Pianoteq on iPad ? (Page 1 of 2)

So when do you release the iPad edition of Pianoteq :-)

Still waiting..

I don't think iPad's CPU is enough for Pianoteq's processing needs at all.

What do you base that assumption on?  iPads do not appear to be slow..

From the FAQ: What kind of CPU (processor) do I need?
PIANOTEQ is CPU intensive software for it computes the sound in real time. However, most modern computers already offer a fully sufficient CPU. We recommend a CPU with dual or multiple cores, such as the Intel Core 2 Duo or AMD X2. By restricting polyphony or the internal sample rate in the Options menu, you can work with less powerful CPU's.

I'm also thinking it might be sufficient, thus the new A6 that's part of the iPhone and the next iPad seems more than sufficient (see anandtech review of it, pretty interesting with a lot of figures to compare to)

In any case, however this discussion about CPU power is not completely relevant, even though it's interesting. One side or the other, it's only wild guess before moddart tells us it has tested it ....

Clock speed isn't everything.  According to Geekbench, the iPad 3 processor is about half speed of the slowest Intel Core 2 Duo, and with a bit of polyphony reduction, effects reduction, a little more latency, and tweaking here and there it should be possible. It isn't as if Pianoteq usually runs with 100% CPU on a normal desktop. Also it might be able to use the GPU (?).

If it was coded in assembler, that would of course be a problem, but why would they have done that? Most compute intensive applications are written in low level languages like C, and iOS applications are usually written in objective-C, which is just C with classes...

This is all speculation, but until we get a definite no from Pianoteq, I remain quite optimistic :-)

Even if the later iPads have just about enough power, do you really want them running full throttle? These things have no fan, and can get warm, or at least drain the battery rather quickly! That processor power is maybe best used in bursts.

It is a well known problem with Android, and it is why all professional music
apps are made for iOS.

The thing is that there are many, really very very good musical instrument apps
available for iOS. They are quite advanced, and there is NO latency. They are a
joy to use.

Some of these apps are made by world famous instrument manufacturers such as
Moog, Yamaha, Korg, and Akaipro. Got to their websites and look for their
iOS apps and demos.

Just to get a feeling on how far this has gone, and how active this area is,
have a look at these links:

The following two performances by DigiEnsemble Berlin are good examples:
http://www.youtube.com/watch?v=_t9t0kZRJ54
http://www.youtube.com/watch?v=SngPaTYTfqw

The "Sonic Touch"-series here is an almost monthly issue on new developments in
iOS music apps. (Youtube link to a series of videos, press the link below the empty black square, that says video not available): http://www.youtube.com/user/sonicstate/...onic+touch

Websites dedicated to iOS music:
http://www.iosmusician.com
http://iosmusicandyou.com

Two of the vendors dedicated to iOS:
http://www.ikmultimedia.com   (notice their iGrand piano app)
http://beepstreet.com

There is also already a lot of music hardware for iOS devices:
http://www.thomann.de/gb/search_dir.html?sw=iOS

This is why I would like to get my favorite software piano on board before it
is too late. I genuinely believe this is the future of mass market electronic
music.

True, most of the sound engines are based on sampled waveforms, but a lot of signal processing is going on.
Some of the synthesizers are based on several types of synthesis engines. The Alchemy Mobile synth is for example base on a mixture of "VA (virtual analog), sampler, granular and additive resynthesis", and the Impaktor Drum synth, that uses microphone sound as input, is based on physical modeling, frequency modulation, and noise shaping.

True, only they will know if they can find a business case for it.

Let's put it this way - I'd rather want Modartt IMPROVE the complexity of their physical model which would demand more powerful CPUs to provide us even better piano sound, rather than DEGRADE it to barely make it run with underpowered tablet CPUs.

Pianoteq is meant to be played with 88 piano-weighted keys with as much CPU power it can get. Tablets are not good enough for this yet.

I'm afraid that not everything in this world is as simple as just "More Power".

That's especially true in the computing world. Intel and AMD have raced for more Ghz sometimes ago, but that era has now ended (10 years ago at least)

A lot of work in the CPU area is now about efficiency, especially power efficiency. And in this area, Apple seems to be pretty good.

