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.

AUM with +24dB doesn't distort - comparing Limiters

The user and all related content has been deleted.

Comments

  • Try on audio signal instead of file?

  • edited December 2022

    AUM must have internal safety measures (limiting) at the audio output (and behind the VU meter) that kick in when levels exceed certain thresholds.
    You can try yourself by loading an AUv3 e.g. sending a sawtooth tone at 0dB and slowly increase the gain. At some point, AUM's output volume won't increase anymore although the VU indicator will show the actual signal level increasing.
    Drambo least does hard clipping at the output.

    I'm sure @j_liljedahl can detail on that for us 😊

  • edited December 2022

    iOS itself has quite an array of (obscure -- or is this documented in detail somewhere?) internal processing before the output, which definitely includes some kind of limiter. @tja maybe try with "Measurement Mode" and see if there's a difference?

    For example, if I run a Web Audio app that generates a +6 dBFS sine wave, it will not distort on iOS Safari, but will distort in Chrome on Linux.

    (it makes sense though what iOS is doing -- after all, multiple apps potentially need to have their outputs mixed and iOS can't just say "well we might have up to 3 apps playing sound so we need to mix each app at only 33% just to be safe", because then everything would be too quiet)

  • The user and all related content has been deleted.
  • The user and all related content has been deleted.
  • At System Level Core Audio has a built-in limiter...
    ...and as Core Audio works with 32-bit floats the headroom is practically unlimited.

    @tja, for 'distortion/overdrive' you really should be using saturators or clippers. Try BLEASS Saturator, It's one of my favorites...

  • @rs2000 said:
    AUM must have internal safety measures (limiting) at the audio output (and behind the VU meter) that kick in when levels exceed certain thresholds.
    You can try yourself by loading an AUv3 e.g. sending a sawtooth tone at 0dB and slowly increase the gain. At some point, AUM's output volume won't increase anymore although the VU indicator will show the actual signal level increasing.
    Same in Drambo.

    I'm sure @j_liljedahl can detail on that for us 😊

    Incorrect. AUM does not do any internal clipping or limiting. (Except for mixbus feedback, or else it could blow up).
    What you're seeing is the behaviour of the CoreAudio driver for the built in audio interface (speaker or wired headphones) of Apple devices (same goes for Mac) which does have an AGC/limiter. Connect an usb audio interface and everything above 0dB will clip. Same if you record to a signed integer format file (not float32).

  • OK thanks @SevenSystems and @j_liljedahl for clearing it up! 😊

  • The user and all related content has been deleted.
  • @tja said:

    I was not aiming for distortion by an App, @Samu - I wanted to compare Limiters and needed a clipping audio file as reference. Distortion by gain was the goal 😅

    A limiter will not 'save' and already clipped file so it's kinda pointless imho.

    The 'clipping' will only happen when you go from floats to integer and the selected integer rage (8,16,24 bits) is too small to cope. As long as floats are used you will not hear any clipping, it's during float > integer conversion the clipping happens.
    (And as mentioned when using the iPad speakers or built-in headphone jack limiting is applied to avoid clipping the DAC and blowing your ears). When using external audio interfaces keeping a check on the output level is essential...
    ...thankfully most DAWs also do limiting on the output so it doesn't sound distorted.

  • The user and all related content has been deleted.
  • edited December 2022

    @tja said:
    ... The clipping was just planned to be used as reference for my comparisions.
    It was not important, I just was wondering why I could not hear any distortions - this is solved now, Core Audio does this by itself.

    2 slightly wrong conclusions... ;)

    A clearly distorted sound is magnitudes above the „distortion artifacts“ that limiting may generate. There is practically no audible similarity.
    As mentioned in your other thread the spectrum of the worst candidate shows distortion overtones (produced by the limiting process) with a maximum of -44dB.
    This is hard to notice (and in fact this limiter was a very common product) on an isolated measuring tone, but it will slightly blur the sound (or accumulate over tracks).
    In some contexts it may even be appreciated, despite it‘s non harmonic character.

    Core audio only controls the final output level, not the processing of plugins producing that output. So if you have a „nasty“ limiter at the end of the chain, it will perform just it‘s way.

  • @tja said:

    What we see in AUM is due to Core Audio limiting, but a DAW should only limit if a Limiter is used.

    Even if there's no limiter enabled in the DAWs higher end audio interfaces have built-in clip-protection and extended headroom. So even when you feed the interface with a 0dbFS input-signal there's always some headroom at the output stage to avoid real clipping which sounds nasty and can damage the connected hardware.

