[LDBaha]

Make sure that the Denon AVR does not digitize its analog inputs. Otherwise, it really defeats the purpose of having an outboard DAC.

As far as DSD512 music, I am not aware of any for sale. If there is, the catalog is extremely small in size.

I see. I'll make sure, I don't really need to hook it to the Denon TBH. I have it connected to an Oppo 203 and also de Denon is one of the new ones so it has all the same bells and whistles.
 
How about connecting it to the Jot through XLR? Since the Jot is both a Preamp and a Headphone amp I think I'm safe.
Ideally I think I would use this Sonica with headphones like 95% of the time. Since the Jotunheim has the balanced inputs I thought "hey what the heck lets go full balance"

My other option is a GUSTARD DAC-X20 or something cheaper like a TEAC but those don't have the other capabilities of the Sonica... I like that I can make the Sonica like a central hub. 

 

[sdolezalek]

As far as DSD512 music, I am not aware of any for sale. If there is, the catalog is extremely small in size.

 
Although the overall catalog may be small here are more than 100 DSD512 titles: www.nativedsd.com/homepage/quad_dsd_music
 
Separately, the way that many of us use the DSD512 capability is by having software such as HQPlayer upsample everything we listen to to DSD512.  In many DACs capable of playing it that is the way to get the very best even out of 16/44 PCM recordings. 

 

[ekrauss]

I don't think upsampling 16/44.1 PCM to DSD would result in any appreciable difference in sound quality.

 

[sdolezalek]

I don't think upsampling 16/44.1 PCM to DSD would result in any appreciable difference in sound quality.

 
That just means that you likely don't fully understand what goes on inside a DAC.  Most DACs convert everything (including PCM) to DSD512 or higher internally.  What filters they use in doing so can greatly impact the music.  Using a software package that carefully lets you choose great filters and filter parametes and move the processing load out of the DAC, to feed it a sonic input it can directly process, can lead to substantially better sound.
 
Unfortunately, far too many people assume that upsampling is either about adding more data (which it can't), creating smoother sine wave curves (which it doesn't), enabling higher frequencies (which it also largely doesn't and you couldn't hear even if it did).  Instead, upsampling is about minimizing the destruction of audio content that occurs in the D/A conversion process.  Each D/A conversion requires use of filters and those filters require the designer to make tradeoffs between frequency, timing accuracy and pre and post-ringing. Upsampling allows you to move many of the conversion artifacts further away from the audible parts of the music spectrum.  Doing your own upsampling also allows you to make the choices that work best for your ears.
 
To bring this back to the Sonica DAC, this is one of the reasons why its DSD512 capabilities matter and is one of the ways to get the most out of this DAC.

 

[ekrauss]

That just means that you likely don't fully understand what goes on inside a DAC.  Most DACs convert everything (including PCM) to DSD512 or higher internally.  What filters they use in doing so can greatly impact the music.  Using a software package that carefully lets you choose great filters and filter parametes and move the processing load out of the DAC, to feed it a sonic input it can directly process, can lead to substantially better sound.

Unfortunately, far too many people assume that upsampling is either about adding more data (which it can't), creating smoother sine wave curves (which it doesn't), enabling higher frequencies (which it also largely doesn't and you couldn't hear even if it did).  Instead, upsampling is about minimizing the destruction of audio content that occurs in the D/A conversion process.  Each D/A conversion requires use of filters and those filters require the designer to make tradeoffs between frequency, timing accuracy and pre and post-ringing. Upsampling allows you to move many of the conversion artifacts further away from the audible parts of the music spectrum.  Doing your own upsampling also allows you to make the choices that work best for your ears.

To bring this back to the Sonica DAC, this is one of the reasons why its DSD512 capabilities matter and is one of the ways to get the most out of this DAC.

I understand perfectly well what occurs within DACs, be they delta-sigma (the variety you were talking about), ladder resistor, or r2r/multibit (which do not operate as you described).

As I said before, there is absolutely no reason why transcoding a 16/44.1 PCM signal to DSD would change the sound in a perceptible manner.

Not only that, but I have done experiments wherein I converted/transcoded 24/44.1 PCM files to DSD64, 128, and higher multiples and could hear absolutely no difference.

So, although I appreciate your comments, I do not agree with them on either a subjective or objective level.

 

[zilch0md]

Spoiler: Three quotes...