Moreover, if moddart was waiting/needing more CPU power to improve their physical model, they would be wrong, as they have actually plenty of CPU power unexploited in current generation of CPU. See, my Intel Core i5 CPU that's 2 years old is used only at 5% to 10% when playing complex music with latest Pianoteq

By the way, Tablets are already good enough to play Pianoteq. See for instance GeekBench benchmark that measures Integer / Floating point CPU and Memory performance (in other words, performance that is essential to pianoteq) :
- my old intel core i5  scores about 9000
- the apple A6 (iphone 5 / ipad 4) scores about 1600 <= that should be enough power

At last, IMHO, the main point of pianoteq is to be a lightweight software that doesn't need tons of GB of storage and could be brought on stage or elsewhere easily (see pianoteq description :"The size of Pianoteq is just 20 MB (MegaBytes) ... You can use Pianoteq on any modern laptop and connect it to your MIDI compatible piano keyboard"

All in all, i think that the future of pianoteq is not to refine the physical model to use more CPU horsepower ...  well if that was the case, moddart would already have done it ... and again, everything is not about power and definitely the current computing world trend is not in that direction. So I really think that, if moddart wants to exist in the future in any other way than as a niche product for some happy few, they need to take the train of new platforms .. and this is now, in a few year it would be too late as a lot of company with, in my opinion a lot less potential than moddart, have already taken a foot in...

Apple has sold 100 millions iPad to people of this world, and I'm pretty sure some of them are musicians or even pianists too

I couldn't agree more.

I don't have a tablet yet myself but I can understand that people would want to have PT running on a tablet, especially when gigging to have a lightweight easy solution.
It's up to Modartt to see if the expected userbase on tablets would be large enough to make a Pianoteq-TAB version for it... If there's enough potential buyers and the app would run good enough, I think they won't mind degrading a specialized version if that will help getting more sales and thus, more resources to work on improving the desktop version....

cheers
Hans

I am using Pianoteq-4  (pianoteq user from version-3) without problems in pentium-m @1.5Ghz. Basically a Pentium-3 with SSE2. So I am deeply interest to optimizations and low cpu usage.

To my understanding iPad-2 computing power is about the same with my laptop. In reality, if the graphics card can co-compute with the cpu (like OpenCL) I am sure it's more powerful. So, it can be done. The first problem - I suppose - is the extra coding hassle. You need to re-write for arm-based proccessor in a new environment (iOS).

The second problem is portability. I play with my laptop + sound card + midi. The scenario to play with ipad + soundcard + midi is not major advance in portability.

If I could connect my midi keyboard via wifi (in logical price) to 'pianoteq-ed' ipad would be major advance in portability. I prefer via OSC

AFAIK, no.

BTW, for those wanting a portable solution, I'm pretty sure that Pianoteq will work just fine without needing ANY special version on Windows Surface

That's the way to go

Pfffft.

I know we feel diferent about operating system, and thats one reason I think Modart realy should extend compability even to work with iOS. iPad is a lot more portable (and cheaper) than a Macbook...

In terms of existing support, the 3 platforms are all x86 which means the same core code can be deployed with only different UI/sound output output/VST code needed for each platforms.  If we strip out just the core command line renderer, I bet that code would recompile on all Mac/Windows/Linux x86 without any changes (many open source examples -- stuff like FFMPEG are cross-platform compileable).  In theory, pure C++ code written for x86 can be compiled on ARM also.  In practice, you need a lot of #IF x86 #IF ARM code blocks.

The question is whether something like C++ would even compile directly on iOS.  The language in iOS is Objective C.  Google up iOS/C++ and you see people talk about downloading "C++ apps" which is pretty strange -- more like an emulator/sandbox instead of a full compiler.  I mean it's not like you can simply copy your source code to an iPad and then compile ... not without jailbreaking your device first.  Any iOS developers can give us the scoop on this?

Because developing a new version for a completely different OS (by implementing an INFERIOR sounding model because ARM still can't compare to i5/i7 CPUs of the world as far as processing power goes) would take precious time that could be better spent to further improving Pianoteq's modeling, offering even better and more detailed sound.

In the same way you could maintain that Google would have built a far superior searching engine, if they just had kept out of advertising.

Hi

And, suppose, that one day we have Ptq for iPad, but the iPad may not output enough power for your midi controller(say  ”dont recognize the device”). Newer iPads do not give so much power out from usb as older(since they made lightning to usb-cable). Some keyboards can not get enough power from iPad usb. I have one(Roland G-600 arranger). Then what is the solution, if you want to have Ptq?

Yes, recognize that. From hear-say I understand Apple build this limitation in software. There are workarounds (using usb hubs), but this seems to be really something to hold against Apple.

I have used VPC1 with an iPhone 5S connected with an Apple camera connection kit and a powered USB hub.  So iPad should work equally well.  It did not work without the powered USB hub.