PC+SW+DSD direct VS Any DAC

[widemediaphotography]

Hello to Everyone,
Can we assert that there is no DAC, of any kind or price range, that can perform better than a system composed of:

• PC
• Specialized software
• Selected Native DSD DAC with the ability to bypass the digital filtering part

Or are there still better solutions available?

For example, even a direct DSD DAC will have a final stage that receives the processed signal (upsampled and filtered) from the PC but will always need to perform the analog conversion, which will always require applying filters to the analog spart to eliminate noise and other issue at this stage.
Have we conducted measurements on this matter here?
Thank you

 

[staticV3]

I don't understand the premise of your argument/question.

DSD has no inherent advantage over PCM and is not required for fully transparent audio playback.

 

[MaxwellsEq]

Hello to Everyone,
Can we assert that there is no DAC, of any kind or price range, that can perform better than a system composed of:

• PC
• Specialized software
• Selected Native DSD DAC with the ability to bypass the digital filtering part

Or are there still better solutions available?

For example, even a direct DSD DAC will have a final stage that receives the processed signal (upsampled and filtered) from the PC but will always need to perform the analog conversion, which will always require applying filters to the analog spart to eliminate noise and other issue at this stage.
Have we conducted measurements on this matter here?
Thank you

I don't think we can make that assertion without hundreds of caveats.

I think it is reasonable to say that a high powered processor plus operating system able to run complex convolutions (such as FIR with a very high number of taps) on a digital audio file has more options than any real-world hardware DAC. It might even be that such an approach (assuming we don't care about latency) may be able to reproduce analogue audio with better in-band performance and out-of-band filtering than a hardware-based DAC. But it's unlikely these benefits are audible.

In terms of DSD, I'm not yet convinced that there are benefits unless the whole chain from microphone to DSD DAC is pure DSD.

 

[DVDdoug]

that can perform better than a system composed of:

Better, no. Equal, yes. (In a proper, scientific, blind ABX test.) In most cases "CD quality" is already better than human hearing and once you are better than human hearing there can be no "improvement".

with the ability to bypass the digital filtering part

Unfiltered, DSD contains tons of ultrasonic junk/noise that's not part of the original analog signal and should be filtered out. You can't hear it but sometimes it can do bad things to an amplifier and you don't need it.

 

[widemediaphotography]

I don't understand the premise of your argument/question.

DSD has no inherent advantage over PCM and is not required for fully transparent audio playback.

In audio enthusiast communities, the practice of performing analog filtering using a native DSD DAC with DSD direct functions is gaining increasing support. Even in the case of simple PCM 16/44.1, any Delta-Sigma DAC performs upsampling/oversampling for filtering and noise shaping. Therefore, by inputting directly with an improved equivalent in DSD 256/512, for example, to one DSD direct DAC performs the conversion directly without performing digital oversampling. The principle is based on the fact that a PC's CPU has a computational capacity far superior to any DAC chip and can therefore do a better job using more efficient real-time filters.

 

[widemediaphotography]

Better, no. Equal, yes. (In a proper, scientific, blind ABX test.) In most cases "CD quality" is already better than human hearing and once you are better than human hearing there can be no "improvement".


Unfiltered, DSD contains tons of ultrasonic junk/noise that's not part of the original analog signal and should be filtered out. You can't hear it but sometimes it can do bad things to an amplifier and you don't need it.

I absolutely agree with you, but from a strictly technical point of view we should still recognize if improvements exist and quantify them, as any scientific approach would require.
This is the crux of my initial question

 

[MaxwellsEq]

Just because lots of people are doing something doesn't make it correct.

This approach makes assumptions about the internal working of a specific DAC design, so may not be constant from DAC to DAC.

I may have missed it, but I've not seen measurements that demonstrate an audible benefit.

 

[staticV3]

The principle is based on the fact that a PC's CPU has a computational capacity far superior to any DAC chip and can therefore do a better job using more efficient real-time filters.

The computational capacity of a DAC chip is more than sufficient for virtually perfect upsampling:

Does a 4 million taps FIR filter sound better than a 16K one? Let's find out.

Claim A claim is frequently made that DAC digital filters with a lot of taps can make a significant difference in sound quality. Some software packages and even hardware manufacturers are touting the benefits of large FIR filters, often with millions of taps, supposedly because these produce a...

www.audiosciencereview.com

Chord quest, Vs Rme adi-2 dac fs + tap count.

Well it looks a lot more advanced than this. But is it really?

www.audiosciencereview.com

 

[widemediaphotography]

Just because lots of people are doing something doesn't make it correct.

This approach makes assumptions about the internal working of a specific DAC design, so may not be constant from DAC to DAC.

I may have missed it, but I've not seen measurements that demonstrate an audible benefit.

It seems to me that @amirm has already been asked to take care of the matter several times, but I don't remember that the matter was followed up...

 

[AnalogSteph]

We've already seen measurements here that indicated that DSD output would take care of the spray of high-order harmonics indicative of the presence of an "ESS hump" - the working theory being that the extra high-frequency noise is basically providing dithering. If I had a DAC with a particularly pronounced hump that reaches as high as -65ish dB THD+N (e.g. Khadas Tone Board, some older Toppings and whatnot), I'd want to make use of DSD, as that figure is getting a bit close for comfort even by my modest standards. If the hump is 20-25 dB lower, I wouldn't bother.

 

[sofrep811]

You mean DAC chips vs what you listed? I can put a PC in a box with a DAC.

 

[widemediaphotography]

You mean DAC chips vs what you listed? I can put a PC in a box with a DAC.

