Loopy Pro: Create music, your way.

What is Loopy Pro?Loopy Pro is a powerful, flexible, and intuitive live looper, sampler, clip launcher and DAW for iPhone and iPad. At its core, it allows you to record and layer sounds in real-time to create complex musical arrangements. But it doesn’t stop there—Loopy Pro offers advanced tools to customize your workflow, build dynamic performance setups, and create a seamless connection between instruments, effects, and external gear.

Use it for live looping, sequencing, arranging, mixing, and much more. Whether you're a live performer, a producer, or just experimenting with sound, Loopy Pro helps you take control of your creative process.

Download on the App Store

Loopy Pro is your all-in-one musical toolkit. Try it for free today.

Recording Perfect Loops in AUM How?

124

Comments

  • @j_liljedahl : in case you missed this thread, there is a bug with regards to the length of recordings made when the loop quantum is greater than 0. At 44.1kHz, recordings are too long. How much too long seems to be a function of the number of bars of the recording and at any BPM, the error is constant as a % of a single measure. People are reporting that when the sample rate is 48 kHz that the loops are slightly too short.

  • edited July 2020

    One thing i really love with the latest updates from Lumbeat’s drum apps is the ability to export perfect loops as audio or midi.

  • edited July 2020
    The user and all related content has been deleted.
  • The user and all related content has been deleted.
  • @ReflectiveHaze said:
    I can’t seem to get a slightly short or long recording...according to AudioShare Trim&Fade, they are spot on. I used an empty audio channel as the only channel. I’ve varied sample rate, bit depth, buffer size and sync quantum (1 bar,2,4,8,16 etc. ). iPhone SE 2016, latest app versions, iOS 13.4.1.
    Here’s a screen shot...

    Interesting. What buffer size are you using? I used the same setup (though iSO 13.5) but my loops are all off. I also was using Audioshare to see the length i terms of beats.

  • The user and all related content has been deleted.
  • Glad to hear you’re getting perfect loops @ReflectiveHaze
    There must be a setting I’m missing...or could it be related to os version perhaps?

  • Drag the out point while in Trim/Fade mode and it will snap to a perfect loop point. Save and problem solved. I had a number of loops slightly over. Did a quick trim in AS and then checked them both in Auditor. Results were consistent. The original loop was slightly long. Trimmed loop was right on. I am on 44.1 devices across the board FWIW.

  • The user and all related content has been deleted.
  • edited July 2020

    Have the same problem when recording in AUM using sample rate 44100 Hz. Tried different buffer sizes and bit depths. But when I switch to 48000 Hz - everything is fine and loops are just perfect.

    Using iPad Pro and OS 13.5.1.

  • The user and all related content has been deleted.
  • @drewinnit said:
    Glad to hear you’re getting perfect loops @ReflectiveHaze
    There must be a setting I’m missing...or could it be related to os version perhaps?

    How many measures had you set the quantum to?

  • @anickt said:
    Drag the out point while in Trim/Fade mode and it will snap to a perfect loop point. Save and problem solved. I had a number of loops slightly over. Did a quick trim in AS and then checked them both in Auditor. Results were consistent. The original loop was slightly long. Trimmed loop was right on. I am on 44.1 devices across the board FWIW.

    Sure there is a workaround. It would be even better if one didn't have to use a workaround. I am guessing that once Jonatan knows about this that he'll be able to address it.

  • I upgraded to iPadOS 13.7 in the hope it would help, but sadly I'm still getting imperfect loops when recording AUM channels.

    I find AUM to be the most fun and creative way to produce on iPad. would love to be able to get some clean loops so I can arrange/mix these stems on my laptop

  • @drewinnit said:
    I upgraded to iPadOS 13.7 in the hope it would help, but sadly I'm still getting imperfect loops when recording AUM channels.

    I find AUM to be the most fun and creative way to produce on iPad. would love to be able to get some clean loops so I can arrange/mix these stems on my laptop

    Try these settings. Never had a problem with loops.

