorance, and cause people to run away screaming.     

While you and some others may not prefer it, this certainly doesn't mean it's obsolete or inadequate. 

I will happily admit to being ignorant when it comes to the actual creation of music, and the magic that goes on in a composer's head.

The recent thread about the musician in Japan who is still creating wonderful-sounding stuff with husic shows that, in the right hands, there's nothing inadequate about MML.

It's just another musical notation.


But ... looking purely at the process that's involved, I can only agree with TailChao ...

Funny sound languages are outdated and I'm going to eventually have to part with my own (or hide it under several carpets).

I tend to point out flaws in all sorts of things ... including my own work.

If you want some "balance", then I'm happy to start ragging on some of the things in deflemask, especially now that I'm getting further into exploring its data format, and how I can process it into something that I can use.

For instance, apart from the half-dozen times that deflemask has crashed while I've been poking around with it ... I got to ask ... why doesn't there seem to be any way to transpose sections of music/pattern data?

I thought that was supposed to be a common occurrence in music data.

Without that, I expect that the converted data will be considerably larger than I'd like. 

[Arkhan]

transpose?  that's pretty easy in MML...  you just type like 2 characters.  lol




I suspect that most of the issue with MML is that the people trying to use it are "musicians" in the sense that they can doodle around in a tracker, and play by ear, but the actual sheet music/theory portion of it has eluded them.

It's like when you find someone that kills it with a flute or something, but if you hand them sheet music they go "uh".

If you can't read sheet music, it seems that MML is extra effort.

[spenoza]

I suspect that most of the issue with MML is that the people trying to use it are "musicians" in the sense that they can doodle around in a tracker, and play by ear, but the actual sheet music/theory portion of it has eluded them.

There's also the baggage of time. Various tracker software packages have been free for a while, whereas good DAW packages typically cost money, and not cheap money, and are a bit more complex and fully featured. For those with less proper musical experience, a more functionally limited and non-expert environment is more useful. Also, folks who actually know music AND are comfortable in DAW environments are likely doing something with music that's more interesting to them than composing tunes for the PCE and then dealing with the environmental limitations (waveforms and envelope shaping). It's the non-trained music hobbyists (maybe gamers first, hobbyist non-musician music crafters second) who will have more time or inclination to make PCE tunes.

So basically, Arkhan is backing MML, which is a better format for musicians, who are, annoyingly, less likely to be interested in our homebrew scene. Elmer is trying to accommodate the less professional tracker scene, which has a lot of hobbyist music makers (often without any education or training in music), but is maybe more likely to be coaxed into composing for a homebrew project.

Basically, Arkhan's a (talented) freak. He has some music training and education AND is interested in older technology. Elmer, I think, is a little more of a pragmatist on this front, and I suspect the direction he's going will be more helpful to homebrew developers who are not Arkhan and will need to rope in others to make music for them. But Elmer's approach would not be helpful to Arkhan for developing music for his own titles, so it's a wash. Both systems will work just great for their respective audiences, and having both present in our homebrew scene is the best solution, because it means that more people will be able to find tools better-suited to their own skills or available talent pools. Win-win!

I know if I were to try and get my wife (music therapist, so LOTS of traditional music education and experience) into composing PCE tracks, a tracker-based setup would be a complete bust for her. She would probably do better working in something that could give her MIDI to convert to MML, and thus Arkhan's music path. Trackers wouldn't suite her at all. And then I would have to work with her to craft proper waveforms and envelopes, because she's not a technologist.

[Arkhan]

I dunno, there are free things:
https://routenote.com/blog/the-10-best-free-daws-available/


But, I can't tell how daunting they are to people.   I find that trackers are more hard to get into because to me they've never exactly been intuitive.   I hated using octamed on my Amiga as a kid.

I think a lot of it simply has to do with this strange chain of events everyone seems to follow.

1) I want to make a chiptune
2) This chiptune was made with a tracker.
3) What is a tracker
4) I found a tracker, time to get good at it
5) I got good at it.  It took awhile, and it was annoying.   I'll forget this detail.
6) What's an MML. 
7) No
DAWs are hard. I like trackers.

I think if you swapped parts 3 and 6, you'd get the same effect in reverse.

lol

[spenoza]

Trackers seem to make an amazing amount of sense to non-musicians, and I'm not entirely sure why. It's like people who don't understand music just don't know what to do with a DAW at all. They go cross-eyed staring at it. But you put them in front of a tracker and they're happy to poke away at shit until they have an ear-bending mess. The world is a strange place.

