Audiobus: Use your music apps together.

What is Audiobus?Audiobus is an award-winning music app for iPhone and iPad which lets you use your other music apps together. Chain effects on your favourite synth, run the output of apps or Audio Units into an app like GarageBand or Loopy, or select a different audio interface output for each app. Route MIDI between apps — drive a synth from a MIDI sequencer, or add an arpeggiator to your MIDI keyboard — or sync with your external MIDI gear. And control your entire setup from a MIDI controller.

Download on the App Store

Audiobus is the app that makes the rest of your setup better.

Still no multi core support for audio processing on iOS?

124

Comments

  • @skrat said:
    I wouldn’t overestimate the effect of multi-core audio processing on overall performance. I remember the times when audio apps on desktop started adding multicore support and I was expecting some performance boost, but there was actually no noticeable effect - at least for Ableton Live which I was using at the time and had projects where I needed to compromise latency to avoid dropouts and multi-core support didn’t help it anyhow.
    Also I remember many people experiencing actually **better performance with multicore turned off. **
    I believe it got better nowadays, but that’s all boiling down to the optimization on the DAW / app side, not the OS itself. As far as I understand also from this discussion, there’s nothing really preventing you on iOS to do your calculations on separate threads, you only need to send the final result to the CoreAudio API, which live on th main thread (which is the same principle as on the desktop). But it’s a tricky business and currently probably no app developer wanted to invest so much time to experiment with this (and eventually breaking their app).
    The AUv3s may be a different beast - it’s very likely that here lies the current software limitation for multi-threading and this is something that Apple could optimise better. But still, as I originally said, don’t expect miracles, the separation of audio and everything else like GUI, memory and disk operations, background OS and other app services, etc... already does quite a lot.

    As far as I understood, on intel based systems DAWS use separate cores for tracks or groups of tracks and then sum them together. I think it should be implemented like this on iOS.

    I have already seen many projects on iOS that used a lot of effects and synths I wouldn’t dream of just 5-7 years ago on desktop. To me it seems performance is not that big deal. You can’t compare it with pro grade performance oriented desktops, but it’s comparable to mid-level laptops. And many successful producers are using such machines and create music no worse than those with top-notch machines ;)

    I compare my latest iPad Pro 2018 to 5 year old i7 laptop, and the laptop is at least twice as powerful, so I doubt your comparison very much.

  • @vov said:
    Everyone, guys, please stay on topic which is:
    If you know any music apps utilizing multi core audio processing - share.
    If you know why exactly it’s so hard to realize on iOS, is it processors’ architecture, iOS itself, or simply lack of knowhow for a relatively young system - please explain.

    Can you explain to all of us why it's better to have multicore support? Can you show us where it would make a tangible difference in the output of the music that is being created on an iOS device? You're asking for answers, but you have not provided an argument as to why it matters. People are creating pretty much whatever they want on the platform given its limitations. If I spec'd you out the exact same hardware and loaded windows 10 on it and pick your daw, will what you spit out be any better? Since you refuse to post anything you've done either in iOS or Desktop multithreaded, I can only assume you wouldn't know the difference.

    So for us, can you paste a link to some multicore music that you have personally created so we can see the advantages of that in the iOS platform? It would be great if you could name the VSTs or AU's on the desktop that you used so we can verify that they are multicore aware. I switch Diva on multicore on my desktop. That's the only one I use that feature on.

    Since you created the topic, show us why it would matter. Otherwise, you are just shouting into the void.

  • U> @drez said:

    @vov said:
    Everyone, guys, please stay on topic which is:
    If you know any music apps utilizing multi core audio processing - share.
    If you know why exactly it’s so hard to realize on iOS, is it processors’ architecture, iOS itself, or simply lack of knowhow for a relatively young system - please explain.

    Can you explain to all of us why it's better to have multicore support? Can you show us where it would make a tangible difference in the output of the music that is being created on an iOS device? You're asking for answers, but you have not provided an argument as to why it matters. People are creating pretty much whatever they want on the platform given its limitations. If I spec'd you out the exact same hardware and loaded windows 10 on it and pick your daw, will what you spit out be any better? Since you refuse to post anything you've done either in iOS or Desktop multithreaded, I can only assume you wouldn't know the difference.

    So for us, can you paste a link to some multicore music that you have personally created so we can see the advantages of that in the iOS platform? It would be great if you could name the VSTs or AU's on the desktop that you used so we can verify that they are multicore aware. I switch Diva on multicore on my desktop. That's the only one I use that feature on.