I mean (e.g) a PC that executes a PCM 16/44.1flac with HQplayer connected to Holo Cyan 2 DAC (used ad DSD direct DAC) versus any Hi-End DAC plugged to any bit-perfect source. For this, there is a long list of recommended DSD direct DACs

 

[MaxwellsEq]

I mean (e.g) a PC that executes a PCM 16/44.1flac with HQplayer connected to Holo Cyan 2 DAC (used ad DSD direct DAC) versus any Hi-End DAC plugged to any bit-perfect source. For this, there is a long list of recommended DSD direct DACs

As I said a few posts ago, I simply don't think anyone can confirm with certainty the audibility of your original conjecture without loading it up with too many caveats.

Within specific constraints and under certain circumstances, you may be right. But for the general case - PCM into the current crop of DACs - I think it's likely there will be no audible benefit of using a PC to convert commercially available PCM content into DSD and feeding that into a DSD DAC. However, for people who have the time and resources, I don't think this approach will make it audibly worse (unless their DAC lacks a low pass filter cutting the HF noise).

 

[widemediaphotography]

As I said a few posts ago, I simply don't think anyone can confirm with certainty the audibility of your original conjecture without loading it up with too many caveats.

Within specific constraints and under certain circumstances, you may be right. But for the general case - PCM into the current crop of DACs - I think it's likely there will be no audible benefit of using a PC to convert commercially available PCM content into DSD and feeding that into a DSD DAC. However, for people who have the time and resources, I don't think this approach will make it audibly worse (unless their DAC lacks a low pass filter cutting the HF noise).

I believe you already know about that Forum with thousands of enthusiasts ready to swear, also supported by measurements, that having the digital filtering performed by a PC is much more efficient than having it performed by the DAC.

 

[Blumlein 88]

In terms of DSD, I'm not yet convinced that there are benefits unless the whole chain from microphone to DSD DAC is pure DSD.

Which rules out all processing. Do you think there are any possible benefits of direct DSD 512 vs PCM 24/96?

 

[Blumlein 88]

I believe you already know about that Forum with thousands of enthusiasts ready to swear, also supported by measurements, that having the digital filtering performed by a PC is much more efficient than having it performed by the DAC.

More efficient.......you must be joking. Very simple measurements indicate that is in the cloud,......cloud cuckoo land. It may have benefits, but efficiency is not one of them.

 

[widemediaphotography]

More efficient.......you must be joking. Very simple measurements indicate that is in the cloud,......cloud cuckoo land. It may have benefits, but efficiency is not one of them.

It is necessary to consider the direct opportunity of DSD in some DACs. The PC can apply more complex filters in real time than any DAC in the digital stage. In this case the DSD direct DAC can receive a DSD file ready for analog filtering only. This is the principle underlying the original question. Any D-S DAC always oversamples to perform the D-A conversion, this oversampling operation can be done better by a PC than any cip DAC, which will only have to worry about performing the filterning in the analogue stage. It's not difficult to understand.

 

[Blumlein 88]

It is necessary to consider the direct opportunity of DSD in some DACs. The PC can apply more complex filters in real time than any DAC in the digital stage. In this case the DSD direct DAC can receive a DSD file ready for analog filtering only. This is the principle underlying the original question. Any D-S DAC always oversamples to perform the D-A conversion, this oversampling operation can be done better by a PC than any cip DAC, which will only have to worry about performing the filterning in the analogue stage. It's not difficult to understand.

So efficiency as long as you don't include energy use, complexity of the processing, complexity of the gear involved and cost. As well as doing all that pre-filtering rather makes it not really DSD anymore. All so you can simplify the one part of the chain by pushing all the complexity everywhere else. Considering that Sigma Delta DACs are very inexpensive, and make everything external totally dead simple and super cheap (while themselves being very cheap) it seems a very false sense of efficiency

 

[Sokel]

It is necessary to consider the direct opportunity of DSD in some DACs. The PC can apply more complex filters in real time than any DAC in the digital stage. In this case the DSD direct DAC can receive a DSD file ready for analog filtering only. This is the principle underlying the original question. Any D-S DAC always oversamples to perform the D-A conversion, this oversampling operation can be done better by a PC than any cip DAC, which will only have to worry about performing the filterning in the analogue stage. It's not difficult to understand.

Theory is nice but it seems that the whole trick is device-specific,level specific,more so I would say whole chain specific and you won't know until you measure your specific rig.
Have a look here:

PCM vs DSD

Did a little comparison and I was wondering if any friend with better gear here has tested this. Results you're about to see is probably narrowed to the specific DAC but it's interesting to see how it handles things differently. I will not try to decipher the results other that DSD seems much...

www.audiosciencereview.com

The DUT (an el cheapo Khadas TB) has a fault that DSD seems to reduce but wouldn't be nicer to choose something without this fault at the first place?
There's also some player comparisons in the thread,HQ player for example doesn't seem to do anything special.

 

[MaxwellsEq]

Which rules out all processing. Do you think there are any possible benefits of direct DSD 512 vs PCM 24/96?

Not really.

My observation over the years since before CD was released is that 16bits should be sufficient for all known music distribution. 24 bits is probably unnecessary for commercial music distribution.

However, 44kHz is a bit close for comfort when it comes to constructing filters which are effective whilst not adding ripple or early roll-off. As a pro studio person, 48kHz would have been better. I think the 44.1kHz/filter is the single weak spot that naysayers can legitimately attack!

But once you get 96kHz, it's hard to imagine what filter artefacts people think they can hear.