I don't think upsampling 16/44.1 PCM to DSD would result in any appreciable difference in sound quality.

 
 

   
That just means that you likely don't fully understand what goes on inside a DAC.  Most DACs convert everything (including PCM) to DSD512 or higher internally.  What filters they use in doing so can greatly impact the music.  Using a software package that carefully lets you choose great filters and filter parametes and move the processing load out of the DAC, to feed it a sonic input it can directly process, can lead to substantially better sound.
 
Unfortunately, far too many people assume that upsampling is either about adding more data (which it can't), creating smoother sine wave curves (which it doesn't), enabling higher frequencies (which it also largely doesn't and you couldn't hear even if it did).  Instead, upsampling is about minimizing the destruction of audio content that occurs in the D/A conversion process.  Each D/A conversion requires use of filters and those filters require the designer to make tradeoffs between frequency, timing accuracy and pre and post-ringing. Upsampling allows you to move many of the conversion artifacts further away from the audible parts of the music spectrum.  Doing your own upsampling also allows you to make the choices that work best for your ears.
 
To bring this back to the Sonica DAC, this is one of the reasons why its DSD512 capabilities matter and is one of the ways to get the most out of this DAC.

 
 

I understand perfectly well what occurs within DACs, be they delta-sigma (the variety you were talking about), ladder resistor, or r2r/multibit (which do not operate as you described).

As I said before, there is absolutely no reason why transcoding a 16/44.1 PCM signal to DSD would change the sound in a perceptible manner.

Not only that, but I have done experiments wherein I converted/transcoded 24/44.1 PCM files to DSD64, 128, and higher multiples and could hear absolutely no difference.

So, although I appreciate your comments, I do not agree with them on either a subjective or objective level.

 

 
 
I can see where both of you are coming from, but you really haven't contradicted each other.  ekrauss made a subjective statement and I don't think he disagrees with any of sdolesalek's explanations of why upsampling makes it easier for a DAC to deal with Redbook having decided that a sampling rate of 44.1 kHz would give sufficient working room for DA conversion above the 40k Nyquist frequency.
 
I've always considered the best measure of a DAC is its ability to do a good job with 44.1/16, as I really believe that's far more challenging than DSD files, where even a $100 DAC can do a pretty good job.  
 
So, in that vein, I would have to agree with ekraus:  "I don't think that [manually] upsampling 16/44.1 PCM to DSD [in advance of feeding it to the DAC] would result in any appreciable difference in sound quality."  Doing so might help with a lesser quality DAC, but not with a DAC that does as a good a job with Redbook as the Sonica DAC.

 

[ekrauss]

I appreciate all of your comments. This is a very complicated process with lots of nuances. I need to clear up a misconception that has been echoed in some posts above: Delta-sigma DACs do not process audio data internally as DSD. They might process the data in a one-bit PWM stream that is DSD-like in certain ways, but they are mathematically different enough that DSD is always manipulated and transcoded in ways that are not much different than a PCM signal would be. In fact, in many--or even most--cases, DSD is converted on-chip to PCM before it is run through the decimation process to result in the internal PWM stream. When that happens, DSD data is subjected to even more mathematical manipulation than PCM!

If you convert your files to DSD first, before sending them to your DAC component and its USB receiver chip and DAC chip, you are not offloading any processing from the DAC chip in any way, shape, or form. Even if you are connecting via I2S, there still will be transcoding of DSD within the chip to PWM or worse, to PCM and then to PWM. It's an attractive notion to believe that doing DSD conversion first will simplify the process and result in better audio performance, but that belief does not jive with how delta-sigma conversion actually works.

 

[ekrauss]

upsampling makes it easier for a DAC to deal with Redbook having decided that a sampling rate of 44.1 kHz would give sufficient working room for DA conversion above the 40k Nyquist frequency.

This statement is true. Upsampling PCM to a higher rate allows gentler antialiasing filters to be used, and can result in better performance in the higher frequencies and with lower level signals. However, DSD is an entirely different animal, and its noise shaping to shift quantization noise to out-of-band frequencies can result in significant phase distortion. I am a proponent of DSD, but not of converting PCM to DSD.