  • @jolico: if you read up-thread, you'll see that there is abi not related to the buffer or sample rate. The longer the loop is, the more extra time there is. On short loops it is inconsequential , but in longer loops it has to be trimmed.

  • @espiegel123 said:
    @jolico: if you read up-thread, you'll see that there is abi not related to the buffer or sample rate. The longer the loop is, the more extra time there is. On short loops it is inconsequential , but in longer loops it has to be trimmed.

    Not for me.

    I record short loops, long loops and entire tracks.
    They all start and stop exactly where they should.

  • @jolico said:

    @espiegel123 said:
    @jolico: if you read up-thread, you'll see that there is abi not related to the buffer or sample rate. The longer the loop is, the more extra time there is. On short loops it is inconsequential , but in longer loops it has to be trimmed.

    Not for me.

    I record short loops, long loops and entire tracks.
    They all start and stop exactly where they should.

    After doing some tests, it is related to sample rate. At 48k, a 64 bar recording is precisely the right length. At 44.1 kHz, the recording is slightly too long. The amount by which it is off is related to the number of measures. Tests done on iPad 6 with iOS 13.7

    @j_liljedahl : it seems like the record end time has some calculation that assumes 48k sample rate.

  • edited September 2020

    @espiegel123 said:

    @jolico said:

    @espiegel123 said:
    @jolico: if you read up-thread, you'll see that there is abi not related to the buffer or sample rate. The longer the loop is, the more extra time there is. On short loops it is inconsequential , but in longer loops it has to be trimmed.

    Not for me.

    I record short loops, long loops and entire tracks.
    They all start and stop exactly where they should.

    After doing some tests, it is related to sample rate. At 48k, a 64 bar recording is precisely the right length. At 44.1 kHz, the recording is slightly too long. The amount by which it is off is related to the number of measures. Tests done on iPad 6 with iOS 13.7

    @j_liljedahl : it seems like the record end time has some calculation that assumes 48k sample rate.

    I can't reproduce this. Tried recording a 24 bar recording, sync quantum 1 bar, at both 44.1k and 48k. It loops perfectly. iPad Pro 1st gen running iOS 14.2 beta, but I'm quite sure it worked fine also on iOS 13.

    PS. The calculation uses the actual sample rate, no assumptions.

  • @j_liljedahl said:

    @espiegel123 said:

    @jolico said:

    @espiegel123 said:
    @jolico: if you read up-thread, you'll see that there is abi not related to the buffer or sample rate. The longer the loop is, the more extra time there is. On short loops it is inconsequential , but in longer loops it has to be trimmed.

    Not for me.

    I record short loops, long loops and entire tracks.
    They all start and stop exactly where they should.

    After doing some tests, it is related to sample rate. At 48k, a 64 bar recording is precisely the right length. At 44.1 kHz, the recording is slightly too long. The amount by which it is off is related to the number of measures. Tests done on iPad 6 with iOS 13.7

    @j_liljedahl : it seems like the record end time has some calculation that assumes 48k sample rate.

    I can't reproduce this. Tried recording a 24 bar recording, sync quantum 1 bar, at both 44.1k and 48k. It loops perfectly. iPad Pro 1st gen running iOS 14.2 beta, but I'm quite sure it worked fine also on iOS 13.

    PS. The calculation uses the actual sample rate, no assumptions.

    Try setting the sync quantum higher. My steps:

    • set sync quantum to 64
    • Tempo 120
    • Set sample rate to 44k
    • Rewind to 0
    • Record enable a track
    • Tap record to start
    • tap record again a moment later so that recording will stop at the end of 64 bars
    • Result length: 256.102 beats
    • Set sample rate to 48 k
    • Rewind
    • Tap record to start
    • Tap record a moment later so that recording will stop at end of 64 bars
    • Result. Length precisely 256 beats

    This is 100% reproducible.

    At 44.1k, 8 beats at 44 k yields a loop 16.006 seconds long 32.013 beats at 120 BPM) at 48k, loop is precisely 32 beats and 32 seconds

    This might seem like a small discrepancy but if you leave those slightly long loops going they will drift noticeably.

