[russdenney]

Hello all,
 
Most of my library is ripped from CDs at AIFF 44.1 kHz sample rate - so essentially CD quality...
 
My question is, is it advisable to only stream to my DAC at the 44.1 kHz (CD quality) setting since this matches my source material? I have heard that up-sampling a lower resolution file buys you nothing and could introduce quantization noise. My DAC supports up to 24/192.
 
Any thoughts or suggestions on the matter?
 
Thank you.
 
Russ

 

[watchnerd]

  Hello all,
 
Most of my library is ripped from CDs at AIFF 44.1 kHz sample rate - so essentially CD quality...
 
My question is, is it advisable to only stream to my DAC at the 44.1 kHz (CD quality) setting since this matches my source material? I have heard that up-sampling a lower resolution file buys you nothing and could introduce quantization noise. My DAC supports up to 24/192.
 
Any thoughts or suggestions on the matter?
 
Thank you.
 
Russ

 
It doesn't really matter one way or the other.
 
My AppleTV feeds a Wyred 4 Sound Remedy Reclocker, just because I want to drop my jitter to <100 ps (for no sane reason).  In the process of doing so, it SRCs the feed to 24bit/96khz.
 
Do I gain anything?  Nope. Does it introduce quantization noise? Not above the noise floor of the rest of the chain, and certainly not audibly.

 

[Roseval]

 I have heard that up-sampling a lower resolution file buys you nothing and could introduce quantization noise.

Depends
The Windows re-sampler does a bad job.
It offer measurable  distortion.
http://archimago.blogspot.nl/2015/11/measurements-windows-10-audio-stack.html
 
Unless your SRC is beyond criticism, I would leave it to the DAC
Hopefully the DAC does it better

 

[watchnerd]

  Depends
The Windows re-sampler does a bad job.
It offer measurable  distortion.
http://archimago.blogspot.nl/2015/11/measurements-windows-10-audio-stack.html
 
Unless your SRC is beyond criticism, I would leave it to the DAC
Hopefully the DAC does it better

 
Windows has some many audio problems I don't know why anyone uses it, TBH.

 

[PleasantSounds]

It would be quite easy to check the upsampled file vs the original and see if you notice any differences. You will not get any additional resolution in the data, but you may decide you like one approach more than the other. For example I have found out that some recordings with particularly harsh treble sound easier on my ears when upsampled to 24/96.
 
With upsampling there's some computational load on the CPU and more data to be pushed to the DAC, but that's not a lot by today's standards and most machines should have no trouble dealing with that, especially if you're using a dedicated audio link like S/PDIF. With USB it could be a bit more complicated if you use some high data rate devices like cameras, printers or hard disks and they happen to be on the same USB bus, but you should avoid this situation regardless of upsampling.

 

[460414]

I am running windows 10: here's some measurements (Auralic Vega) with a 1k sine . The sine is at 13.16db as shown in the top right. So minus 13 from x value to get the "correct" relative value.




The top image is 16/44.1 which is the "recording" of the sine. The bottom is upsampled via default windows "sound" setting "panel" (not sure what you would call it) to 24/44.1.

I tried a 19+20khz sine (don't recall the bit/sample rate) and I certainly didn't get any of the shown issues?

With the up-sampled measurements I cant say if all the miniature "peaks" are from the measuring equipment or its caused by windows. The measuring device is rated at approximately -115db from the top value, which is 13db. So its rated at ~-102db in this setup, which is not far from where the "peaks" start.

 

[spruce music]

  I am running windows 10: here's some measurements (Auralic Vega) with a 1k sine . The sine is at 13.16db as shown in the top right. So minus 13 from x value to get the "correct" relative value.
 
 


 
The top image is 16/44.1 which is the "recording" of the sine. The bottom is upsampled via default windows "sound" setting "panel" (not sure what you would call it) to 24/44.1. 
 
I tried a 19+20khz sine (don't recall the bit/sample rate) and I certainly didn't get any of the shown issues?
 
With the up-sampled measurements I cant say if all the miniature "peaks" are from the measuring equipment or its caused by windows. The measuring device is rated at  approximately -115db from the top value, which is 13db. So its rated at ~-102db in this setup, which is not far from where the "peaks" start.

If you are going to feed a DAC from windows you want something that uses WASAPI or ASIO.  You can use WASAPI with Foobar by getting a plugin for it. 
 