 

[sdolezalek]

Here is a quote directly from the developer of HQPlayer: "All modern DAC chips are delta-sigma designs, meaning that PCM input needs to be converted to high speed bit-stream. As said earlier, DAC chips typically convert PCM input first to 352.8 or 384 kHz PCM using a digital filter of varying quality and then use stupid sample-copying (sample-and-hold/zero-order-hold) to take the rate up to typical 5.6/6.1 MHz speed for the delta-sigma modulator to produce the bit-stream for the actual D/A conversion process. What HQPlayer is doing is not adding any additional processing to the chain, but replacing the processing performed by DAC chip with better implementation done in software. So HQPlayer performs high quality digital filters taking the sampling frequency straight to 5.6/6.1 MHz or even higher 11.3/12.2 or 22.6/24.6 MHz, without any quality compromising sample-and-hold stages and then converts it to bit-stream for the actual D/A conversion using high quality dithered delta-sigma modulator."
 
Although I haven't tried the Sonica DAC, I strongly believe that this upsampling benefits the Oppo-BDP105D's DAC. 

 

[ekrauss]

Yes, that is in complete agreement with what I said above about upsampling redbook PCM to PCM with a higher sample rate. You understand that paragraph above is talking about PCM upsampling, right? It does not have anything to do with DSD. The author is fudging with the facts a little, too, as there is no way to bypass the internal on-chip conversion from PCM to delta-sigma PWM (or DSD to delta-sigma PWM, for that matter).

 

[ekrauss]

Look at the post by Norton here for a good summary:

https://www.computeraudiophile.com/forums/topic/29526-upsampling-to-anything-other-than-your-dacs-internal-conversion-rate/

 

[sdolezalek]

Thanks, but that is just referring me back to my own post at Computer Audiophile. )  I agree that there is a lot of complexity involved here and as far as I know there hasn't really been a good clear discussion of all that is involved.  Norton is one of the folks who prefers the internal conversions of his own DAC to upsampling in HQPlayer or otherwise.  Many others on CA believe exactly the opposite.  It is clearly also DAC dependent.  Chord DACS, for example, seem to do better when fed PCM material.  The T+A DAC8 DSD seems to do especially well when fed with DSD512.  
 
There is also precision of language needed.  My own reference to internal upconverting to DSD was an attempt to stay with just the PCM/DSD terminology, but, as you correctly point out, the real conversion is to a DSD-like 1-bit PWM stream (which then leads most readers to say "what on earth is PWM, I still trying to figure out PCM/DSD.") 
 
I think what we can probably agree upon is that you want to do as much of the filtering as possible outside of audible bandwidths and doing that with a plain 16/44 signal is difficult.  At that point the question becomes two-fold: a) does my DAC do it better than an external piece of software can, and b) if done externally can that signal be fed to the DAC in a way that bypasses the DAC's own upsampling/filtering?  
 
I certainly believe that there are DACs that do this extremely well internally and have been very carefully designed to do so.  Most of them are quite expensive.  Others, like the consumer-priced iFi DACs seem to benefit greatly by having this processing done in HQPlayer.  Given the cost of the Oppo Sonica DAC, the real question is whether software can better the combination of what Oppo and ESS/Sabre have done already in DAC and chip? 

 

[ekrauss]

Actually, I referring you to *Norton's* post in your thread at Computer Audiophile! Basically, what he says is right--there is no way, period, to bypass a delta-sigma DAC chip's internal conversion to PWM. Transcoding PCM to DSD will only incur additional processing overhead and does not offload any of the DAC's work to HQPlayer. Instead, your music will undergo double the amount of processing! Again, it would be neat if you could avoid the DAC chip's internal processing and use a program like HQPlayer to do it instead, but unfortunately, that's impossible.

 

[sdolezalek]

If what you are suggesting were true then playing hi-res PCM, DXD or DSD files directly on the Sonica would be a waste of time as it would treat all of them the same as 16/44 and start from there. So to be clear, are you suggesting that a) an ESS/Sabre DAC doesn't treat an incoming 16/44 file differently than an incoming 24/384 file (does it first resample the 384 file to 44 before reupsampling it)? b) an ESS/Sabre DAC can't benefit from DSD files because it converts them all to PCM anyway, so you might as well start with PCM material only?  
 
Or are you suggesting that the DAC can tell the difference between a file that started as 16/44 and was upsampled to 24/384 or to DSD512 and one that was 24/384 or DSD512 to start with? 

 

[zilch0md]