    Since you created the topic, show us why it would matter. Otherwise, you are just shouting into the void.

    You’re a very conceited person. Please stop make any posts in my thread. I don’t need your off-topic or personal attacks.

  • @vov said:
    U> @drez said:

    @vov said:
    Everyone, guys, please stay on topic which is:
    If you know any music apps utilizing multi core audio processing - share.
    If you know why exactly it’s so hard to realize on iOS, is it processors’ architecture, iOS itself, or simply lack of knowhow for a relatively young system - please explain.

    Can you explain to all of us why it's better to have multicore support? Can you show us where it would make a tangible difference in the output of the music that is being created on an iOS device? You're asking for answers, but you have not provided an argument as to why it matters. People are creating pretty much whatever they want on the platform given its limitations. If I spec'd you out the exact same hardware and loaded windows 10 on it and pick your daw, will what you spit out be any better? Since you refuse to post anything you've done either in iOS or Desktop multithreaded, I can only assume you wouldn't know the difference.

    So for us, can you paste a link to some multicore music that you have personally created so we can see the advantages of that in the iOS platform? It would be great if you could name the VSTs or AU's on the desktop that you used so we can verify that they are multicore aware. I switch Diva on multicore on my desktop. That's the only one I use that feature on.

    Since you created the topic, show us why it would matter. Otherwise, you are just shouting into the void.

    You’re a very conceited person. Please stop make any posts in my thread. I don’t need your off-topic or personal attacks.

    I’m asking for evidence from you to support your own topic. You are asking why there’s no multithreaded apps. I’m asking you to show me why it matters. It’s not conceited, it’s asking for proof.

    Show us some.

  • wimwim
    edited September 2019

    Why does someone need to provide “proof” for asking a question?

    Granted, this is a useless thread. The OP isn’t going to accomplish anything with it. Apple doesn’t care. The questions have already been answered as best as they’re going to be. No developer is going to significantly change how they’re doing things based on it.

    But why all the push-back for asking the question?

    This thread going on so long because people for some reason I don’t understand are touchy about the subject. Question asked. Question answered. It could be done by now.

  • @drez said:

    @vov said:
    U> @drez said:

    @vov said:
    Everyone, guys, please stay on topic which is:
    If you know any music apps utilizing multi core audio processing - share.
    If you know why exactly it’s so hard to realize on iOS, is it processors’ architecture, iOS itself, or simply lack of knowhow for a relatively young system - please explain.

    Can you explain to all of us why it's better to have multicore support? Can you show us where it would make a tangible difference in the output of the music that is being created on an iOS device? You're asking for answers, but you have not provided an argument as to why it matters. People are creating pretty much whatever they want on the platform given its limitations. If I spec'd you out the exact same hardware and loaded windows 10 on it and pick your daw, will what you spit out be any better? Since you refuse to post anything you've done either in iOS or Desktop multithreaded, I can only assume you wouldn't know the difference.

    So for us, can you paste a link to some multicore music that you have personally created so we can see the advantages of that in the iOS platform? It would be great if you could name the VSTs or AU's on the desktop that you used so we can verify that they are multicore aware. I switch Diva on multicore on my desktop. That's the only one I use that feature on.

    Since you created the topic, show us why it would matter. Otherwise, you are just shouting into the void.

    You’re a very conceited person. Please stop make any posts in my thread. I don’t need your off-topic or personal attacks.

    I’m asking for evidence from you to support your own topic. You are asking why there’s no multithreaded apps. I’m asking you to show me why it matters. It’s not conceited, it’s asking for proof.

    Show us some.

    Agree with @drez (and just about everyone else). What apps are you using that cause crashing and poor performance? You’re telling the doctor you’re sick and it’s his fault but you won’t say exactly what’s wrong.

  • vovvov
    edited September 2019

    @wim said:
    Why does someone need to provide “proof” for asking a question?

    Granted, this is a useless thread. The OP isn’t going to accomplish anything with it. Apple doesn’t care. The questions have already been answered as best as they’re going to be. No developer is going to significantly change how they’re doing things based on it.

    But why all the push-back for asking the question?

    This thread going on so long because people for some reason I don’t understand are touchy about the subject. Question asked. Question answered. It could be done by now.