[elmer]

Basically, Arkhan's a (talented) freak. He has some music training and education AND is interested in older technology.

This is the problem, in a nutshell ...

1: You take the folks that learned to play sheet music at school, and who probably played in the band.
2: Take the few that actually still want to keep on playing when they're older.
3: Take the few of those that want to compose instead of just playing in their garage band or cover band.
4: Take the few of those that have a passion for old videogame consoles.
5: Take the few of those that have the technical skills to deal with the limitations of PSG chips.
6: Take the few of those that have the detailed technical desire to deal with waveforms and volume envelopes.
7: Take the few of those that are willing to enter those waveforms and envelopes in a very-specific format in a text editor.
8: Take the few of those that hang around in English-speaking forums that are interested in old consoles that failed badly in the commercial marketplace in the Western World, and are barely known.

... and you've got a very small pool of people to go to.

Both systems will work just great for their respective audiences, and having both present in our homebrew scene is the best solution, because it means that more people will be able to find tools better-suited to their own skills or available talent pools. Win-win!

That's the hope.

Trackers seem to make an amazing amount of sense to non-musicians, and I'm not entirely sure why. It's like people who don't understand music just don't know what to do with a DAW at all. They go cross-eyed staring at it. But you put them in front of a tracker and they're happy to poke away at shit until they have an ear-bending mess. The world is a strange place.

I suspect that it's like most creative-endeavors.

Getting *started* is often the hardest thing, and lowering that hurdle to make it as easy-as-possible to start creating a noise, no matter how horrible that noise is ... at least gives people some quick feedback that they start iterating on to improve it.

Trackers provide a simple-but-rigid framework that give folks somewhere to start.

I'm sure that it can/would feel like a straight-jacket to someone like Arkhan ... but for others, it would allow them to just jump right in.

At the end of the day, it's not the tool that makes the song, it's the musician.

I've heard horrible-sounding MML tracks, and I've heard horrible-sounding Deflemask tracks.

What is actually surprising to me, is that there are a couple of people out there, like Michirin9801 and JIR-O, that have created really excellent-sounding music within the limitations of the Deflemask sound engine, and have worked-out some really creative ways around those limitations.


**********************

For instance, apart from the half-dozen times that deflemask has crashed while I've been poking around with it ... I got to ask ... why doesn't there seem to be any way to transpose sections of music/pattern data?

I thought that was supposed to be a common occurrence in music data.

Without that, I expect that the converted data will be considerably larger than I'd like. 

I've done the 1st tests at converting deflemask pattern data into something semi-efficient that I'll be able to write out and have my sound driver process, and I'm quite happy with the results.

It's definitely bigger than the sort of hand-optimized compiled-MML-descendant that you'd have seen back-in-the-day, but with some careful encoding, it's not bad at all.

Michirin9801's "Misty Blue" cover is using approx 3.5KB of pattern data (for notes, effects and timing).

The largest track that I've seen so far uses approx 5KB of pattern data.

The total usage will get bigger as things like waveforms and effects-envelopes, and samples are added, but I don't see why a complex tune (excluding samples) shouldn't easily fit in less than 8KB.

I don't really see (yet) why the sound driver itself should get much larger than its current 2KB size ... but lets allow for 3KB in order to add the sample-playback code ... hmmm, maybe 4KB max if I add the code for the ADPCM chip in the CD hardware.

BTW ... the driver allows for a complete set of sound effect channels, with per-channel override, so when it's all done, it should be immediately-usable in a game.

[Gredler]

I imagine it's the difference between using a pre-developed engine vs coding your own; modern pre-developed engines are full of limitations but the barrier to entry is very low.

whoever you have that has issues with Squirrel, have them get in touch with me.   They might just be making one tiny stupid error

The dude's so burned out on trying and keeps telling me we should be making it a mobile game with micro transactions. I don't have a friend who is interested in TG16 development, I was just trying to get a favor and convince them to do so.

[touko]

I think the goal,is to have 2/3 drivers, you'll have just to include the desired one as you did with sgx or AC support .

[TailChao]

At the end of the day, it's not the tool that makes the song, it's the musician.

This is a fairly important point which I don't see addressed much.

Creating a game soundtrack is already a significant time investment, making it within the limitations of old hardware can be worse. Sometimes it can be easier to split the job into composition and sound programming with some back and forth between both parties.

Really - it's not whether trackers, sound languages, or whatever driver is best that matters. Just use what works best for your group. You're all lucky enough to be working with a console with a CD upgrade and can just toss all these dumb requirements out the window anyway.