Become a Site Supporter and Never see Ads again!

Author Topic: Attn SAM users: different Sample Rate Conversion results via different menus  (Read 2600 times)

0 Members and 1 Guest are viewing this topic.

Offline Brian Skalinder

  • Complaint Dept.
  • Trade Count: (28)
  • Needs to get out more...
  • *****
  • Posts: 18886
  • Gender: Male
Found this out on the Samplitude forums and wanted to share with fellow TSers.

SAM SE v8.3 (and I assume other versions of SAM) includes multiple options for performing Sample Rate Conversion:

    <1>  Offline Effects | Change Sample Rate
    <2>  Tools | Track Bouncing
    <3>  File | Convert Audio | Save in Format
    <4>  File | Batch Processing
    <5>  File | Export Audio | Wave (no quality options, I believe I tested previously and it uses the Very High setting)

I've always used SRC via Tools | Track Bouncing and assumed (reasonably, I think) SAM applies the same SRC algorithm regardless of the menu selection, as long as one always selects the same quality option (High, Very High, Ultra High 1, or Ultra High 2).  Apparently, this is not the case!

Options <2-4> all produce identical results when the quality setting's the same.  But option <1> produces a different result.  Some of the users on the SAM forums indicate they hear a difference, favoring option <1>, when using the same quality setting with options <2-4> v. option <1>.  I've not done any listening tests yet, though I have confirmed <2-4> generate the same results while <1> generates a different result.  Thought others may like to know.

Sidenote:  If <1> sounds significantly better, then I'm a little cranky.  One of the things I've grown to really enjoy about SAM is the workflow - I can make all my edits virtually, then SRC and dither as I bounce.  If my ears prefer <1>, then I'll have to revert back to a more manual and time-intensive workflow, as I experienced with other apps I've used.
Milab VM-44 Links > Fostex FR-2LE or
Naiant IPA (tinybox format) >
Roland R-05

Offline boojum

  • Trade Count: (1)
  • Needs to get out more...
  • *****
  • Posts: 3633
  • Gender: Male
The only time I change the sampling rate is when I export the finished file.  That is option 5. 
Nov schmoz kapop.

Offline ghellquist

  • Trade Count: (0)
  • Taperssection Member
  • ***
  • Posts: 477
  • Gender: Male
The SRC difference was fixed in version 9 of Sam.

I love Sam as well for its workflow. Load the wav file, set track markers and burn CD in one go. No slow bouncing, no external burners. It also bounces all the songs to .mp3-s ready to put up on the website in one command.

Gunnar

Offline Brian Skalinder

  • Complaint Dept.
  • Trade Count: (28)
  • Needs to get out more...
  • *****
  • Posts: 18886
  • Gender: Male
Thanks for the info, Gunnar - restoring my workflow while utilizing the highest quality SRC is worth the upgrade price alone (well, really re-purchase price since I don't think it's possible to upgrade from SE V8 to SE v9).
Milab VM-44 Links > Fostex FR-2LE or
Naiant IPA (tinybox format) >
Roland R-05

Offline Brian Skalinder

  • Complaint Dept.
  • Trade Count: (28)
  • Needs to get out more...
  • *****
  • Posts: 18886
  • Gender: Male
The SRC difference was fixed in version 9 of Sam.

I just tried SAM SE v9 and this appears not to be the case for me.  Using SRC quality Ultra High 2 via Track Bouncing and Offline Effects | Change Sample Rate yields different results.  In case I'm doing something wrong, here are my precise steps:

Track Bouncing

[01]  Open a new VIP
[02]  Add a 32/48 WAV
[03]  Select from the main menu Tools | Track Bouncing
[04]  Set Format to 32bfp
[05]  Set Sample Rate to 44100
[06]  Set Quality to Ultra High 2
[07]  Set Dithering to No Dithering, math. rounding of sample value
[08]  Click OK, specify filename, and process/save

Change Sample Rate (Option 1)

[01]  Load (Load / Import | Load Audio File) the same 32/48 WAV as above in Track Bouncing [2]
[02]  Select from the main menu Offline Effects | Change Sample Rate
[03]  Set New Sample Rate to 44.1 kHz
[04]  Set Quality to Ultra High 2
[05]  Click OK to process
[06]  Select from the main menu File | Export | Save in Format
[07]  Set Format to 32bfp
[08]  Set Sample Rate to 44100 *
[09]  Set Quality to Ultra High 2 *
[10]  Set Dithering to No Dithering, math. rounding of sample value #