    Thanks, but unfortunately I still haven’t found out why exactly multithreaded audio processing is so hard to realize on iOS, is it processors’ architecture, iOS itself, or simply lack of knowhow for a relatively young system.
    It can be either of those, but noone seems to know for sure. And how long the situation will continue depends a lot on which of those reasons is the case.

  • edited September 2019

    this is pretty busy thread:-)

    @vov
    I compare my latest iPad Pro 2018 to 5 year old i7 laptop, and the laptop is at least twice as powerful, so I doubt your comparison very much.

    How you compared that ? What methodology you did used ? Because if you want do benchmark to compare two devices / two CPUs you have to use same conditions (same apps, same settings like buffer size). Incan hardly imagine how it will be possible to compare iOS with desktop OS that way.

    Based on Geekbench current iPad models with A12 CPU do have approx same performance like above average dual core i7 2010 cpu

    (Btw Geekbench is VERY complex and relevant benchmark, relative numbers from geekbench between iPad CPU generations A9 - A10 - A11 - A12 are exactly matching my own tests which i did
    with same conditions, same apps on those devices)

    MacBook Pro 2010, i7 dual core 2 (my old macbook)
    https://browser.geekbench.com/macs/59
    single core: 457
    multi core: 1010

    MacBook Pro 15, i7 quad core 2.88 (my current macbook)
    https://browser.geekbench.com/macs/385
    single core: 897
    multi core: 3208

    (note that base on my own tests in Logic, real gain between those two models is very much equivalent to those geekbemch numbers, so they are true expressio of real-world performance)

    iPad Air 1
    https://browser.geekbench.com/ios_devices/30
    single core: 276
    multi core: 522

    iPad Pro 12.9 2019
    https://browser.geekbench.com/ios_devices/58
    single core: 1112 <<<
    multimcore: 4604

    iPad MINI 2019
    https://browser.geekbench.com/ios_devices/59
    single core: 1109 <<<
    multi core: 2773

    Important note : based on tests we did in other thread here at AB, real performance (number of instances of same apps with same buffer size until you hit 80-90% of cpu load) is almost same on 2018 PRO model vs. 2019 MINI/AIR model .

    which is also direct proof that audio apps on iOS devices are NOT utilising multicore performance at all - because both those models have almost same single core performance, just PRO models have more cores)

  • Having cut my teeth in computer audio at the very beginnings of OS X (I recognize more than a few of you from my days moderating osxaudio.com forums) and before that desktop publishing and photography, it never ceases to amaze me the amount of entitlement that follows every new bit of technological achievement.

    Even to this day users expect some kind of infinite performance of desktop workstations, yet I can easily bring a MBP or even Mac Pro to its knees if I do not respect their limitations. There is always a limit. Limits always lead to creative solutions for those that are willing to expend the mental capital necessary to find them.

    It’s a false narrative to point at apples literature regarding performance improvements. They are pretty specific about where the gains actually are. When it’s graphics they point it out. When it’s new wireless connectivity, they tell you. When it’s storage size and access time they tell you. When it’s video or audio performance they tell you. Battery life, display improvements, JavaScript performance for web apps; you get my point.

    The answer is always the same, if the tool that you are using does not fit the need of the job, replace it with the right tool.

  • wimwim
    edited September 2019

    @vov said:

    @wim said:
    Why does someone need to provide “proof” for asking a question?

    Granted, this is a useless thread. The OP isn’t going to accomplish anything with it. Apple doesn’t care. The questions have already been answered as best as they’re going to be. No developer is going to significantly change how they’re doing things based on it.

    But why all the push-back for asking the question?

    This thread going on so long because people for some reason I don’t understand are touchy about the subject. Question asked. Question answered. It could be done by now.

    Thanks, but unfortunately I still haven’t found out why exactly it’s so hard to realize on iOS, is it processors’ architecture, iOS itself, or simply lack of knowhow for a relatively young system.
    It can be either of those, but noone seems to know for sure. And how long the situation will continue depends a lot on which of those reasons is the case.

    I think you’ve gotten all the answers you’re going to get. IOS is what it is. It will certainly incrementally improve, but very slowly. Sorry to be discouraging, but you’re not going to change that by anything you’re trying to accomplish here. My friendly suggestion is to let it go and go have some fun making music within the constraints you have.