  • @[Deleted User] said:
    I got a really nice guitar Arp going with ST2 wanna record it as a loop but I can’t get it right. Help please. Thanks!

    OPEN AUDIOBUS
    OPEN LOOPY IN OUTPUT
    OPEN AUM
    OPEN CHANNEL IN AUM
    OPEN OUTPUT OF AUM AND SELECT LOOPY
    PUT SOURCE IN INPUT OF AUM INTO LOOPY OUTPUT (Group the Loop and other Loop apps applicable)

    LOOPING

    ALSO- IF YOU HAVE AN INTERFACE- PUT BLOCS WAVE IN THE INPUT SLOT OF AUM. BLOCS WILL RECORD INCOMING AUDIO IN PERFECT TIME.

  • @espiegel123 said:

    @j_liljedahl said:

    @espiegel123 said:

    @jolico said:

    @espiegel123 said:
    @jolico: if you read up-thread, you'll see that there is abi not related to the buffer or sample rate. The longer the loop is, the more extra time there is. On short loops it is inconsequential , but in longer loops it has to be trimmed.

    Not for me.

    I record short loops, long loops and entire tracks.
    They all start and stop exactly where they should.

    After doing some tests, it is related to sample rate. At 48k, a 64 bar recording is precisely the right length. At 44.1 kHz, the recording is slightly too long. The amount by which it is off is related to the number of measures. Tests done on iPad 6 with iOS 13.7

    @j_liljedahl : it seems like the record end time has some calculation that assumes 48k sample rate.

    I can't reproduce this. Tried recording a 24 bar recording, sync quantum 1 bar, at both 44.1k and 48k. It loops perfectly. iPad Pro 1st gen running iOS 14.2 beta, but I'm quite sure it worked fine also on iOS 13.

    PS. The calculation uses the actual sample rate, no assumptions.

    Try setting the sync quantum higher. My steps:

    • set sync quantum to 64
    • Tempo 120
    • Set sample rate to 44k
    • Rewind to 0
    • Record enable a track
    • Tap record to start
    • tap record again a moment later so that recording will stop at the end of 64 bars
    • Result length: 256.102 beats
    • Set sample rate to 48 k
    • Rewind
    • Tap record to start
    • Tap record a moment later so that recording will stop at end of 64 bars
    • Result. Length precisely 256 beats

    This is 100% reproducible.

    At 44.1k, 8 beats at 44 k yields a loop 16.006 seconds long 32.013 beats at 120 BPM) at 48k, loop is precisely 32 beats and 32 seconds

    This might seem like a small discrepancy but if you leave those slightly long loops going they will drift noticeably.

    Btw, I just checked and the sync quantum isn’t relevant. I know that this length difference seems trivial but the result is sometimes picking up the beginning of new loop cycle that results in a click (which doesn’t happen at 48k) and that if you loop it for a long time (as in a jam), the loop will become noticeably out of sync.

    16 measure loop at 120bpm In AUM ends up 32.000 seconds (64.000 beats) at 48k and 32.012 secs (64.025 beats) at 44.1k.

    At 44.1k, a 64 measure recording will have a length that is 256.102 beats 2:08.051 secs
    At 48k, 64-measure recording is 256.0 beats: 2:08.00 secs

    If you change the tempo to 110 bpm. Number of beats at 44.1k is 256.103
    At 240 bpm, the number of beats is 256.102 when 44.1k

    Measurement mode and buffer sizes don’t matter

    @j_liljedahl : do you get a different result?

    Running iOS 13.7 on an iPad 6. The results were the same under iOS 12.x

  • @espiegel123 said:

    @espiegel123 said:

    @j_liljedahl said:

    @espiegel123 said:

    @jolico said:

    @espiegel123 said:
    @jolico: if you read up-thread, you'll see that there is abi not related to the buffer or sample rate. The longer the loop is, the more extra time there is. On short loops it is inconsequential , but in longer loops it has to be trimmed.

    Not for me.

    I record short loops, long loops and entire tracks.
    They all start and stop exactly where they should.