You might check on this software player if you want to do upsampling.
 
http://www.signalyst.com/consumer.html
 
It does upsampling with excellent filtering and I have measured genuine improvements (though minor and likely inaudible) vs simple bit perfect playback of 44 khz files. 

 

[460414]

My point being is that what is shown in the link (http://archimago.blogspot.nl/2015/11/measurements-windows-10-audio-stack.html) is not true. I wasn't using foobar like they were, I was just using windows media player though.
 
 
Quote:

 
If you are going to feed a DAC from windows you want something that uses WASAPI or ASIO.  You can use WASAPI with Foobar by getting a plugin for it. 
 
You might check on this software player if you want to do upsampling.
 
http://www.signalyst.com/consumer.html
 
It does upsampling with excellent filtering and I have measured genuine improvements (though minor and likely inaudible) vs simple bit perfect playback of 44 khz files. 

Why is that better than the default windows one? I'm not saying it isn't I'm just curious as to how?

 

[Roseval]

   
Windows has some many audio problems I don't know why anyone uses it, TBH.

Really?
Use it a decade now for audio without any problem.
As Spruce pointed out, all you need is a driver bypassing the Win audio stack like ASIO or WASAPI.

 

[castleofargh]

the logical way to answer that type of question would be to test it on the specific system used (by ear or measurements). too many things depend on circumstances otherwise.
some NOS DACs with badly implemented filters could absolutely benefit from not using the filter for 44.1khz signal, while many other devices use upsampling/oversampling internally for specific reasons and it would serve no purpose or could even be slightly detrimental to do it before the DAC does its own salsa anyway.

 

[460414]

 
some NOS DACs with badly implemented filters could absolutely benefit from not using the filter for 44.1khz signal

 
I have a decent understanding of the analogue side, but when it comes to the digital (more specifically software) stuff my knowledge is lacking (only really have had an interest in audio for about 2 months). How would not using a filter benefit the dac? Would have the huge tones above 22khz which would have at least some (possibly) indirect negative effect?
I understand some filters are quite terrible such as:

 
   vs
 

 
(same dac used, just different filters)
 
But still..?
Edit: What does no filter look like anyway? I would have though rather nasty?
 
 
Also: Aren't filters are normally done internally (in the dac vs in windows)? In the possibly not so credible link above http://archimago.blogspot.com.au/2015/11/measurements-windows-10-audio-stack.html how can you do this in software?:
 

 
I understand how you could add in a filter via windows but not remove the effects of one like shown? As I said "when it comes to the digital (more specifically software) stuff my knowledge is lacking". Also wouldn't running at a non native sample rate (eg oversampling to 192khz) just change the filter used within the dac? Which would not cause the above?

 

[castleofargh]

I didn't mean using no filter for the bad NOS argument, just not the filter for 44.1khz signal. if you send 96khz music, the DAC will not apply the same low pass and the roll off will start at higher frequencies.
does it make sense said like that?

 

[460414]

  I didn't mean using no filter for the bad NOS argument, just not the filter for 44.1khz signal. if you send 96khz music, the DAC will not apply the same low pass and the roll off will start at higher frequencies.
does it make sense said like that?

 
Yep, though you meant no filter at all.

 

[watchnerd]

  Really?
Use it a decade now for audio without any problem.
As Spruce pointed out, all you need is a driver bypassing the Win audio stack like ASIO or WASAPI.

 
Read what you just wrote: you have to bypass the native audio stack.
 
In other OSes, you don't need to do that at all. No drivers needed.

 

[spruce music]

  Why is that better than the default windows one? I'm not saying it isn't I'm just curious as to how?

When you do sample rate conversion the quality of that conversion, especially the digital filtering varies.  The conversion in Windows is not a terrific algorithm.  The filters in the player I suggested and also some other players is of far higher quality. 
 
http://src.infinitewave.ca/
 
Look here.  Though these are down-conversions you can see quite a range of quality in the conversion depending upon what is used. 
 
In the case of HQ player the filtering is complex and of high quality using many taps.  If you upsample to the max rate your DAC can do then it does less internal upsampling and filtering.  This lets the player software use better filtering for conversion of 44 khz material.  Less aliasing, less noise, better results.  If your DAC has a very poor analog section it may swamp the benefit, but most decent ones are good enough.