Page 1 of 2 12 LastLast
Results 1 to 10 of 16

Thread: WARNING - some very important/critical settings for Adobe Audition

  1. #1
    Join Date
    Aug 2006
    Location
    Italy
    Posts
    743
    Thanks
    590
    Thanked 952 Times in 20 Posts

    Post WARNING - some very important/critical settings for Adobe Audition

    Hi, i also use Adobe Audition for some things and perhaps i should have written this post years ago.
    This is a powerful software but when you install it on your PC it has a few default settings which are truly harmful for your audio and so, to seriously work with Adobe Audition, you should disable/modify a few of these settings.

    ------------------------------

    Go to preferences and then choose the 'data' tab.

    1) interpret 32bit-PCM .wav files as 16.8 float
    please, DISABLE it immediately.
    That feature is meant to make wav files compatible with the very old "Cool Edit" software from late 90's and so you have only 16 bit resolution even if your audio is at 24bit.
    Moreover it's a weird non-standard sample format which has no reason to be used anymore.

    2) Dither Transform Results (increases dynamic range)
    disable it.
    This feature is described like a good thing... really it means that anytime you make any editing there is a dithering applied, for example you edit/process only a small segment and you have that segment with dither too... or you edit/process something twice or more times and you have dither applied twice or more times too...
    Very nasty idea!

    3) Use Symmetric Dithering
    disable it.
    as above.

    4) Smooth Delete and Cut boundaries over 2ms
    DISABLE IT!!!
    if you cut or delete a segment, you will automatically have a very short fade-in/fade-out applied... and while it may be useful if we're cutting and pasting music loops to make a song, it is a true disaster while restoring/fixing a recording as we would add tiny drop-outs a go-go.

    5) Smooth all edit boundaries by crossafading 5ms
    DISABLE IT!!!
    same as above.

    Regarding the Auto-convert all data to 32 bit upon opening feature, it depends on what you must do...
    if you only need to open a file to listen/look/check it then it isn't needed to convert it to 32bit while opening, but if you are going to work on it then it's useful to have it converted to 32bit just while opening it.
    (Anyway, you may wish to keep a backup copy of your source files untouched).

    ---------------------------

    Then go to preferences and then choose the 'multitrack' tab and set both recording and mixdown resolutions to 32bit.

    ---------------------------

    Last, if you happen to edit a file on Audition, my advice is to save/export/render it to 32bit and to save WAV files only as IEEE "type 3" normalized 32-bit floating point which is the only equivalent format as the common non-dithered "32bit float" you can find on other software editors (i.e. Wavelab)... the other 24 or 32 bit saving formats available on Audition are mostly non standard stuff.

    Of course, to make the final 16bit or 24bit dithered files, my advice is NOT to use Audition but to make the finishing step with something else (i.e. Wavelab) and this will be the only right moment to apply dithering so that the 32bit float work will be "finished" to 16 or 24bit resolutions.


    If you did just set Adobe Audition this way then you can ignore this post.

    For all the others, I hope it helps,


    Vince.
    Last edited by vince666; 2016-08-22 at 02:01 PM.

  2. #2
    Join Date
    Jul 2006
    Location
    Buffalo, New York
    Posts
    2,195
    Thanks
    470
    Thanked 3,955 Times in 101 Posts

    Default

    Thank you for the useful recommendations. Without knowing how to fully utilize AA, like me, these settings will be handy to refer to if and when I need to work on some audio, which I don't seem to do too often these days. I wish I had this back in the day when I was trying to work on some of my old Floyd projects and recordings that I made. Thanks again!
    Last edited by buffalofloyd; 2016-08-22 at 04:47 PM.

    Click here to access my Pink Floyd lists!

  3. #3
    Join Date
    Mar 2007
    Location
    My imagination most of the time...
    Posts
    2,472
    Thanks
    222
    Thanked 6,979 Times in 93 Posts

    Default

    I'm pretty convinced that default settings like that are sometimes used as a form of copy protection when it comes to audio software. Same deal with media players. I also think sound presets on instrument synth modules (for keyboards and such) are designed to make amateur musicians sound foolish until they realize they need to dial up their own sounds.

    If you are getting serious about tinkering with audio, you should check out the DAW Reaper.
    It's stunningly affordable. Shockingly for the price, it's become the new flagship DAW IMHO. It was a huge upgrade from Pro Tools HD!

    It might be a bit complex to jump into at first but well worth it.