    After doing some tests, it is related to sample rate. At 48k, a 64 bar recording is precisely the right length. At 44.1 kHz, the recording is slightly too long. The amount by which it is off is related to the number of measures. Tests done on iPad 6 with iOS 13.7

    @j_liljedahl : it seems like the record end time has some calculation that assumes 48k sample rate.

    I can't reproduce this. Tried recording a 24 bar recording, sync quantum 1 bar, at both 44.1k and 48k. It loops perfectly. iPad Pro 1st gen running iOS 14.2 beta, but I'm quite sure it worked fine also on iOS 13.

    PS. The calculation uses the actual sample rate, no assumptions.

    Try setting the sync quantum higher. My steps:

    • set sync quantum to 64
    • Tempo 120
    • Set sample rate to 44k
    • Rewind to 0
    • Record enable a track
    • Tap record to start
    • tap record again a moment later so that recording will stop at the end of 64 bars
    • Result length: 256.102 beats
    • Set sample rate to 48 k
    • Rewind
    • Tap record to start
    • Tap record a moment later so that recording will stop at end of 64 bars
    • Result. Length precisely 256 beats

    This is 100% reproducible.

    At 44.1k, 8 beats at 44 k yields a loop 16.006 seconds long 32.013 beats at 120 BPM) at 48k, loop is precisely 32 beats and 32 seconds

    This might seem like a small discrepancy but if you leave those slightly long loops going they will drift noticeably.

    Btw, I just checked and the sync quantum isn’t relevant. I know that this length difference seems trivial but the result is sometimes picking up the beginning of new loop cycle that results in a click (which doesn’t happen at 48k) and that if you loop it for a long time (as in a jam), the loop will become noticeably out of sync.

    16 measure loop at 120bpm In AUM ends up 32.000 seconds (64.000 beats) at 48k and 32.012 secs (64.025 beats) at 44.1k.

    At 44.1k, a 64 measure recording will have a length that is 256.102 beats 2:08.051 secs
    At 48k, 64-measure recording is 256.0 beats: 2:08.00 secs

    If you change the tempo to 110 bpm. Number of beats at 44.1k is 256.103
    At 240 bpm, the number of beats is 256.102 when 44.1k

    Measurement mode and buffer sizes don’t matter

    @j_liljedahl : do you get a different result?

    Running iOS 13.7 on an iPad 6. The results were the same under iOS 12.x

    I tried a sync quant of 16 and recorded one cycle:

    44.1 kHz resulted in a file of 1 411 197 frames, which is 1411197/44100 = 31.9999319728 seconds.
    Expected number of frames would be 1 411 200, so it's missing 3 samples there. Not as big difference as you got, and I don't hear it when looping.

    48 kHz gave exactly the expected 1 536 000 frames.

    I'll look into if I can improve this, but I have a vague memory that it wasn't entirely easy, and it might be because there is slight jitter from the clock coming from Ableton Link (which is used all the time even if it's not currently enabled and syncing).

  • @j_liljedahl said:

    @espiegel123 said:

    @espiegel123 said:

    @j_liljedahl said:

    @espiegel123 said:

    @jolico said:

    @espiegel123 said:
    @jolico: if you read up-thread, you'll see that there is abi not related to the buffer or sample rate. The longer the loop is, the more extra time there is. On short loops it is inconsequential , but in longer loops it has to be trimmed.

    Not for me.

    I record short loops, long loops and entire tracks.
    They all start and stop exactly where they should.

    After doing some tests, it is related to sample rate. At 48k, a 64 bar recording is precisely the right length. At 44.1 kHz, the recording is slightly too long. The amount by which it is off is related to the number of measures. Tests done on iPad 6 with iOS 13.7

    @j_liljedahl : it seems like the record end time has some calculation that assumes 48k sample rate.

    I can't reproduce this. Tried recording a 24 bar recording, sync quantum 1 bar, at both 44.1k and 48k. It loops perfectly. iPad Pro 1st gen running iOS 14.2 beta, but I'm quite sure it worked fine also on iOS 13.

