INTERACT FORUM

Please login or register.

Login with username, password and session length
Advanced search  
Pages: [1]   Go Down

Author Topic: Different downsampling rates severely effect performance.  (Read 2245 times)

kr4

  • MC Beta Team
  • Citizen of the Universe
  • *****
  • Posts: 741
Different downsampling rates severely effect performance.
« on: July 18, 2016, 12:54:14 pm »

I am re-posting this here for clarity (I am using MC21) and at Jim H's suggestion.

The issue is strange and has to do with the downsampling of DSD files that I do in order to have roomEQ from DiracLive (which works only in PCM up to 192k).
 
This is with files on a NAS connected to the same switch as the server, with files from a directly-connected USB drive and with files on an internal SSD:
 
1. With a DSD256 5 channel file converted to PCM and DiracLive, success depends on the target resolution:
a. Converted to 192k or 96k, loading is slow and there are interruptions and bufferings every 30 seconds or so.  Dirac on/off makes no difference.
b. Converted to 176k or 88k, everything plays fine without interruption.  Dirac on/off makes no difference.
 
2. In all cases, the CPU/memory usage seems similar (5-22%, mostly in the teens, and memory usage at 2-4 GB).  However, when I look at the individual cores, only 4 are running and the others are “Parked.”  
 
3. On the other hand, network behavior varies strangely.  In case 1b (good playback), network usage averages 2% but peaks in the 85-100mb/s range.  In case 1b (interrupted playback), the average is the same but the peaks are in the 50-60mb/s range.  Changing from direct to play from memory changes the network activity pattern but not the peak values.  
 
I do not know what all this means or what may be significant.  I have a colleague with a very similar setup (i7, 16Gb RAM) who has the same observations.  Sure, conversions to 88k and 176k are arithmetically simpler than to 96 and 192 but the limited core usage puzzles me.
Logged
Kal Rubinson
"Music in the Round"
Senior Contributing Editor, Stereophile

Matt

  • Administrator
  • Citizen of the Universe
  • *****
  • Posts: 42444
  • Shoes gone again!
Re: Different downsampling rates severely effect performance.
« Reply #1 on: July 18, 2016, 06:08:23 pm »

Hi there.

Could you try MC22 with the new SoX resampling option?  I'm just curious if it might help.

Options > Audio > Use SoX for resampling

Thanks.
Logged
Matt Ashland, JRiver Media Center

kr4

  • MC Beta Team
  • Citizen of the Universe
  • *****
  • Posts: 741
Re: Different downsampling rates severely effect performance.
« Reply #2 on: July 18, 2016, 07:16:55 pm »

I have said this before but................MC22 with SoX works!  I had suspected it might make a difference when I posted this query first in the MC22 forum.  I have not yet tried it with MC22 w/o SoX but I will soon.  And, fwiw, all cores are in action. 

 :)
Logged
Kal Rubinson
"Music in the Round"
Senior Contributing Editor, Stereophile

Matt

  • Administrator
  • Citizen of the Universe
  • *****
  • Posts: 42444
  • Shoes gone again!
Re: Different downsampling rates severely effect performance.
« Reply #3 on: July 18, 2016, 07:26:06 pm »

So does that mean we've found a good working solution (please say yes!)?

Thanks.
Logged
Matt Ashland, JRiver Media Center

kr4

  • MC Beta Team
  • Citizen of the Universe
  • *****
  • Posts: 741
Re: Different downsampling rates severely effect performance.
« Reply #4 on: July 18, 2016, 09:56:25 pm »

I think so.  I tested it out only briefly but with selections that should have been problematic.  Gimme 24 hours to wring it out.
Logged
Kal Rubinson
"Music in the Round"
Senior Contributing Editor, Stereophile
Pages: [1]   Go Up