Sampler: Retain volume of resampled master output

edited December 2017 in Feature Requests

When I resampling the main outs, the volume level of the recorded sample is lower than the source material. This makes it difficult to use the resampled loop, as I need to match the volume levels before I can bring it in.

Comments

  • The user and all related content has been deleted.
  • Thanks, I tried that and it makes an improvement but it’s still not an exact match:

  • I've noticed it's possible to 'distort' the output and that 'distortion' is not present when re-sampling or exporting a track. (To me it sounds almost like some low-pass filter is applied to avoid the harder edges).

    One suggestion here might be to normalise the exported audio during export?

  • edited December 2017

    dude the sampler volume in the sample edit window defaults to -6db turn it up to 0db with the "gain" control next to fine tuning control and should match your original volume level.

  • The default gains for file-browser preview and pads can be set in the preferences.
    The -6db is simply there to give more head-room when playing polyphonic sounds.

    Korg Electribe Sampler defaults to -12db in comparison so it won't distort so easily when playing all 24 voices at once.

    If one mainly uses one-shots it's quite safe to set both gains to 0db.

  • edited December 2017

    In the second video, the Sampler gain is set to 0dB at 0:06 and 0:30. I’ve also tried normalising the recorded sample in the edit screen. None of these tactics result in matched levels. The source material peaks over 0dB, the recorded material peaks under -2dB.

  • If this in the videos, sorry, no time to watch right now...

    1 What is the source (Note a recording of a synth will never match a synth)

    2 You say the source peaks over 0db, you arent trying to record that are you ?

    3 Can you post a session that does this for us to take a look at, point the developers at etc.

  • edited December 2017

    A summary of the workflow shown in the videos, and the preceding steps:

    Previously:

    1. I created a pattern in the Korg Gadget IAA plugin
    2. I created a -6dB recording/resampling of the Korg Gadget audio.
    3. I copied the recording across 4 pads, using different loop points to create a combination rhythm.

    Video 2:

    1. The combination loop peaks slightly above 0dB on the main outs
    2. I set the global pad gain to 0dB as per the suggestion in this thread
    3. I created a new bank and recorded the main outs to a pad
    4. I verified that the recorded sample had a gain of 0dB
    5. The recorded sample peaks at around -3dB. As a result it sounds quieter than the Korg Gadget recording.
    6. I normalised the recording to 0dB. This had no effect as the setting was already 0dB as per the global setting

    Questions:

    1. Why is the recording ~3dB quieter than the source (even allowing for a max volume of 0dB)
    2. Why is it not possible to record signals greater than 0dB?

    Thanks
    Dan

  • A sin/cos panning law with -3dB center is applied on internal buses. That is probably why!

  • Thanks @mathieugarcia . I’ve put a 9dB make-up gain on my resampling bank and the levels match perfectly. :)

  • We have a 0dB center option as well, it seems super advanced to offer this option to the end-user, but I might be wrong :)

  • @mathieugarcia said:
    We have a 0dB center option as well, it seems super advanced to offer this option to the end-user, but I might be wrong :)

    Save that for an 'Advanced Preferences' tab ;)

    I too can sometimes get 'confused' when the re-sampled sound doesn't match the output I'm hearing.
    And most of the time it's due to extreme clipping that causes an extra overdriven signal (Boosting EQ/Filter with insane amounts as an example). Meaning the signal really 'clips'(peaks at > +13db in the mixer) and the recorded audio feels like it's been run thru a low-pass filter to avoid the harsh 'clipped edges' which would add some character to the sound even at lower volumes.

  • edited December 2017

    I do a lot of live resampling, so anything that improves the workflow / lets me stay in the mix/zone is good.

  • @mathieugarcia said:
    A sin/cos panning law with -3dB center is applied on internal buses. That is probably why!

    cheers for the info @mathieugarcia i was trying to work this out ..

  • It would probably make sense to start having settings for mixer. It's been requested to have classical coloring for volumes, to read saturating levels straight away. Having the pan law option in there, while super advanced, will be a good place.

  • edited December 2017

    To be fair its not that advanced anymore really, go to any DAW forum and search 'Pan law' and 'Null test' at least 1/3 of DAW forum topucs contain pan law conversation.

    Next we will need width control too haha.

  • Definitely need normal colors for mixer meters.

Sign In or Register to comment.