    I think that’s what gets under people’s skin here. We’re all having great fun. Most wouldn’t be hanging out on a forum like this if it wasn’t our chosen platform, or part of it. You kind of picked the wrong place if you want a lot of sympathetic ears. :# o:)

  • @dendy said:
    this is pretty busy thread:-)

    @vov
    I compare my latest iPad Pro 2018 to 5 year old i7 laptop, and the laptop is at least twice as powerful, so I doubt your comparison very much.

    How you compared that ? What methodology you used ? Because if you want do benchmark to compare two devices / two CPUs you have to use same
    conditions (same apps, same settings like buffer size). Incan hardly imagine how it will be possible to compare iOS with desktop OS that way.

    Based on Geekbench (which is VERY complex and relevant benchmark, relative numbers from geekbench between iPad CPU generations A9 - A10 - A11 - A12 are exactly matching my own tests which i did
    with same conditions, same apps on those devices)

    Regarding Geedbench:
    MacBook Pro 2010, i7 dual core 2 (my old macbook)
    https://browser.geekbench.com/macs/59
    single core: 457
    multi core: 1010

    MacBook Pro 15, i7 quad core 2.88 (my current macbook)
    https://browser.geekbench.com/macs/385
    single core: 897
    multi core: 3208

    (note that base on my own tests in Logic, real gain between those two models is very much equivalent to those geekbemch numbers, so they are true expressio of real-world performance)

    iPad Air 1
    https://browser.geekbench.com/ios_devices/30
    single core: 276
    multi core: 522

    iPad Pro 12.9 2019
    https://browser.geekbench.com/ios_devices/58
    simgle core: 1112
    multimcore: 4604

    iPad MINI 2019
    https://browser.geekbench.com/ios_devices/59
    single core: 1109
    multi core: 2773

    based on tests we dis in other thread here at AB, real performance (number of instances of same apps with same buffer size until ypu hit 80-90% of cpu load) is almost same on 2018 PRO model vs. 2019 MINI/AIR model - which is also direct proof that audio apps on iOS devices are NOT utilising multicore performance at all

    Where it says that geekbench is relevant to music production?

  • vovvov
    edited September 2019

    I have to repeat myself because some people didn’t understand.
    If you know any music apps utilizing multi core audio processing - share.
    If you know why exactly it’s so hard to realize on iOS, is it processors’ architecture, iOS itself, or simply lack of knowhow for a relatively young system - please explain.
    No need for other stuff in this thread.

  • edited September 2019

    As i said, geekbemch numbers between various models are +/- matching my own comparision tests with real audio apps.. they are matching also with other various tests we did with other users.. based on this i tend to very much believe ti those geekbench numbers, they are expressing relative performance of devices very precisely...

    i have old AIR and MINI 5.. based on geekbench air single core performance is 276 and MINI 5 sc performamce is 1100.. which means in average 4x more performance - which exaclty matches my plugins tests, i'm able to run in average 4x as much plugins on mini than on air

    don't get me wrong, i'm not trying to argue here or to be offensive .. i just wrote some numbers, facts. I spend lot of time by benchmarking and comparing various ipad models and gathering informations from other users

  • @dendy said:
    I said geekbemch numbers between various models +/- matching wmy own comparision tests with real audio apps.. also with other various tests we did alrady here with other users.. based on this i tend to very much believ those geekbench numbers, they are expressing relative performance of devices very precisely..

    i have old AIR and MINI 5.. based on geekbench air single core performance is 276 and MINI 5 sc performamce is 1100.. which means in average 4x more performance - which exaclty matches my plugins tests, i'm able to run in average 4x as much plugins on mini than on air

    don't get me wrong, i'm not trying to argue here or to be offensive .. i just wrote some numbers, facts. I spend lot of time by benchmarking and comparing various ipad models and gathering informations from other users

    I don’t believe geekbench but if you did tests by maxing out your current laptop and iPad Pro to a halt with exactly the same apps and iPad beats laptop, I would be very surprised and really wish to see it, but anyway it’s offtopic.

  • We're trying to make a motorcycle buyer deal with the fact that it can't carry 4 passengers when the same company makes a sedan for more money. If he spends more money he can add a code car but that's only 3.

    Caveat emptor.

    We have coached many on the workflows required to use their iPads for music like freezing audio one track at a time in Cubasis to make something you could just run in a laptop. Anyone interested in results considers it a puzzle to be solved and acts accordingly.

    iPad multi-threading in IOS could break a lot of apps. All these apps have resource requirements and IOS has to manage them for the best user experience.

