any insider news on Pianoteq for iOS?

[miden]

Just wondering if anyone here has contacts or knows of any news regarding the development of Pianoteq for iOS? Apart from the speculation stuff we already have heard about.

 

Or, is it an urban myth and never going to happen?

[ElmerJFudd]

Intel Core 2 Duo / AMD Athlon 64 X2 or higher, 4GB RAM or more recommended

 

iPads and phones definitely have enough processing power to run Pianoteq. But I believe only iPad Pro has 4gb of RAM. That market perhaps is perhaps too small for the investment? And of course, the developer would have to be OK with giving Apple their cut through the App Store. That said, I think Modartt would now be able to offer sales of their additional instrument models directly from their website and not only limited to IAP.

[miden]

Intel Core 2 Duo / AMD Athlon 64 X2 or higher, 4GB RAM or more recommended

 

iPads and phones definitely have enough processing power to run Pianoteq. But I believe only iPad Pro has 4gb of RAM. That market perhaps is perhaps too small for the investment? And of course, the developer would have to be OK with giving Apple their cut through the App Store. That said, I think Modartt would now be able to offer sales of their additional instrument models directly from their website and not only limited to IAP.

 

Yes I tend to agree...although when I heard the rumours, I figured Modartt must have figured a way to reduce the level of CPU "grunt" needed, and perhaps had a few different models for varying iOS devices. But all requiring a minimum of say an iPad 7/8th gen? Just guessing really. And there has been no hints at the forum or website. And with NAMM now postponed to June any pre-release promo space is gone (always online though I guess )

[AnotherScott]

Intel Core 2 Duo / AMD Athlon 64 X2 or higher, 4GB RAM or more recommended

 

iPads and phones definitely have enough processing power to run Pianoteq. But I believe only iPad Pro has 4gb of RAM. That market perhaps is perhaps too small for the investment? And of course, the developer would have to be OK with giving Apple their cut through the App Store.

A few thoughts about those things...

 

RAM requirements are not fixed, they inherently can vary with the OS, most significantly here because what matters to an app isn't really total RAM but available RAM. An iOS device with less than 4 GB total RAM may still have more available RAM than a 4 GB Mac/PC just because the OS itself is smaller and takes up less RAM when loaded.

 

But I'm pretty sure that Apple's developer rules are that you can only segregate your app by device (runs on iPads, iPhones, or both), and by OS, and not by RAM (which isn't even a spec they officially provide). Meaning, if you offer an iPad app that runs on an iPad under iOS 14 or whatever, it is supposed to be able to run on EVERY model that can run that OS, not just ones with certain amounts of memory. The app store provides no other way to permit/prevent someone from downloading an app. I can't remember the name of the app, but there is some app which only works on some models (possibly because of RAM requirements, I don't remember), and they explained on their web site that, even though the app store will let you download the app on a device that can't run it (they have no way to prevent it), it won't run (or only very poorly), and they tell you that if that's what happened to you, your only recourse is to request a refund from Apple. So I guess what I'm saying here is that, while that may kind of be a way around it, it's not "Apple approved" to limit apps based on the RAM in the device, and it's not the user experience Apple wants to provide, where people can download apps that don't work on their devices. So it's not an ideal way to market the product.

 

As for the Apple cut, I don't think that's an issue. B3X kind of proved that there is a market for higher priced apps, and if they have to give 30% to Apple, that's probably about the same that Sweetwater gets to keep for themselves when you buy Pianoteq from them, too.

[Docbop]

Intel Core 2 Duo / AMD Athlon 64 X2 or higher, 4GB RAM or more recommended

 

iPads and phones definitely have enough processing power to run Pianoteq. But I believe only iPad Pro has 4gb of RAM. That market perhaps is perhaps too small for the investment? And of course, the developer would have to be OK with giving Apple their cut through the App Store. That said, I think Modartt would now be able to offer sales of their additional instrument models directly from their website and not only limited to IAP.