    PS. The calculation uses the actual sample rate, no assumptions.

    Try setting the sync quantum higher. My steps:

    • set sync quantum to 64
    • Tempo 120
    • Set sample rate to 44k
    • Rewind to 0
    • Record enable a track
    • Tap record to start
    • tap record again a moment later so that recording will stop at the end of 64 bars
    • Result length: 256.102 beats
    • Set sample rate to 48 k
    • Rewind
    • Tap record to start
    • Tap record a moment later so that recording will stop at end of 64 bars
    • Result. Length precisely 256 beats

    This is 100% reproducible.

    At 44.1k, 8 beats at 44 k yields a loop 16.006 seconds long 32.013 beats at 120 BPM) at 48k, loop is precisely 32 beats and 32 seconds

    This might seem like a small discrepancy but if you leave those slightly long loops going they will drift noticeably.

    Btw, I just checked and the sync quantum isn’t relevant. I know that this length difference seems trivial but the result is sometimes picking up the beginning of new loop cycle that results in a click (which doesn’t happen at 48k) and that if you loop it for a long time (as in a jam), the loop will become noticeably out of sync.

    16 measure loop at 120bpm In AUM ends up 32.000 seconds (64.000 beats) at 48k and 32.012 secs (64.025 beats) at 44.1k.

    At 44.1k, a 64 measure recording will have a length that is 256.102 beats 2:08.051 secs
    At 48k, 64-measure recording is 256.0 beats: 2:08.00 secs

    If you change the tempo to 110 bpm. Number of beats at 44.1k is 256.103
    At 240 bpm, the number of beats is 256.102 when 44.1k

    Measurement mode and buffer sizes don’t matter

    @j_liljedahl : do you get a different result?

    Running iOS 13.7 on an iPad 6. The results were the same under iOS 12.x

    I tried a sync quant of 16 and recorded one cycle:

    44.1 kHz resulted in a file of 1 411 197 frames, which is 1411197/44100 = 31.9999319728 seconds.
    Expected number of frames would be 1 411 200, so it's missing 3 samples there. Not as big difference as you got, and I don't hear it when looping.

    48 kHz gave exactly the expected 1 536 000 frames.

    I'll look into if I can improve this, but I have a vague memory that it wasn't entirely easy, and it might be because there is slight jitter from the clock coming from Ableton Link (which is used all the time even if it's not currently enabled and syncing).

    There is something odd here. We are getting different file lengths when using identical params.

    At 44.1k, 120 bpm, sync quant 16, my files are always 32.012 seconds long and 1,411,765 sample frames. I get the same result independent of whether or not Ableton link is enabled.

    So, my files are more than 500 frames longer than yours with the same parameters.

    When this was first reported, someone else was getting the same numbers as I was. And someone else was getting expected numbers.

    This is strange.

  • @espiegel123 said:

    @j_liljedahl said:

    @espiegel123 said:

    @espiegel123 said:

    @j_liljedahl said:

    @espiegel123 said:

    @jolico said:

    @espiegel123 said:
    @jolico: if you read up-thread, you'll see that there is abi not related to the buffer or sample rate. The longer the loop is, the more extra time there is. On short loops it is inconsequential , but in longer loops it has to be trimmed.

    Not for me.

    I record short loops, long loops and entire tracks.
    They all start and stop exactly where they should.

    After doing some tests, it is related to sample rate. At 48k, a 64 bar recording is precisely the right length. At 44.1 kHz, the recording is slightly too long. The amount by which it is off is related to the number of measures. Tests done on iPad 6 with iOS 13.7

    @j_liljedahl : it seems like the record end time has some calculation that assumes 48k sample rate.

    I can't reproduce this. Tried recording a 24 bar recording, sync quantum 1 bar, at both 44.1k and 48k. It loops perfectly. iPad Pro 1st gen running iOS 14.2 beta, but I'm quite sure it worked fine also on iOS 13.

    PS. The calculation uses the actual sample rate, no assumptions.