    If you want to consider a developer that has addressed these issues with the best, proven solution consider NanoStudio2. It can run many, many tracks in parallel without issues. The internal synth has the ability to load and render 3 layers of sampled apps (use SynthJacker to auto-sample any AUv3... IAA seems to be in beta). NOTE: Still waiting for NanoStudio to record and process audio in an update.

    iPads have cores but not all cores are equal in many of these mobile chips. They are often design for some specific system workload.

    iPads have reasonable clock speeds but nothing like the air cooled desktops can offer.

    IOS as an operating system is focused on stability. Most of us are more frustrated with the file system and get our work done with the systems we own. Do AUv3 FX's "crash" in the DAW? Yes. Can we only run 2 copies of the most resource hungry apps like Ravenscroft and only 1 Crudebyte's app? True. I'm just glad we can have those sounds.

    Sell the motorcycle.

    Getting Apple to added 2 more seats is just bad engineering. Then the majority here would be ready to storm Cupertino. Until then your trying to gather facts to support your assumptions that there isn't a reason it can't work like your I7 desktop which had more RAM, faster system buses, more CPU caching, co-processors for math calculations and many other standard features of desktop systems.

    How is the battery life on that I7? A car battery with a voltage inverted probably works for a few hours on the bus.

  • vovvov
    edited September 2019

    @McD said:
    We're trying to make a motorcycle buyer deal with the fact that it can't carry 4 passengers when the same company makes a sedan for more money. If he spends more money he can add a code car but that's only 3.

    Caveat emptor.

    We have coached many on the workflows required to use their iPads for music like freezing audio one track at a time in Cubasis to make something you could just run in a laptop. Anyone interested in results considers it a puzzle to be solved and acts accordingly.

    iPad multi-threading in IOS could break a lot of apps. All these apps have resource requirements and IOS has to manage them for the best user experience.

    If you want to consider a developer that has addressed these issues with the best, proven solution consider NanoStudio2. It can run many, many tracks in parallel without issues. The internal synth has the ability to load and render 3 layers of sampled apps (use SynthJacker to auto-sample any AUv3... IAA seems to be in beta). NOTE: Still waiting for NanoStudio to record and process audio in an update.

    iPads have cores but not all cores are equal in many of these mobile chips. They are often design for some specific system workload.

    iPads have reasonable clock speeds but nothing like the air cooled desktops can offer.

    IOS as an operating system is focused on stability. Most of us are more frustrated with the file system and get our work done with the systems we own. Do AUv3 FX's "crash" in the DAW? Yes. Can we only run 2 copies of the most resource hungry apps like Ravenscroft and only 1 Crudebyte's app? True. I'm just glad we can have those sounds.

    Sell the motorcycle.

    Getting Apple to added 2 more seats is just bad engineering. Then the majority here would be ready to storm Cupertino. Until then your trying to gather facts to support your assumptions that there isn't a reason it can't work like your I7 desktop which had more RAM, faster system buses, more CPU caching, co-processors for math calculations and many other standard features of desktop systems.

    How is the battery life on that I7? A car battery with a voltage inverted probably works for a few hours on the bus

    1.iOS as of now is the most unstable OS I ever used.
    2.Post more off topic if you want to prove how brainwashed you are.

  • @dendy said:
    As i said, geekbemch numbers between various models are +/- matching my own comparision tests with real audio apps.. they are matching also with other various tests we did with other users.. based on this i tend to very much believe ti those geekbench numbers, they are expressing relative performance of devices very precisely...

    i have old AIR and MINI 5.. based on geekbench air single core performance is 276 and MINI 5 sc performamce is 1100.. which means in average 4x more performance - which exaclty matches my plugins tests, i'm able to run in average 4x as much plugins on mini than on air

    don't get me wrong, i'm not trying to argue here or to be offensive .. i just wrote some numbers, facts. I spend lot of time by benchmarking and comparing various ipad models and gathering informations from other users

    What do you think of the new 2019 iPad specs, do you think it’ll have the same issues as the reported A12 crackles?

  • @vov said:
    I have to repeat myself because some people didn’t understand.
    If you know any music apps utilizing multi core audio processing - share.
    If you know why exactly it’s so hard to realize on iOS, is it processors’ architecture, iOS itself, or simply lack of knowhow for a relatively young system - please explain.
    No need for other stuff in this thread.