 

Yes I tend to agree...although when I heard the rumours, I figured Modartt must have figured a way to reduce the level of CPU "grunt" needed, and perhaps had a few different models for varying iOS devices. But all requiring a minimum of say an iPad 7/8th gen? Just guessing really. And there has been no hints at the forum or website. And with NAMM now postponed to June any pre-release promo space is gone (always online though I guess )

 

 

I just looked and PianoTeq's system requirements are really low it will even run on a tiny Raspberry Pi and they say bare minimum 256MB RAM. So no iOS is about investing the time to port to a new platform. It could be ugly if Pianoteq has a lot of system level code require porting to another chipset.

[ElmerJFudd]

Interesting. However, I think we had some posts a while back of pianoteq running on weak hardware and its performance was not impressive. High polyphony, pedal use causing audio breakup. However, I see here on YouTube a more robust Rasberry 4 Pi B 4gb, 64bit OS+ HiFiBerry DAC+ Pro XLR shared in March 2021.

Performance index of 32 stress tested using Beethoven Sonata No 14 in C# minor.

 

 

[video:youtube]

 

 

Here is the Python script he used to get it running.

https://github.com/youfou/pianoteq-pi

 

It"s a neat idea of pairing this hardware with say a Yamaha P, Roland FP, Kawai ES110.

 

I wonder what else this hardware can handle. If it could also run UVI (or heck, maybe Kontakt). Well now we"re talking. Someone should open a little shop with the Rasberry Pi prebuilt for VST - 64bit OS, DAC and io already installed and working. User just needs to buy and install their VSTs.

 

Korg is using Rasberry Pi's in their Modwave, Opsix and Wavestate. Similar to what they were doing with stock hardware on the Kronos/Nautilus - pairing it up with their own audio hardware.

 

These guys are working on getting something going.

https://elk.audio/

 

Back on topic - obviously Pianoteq and other full 'pro' versions of desktop OS instruments running on iOS is a lot more user friendly for keyboard players.

[analogika]

I just looked and PianoTeq's system requirements are really low it will even run on a tiny Raspberry Pi and they say bare minimum 256MB RAM. So no iOS is about investing the time to port to a new platform. It could be ugly if Pianoteq has a lot of system level code require porting to another chipset.

It's already M1-native as of version 7.1, so porting to ARM is not the issue.

[Reezekeys]

I guess I'm gonna be the debbie downer here. I don't want to downplay or diss the effort and time involved getting PT to run on that Pi. But based on what I'm hearing in this YT video, no pianist I know of would consider "pairing this hardware with say a Yamaha P, Roland FP, Kawai ES110" â imo, those keyboards' piano sounds are way preferable to this example if only for the fact that they're all stereo samples, and the piano in that video is mono. The last mono piano I used extensively was the Roland MKS20 back in the late 1980s. Well, it's 2022 â thirty years later â and a solo monophonic piano ain't happening, sorry â not as anything to replace what even a budget slab can do right now. Add to that the hyper-quantized "performance" that's just grating on the ear â hell, that might be a midifile originally entered in step time. Ugh. Again, this is not a comment on the technical achievement of having piano modeling software running on a small and inexpensive computer platform. That's something impressive for sure â but it's not gonna replace anybody's virtual acoustic piano in a real-world musical situation. For all I know, whoever made that video might agree with me completely. I'm sure the situation will change in the future, and of course this is all "IMO"... Happy New Year!

[Groove On]

Pianoteq needs 256MB RAM (1/4 of a GB). I have Pianoteq running on a Raspberry Pi type computer with 2GB of RAM; total RAM usage is 1.2GB RAM (1GB for Linux, 256MB for Pianoteq). Incidentally, the computer boots from a 16GB microSD card, same card you'd use in a camera, so, I really don't think Apple hardware performance is the reason we haven't seen Pianoteq on iOS.

 

SPECULATION

I think the main issue with porting Pianoteq is needing to re-design the User Interface for iOS mobile devices. I understand that Pianoteq is coded on the JUCE/C++ platform. And since software written in JUCE/C++ is easy to port to other desktops, Modartt can easily support 4 desktop versions (PC, MacOS, Linux and Linux ARM).