    Try setting the sync quantum higher. My steps:

    • set sync quantum to 64
    • Tempo 120
    • Set sample rate to 44k
    • Rewind to 0
    • Record enable a track
    • Tap record to start
    • tap record again a moment later so that recording will stop at the end of 64 bars
    • Result length: 256.102 beats
    • Set sample rate to 48 k
    • Rewind
    • Tap record to start
    • Tap record a moment later so that recording will stop at end of 64 bars
    • Result. Length precisely 256 beats

    This is 100% reproducible.

    At 44.1k, 8 beats at 44 k yields a loop 16.006 seconds long 32.013 beats at 120 BPM) at 48k, loop is precisely 32 beats and 32 seconds

    This might seem like a small discrepancy but if you leave those slightly long loops going they will drift noticeably.

    Btw, I just checked and the sync quantum isn’t relevant. I know that this length difference seems trivial but the result is sometimes picking up the beginning of new loop cycle that results in a click (which doesn’t happen at 48k) and that if you loop it for a long time (as in a jam), the loop will become noticeably out of sync.

    16 measure loop at 120bpm In AUM ends up 32.000 seconds (64.000 beats) at 48k and 32.012 secs (64.025 beats) at 44.1k.

    At 44.1k, a 64 measure recording will have a length that is 256.102 beats 2:08.051 secs
    At 48k, 64-measure recording is 256.0 beats: 2:08.00 secs

    If you change the tempo to 110 bpm. Number of beats at 44.1k is 256.103
    At 240 bpm, the number of beats is 256.102 when 44.1k

    Measurement mode and buffer sizes don’t matter

    @j_liljedahl : do you get a different result?

    Running iOS 13.7 on an iPad 6. The results were the same under iOS 12.x

    I tried a sync quant of 16 and recorded one cycle:

    44.1 kHz resulted in a file of 1 411 197 frames, which is 1411197/44100 = 31.9999319728 seconds.
    Expected number of frames would be 1 411 200, so it's missing 3 samples there. Not as big difference as you got, and I don't hear it when looping.

    48 kHz gave exactly the expected 1 536 000 frames.

    I'll look into if I can improve this, but I have a vague memory that it wasn't entirely easy, and it might be because there is slight jitter from the clock coming from Ableton Link (which is used all the time even if it's not currently enabled and syncing).

    There is something odd here. We are getting different file lengths when using identical params.

    At 44.1k, 120 bpm, sync quant 16, my files are always 32.012 seconds long and 1,411,765 sample frames. I get the same result independent of whether or not Ableton link is enabled.

    So, my files are more than 500 frames longer than yours with the same parameters.

    When this was first reported, someone else was getting the same numbers as I was. And someone else was getting expected numbers.

    This is strange.

    While I was developing the first version of AUM I noticed that some device models had much more jitter between the audio timestamps mSampletime and mHosttime (IIRC) than other models. So this might be related. Even though we use the same software and parameters, you're probably on a different device model? I have an iPad Pro 12.9" 1st gen. I'll do the same test on my iPhone 7 as well and report back.

  • @j_liljedahl said:

    @espiegel123 said:

    @j_liljedahl said:

    @espiegel123 said:

    @espiegel123 said:

    @j_liljedahl said:

    @espiegel123 said:

    @jolico said:

    @espiegel123 said:
    @jolico: if you read up-thread, you'll see that there is abi not related to the buffer or sample rate. The longer the loop is, the more extra time there is. On short loops it is inconsequential , but in longer loops it has to be trimmed.

    Not for me.

    I record short loops, long loops and entire tracks.
    They all start and stop exactly where they should.

    After doing some tests, it is related to sample rate. At 48k, a 64 bar recording is precisely the right length. At 44.1 kHz, the recording is slightly too long. The amount by which it is off is related to the number of measures. Tests done on iPad 6 with iOS 13.7

    @j_liljedahl : it seems like the record end time has some calculation that assumes 48k sample rate.

    I can't reproduce this. Tried recording a 24 bar recording, sync quantum 1 bar, at both 44.1k and 48k. It loops perfectly. iPad Pro 1st gen running iOS 14.2 beta, but I'm quite sure it worked fine also on iOS 13.

