Optimal Midi Architecture

[Christoph Hart]

HISE doesn't support MIDI 2.0

I'm living under a rock here. How wide is the DAW industry adoption of this? All of the hosts I'm using don't care at all about MIDI 2.0.

[Bart]

@HISEnberg

Wow COOL!

I think the MIDI Player is whats needed also... it just wants an upgrade into the MIDI 2.0 future

DAW's do in fact purpose Sequencers and Piano Rolls effectively, but I think they're a little lacking in realm of fun easy fast inspirational music composition, so I want to fill the void.

Took a video but the files too large, next time

[HISEnberg]

@Christoph-Hart Lol it's a fairly interesting yet comical story from what I know. Long story short - no one is using it. JUCE announced they will start supporting it soon and there is a handful of controllers on the market that use it.

I think the MMA (not the wrestlers, the Midi Manufacturing Association) has been trying to make it industry standard for quite a number of years now. MIDI 1.0 was only ever adopted because the major music companies of the 80s (Yamaha, Roland, etc.) came together and standardized it. Companies don't want to collaborate like that anymore, so we are stuck in a limbo of MIDI 2.0 always coming soon and never arriving.

So probably something to keep on your radar but not worth the time to investigate just yet. I'd be more interested to know if there are any plans for CLAP and ARA support in the distant future but that's another topic and is also not a huge deal immediately. All that being said is MIDI 2.0 sounds awesome and should be adopted....

[Bart]

@Christoph-Hart said in Optimal Midi Architecture:

I'm living under a rock here. How wide is the DAW industry adoption of this? All of the hosts I'm using don't care at all about MIDI 2.0.

It's only just being rolled out across operating systems and will indeed take some time for DAW's to catch up, although predicting Bitwig will be all over it soon enough.

I guess for me it's always been a fantasy to have more control and resolution, particularly regarding per note automation, next level music... MPE has always been a limited workaround.

Internal synths could could make use of the upgrade

Oh look Bitwig are already in the game.... yas

[David Healey]

@Bart said in Optimal Midi Architecture:

Oh look Bitwig are already in the game.... yas

I don't trust AI "search" results. No DAW yet supports MIDI 2 from what I've seen.

[Bart]

@David-Healey

Rule #1

Dont trust anything AI

[ulrik]

@Bart Hi Bart, my experience with midi is that you can work without the MidiPlayer module, this is a video on one of my projects where you can have several independent sequences running in parallel.
The whole concept is built around the TransportHandler (internal or external clock), synced or independent to a daw clock.

video

So I set the TransportHandler to trigger the clock every 1/64T note (the lowest value available)
and based on that, I can make calculations for each note value for position and length etc.. using this function

TH.setEnableGrid(true, 23); // 1/96
TH.setOnGridChange(true, gridChange);

//  nv = note value (part of a quarter note), wheel = sequence
inline function gridChange(clock, arg2, arg3)
{
	if (!ENABLEPLUGIN.EnablePluginPnl.getValue())
		return;
	local nv = NVLookUpNum[Canvas.getValue().notevalue];
	
	for (wheel in Canvas.getValue().wheels)
	{
		local steps = wheel.steps;
		local length = nv * steps;
		
		local beat = clock % (96*length);
		local idx = Canvas.getValue().wheels.indexOf(wheel);
		
		Canvas.data["beat"+idx] = Math.floor(clock / (96*nv)) % steps;
		
		if (clock % (96*nv) == 0)
		{
			play(wheel, Canvas.data["beat"+idx], clock/96*4);
		}
		
		local cycle = beat / (96*nv*steps);
		Canvas.data["pointer"+idx] = cycle;
		Canvas.changed();
	}
};

The video is a tutorial for setting up key switches you can skip to around 3.00 to see at least 8 independent sequences running

[Bart]

I'm hitting the pillow anyway, guess this projects on hold for now. Sounds like working without the MIDI Player to achieve the objectives might be too much to take on myself atm.

Cheers doods.

[Bart]

@ulrik

Oh wow, thanks ulrik

I'll have a look in the morrow, my eyes are falling out right now.

[Bart]

@ulrik said in Optimal Midi Architecture:

So I set the TransportHandler to trigger the clock every 1/64T note (the lowest value available)
and based on that, I can make calculations for each note value for position and length etc.. using this function

Nice plugin ulrik

I'm likely aiming for a higher PPQ of 960, so not sure it's possible that way..

Chatbot stating: Running a JS loop 960 times per quarter note per sequencer lane will cause hell. And I think you're stating that 96 PPQ is the highest that can go?.

Supposedly the Hise MIDI Player has an internal PPQ resolution of 960 PPQ, handled by sample accurate C++ timing. That's what I need

The MIDI Player could work, but I don't see a way to use it without initially clogging up the plugin with 50 - 100+ MIDI Players, nor finding a way to set them to other MIDI Channels.

Info pointing toward a lot of work to get higher resolution timing without using the MIDI Player..

Stuck, but there must be a way in Hise.