Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[jason3]

Like the title says I am trying to do DSD (.DSF Files) to PCM on the fly with DSP studio.  Is that even possible?  Or is there another way?  Or do I have to convert my files individually before I play them?  Thanks.  (Hope this hasn't been covered elsewhere, I searched but came up with nothing)
Thanks!!!
Jason

[Awesome Donkey]

MC should automatically convert DSD to PCM unless a) you have DSD bitstreaming enabled or b) have DSP Studio > Output Format set incorrectly.

[Hendrik]

You may have to set MC to downsample high samplerates. Say anything above 192k down to 192k. Or maybe even lower. Depends on what your device supports.

[jason3]

 
Thank you both!!

[dtc]

When Media Center converts DSD to PCM it does so by a factor of 8. So, 2.8 MHz DSD goes to 352 MHz PCM, 5.6 MHz DSD to 705 MHz PCM, and 11.2 MHz to 1410 MHz.  In DSP Studio - Output Formatting you then need to set the conversion to your final desired sample rate. So, if you want 176 Khz PCM as your final rate, you need to put 176 KHz in the conversion lines for 352 Khz and 705 KHz and also in the >768 KHz line.  I would suggest 176 Khz or 88 Khz, since they are integer multiples of the original PCM rate and that makes the conversion more efficient than a non-integer conversion like 192 Khz.

You should also check to be sure SoX is selected for downsampling in the Audio options section. I am not sure if it currently defaults to that or not. It is much more efficient than the old converter in doing the PCM conversion. 

As part of the conversion, a large amount of high frequency noise is generated that has to be filtered out. By default, a low pass filter is applied. The default is a filter at 24 Khz with a slope of 48 dB. This is a conservative filter and works for most people. The Sony standard was 50 Khz with a slope of 48 KHz. Unfortunately, that is not an option, but you can set one like that in DSP Studio, if you turn  off the default  filter in Audio - Advanced - Configure Input Plugin - DSD.  Only do that if you set an appropriate low pass filter in DSP Studio.  If none of this makes sense, just leave the default alone.

[jason3]

Thanks!!  I see you CAN set a similar filter in the advanced settings.  It is called "permissive" and is 50Khz with a slope of 24 dB.  Would that be better?  (By the way I don't have any idea what all this means.)
I did set the SoX and set the conversions to 176 Khz.

[dtc]

The permissive filter is not step enough to filter out the noise at higher frequencies.  The slope is 24 dB/octave versus the Sony one of 48 dB/octave.  It keeps the noise down, but is not as low as it should be. 

If you want to have frequencies above 24 KHz, I would suggest setting a low pass filter in the Parametric Equalizer  in DSP Studio.  Just set a low pass filter at 50 Khz with a slope of 48 dB/octave.

Give it a try and if you don't hear a difference, just go with the default.

[jason3]

Any chance of hurting my DAC?

[Hendrik]

I would suggest 176 Khz or 88 Khz, since they are integer multiples of the original PCM rate and that makes the conversion more efficient than a non-integer conversion like 192 Khz.

As a general note, this is really not a factor with a complex resampling algorithm as used by SSRC or SoX (the two options in MC). There are no quality or efficiency differences in any measurable range, the performance primarily depends on the input and output sample rate, the higher each of those are, the more it has to work, because more audio is being read or written. So 705k -> 96k is slightly slower then 705k -> 88.2k, but not because the factor is so nice, but just because 96k is more audio per second.

[dtc]

The filter is  there to prevent the high frequencies hurting the electronics. The 24 dB/octave filters will probably provide enough filtering, but the 48 dB/octave standard was set to be sure there was no damage. Sony developed DSD (SACD) so I follow their recommendations.

[jason3]

Cool.  I am learning a ton of stuff!!

[dtc]

As a general note, this is really not a factor with a complex resampling algorithm as used by SSRC or SoX (the two options in MC). There are no quality or efficiency differences in any measurable range, the performance primarily depends on the input and output sample rate, the higher each of those are, the more it has to work, because more audio is being read or written. So 705k -> 96k is slightly slower then 705k -> 88.2k, but not because the factor is so nice, but just because 96k is more audio per second.

I see as much as a 2 time difference in converting to 192 Khz versus 176 KHz starting from the same DSD file - using SoX.  That is significant to me, especially when converting whole albums.

[jason3]

I am guessing leaving the high pass filter on all the time won't affect my flac files and videos?  They don't have information above 50 khz right?

[dtc]

Correct - but if you are in doubt, use the default, unless you hear a difference.