    PS. The calculation uses the actual sample rate, no assumptions.

    Try setting the sync quantum higher. My steps:

    • set sync quantum to 64
    • Tempo 120
    • Set sample rate to 44k
    • Rewind to 0
    • Record enable a track
    • Tap record to start
    • tap record again a moment later so that recording will stop at the end of 64 bars
    • Result length: 256.102 beats
    • Set sample rate to 48 k
    • Rewind
    • Tap record to start
    • Tap record a moment later so that recording will stop at end of 64 bars
    • Result. Length precisely 256 beats

    This is 100% reproducible.

    At 44.1k, 8 beats at 44 k yields a loop 16.006 seconds long 32.013 beats at 120 BPM) at 48k, loop is precisely 32 beats and 32 seconds

    This might seem like a small discrepancy but if you leave those slightly long loops going they will drift noticeably.

    Btw, I just checked and the sync quantum isn’t relevant. I know that this length difference seems trivial but the result is sometimes picking up the beginning of new loop cycle that results in a click (which doesn’t happen at 48k) and that if you loop it for a long time (as in a jam), the loop will become noticeably out of sync.

    16 measure loop at 120bpm In AUM ends up 32.000 seconds (64.000 beats) at 48k and 32.012 secs (64.025 beats) at 44.1k.

    At 44.1k, a 64 measure recording will have a length that is 256.102 beats 2:08.051 secs
    At 48k, 64-measure recording is 256.0 beats: 2:08.00 secs

    If you change the tempo to 110 bpm. Number of beats at 44.1k is 256.103
    At 240 bpm, the number of beats is 256.102 when 44.1k

    Measurement mode and buffer sizes don’t matter

    @j_liljedahl : do you get a different result?

    Running iOS 13.7 on an iPad 6. The results were the same under iOS 12.x

    I tried a sync quant of 16 and recorded one cycle:

    44.1 kHz resulted in a file of 1 411 197 frames, which is 1411197/44100 = 31.9999319728 seconds.
    Expected number of frames would be 1 411 200, so it's missing 3 samples there. Not as big difference as you got, and I don't hear it when looping.

    48 kHz gave exactly the expected 1 536 000 frames.

    I'll look into if I can improve this, but I have a vague memory that it wasn't entirely easy, and it might be because there is slight jitter from the clock coming from Ableton Link (which is used all the time even if it's not currently enabled and syncing).

    There is something odd here. We are getting different file lengths when using identical params.

    At 44.1k, 120 bpm, sync quant 16, my files are always 32.012 seconds long and 1,411,765 sample frames. I get the same result independent of whether or not Ableton link is enabled.

    So, my files are more than 500 frames longer than yours with the same parameters.

    When this was first reported, someone else was getting the same numbers as I was. And someone else was getting expected numbers.

    This is strange.

    While I was developing the first version of AUM I noticed that some device models had much more jitter between the audio timestamps mSampletime and mHosttime (IIRC) than other models. So this might be related. Even though we use the same software and parameters, you're probably on a different device model? I have an iPad Pro 12.9" 1st gen. I'll do the same test on my iPhone 7 as well and report back.

    iPad gen 6 here. One notable thing is that the discrepancy only happens at 44.1 kHz.

    And the number of samples recorded has been the same with every test which have been done a few different times over a few months.

  • @j_liljedahl said:

    @espiegel123 said:

    @j_liljedahl said:

    @espiegel123 said:

    @espiegel123 said:

    @j_liljedahl said:

    @espiegel123 said:

    @jolico said:

    @espiegel123 said:
    @jolico: if you read up-thread, you'll see that there is abi not related to the buffer or sample rate. The longer the loop is, the more extra time there is. On short loops it is inconsequential , but in longer loops it has to be trimmed.

    Not for me.

    I record short loops, long loops and entire tracks.
    They all start and stop exactly where they should.

