[audiojun]

Yes you'll be burning it in. But it probably wont do much to the sound.

Can you burn in a DAC by playing music but shutting off the headphone amplifier? I prefer not to leave the Valhalla on for 48+ hours straight...

 

[traehekat]

Quote:

Yes you'll be burning it in. But it probably wont do much to the sound.

 
Okay cool, thanks. Yeah, I figure if it does anything to the sound might as well speed it up a bit. 
 
So I am using Audirvana and I am pretty new it, but does this look right? 
 
 

 
I'm referring specifically to the fact that the file is ALAC 16/44.1 but it looks like the DAC is set to 24/44.1 - should the DAC be set to 16/44.1 or does that not matter? 

 

[Erukian]

Quote:

 
Okay cool, thanks. Yeah, I figure if it does anything to the sound might as well speed it up a bit. 
 
So I am using Audirvana and I am pretty new it, but does this look right? 
 
 

 
I'm referring specifically to the fact that the file is ALAC 16/44.1 but it looks like the DAC is set to 24/44.1 - should the DAC be set to 16/44.1 or does that not matter? 

 
Does not matter, if you have 16bit material the stream is padded to 24bit with zero's and is still bit perfect.

 

[Mediahound]

Does not matter, if you have 16bit material the stream is padded to 24bit with zero's and is still bit perfect.

If you want it to be bit perfect, you need to set it to play at 16 bit if that's what the source material is. Not 24.

Playing a 16 bit track at 24 bit is not bit perfect.

Whether or not you ears will notice a difference is debatable but from a technical standpoint, some quality will be lost in the upsample.

 

[kn19h7]

Quote:

If you want it to be bit perfect, you need to set it to play at 16 bit if that's what the source material is. Not 24.

Playing a 16 bit track at 24 bit is not bit perfect.

Whether or not you ears will notice a difference is debatable but from a technical standpoint, some quality will be lost in the upsample.

No. There is no upsample involved by padding 0's to make 16bit to 24bit, just no.

 

[judmarc]

Correct, zero padding is not upsampling, it is simply adding places to the "word length." This is not the rate at which sampling is done, it is the amplitude (loudness) variation range of the samples. So whether you say amplitude is -.5 or -.500000000, no actual difference in loudness.

If you do upsample, you need longer word lengths. Whether you have a -.5 amplitude sample once in a second, or once every half second for two samples in a row, your graph will look just the same. But if you have a -.25 amplitude sample the first half second and a -.50 amplitude after one second, you've got a different, sloping graph.

Finally: Almost no DACs, includind Bifrost and Gungnir, are "bit perfect" through the D/A conversion stage. Nearly all of them have DAC chips that send the signal through a series of 2x upsampling steps to get to the 8x rates (352.8/384kHz) where D/A conversion is usually done. Although 2x upsampling just sounds like multiplication and therefore capable of being done exactly, it actually involves more complex math called Fourier transforms and cannot be done with exact accuracy. So depending on the algorithm that's used, using good player software like the iZotope software bundled with Audirvana to upsample to the max 192kHz input may actually give a more accurate final result than just leaving the signal bit perfect all the way to the DAC input.

Not saying everyone should start upsampling, just don't get too "religious" about bit perfect - listen to both if you have a chance and see what you like.

 

[Jason Stoddard]

Quote:

Although 2x upsampling just sounds like multiplication and therefore capable of being done exactly, it actually involves more complex math called Fourier transforms and cannot be done with exact accuracy.

 
Actually, it can. True closed-form digital filters are possible. Stay tuned.

 

[Anathallo]

Quote:

 
Actually, it can. True closed-form digital filters are possible. Stay tuned.

 
Such a tease!

 

[judmarc]

Actually, it can. True closed-form digital filters are possible. Stay tuned.

Now *that* is interesting. So a couple decades after Mike showed everyone how to do digital filtering the first time, you guys are gonna show how to do it better, eh?

What, one digital-to-analog conversion revolution ain't enough for one lifetime?

 

[Erukian]

Quote:

Now *that* is interesting. So a couple decades after Mike showed everyone how to do digital filtering the first time, you guys are gonna show how to do it better, eh?

What, one digital-to-analog conversion revolution ain't enough for one lifetime?

Sounds like a potential DAC upgrade to me. 

 

[azteca x]

Quote:

 
Actually, it can. True closed-form digital filters are possible. Stay tuned.

DADDY LIKE

 

[JWahl]

Very interesting information about the filtering.  I recently had a Resonessence Concero that I had to sell for the money and temporarily replaced it with a Modi (for the 2nd time).  Interestingly, owning the Concero has really taught me the importance of well executed digital filters as the Concero simply uses an ES9023 chip with it's integrated analog output, but adds an FPGA with their own upsampling/filtering algorithms.  Between their minimum phase IIR and linear phase apodizing filters, I generally preferred the apodizing filter.  It's also powered by USB bus power like the Modi and also 6x the price of the Modi. 
 
Another interesting thing I learned, is that at least from my PC, the windows processor scheduling, and background tasks actually has a significant negative impact on sound quality in my experience.  Using a program like Fidelizer, re-prioritizes processor scheduling for audio priority and kills unnecessary background tasks, significantly improved the sound of the Concero.  The effect was even more dramatic with the Modi.  With Fidelizer enabled, the Modi comes dangerously close to the Concero at 1/6th the price.  And actually has a slightly more natural midrange presentation.  I'm actually feeling the Modi will be satisfying enough in my source critical setup until the statement DAC rolls around.  Though it does make me curious of what the Bifrost Uber, or Gungnir USB would be capable of when fed a highly prioritized USB stream.

 

[xpersona]

To the left you may see the upgrade to USB Gen 2 which I have just received (had to order one because my 'old' board could not survive the short circuit received via USB from my Mac Mini (RIP) which in turn got it from the TV (burn in hell!) connected via HDMI).
 
To the right you may see the USB board which i found in my Bifrost which was shipped to me one or two weeks before the upgrade was announced.
 
Could you please tell me if Schiit guys shipped me old stock or they were kind earlier to send me the new board before it was even announced?
 
 

 
Also installed Uber upgrade - can't tell if my ears can hear the difference on my HD650 so far.
 
Hurray to Bifrost's module design!

 

[Alexnova]

edit: going to start a new thread on this.

 

[Defiant00]

Quote:

To the left you may see the upgrade to USB Gen 2 which I have just received (had to order one because my 'old' board could not survive the short circuit received via USB from my Mac Mini (RIP) which in turn got it from the TV (burn in hell!) connected via HDMI).
 
To the right you may see the USB board which i found in my Bifrost which was shipped to me one or two weeks before the upgrade was announced.
 
Could you please tell me if Schiit guys shipped me old stock or they were kind earlier to send me the new board before it was even announced?
 
 
<picture snip>
 
 
Also installed Uber upgrade - can't tell if my ears can hear the difference on my HD650 so far.
 
Hurray to Bifrost's module design!

 
Schiit has said that they were using the new USB boards in DACs for a few weeks before they announced the upgrade.