Topic: ARM64 Linux build?

Is there an ARM64 Linux build in the pipe, like with Pianoteq? It would be highly appreciated, provided the Raspberry Pi 4 is enough to handle the load of Organteq. It would simplify building of a standalone organ cabinet.

Re: ARM64 Linux build?

Organteq can actually be way more CPU intensive than Pianoteq, so I'm not sure there'll be a RPi build... There's just not enough juice in RPi for this.

Hard work and guts!

Re: ARM64 Linux build?

Doesn't seem to use much more CPU power on my desktop computer than Pianoteq does. Besides, the Pi4 is much more powerful than the previous versions which could host Pianoteq without much trouble.

Re: ARM64 Linux build?

Just try to assign all stops, then enable Tutti and a few couplers and then play a chord and see that RPi cry and glitch in shame.

Hard work and guts!

Re: ARM64 Linux build?

I'm very interested too for the ARM structure.
In the future Raspberry will be more and more powerful and the processor capacities enlighted so it could be a good choice for integrated solution. Never for tones of samples like Hauptwerk I suppose...

Up !

Re: ARM64 Linux build?

It would be interesting in the long run, if there were an ability to decentralize the processing for Organteq, much like a real organ is decentralized--especially processing of some electric organs which have dedicated hardware boards for each stop which are blended together.  If you could you divide processing for different divisionals either through midi or some processor+network solution, you could build a cluster of machines (or phones/RPis) and be able to design infinitely large organs.

Right now, Organteq seems unable to reach the size of the largest Symphonic Organs of 10,000 - 40,000 pipes on consumer-accessible hardware; whereas, if processing could be clustered in some way (which is theoretically doable already now if you physically route MIDI to three separate high-cpu computers running different divisionals/elements of Organteq on the same license--something definitely worth experimenting with) those extraordinarily large installations become possible with live-performance-friendly latency.

Along the same logic, if ported to ARM and RPis, you could link 5-10 phones or RPis together where any one of them can comfortably handle 5 or so stops, and that would you give you a small to midsize organ in a (fairly) small and versatile container; this could have extensive applications in building organ consoles that are Organteq-enhanced (or dedicated), since hardware of ARM size can be deployed in everything from small keyboards to large multi-manual consoles and even theoretically be able to have Plug-and-play additions made in the future as new ARM/RPi hardware that's faster or more powerful becomes available.  I see this trend and technology only increasing in the next decade.

Currently, there are a lot of 1970s and 1980s electric organs on the market that need extensive replacement of proprietary, dedicated stop simulation hardware which simply isn't on the market anywhere and must be custom-made to allow for any repairs.  By introducing a runs-on-anything sort of hardware solution like Organteq-on-ARM, these professional-quality giant electric organ consoles in concert halls and churches that currently cost $30-60K USD to repair could be repaired (or even built from scratch) at better quality for $4-5K USD or less.

Spotify: https://open.spotify.com/artist/2xHiPcCsm29R12HX4eXd4J
Pianoteq Pro & Organteq 2
Steinways, Grotrian, Steingraeber, Bechstein, Petrof, Blüthner, K2, Karsten, & Kremsegg
Casio GP300

Re: ARM64 Linux build?

tmyoung wrote:

It would be interesting in the long run, if there were an ability to decentralize the processing for Organteq, much like a real organ is decentralized--especially processing of some electric organs which have dedicated hardware boards for each stop which are blended together.  If you could you divide processing for different divisionals either through midi or some processor+network solution, you could build a cluster of machines (or phones/RPis) and be able to design infinitely large organs.

Right now, Organteq seems unable to reach the size of the largest Symphonic Organs of 10,000 - 40,000 pipes on consumer-accessible hardware; whereas, if processing could be clustered in some way (which is theoretically doable already now if you physically route MIDI to three separate high-cpu computers running different divisionals/elements of Organteq on the same license--something definitely worth experimenting with) those extraordinarily large installations become possible with live-performance-friendly latency.

Along the same logic, if ported to ARM and RPis, you could link 5-10 phones or RPis together where any one of them can comfortably handle 5 or so stops, and that would you give you a small to midsize organ in a (fairly) small and versatile container; this could have extensive applications in building organ consoles that are Organteq-enhanced (or dedicated), since hardware of ARM size can be deployed in everything from small keyboards to large multi-manual consoles and even theoretically be able to have Plug-and-play additions made in the future as new ARM/RPi hardware that's faster or more powerful becomes available.  I see this trend and technology only increasing in the next decade.

Currently, there are a lot of 1970s and 1980s electric organs on the market that need extensive replacement of proprietary, dedicated stop simulation hardware which simply isn't on the market anywhere and must be custom-made to allow for any repairs.  By introducing a runs-on-anything sort of hardware solution like Organteq-on-ARM, these professional-quality giant electric organ consoles in concert halls and churches that currently cost $30-60K USD to repair could be repaired (or even built from scratch) at better quality for $4-5K USD or less.

Good idea. The present limit of 10 stops/manual is not enough in my opinion.
It occurs to me that if a Rpi version were available, even with restricted stop capability, an extra Rpi running, for instance, something like mididings, could handle several of them, making a nice flexible system. It should be possible to use up to 16, each set to a different channel, preset with all manuals coupled and maybe just use 3 or 4 stops on each.
It would also offer the benefit of effectively unlimited channels, which in turn removes the electronic nulling effect when two or more similar waveforms move in and out of phase. If sent to different channels the result is AUDIO mixing which, because of random reflections cannot cause the same degree of nulling. (Apps like Hauptwerke and Grand Orgue allow any pipe to be allocated to any channel, which at present Organteq cannot.)
That said, Organteq is very impressive!
jzz

Re: ARM64 Linux build?

jzz wrote:

Good idea. The present limit of 10 stops/manual is not enough in my opinion.
It occurs to me that if a Rpi version were available, even with restricted stop capability, an extra Rpi running, for instance, something like mididings, could handle several of them, making a nice flexible system. It should be possible to use up to 16, each set to a different channel, preset with all manuals coupled and maybe just use 3 or 4 stops on each.
It would also offer the benefit of effectively unlimited channels, which in turn removes the electronic nulling effect when two or more similar waveforms move in and out of phase. If sent to different channels the result is AUDIO mixing which, because of random reflections cannot cause the same degree of nulling. (Apps like Hauptwerke and Grand Orgue allow any pipe to be allocated to any channel, which at present Organteq cannot.)
That said, Organteq is very impressive!
jzz

Has the 8-channel output in 1.5 made it possible to send any pipe to any channel (as long as it's on the division for that channel)?  I wonder if that means we're closer to some of this.

Also, if Nvidia actually acquires ARM, that could also lead to greater interoperability between Organteq/Pianoteq, Graphics Cards, and future/larger ARM processors if Nvidia creates a universal API that can speak to ARM or Graphics Cards interchangeable (which would have all kinds of crazy applications).

Spotify: https://open.spotify.com/artist/2xHiPcCsm29R12HX4eXd4J
Pianoteq Pro & Organteq 2
Steinways, Grotrian, Steingraeber, Bechstein, Petrof, Blüthner, K2, Karsten, & Kremsegg
Casio GP300

Re: ARM64 Linux build?

Sadly the 8 channel feature only allows each manual to feed its own stereo pair, unless I’m missing something.
This is a limited improvement but better than nothing! I'm hoping that a system with the potential ( I believe) of Organteq will eventually match H/W & Grand Orgue for the ability to allocate pipes to speakers.
jzz