    On the pitch/time correction front, it has the best algorithm (Elastique Pro) integrated (codec style). You can use markers and visually drag audio item edges to correct pitch. The linked pitch and time correction (ie. the 'classic' linked pitch/time where they move together like the pitch knob on an analog tape deck - ie. time slowed down = pitch lowered and vice verse) is 100% artifact free over 100 iterations (I did test that).


    The dither advice above is spot on.
    Dither can be effectively used to "clean up" the low end of the dynamic range when packing full program into a format with limited dynamic range. Like the old 16 bit format for audio CD which has just enough dynamic range for full range program to begin with if used carefully.

    But you absolutely want it off anywhere and everywhere else! It is a noise signal after all!
    It has pretty much no use outside of supporting the old CD format. It doesn't come up with 24 bit audio. What I'm saying is, if you can't get your program into a 24 bit container without pushing something beyond that up from the decimal dust with noise, then you should go back to the mixing board!
    Last edited by jimfisheye; 2016-08-22 at 04:39 PM.

  4. #4
    Join Date
    Aug 2006
    Location
    Italy
    Posts
    743
    Thanks
    590
    Thanked 952 Times in 20 Posts

    Default

    thank you for further details, jim.

    i know there is more modern and better stuff regarding audio software but, since Audition is one of those which is still used by some, then i thought it did worth the effort to make clear these few default settings which make Audition sort of a toy if not disabled.

    Cheers,

    Vince.

  5. #5
    Join Date
    Mar 2007
    Location
    My imagination most of the time...
    Posts
    2,472
    Thanks
    222
    Thanked 6,979 Times in 93 Posts

    Default

    Quote Originally Posted by vince666 View Post
    thank you for further details, jim.

    i know there is more modern and better stuff regarding audio software but, since Audition is one of those which is still used by some, then i thought it did worth the effort to make clear these few default settings which make Audition sort of a toy if not disabled.

    Cheers,

    Vince.
    I don't know about better... just more features. There's nothing wrong with Audition for it being simple.

    I know you intended that specifically for Audition but it's a good general guide for any DAW.

    I mention Reaper especially for anyone comfortable enough to go after speed correction and related stuff.
    You can take 3 different recordings in 3 different sample rates. All in flac files. Drag them into the same project. No decoding or format conversions needed.
    Make splits in one of them and drag the edges (in pitch expand/compress drag mode) to just quickly visually line everything up.
    Everything is 100% artifact free audio quality the whole time. No "preview" modes to deal with like in the past.

    There are a few defaults to change in Reaper too. But if that gets your attention...

  6. #6
    Join Date
    Jul 2006
    Location
    Nijmegen, The Netherlands
    Posts
    2,453
    Thanks
    971
    Thanked 5,735 Times in 86 Posts

    Default

    Littlepieces, who's offered me a lot of assistance and support when he was still active here during my learning curve when it comes to processing audio, alerted me to these settings a year or five ago. I still miss that guy, hope he's happy and succesfull with whatever he's doing these days.

  7. #7
    Join Date
    Aug 2006
    Location
    Italy
    Posts
    743
    Thanks
    590
    Thanked 952 Times in 20 Posts

    Default

    Thank you Jim, for further infos about Reaper... now i am curious to try it!
    But i tend to not believe to the 100% free artifacts point... just because 100% would be an ideal percentage which isn't easy to match in real life.

    Yes, Bert, i also miss littlepieces... and i'm happy he did pass these infos about Audition to someone else too...
    actually, it was just me to give these infos to him in the first moment but i forgot to make a proper public post about this matter at that time... but better late than never.

    Btw, years ago i did also found a few problems in Samplitude... that software used to eat small bits of audio while going through segments with different speed correction ratios... not a small issue.


    Cheers,

    Vince.

  8. #8
    Join Date
    Jul 2006
    Location
    Nijmegen, The Netherlands
    Posts
    2,453
    Thanks
    971
    Thanked 5,735 Times in 86 Posts

    Default

    What about the "Auto-convert all data to 32 bit upon opening" feature? Do you recommend opening 16 bit audio in 32 bit for processing, except editing? A small inconvenience would be that you'd have to re-convert to 16 bit upon saving.