 

But those desktop GUI elements don't transfer well to a mobile device. it you try to port to iOS/mobile devices (and JUCE/C++ platform does support this) - you get a tiny version of the Pianoteq desktop GUI, except all those desktop GUI elements which are awkward on a desktop, become ridonkulous on a mobile screen. Think of all those times you've tried to search for a Pianoteq preset in the drop-down menu and had it explode all over (and sometimes off) the screen, now imagine that happening on a mobile screen and you can see they've got some work to do. I honestly think Modartt tried this to see if it works (and it probably did) but quickly figured out what a mess it would be.

 

This is probably why Modartt has been so resistant to an iOS version. It requires a new separate code base to be maintained OR they have to re-write the entire code base so it functions well on both mobile and desktop devices. And I don't want to think of the headache of choosing whether to stay on JUCE/C++ or switch to SWIFT (which works better and is native to MacOS/iOS).

 

Not a lot of fun choices here for Modartt ... but even then, I'm willing to pay the premium to get Pianoteq on iOS, it would be that cool.

[ElmerJFudd]

This makes a lot of sense, as I have a few friends that use JUCE/C++ to maintain their audio apps across platforms (on iOS/Android it"s very common). And with supporting iOS and OSX as an AU, indeed - the GUI is a complete rewrite. Nothing translates in how desktop interface and tablet interface function. Furthermore, going from iPad to iPhone requires a re-design and redraw of the interface just for screen size and shape. Not to mention keeping up with Apple"s rapid OS and hardware update cycles.

 

 

That said, it is doable. But I suspect it would mean hiring (or outsourcing to freelance) an iOS dev that specializes in GUI design or cracking the books and getting their existing devs up to speed on SWIFT. Inconvenient either way. So, the additional income from the App Store may or may not be enough of an incentive.

[Groove On]

I wonder what else this hardware can handle. If it could also run UVI (or heck, maybe Kontakt) ... Someone should open a little shop with the Rasberry Pi prebuilt for VST - 64bit OS, DAC and io already installed and working. User just needs to buy and install their VSTs.

Most commercial audio software developers have shown zero interest in porting their software to Linux; so the number who actually sell anything is disappointingly small (Pianoteq, U-He Synths, Reaper, Bitwig etc.). Somebody will always point out the hundreds of open-source software audio apps, but none of it comes close to the commercial offerings. This has always been the problem with audio on Linux, it quickly comes to a dead-end, so you end up with a dedicated Pianoteq or Reaper box, but not much beyond that.

[Groove On]

... no pianist I know of would consider "pairing this hardware with say a Yamaha P, Roland FP, Kawai ES110" â imo, those keyboards' piano sounds are way preferable to this example if only for the fact that they're all stereo samples, and the piano in that video is mono ... but it's not gonna replace anybody's virtual acoustic piano in a real-world musical situation ...

Reezekeys what's your impression of this Pianoteq performance? He's using a Roland HP605 digital piano.

[video:youtube]

[Reezekeys]

That is a great performance and definitely showcases what Pianoteq can do â then again it seems obvious the software is not running on a Raspberry Pi! This is the retail version running on a Mac or PC, right?

 

It would feel weird for me to play on a piano and hear myself from the audience as this example is done - I would be more comfortable using a player's perspective for the initial performance, then rendering a new performance in post with a different setting. Maybe that's what Phil did here - I haven't read all the comments. It really sounds fantastic and he's obviously an accomplished player.

 

I tried a demo of Pianoteq years ago - it was impressive but had a somewhat "sterile" sound, imo. If I wasn't 100% happy with what I use now I would definitely be checking it out again. Thanks for that post!

[AnotherScott]

Yes, at least based on another video, Phil is running regular Pianoteq on computer, as you'd expect. As for whatever modifications the Pi-user made to get it to run smoothly on a less capable system, you mentioned it being mono, but lots of gigging players play in mono, so for them, the question then becomes, does this sound better than their current slab sounds like in mono.