    It’s clear to me after reading this long thread that unfortunately nobody knows the answer, not even the devs that commented, or at least there’s no consensus at the moment. There was another thread where a couple of devs stated that multicore can not be used for music apps on IOS and one dev that argued it is actually possible, but they didn’t reach any valid conclusion to my understanding.

    I’d also like to know the answer, but I’m afraid you won’t get a definitive answer on this forum.

  • vovvov
    edited September 2019

    @MonzoPro said:

    @dendy said:
    As i said, geekbemch numbers between various models are +/- matching my own comparision tests with real audio apps.. they are matching also with other various tests we did with other users.. based on this i tend to very much believe ti those geekbench numbers, they are expressing relative performance of devices very precisely...

    i have old AIR and MINI 5.. based on geekbench air single core performance is 276 and MINI 5 sc performamce is 1100.. which means in average 4x more performance - which exaclty matches my plugins tests, i'm able to run in average 4x as much plugins on mini than on air

    don't get me wrong, i'm not trying to argue here or to be offensive .. i just wrote some numbers, facts. I spend lot of time by benchmarking and comparing various ipad models and gathering informations from other users

    What do you think of the new 2019 iPad specs, do you think it’ll have the same issues as the reported A12 crackles?

    I wish I could help, I don’t know really, in my case it’s quiet noise and crackles, something you would expect when your processor grinds to a halt. A12x, 4Gb RAM.

  • @Rodolfo said:

    @vov said:
    I have to repeat myself because some people didn’t understand.
    If you know any music apps utilizing multi core audio processing - share.
    If you know why exactly it’s so hard to realize on iOS, is it processors’ architecture, iOS itself, or simply lack of knowhow for a relatively young system - please explain.
    No need for other stuff in this thread.

    It’s clear to me after reading this long thread that unfortunately nobody knows the answer, not even the devs that commented, or at least there’s no consensus at the moment. There was another thread where a couple of devs stated that multicore can not be used for music apps on IOS and one dev that argued it is actually possible, but they didn’t reach any valid conclusion to my understanding.

    I’d also like to know the answer, but I’m afraid you won’t get a definitive answer on this forum.

    Yeah, looks like it. Secrecy at its worth, probably to sell more stuff and avoid refunds.
    But there’re devs who use other systems, hopefully some of them will chime in.

  • edited September 2019

    @vov said:

    @MonzoPro said:

    @dendy said:
    As i said, geekbemch numbers between various models are +/- matching my own comparision tests with real audio apps.. they are matching also with other various tests we did with other users.. based on this i tend to very much believe ti those geekbench numbers, they are expressing relative performance of devices very precisely...

    i have old AIR and MINI 5.. based on geekbench air single core performance is 276 and MINI 5 sc performamce is 1100.. which means in average 4x more performance - which exaclty matches my plugins tests, i'm able to run in average 4x as much plugins on mini than on air

    don't get me wrong, i'm not trying to argue here or to be offensive .. i just wrote some numbers, facts. I spend lot of time by benchmarking and comparing various ipad models and gathering informations from other users

    What do you think of the new 2019 iPad specs, do you think it’ll have the same issues as the reported A12 crackles?

    I wish I could help, I don’t know really, in my case it’s quiet noise and crackles, something you would expect when your processor grinds to a halt. A12x, 4Gb RAM.

    Blimey, that’s surprising, I don’t get any of that with my old Air2 - which is a lot lower spec. I’m looking at the new iPad that’s coming out as a replacement, hopefully it doesn’t have those issues.

  • wimwim
    edited September 2019

    @vov said:

    @MonzoPro said:

    @dendy said:
    As i said, geekbemch numbers between various models are +/- matching my own comparision tests with real audio apps.. they are matching also with other various tests we did with other users.. based on this i tend to very much believe ti those geekbench numbers, they are expressing relative performance of devices very precisely...

    i have old AIR and MINI 5.. based on geekbench air single core performance is 276 and MINI 5 sc performamce is 1100.. which means in average 4x more performance - which exaclty matches my plugins tests, i'm able to run in average 4x as much plugins on mini than on air

    don't get me wrong, i'm not trying to argue here or to be offensive .. i just wrote some numbers, facts. I spend lot of time by benchmarking and comparing various ipad models and gathering informations from other users

    What do you think of the new 2019 iPad specs, do you think it’ll have the same issues as the reported A12 crackles?