    I always leave plenty of headroom to catch potential accidents also protect my ears...
    Instead of pushing everything to 0dbFS it's in my opinion better to increase the listening levels.

    But again, that's just my way of seeing this.

    Cheers!

  • The user and all related content has been deleted.
  • The user and all related content has been deleted.
  • @Samu said:

    ...thankfully most DAWs also do limiting on the output so it doesn't sound distorted.

    I would certainly not want my DAW to do any limiting unless I add a limiter! And I don't know of any DAW that does this, do you?

  • @j_liljedahl said:

    @Samu said:

    ...thankfully most DAWs also do limiting on the output so it doesn't sound distorted.

    I would certainly not want my DAW to do any limiting unless I add a limiter! And I don't know of any DAW that does this, do you?

    I could have mistaken this...
    ...but when I do export in Logic there's pop-up for overload protection only.

    During normal usage I've never been able to make the output audibly clip even thought the output meter on my Audient ID4mk2 indicates that the input from the computer is clipping, but no super audible clipping even at higher playback volumes.

    I honestly have no need for peaks up to 0dbFS so the only cases I've pushed it is to test if I can hear some kind of degradation.

    However I can easily hear at even lower volumes when the audio file has 'clipped'.

    In other apps I use (Renoise) I have an option to reduce the overall output volume when clipping is detected (it does this by calculating the level of each frame buffer before playback) which has saved my ears a couple of times.

    As long as one is aware of the whole signal chain I have no issues with a master limiter to control the overall output level.

    Cheers!

  • edited December 2022

    @tja said:
    I am not sure what you mean here.
    I did not post any conclusions.

    It doesn‘t matter what you call it, your idea (of comparison) doesn‘t work because the loudest artifact from processing is 44dB lower than the original signal.
    (equivalent to roughly the 1/150th part of the source, as each -6 dB step halves the value)
    The figures were taken from a decade(s) old example, with at least values in a range perceivable. Modern versions are unlikely to exhibit such behaviour.

    I was wondering why AUM did not audibly distort with a high gain.

    as @Samu wrote: a floating point engine doesn‘t clip, this only happens during the conversion to integer (f.e. to write a file in integer format or drive a DAC) if the integer‘s range is exceeded.

    And as it seems, this will be prevented by Core Audio in the OS, before reaching the speakers.

    Yes, that’s an example for the point above.

    And again, this was only done as first step to compare limiters - I wanted to see (hear) distorted audio first and then apply Limiters.

    Understandable, but as mentioned: the level of such artifacts is just too low to compare to a distortion as experienced in an overdriven amp.

    Consider that the example is ages old and the 2nd candidate only produced artifacts below -90dB.
    Today it‘s safe to assume that no such type of limiter produces any noticeable distortion at all.
    There may be intentional drive to emulate analog circuits, though, in some products.
    (which usually is mentioned by the makers in specs and ads)

  • edited December 2022
    The user and all related content has been deleted.
  • edited December 2022
    The user and all related content has been deleted.
  • To be slightly more precise: A limiter with threshold @ -3 dB should give the same results as the following chain:

    [+3 dB] -> [limiter with threshold @ 0 dB] -> [-3 dB]

  • edited December 2022

    @tja I suspect that whatever is inserted in channel 2 and 3 is not a pure limiter. Many plug-ins call themselves "Limiter" but do other stuff as well, like waveshaping.

    (the differences are also quite small at -40 dB, but I agree it's odd that there are any. Is all of this phase-accurate, i.e. does routing through buses not introduce latency?)

    (the extremely straight slope of the spectrum towards the high end also looks very suspicious, it's almost exactly 12 dB / octave... almost as if one of the limiters applied a low pass during its processing...)

  • The user and all related content has been deleted.
  • edited December 2022
    The user and all related content has been deleted.
  • edited December 2022
    The user and all related content has been deleted.
  • edited December 2022
    The user and all related content has been deleted.
  • edited December 2022

    I've looked in the manual for Fabfilter Pro-L 2, and the "Gain" slider on the left is effectively its threshold control, not "Out:" on the right. If you set "Gain" to +1.0, it probably will set the threshold to -1.0, and then add +1.0 of gain at the end to compensate for the lower threshold.

    So, what you'd have to do to "emulate" a "clean" -3 dB threshold is to set its "Gain" to +3 and then set "Out" to -3.

    It's hard to say and compare though because what constitutes a "Limiter" is not really standardized.

  • edited December 2022
    The user and all related content has been deleted.
  • edited December 2022
    The user and all related content has been deleted.
Sign In or Register to comment.