 

That said, there's no depth to the lower register in either video. And all the streams of fast notes make it hard to get a sense of the quality and "shape" of the decays, which is where most pianos most exhibit their weaknesses.

[Reezekeys]

As for whatever modifications the Pi-user made to get it to run smoothly on a less capable system, you mentioned it being mono, but lots of gigging players play in mono, so for them, the question then becomes, does this sound better than their current slab sounds like in mono.

It's very hard to tell with the example in that video, but the mono sound from Pianoteq on the Pi is not doing anything for me at all - it kinda reminds me of a rompler piano from the 90s like the Roland "Nice Piano" or the Kurz Micropiano, or heaven forbid the Korg M1 piano (OK it's not that bad!). Again - a cool technical achievement, but not a viable sound to gig with. I'll bet that anyone would choose the mono sound of their slab over this. Of course I've been wrong before, and to each his/her own, etc. Anyone reading this thread care to tell us they'd prefer playing that Raspberry Pi Pianoteq over their current piano? Let's hear from you!

That said, there's no depth to the lower register in either video. And all the streams of fast notes make it hard to get a sense of the quality and "shape" of the decays, which is where most pianos most exhibit their weaknesses.

Phil said this in a comment: "It's the NY Steinway very close to the default preset but with the mics moved into audience position before rendering, and a little bass and low-mid cut to EQ." I'm not sure what you mean by "no depth to the lower register" but if you're talking less low freq content than you like, imo it's a very good approximation of what an acoustic piano sounds like playing classical music, when you're in the audience in a concert hall. I agree that the density of the notes might prevent one from evalutating how well the software does with decays. I'm sure there are plenty of other pieces in the classical repertoire that would do this, maybe Phil has posted one of those. Actually, not to toot my own horn but I'd be curious to hear my jazz/braz excursion from my recent "piano shootout" thread rendered on a Pianoteq piano. I deliberately held out some notes longer than I might, to hear how they decay. Maybe that's for another thread though, sorry for going OT.

[Lady Gaia]

Most commercial audio software developers have shown zero interest in porting their software to Linux; so the number who actually sell anything is disappointingly small (Pianoteq, U-He Synths, Reaper, Bitwig etc.)

 

It's not just audio, but a lot of commercial software. Too often, there's not a large enough audience willing to pay for Linux software to make the return on investment worthwhile. Linux is decent at any number of things, but what it's legitimately great at? Giving people options and saving them money. For commercial software the former is a support problem and the latter is a revenue problem.

[ElmerJFudd]

What's the story with Windows on cheap hardware like the Raspberry Pi? Is it too bloated to run real time instruments with high poly, low latency?

[Docbop]

Most commercial audio software developers have shown zero interest in porting their software to Linux; so the number who actually sell anything is disappointingly small (Pianoteq, U-He Synths, Reaper, Bitwig etc.)

 

It's not just audio, but a lot of commercial software. Too often, there's not a large enough audience willing to pay for Linux software to make the return on investment worthwhile. Linux is decent at any number of things, but what it's legitimately great at? Giving people options and saving them money. For commercial software the former is a support problem and the latter is a revenue problem.

 

Linux and other Unix flavors are great for servers. I worked on Unix and Linux a long time before starting to use it on my personal computers. Doing music is what brought me back to using Mac's. On personal Linux there is a lot of FOSS that is very good for doing all the basics uses. I've use Libre Office and other FOSS software on my Macs.

 

A lot of Linux was used in video for big render farms for movies. I lived close to Digital Domain and the Linux user group I was a member of we got permission to use the big projection room for our meetings. They would show us around now and then to see the miniatures and the server farm for movies like Titanic they worked on. It was also great for our meetings we could plug our Linux laptop into their theater projection system and use it for a giant theater sized monitor. I hear that video is still a strong market for Linux, music software just doesn't get much attention from the FOSS community.

[AnotherScott]

What's the story with Windows on cheap hardware like the Raspberry Pi? Is it too bloated to run real time instruments with high poly, low latency?

Full Windows is still Intel-only AFAIK. Their lurches toward ARM may someday amount to something...