    I wish I could help, I don’t know really, in my case it’s quiet noise and crackles, something you would expect when your processor grinds to a halt. A12x, 4Gb RAM.

    Apologies if you’ve been asked this and answered it before, but ...

    Crackling under lighter than expected load is most often related to buffer settings. Lower buffers = higher CPU usage and more crackling (which is caused by buffer overruns). Higher buffer settings = lower CPU usage and less overruns, but higher latency.

    Just for sake of information, what host are you using, and what are the buffer settings? Have you tried increasing the buffers / latency to see if you can get more instances that way?

  • @vov: you aren't going to get more of an answer than you have. The devs who have the knowledge are unlikely to spend more time than they have elucidating the issues. The non-devs may or may not have technical knowledge.

    And in the end, why does it matter? The apps an OS are what they are today.

    Why not focus on figuring out how to make the most of what you have?

  • @vov said:
    iOS as of now is the most unstable OS I ever used.

    I can totally agree and still use IOS devices as the best products for mobile music at this price point and considering the types of music I want to make. I can velcro my iPhone to my guitar and play while taking walks. My iPad sounds better and does more than my $1200 Digital Stage piano. I can sit on the couch and in 30 minutes upload finished guitar generated music than includes MIDI synths to SoundCloud without getting up. The benefits outweigh these frustrations of expectations of scaling.

    You could make music if you have the will to accept and use the best advice of this forum. Calling members trolls for their frustrations with you just makes the case that you should be ignored or at least not poked with opinions that don't serve your agenda.

    There isn't anyone here that can't make something on the iPad crash by overloading the system with parallel work. There's often anger and complaints about Apple's engineering in these cases. Old news, my friend: Misplaced expectations. Move forward and get unstuck.

    Motorcycles fall over and have no airbags. Plan according to your needs. You want safety and the ability to carry passengers safely at high speeds.

    Sell the iPad and accept the lesson of misplaced expectations.

    We are running in circles here but I'm learning a lot so it's time well spent. There's a Rooms! manual!
    I'd love to know more about how IR processing can be batched but that's just because I like to learn how
    machines really work. Not so I can tell every developer to add multi-threading to their stable apps or tell Apple to magically take the existing apps and run them all in parallel without conflict in the next release.

  • edited September 2019

    @vov said:
    1.iOS as of now is the most unstable OS I ever used.

    Someone never used Windows Vista apparently. :D

  • edited September 2019

    Sorry but OP is just :trollface:

    He’s asking the same question over and over even though it’s been answered because he can, not because he cares. It’s been answered that there are ways to do it and have been done by at least one developer in this thread. As he states in the starting post...

    @vow Said

    I read that all the audio processing of all the music apps at any given moment is locked to only one core. As I understand it’s the main reason why all the modern iPads with their powerful processors underperform intel based systems by so much.

    He’s stating that this is the reason why iPads underperform intel even though for some reason he’s thinking that an apple mobile processor is in some way inferior to Intel desktop or even laptop procs simply because they don’t use multithreading I’ve got an A10X proc which came out in 2017. Has the OP even looked at the actual specs of these processors?

    Look at this:
    https://en.m.wikipedia.org/wiki/Apple_A10X

    This processor is in no way close to competing with anything at all in the 2017 Intel processor class that isn’t in the extreme lower end of modern Intel CPU’s. To find a comparable Intel chip, you’ve got to go to the way back machine, which makes no sense. Why would you have go to back several+ years to do a comparison? Because the A10X max’s out at 2.3Ghz on its 3 workhorse chips (Hurricane) and the Zephyr chips are what it uses when it’s saving battery for lightweight processing. These A10X processors just aren’t as beefy.

    OP even confirms this with his other statement:
    @vow Said

    I compare my latest iPad Pro 2018 to 5 year old i7 laptop, and the laptop is at least twice as powerful, so I doubt your comparison very much.

    If you are complaining about performance and are saying that your 5 year old laptop is at least twice as powerful... why didn’t you look at the specs of the iPad you are buying so you’ll know that before you lay the cash down? If you expected one thing and didn’t get it because of reality, then that’s the fault of the buyer not the seller.

    I’ve got a piece of crap 3770K laying around from 2012 that has 4 cores at 3.5GHz that boost to 3.9. That is 7 years old and it’s more powerful from a pure geekbench processing standpoint than the A10X.

    Anyway, to swing this around, the OP doesn’t even seem to consider that iOS audio is actually squeezing an unbelievable amount of audio processing capability on these A1X class chips!