Change Sample Rate (Option 2)

[01]  Open a new VIP
[02]  Add the same 32/48 WAV as above in Track Bouncing [2]
[03]  Select from the main menu Offline Effects | Change Sample Rate
[04]  Set Format to 32bfp
[05]  Set Sample Rate to 44100
[06]  Set Quality to Ultra High 2
[07]  Set Dithering to No Dithering, math. rounding of sample value
[08]  Click OK to specify filename, start processing, and save file

Comparison

[01]  Open a new VIP with 2 tracks
[02]  Add the WAV result from above Track Bouncing process to Track 1
[03]  Add the WAV result from above Change Sample Rate process to Track 2
[04]  Select Track 2 only and select from the main menu Realtime Effects | Invert Phase | Both
[05]  Select from the main menu Tools | Track Bouncing
[06]  Set Format to 32bfp *
[07]  Set Sample Rate to 44100 *
[08]  Set Quality to Ultra High 2 *
[09]  Set Dithering to No Dithering, math. rounding of sample value #
[10]  Click OK, specify filename, and process/save

*  Assumption:  if the export format is the same as the actual data format, SAM won't apply any SRC or dithering.  Perhaps that's a faulty assumption?
#  It's my understanding that SAM won't apply any dithering, anyway, when outputting 32bfp, since both its internal processing and the output file are the same format, i.e. 32bfp.

Results

Comparison of Change Sample Rate (Option 1) and Change Sample Rate (Option 2) produces a null result (as expected).  Comparison of either Change Sample Rate option with the Trackbouncing process produces a WAV with data, i.e. the mixdown does not null, which indicates that performing SRC via Trackbouncing and Change Sample Rate options do not produce the same results even when one selects the same sample rate and quality options.

Am I missing something?  Doing something wrong in my testing?

FWIW, I've not yet listened to the different 32/44 files to determine whether I hear a difference or not.  If I prefer the Change Sample Rate option, though, I'll be a little grumpy that I need to add a manual SRC step to my workflow.
Milab VM-44 Links > Fostex FR-2LE or
Naiant IPA (tinybox format) >
Roland R-05

Offline Gutbucket

  • record > listen > revise technique
  • Trade Count: (14)
  • Needs to get out more...
  • *****
  • Posts: 13651
  • Gender: Male
Did you resolve this issue Brian?
musical volition > vibrations > voltages > numeric values > voltages > vibrations> virtual teleportation time-machine experience
Better recording made easy - >>Improved PAS table<< | Made excellent- >>click here to download the Oddball Microphone Technique illustrated PDF booklet<<

Offline Brian Skalinder

  • Complaint Dept.
  • Trade Count: (28)
  • Needs to get out more...
  • *****
  • Posts: 18886
  • Gender: Male
Did you resolve this issue Brian?

Not resolved.  I've just taken to offline SRC, for now.   :-\
Milab VM-44 Links > Fostex FR-2LE or
Naiant IPA (tinybox format) >
Roland R-05

Offline Gutbucket

  • record > listen > revise technique
  • Trade Count: (14)
  • Needs to get out more...
  • *****
  • Posts: 13651
  • Gender: Male
Finally loaded up my new copy of Sam V10 Master over the weekend and I'm wading into the learning curve.  Thinking through the work flow.   Searching the site I found this thread again.  Once I get deep enough I'll let you know if anything changed in v10 that I can tell. 
musical volition > vibrations > voltages > numeric values > voltages > vibrations> virtual teleportation time-machine experience
Better recording made easy - >>Improved PAS table<< | Made excellent- >>click here to download the Oddball Microphone Technique illustrated PDF booklet<<

Offline boojum

  • Trade Count: (1)
  • Needs to get out more...
  • *****
  • Posts: 3633
  • Gender: Male
GB - the answer will be intersting but mostly in an academic sense.  Can anyone double blind the difference in the different methoids??  I doubt it.  And a no-brainer would be to write SAM support on the question.  Germans love these kinds of questions.  Not this one, though.  The real ones, those in Germany, seem to.  I know other German companies have always been very helpful to me when I have asked for support.

Cheers

BTW - SAM10 Master is pretty nice.  I like working with mine a lot.  But I have a lot more to learn.  Next lesson: mixing four tracks.  Tomorrow.
Nov schmoz kapop.

 

RSS | Mobile
Page created in 0.109 seconds with 50 queries.
© 2002-2020 Taperssection.com
Powered by SMF