[ElmerJFudd]

Article suggests possibility of Win11 on Pi4 or better...

 

https://www.pcgamer.com/windows-11-raspberry-pi

[Theo Verelst]

For commercial software developers it might be better to run on a more advanced (storage, device and module architecture) and potent (multi processing) and better security (users, file systems) system, but it's true it takes another form of professional software development to get your software to work in stanfard ways (UI, device management, audio support type).

 

There is lot of audio sw for Linux, but not the piles of bloated copycat software/plugin stuff, it's not automatically clear what works for which purpose. A lot of windows audio processing will pass as containing bad ideas, on Linux there's technical audio processing software which makes sense at a low level of aggregation without promising grand sound results or pretence. From a global scientific point of view, the technical buildup is sane, and it takes a lot of accurate signal graph buildup before high quality audio can be achieved.

 

The RPi is fun, I have an older one, but no direct match for a powerful PC, which is reflected in it's size, price and power use.

 

T

[Lady Gaia]

Full Windows is still Intel-only AFAIK. Their lurches toward ARM may someday amount to something...

 

I think that's still true for retail releases, but the preview program has had Arm variants available for some time. It runs surprisingly well in a virtualized environment on my M1 MacBook Pro. There have been claims of an exclusivity agreement with Qualcomm regarding official retail releases, but I haven't followed up to see if there's any official news on the subject. In any event, software written for Windows still needs to be compiled for Arm or it runs in an automatically translated form, much like Rosetta on macOS.

[ElmerJFudd]

Apparently there are wrapper/bridge programs for running Windows VSTs on Linux.

However getting involved with it is hit or miss due to how well the VST developers conform to standards (like the Steinberg VST3 spec) and what type of copy protection they are using.

So no easy solutions.

[Reezekeys]

Apparently there are wrapper/bridge programs for running Windows VSTs on Linux.

Isn't that what Muse did with their Receptors? That did not turn out too well, as I remember.

[ElmerJFudd]

Muse learned being a single purpose PC hardware vendor and having to negotiate with every VST developer large and small to allow their software to run on their platform is a nightmare. What Muse users got was a small selection of instruments, with developer limited feature sets, and no desire from developers to support them.

 

I feel developers are leery of iOS as well. They feel the need to be there, they like the idea of the number of iOS users they can reach. But they must work within the limitations of the weakest device in the lineup and don"t want to cannibalize their desktop versions which have higher profit margin even if they don"t sell as many copies.

 

Back on topic, in Pianoteq"s case. Not really a related issue since they have a Linux version. U-he and a few others support Linux as well.

[miden]

thanks Elmer, was wondering how this thread got to be about Linux and Windows???

[Dr Mike Metlay]

[sees new thread on Pianoteq]

 

[watches thread run all over the place with random speculation, getting nowhere]

 

[watches thread wander off through Raspberry Pi, Linux, and Windows, with iOS no longer anywhere to be found]

 

[shakes head wearily and goes to write email to Modartt]

 

[lets people speculate away while waiting for response]

[Dr Mike Metlay]

As of about six hours ago:

 

"Hello Mike,

 

Thank you and Best Wishes for the New Year!

 

Yes, we are working on an iOS version, and we hope to test it soon with a dedicated beta team.

 

There is no release date yet, but it should definitely be out this year if all goes well.

 

Best regards,

[...]"

 

So there you go.

[drawback]

Care to write Tim Cook about the headphone jack iPad thing? :grin:

[miden]

As of about six hours ago:

 

"Hello Mike,

 

Thank you and Best Wishes for the New Year!

 

Yes, we are working on an iOS version, and we hope to test it soon with a dedicated beta team.

 

There is no release date yet, but it should definitely be out this year if all goes well.

 

Best regards,

[...]"

 

So there you go.

 

Yes I got a similar email from them several months ago Mike

 

My OP was simply to ask if anyone had heard any rumours that could illuminate progress. Knowing that any further correspondence with Modartt would result in the same response, previously obtained. Sometimes people in the industry over there hear things. It was merely a bit of a whimsical thread tbh. What followed was NOT expected!