    From OP in this thread:

    @vow said:

    The problem is I top out my latest iPad Pro with only about 5 well processed tracks, not hundreds, without any other apps running. I tried different DAWs for that. So if multicore audio processing were in place on iOS it would give me some breathing space.

    Clearly this is a problem between the OP’s iPad and the chair and nothing to do with multithreading.

    I’ve got an actual song in NS2 right now (not some test tone per track) on my A10X that has the following:

    65 Channels

    Effects Processors:
    220 native NS2 effect processors (EQ’s, filters, reverbs, filters, delays, reverbs, multifx, compressors, etc)
    2 Fabfilter Pro-MB
    3 Fabfilter Pro-Q3
    5 Fabfilter Pro-Q2
    1 Audio Damage Dubstation
    1 Audio Damage Discord 4
    1 Audio Damage EOS2
    1 Audio Damage Grind
    1 Audio Damage Fuzzplus 3
    1 Klevgrand Svep
    1 Klevgrand Korpressor
    2 K-Devices Shaper
    1 FAC Transient

    Sound Modules:
    1 Ravenscroft Piano
    1 SynthMaster
    4 Slates
    33 Obsidian

    At High latency, my CPU bumps up to 75% at playback. Single threaded iOS audio Apps are performing just fine :lol:

    So OP can state how the apple chips underperform Intel chips by so much because of lack of multi threading or such pure speculation, but i’m providing evidence against that in real world use. If he’s gonna ask the question and make a statement, then give some evidence to support the claim. OP asked a question and then claims one is better than an another because of something he read. Show me why that statement is true. If the evidence is my iPad only run 5 channels, then that’s obviously operator error and if they were interested in fixing it, they would. Otherwise...it’s shouting to the void and :trollface: orama.

  • @brambos said:

    @vov said:
    1.iOS as of now is the most unstable OS I ever used.

    Someone never used Windows Vista apparently. :D

    Don’t forget Windows ME (millennium Edition). Windows 2000 wasn’t that great either .

  • I can see value in talking about the technical issues without needing to convince anyone.

    Here's something to consider in the case of multi-core systems.

    CPU's don't grind to a halt. If fact, the do something with every clock cycle. If the Operating system does queue up some instructions to run they just perform "No Op" instructions. They have to run some type of instruction with every clock tick.

    When you get a crackle some core didn't get instructions in time to keep the audio buffer full. Even if everything is running on just one core let alone multi-cores.

    For heat and battery concerns Apple does throttle down the clocks and make some cores idle which will only increase the likelihood you will experience crackling or slow ramp up times for bursty audio events.

    One develop has mentioned he implemented multi-core processing in an IR reverb but also explained that
    if all the cores didn't finish within a precise interval the app would not perform as normal. It might be you can hack a substitute output that doesn't really cause that type of app to crackle. Remember those early internet phone apps that when hit with dropped packets would just loop the most recent one to keep something moving and not suffer the zero to 16,000 jump that makes a speaker coil bounce off it's frame
    with a mechanical click. It's also dangerous to send a fixed number to a speaker (which implies a direct current) since that burns out the coils. Speaker must be in constant movement or completely idle without current flowing through them. And changes in voltage levels should be within specific tolerances.

    Not great fidelity and with Multi-core running rampant it will probably only make the experience less predictable for insuring every core has instructions and meets the timing expectations of every app currently being processed.

    With a single core you essentially remove the potential for timing race conditions between threads and parallel apps and get a more predictable result.

    You don't have to be a programmer to appreciate the complexity involved to understand there isn't a definitive answer for complex systems like IOS, iPad and IOS DAW's and Apps. Test and workaround limitations. It's more like an airplane design that is built for specific needs. There are limits in the design.

    Lear Jet vs 747.

  • wimwim
    edited September 2019

    Humm. Start off by concluding someone is a troll. Then proceed with an epic post feeding said troll? Kinda confused as to the logic here.

    Sorry for the OT post. I’ll shut up now.

  • PadOS13 pretty much improved performamce of music apps, especially those with high memory requirements when running with multiple instances (iSymphonic, Ravencroft, etc)

    check this thread, it is true
    https://forum.audiob.us/discussion/34746/ipados13-beta-many-more-auv3-instances-can-be-run-has-the-ram-limit-per-au-instance-been-amended

Sign In or Register to comment.