    After doing some tests, it is related to sample rate. At 48k, a 64 bar recording is precisely the right length. At 44.1 kHz, the recording is slightly too long. The amount by which it is off is related to the number of measures. Tests done on iPad 6 with iOS 13.7

    @j_liljedahl : it seems like the record end time has some calculation that assumes 48k sample rate.

    I can't reproduce this. Tried recording a 24 bar recording, sync quantum 1 bar, at both 44.1k and 48k. It loops perfectly. iPad Pro 1st gen running iOS 14.2 beta, but I'm quite sure it worked fine also on iOS 13.

    PS. The calculation uses the actual sample rate, no assumptions.

    Try setting the sync quantum higher. My steps:

    • set sync quantum to 64
    • Tempo 120
    • Set sample rate to 44k
    • Rewind to 0
    • Record enable a track
    • Tap record to start
    • tap record again a moment later so that recording will stop at the end of 64 bars
    • Result length: 256.102 beats
    • Set sample rate to 48 k
    • Rewind
    • Tap record to start
    • Tap record a moment later so that recording will stop at end of 64 bars
    • Result. Length precisely 256 beats

    This is 100% reproducible.

    At 44.1k, 8 beats at 44 k yields a loop 16.006 seconds long 32.013 beats at 120 BPM) at 48k, loop is precisely 32 beats and 32 seconds

    This might seem like a small discrepancy but if you leave those slightly long loops going they will drift noticeably.

    Btw, I just checked and the sync quantum isn’t relevant. I know that this length difference seems trivial but the result is sometimes picking up the beginning of new loop cycle that results in a click (which doesn’t happen at 48k) and that if you loop it for a long time (as in a jam), the loop will become noticeably out of sync.

    16 measure loop at 120bpm In AUM ends up 32.000 seconds (64.000 beats) at 48k and 32.012 secs (64.025 beats) at 44.1k.

    At 44.1k, a 64 measure recording will have a length that is 256.102 beats 2:08.051 secs
    At 48k, 64-measure recording is 256.0 beats: 2:08.00 secs

    If you change the tempo to 110 bpm. Number of beats at 44.1k is 256.103
    At 240 bpm, the number of beats is 256.102 when 44.1k

    Measurement mode and buffer sizes don’t matter

    @j_liljedahl : do you get a different result?

    Running iOS 13.7 on an iPad 6. The results were the same under iOS 12.x

    I tried a sync quant of 16 and recorded one cycle:

    44.1 kHz resulted in a file of 1 411 197 frames, which is 1411197/44100 = 31.9999319728 seconds.
    Expected number of frames would be 1 411 200, so it's missing 3 samples there. Not as big difference as you got, and I don't hear it when looping.

    48 kHz gave exactly the expected 1 536 000 frames.

    I'll look into if I can improve this, but I have a vague memory that it wasn't entirely easy, and it might be because there is slight jitter from the clock coming from Ableton Link (which is used all the time even if it's not currently enabled and syncing).

    There is something odd here. We are getting different file lengths when using identical params.

    At 44.1k, 120 bpm, sync quant 16, my files are always 32.012 seconds long and 1,411,765 sample frames. I get the same result independent of whether or not Ableton link is enabled.

    So, my files are more than 500 frames longer than yours with the same parameters.

    When this was first reported, someone else was getting the same numbers as I was. And someone else was getting expected numbers.

    This is strange.

    While I was developing the first version of AUM I noticed that some device models had much more jitter between the audio timestamps mSampletime and mHosttime (IIRC) than other models. So this might be related. Even though we use the same software and parameters, you're probably on a different device model? I have an iPad Pro 12.9" 1st gen. I'll do the same test on my iPhone 7 as well and report back.

    Ok, I tried on my iPhone 7 and got exactly the same result as I got on my iPad Pro.

  • edited April 2022

    Ancient thread bump alert. I ran across this issue tonight. With sync quantum set to 1 bar in AUM Im seeing audio loops in AS that are a hair too long. Trimming em up in AS is not the end of the world but I’d love to skip that step if possible. First world problem I know. 🙂

Sign In or Register to comment.