  9. #9
    Join Date
    Aug 2006
    Location
    Italy
    Posts
    743
    Thanks
    590
    Thanked 952 Times in 20 Posts

    Default

    Quote Originally Posted by Bert13 View Post
    What about the "Auto-convert all data to 32 bit upon opening" feature? Do you recommend opening 16 bit audio in 32 bit for processing, except editing? A small inconvenience would be that you'd have to re-convert to 16 bit upon saving.
    if i only need to look/listen/check a file then i would not convert it upon opening (no need to waste time and HD space for no reason).

    but if i am going to process it then it's useful to directly convert it to 32bit just from the start as, after all, I still need to save it as a 32bit float file (the undithered "type3" mode on Audition) after processing, regardless if the source/raw file was at 16 or 24bit.

    so it depends of what I need to do.

    what i forgot to check at that time was IF Audition keeps temporarily processed audio as 32bit regardless of if the opened file is at 16bit or any other resolution or IF it keeps temporary bits like the opened file... i hope it keeps processed bits always at 32bit (as quite any other software actually does) and, if so, the "Auto-convert all data to 32 bit upon opening" setting has no true influence in this exact regard (but why risking if we can have the file converted to 32bit just while opening? even because i am not too sure about what happens really).

    But what i easily checked is that if your file isn't at 32bit (or opened/converted as 32bit) then you cannot save it as a 32bit file (example, if you open a 16bit file while keeping it at 16bits then you can save it only as a 16bit file regardless if you did process it or not)... so you just need to convert it at 32bits at some moment... and at this point it's better to do it just while opening the file (or eventually applying a "convert sample type" to 32bit just after opening the file as is).

    Another point which is easy to notice is that Audition doesn't see 24bit files as "24bit files"... i mean that if I open a common 24bit file then it would be opened/handled as a 32bit file.
    (possibly Audacity makes the same about that).


    As a side note about the "different" way of calling sample formats on Audition...
    The IEEE "type 3" normalized 32-bit floating point format (which has dither disable and not useable, of course) is the only true equivalent as the common not-dithered "32 bit float" you can find quite on any other software... so, if you wish to have proper sample-format compatibility then this is the only right way to save files on Audition (apart che classic 16bit format, of course).
    Infact, like happens on the typical 32bit float format, also this "type3" 32bit format from Audition correctly keeps the waveform even when it goes over the 0dB FS... try it... you have a peak which goes over the 0dB but it will be still correctly saved into a 32bit float file (without flattening the overloaded peaks) , even if while listeing it will give you distortion as the DA converters work at fixed resolutions and cannot handle overloads... so only the "type3" 32bit format on Audition behaves this useful way as the typical 32 bit float.



    Cheers,

    Vince.
    Last edited by vince666; 2016-08-23 at 02:46 PM.

  10. #10
    Join Date
    Mar 2007
    Location
    My imagination most of the time...
    Posts
    2,472
    Thanks
    222
    Thanked 6,979 Times in 93 Posts

    Default

    Quote Originally Posted by Bert13 View Post
    What about the "Auto-convert all data to 32 bit upon opening" feature? Do you recommend opening 16 bit audio in 32 bit for processing, except editing? A small inconvenience would be that you'd have to re-convert to 16 bit upon saving.
    A 32 bit floating point container can fully hold 24 bit fixed data (and obviously anything lower). Zero loss there.
    24 bit fixed files are actually "converted" to 32 bit float for handling because floating point math is the default and more efficient. Processing is done in a DAW with 64 bit fp math. (There might be one or two stragglers out there still running their mix buses at 32 bit float.)

    The quick answer to "WTF?":
    Digital 101: audio quality goes down with the volume. Make a low recording to 16 bit format and you use less than the full 16 bits. Half volume of the system? Only 8 bits! You need a "base" in digital that you basically treat as the noise floor.
    The digital mixing board (the DAW) preserves the original 24 bit sources completely even when you pull a fader for a particular channel way down. Because you have a 64 bit float "wire" for that digital signal to use, you can preserve the full signal even when attenuated.

    Rule of thumb:
    Use 24 bit fixed (at any sample rate) as the standard for finished files.
    If some source material is 16 bit, keep it there if it is only to be edited. But if there is some processing or mastering work being done, keep the now 24 bit files you produced from that.
    The only reason to ever reduce anything higher to 16 bit is to satisfy the now very old CD format. Which should be an aside to the 24 bit master.

    If you work with, for example, a 16 bit source in a DAW. If you only make some edits and no change to the audio, your digital signal on the output will still only be 16 bit resolution. Knowing this (and confirming with a bitdepth meter), you can truncate the zeros off and print a 16 bit file. Inside the DAW, that file was read into a 32 bit float container which was then passed through 64 bit float "wires".

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •