World’s First Valid Comparison of PCM versus DSD

[microstrip]

I think a more proper comparison would be using both sides' best units. So, what's the best DSD-capable ADC out there? And which one's the best for PCM?
Use them both, and then play the files back on 3 different DACs, like a Vivaldi, an MSB and a Playback Designs.

microstrip,

Assuming no ADC or DAC is 100% transparent, there is no way to compare the formats without taking into account the "signature" of the device itself, no? Do you see a way that the format itself can be compared, isolating the device?


alexandre

Alexandre,

We will never have an absolute comparison and an absolute answer. We can however try to recreate conditions that most people find correct and debate the findings of the people involved. For the encoders I would rely on Bruce and other recording experts and pick a few of their suggestions as the best in each format. For the DACs I would be very happy with your list, but adding a DAC from a manufacturer that feels that adding native DSD would compromise his PCM design, such as the Alpha Berkeley. I would however never make an opinion on PCM based on listening to the Playback Designs - don't you agree with me?

 

[microstrip]

The Korg was the only digital component used. You can't get more simpler than that!

I own it and have listened to it using very good quality needle drops made by WBF members and a few of my experiences. I found its performance in PCM playback very poor, compared to the the other PCM DACs I have experience with. Unless some one I can trust tells me that by miracle its PCM ADC is perfect and sounds great I will have doubts on its performance. IMHO, you should borrow a DCS Vivaldi for the good of audio science (and our reading pleasure ).

 

[Bruce B]

I own it and have listened to it using very good quality needle drops made by WBF members and a few of my experiences. I found its performance in PCM playback very poor, compared to the the other PCM DACs I have experience with. Unless some one I can trust tells me that by miracle its PCM ADC is perfect and sounds great I will have doubts on its performance. IMHO, you should borrow a DCS Vivaldi for the good of audio science (and our reading pleasure ).

There is NO A/D converter that is "best" at both.

 

[microstrip]

There is NO A/D converter that is "best" at both.

It is why I say DSD partisans should pick their best and the PCM ones should do the same. Why using the same converter? We only Want The Best of each class!

 

[asiufy]

Alexandre,

We will never have an absolute comparison and an absolute answer. We can however try to recreate conditions that most people find correct and debate the findings of the people involved. For the encoders I would rely on Bruce and other recording experts and pick a few of their suggestions as the best in each format. For the DACs I would be very happy with your list, but adding a DAC from a manufacturer that feels that adding native DSD would compromise his PCM design, such as the Alpha Berkeley. I would however never make an opinion on PCM based on listening to the Playback Designs - don't you agree with me?

Sure, same as we're picking "best" ADCs for each format, why not pick best DACs too?

So, perhaps a good "fight" would be the Berkeley on PCM, Playback on DSD, as far as DACs go. Now, on to find someone who has access to both

And regarding the Playback, I'm actually curious about it. I had it briefly here, and it did quite well, on Redbook (PCM)... Something tells me I should investigate further!

alexandre

 

[Bruce B]

And regarding the Playback, I'm actually curious about it. I had it briefly here, and it did quite well, on Redbook (PCM)... Something tells me I should investigate further!

alexandre

I'm using the Playback here for all my DAC work.

 

[Bruce B]

It is why I say DSD partisans should pick their best and the PCM ones should do the same. Why using the same converter? We only Want The Best of each class!

Well this is what I did for the Wilson transfers. I had the best of both worlds.

PCM:
MSB Studio Platinum ADC
Weiss ADC 2
DAD AX24

DSD:
Grimm AD1
DAD AX24
EMM labs ADC8 Mk IV

In every blind listening test the Wilsons did, they would always pick the DSD as their preference. They should know, Dave did the recordings and had the master tapes!

 

[microstrip]

Well this is what I did for the Wilson transfers. I had the best of both worlds.

PCM:
MSB Studio Platinum ADC
Weiss ADC 2
DAD AX24

DSD:
Grimm AD1
DAD AX24
EMM labs ADC8 Mk IV

In every blind listening test the Wilsons did, they would always pick the DSD as their preference. They should know, Dave did the recordings and had the master tapes!

A great answer always brings more questions. And what were the preferred DACs for each format?

 

[Bruce B]

A great answer always brings more questions. And what were the preferred DACs for each format?

I only listen to files through a PBD MPS-5. I use this as my reference and it doesn't change. I had no preference because I didn't know what the original performance sounded like. Besides, a DAC doesn't change the sound of the recording... only the A/D can change the sound of the recording.

I don't know what DAC they were using for their preference.

 

[microstrip]

Besides, a DAC doesn't change the sound of the recording... only the A/D can change the sound of the recording.

OK, we are now in semantics - my fault, I should have been more clear - I am addressing the playback of the recording. There are large differences between the sound on the DACs I have used, and IMHO can not we exclude the possibility that a DAC that operates a transcoding to DSD of PCM data changes sound attributes. According to Andreas Koch : In our devices we use the 44.1 kHz sample frequency as a basis, convert the digital signal into 16FS PCM, make a DSD stream of it in 5.6 MHz, multiply the sample rate by another factor of 8 and finally, at a sample frequency of around 45 MHz, we decode the signal into analogue sound.

I understand that here are very few real native DACs in both families, but the PB seems a real DSD unit.

 

[asiufy]

I'm using the Playback here for all my DAC work.

And how would you rate it against the MSB Diamond DAC, specially in Redbook/PCM ?


alexandre

 

[Bruce B]

And how would you rate it against the MSB Diamond DAC, specially in Redbook/PCM ?
alexandre

What Winston and I heard from his CD test pressings is that the Playback was more true to the original mastered files and the MSB was more musical or more enjoyable to listen to.

 

[Orb]

What Winston and I heard from his CD test pressings is that the Playback was more true to the original mastered files and the MSB was more musical or more enjoyable to listen to.

1st Bruce thanks for keeping up the information and it is both really usesful and interesting.
However the above quote is a bit ironic (and I respect that is what your hearing) because the Playback changes the PCM format in every way more so than MSB, so in theory it is nigh on impossible for Playback to be more true to the original PCM source files.
But the key is also ensuring what filter is used, as I mentioned this has a great impact on what may be deemed closer to the original master files, and a technical headache to decide which filter/co-efficients are closer to the music.
I understand Playback use an apodizing filter for the PCM?
If so it does have some strong advantages along with a few cons, technically it is still not the perfect impulse response and that is where a few of the cons come in (but can be said about nearly every PCM reconstruction filter).

It is a pain to do but a worthwhile test; try also creating original PCM source file to be 96kHz and also 88.2Khz and compare again on both MSB and the Playback - it is worth doing as some hardware chipset or clocking architecture may work better with these native rates.

Tailspn good question regarding DAD AX32 and ESS Sabre but only know about the DA.
Cheers
Orb

 

[ack]

Don't forget, two different analog output sections in those two digital processors...

 

[Bruce B]

1st Bruce thanks for keeping up the information and it is both really usesful and interesting.
However the above quote is a bit ironic (and I respect that is what your hearing) because the Playback changes the PCM format in every way more so than MSB, so in theory it is nigh on impossible for Playback to be more true to the original PCM source files.
But the key is also ensuring what filter is used, as I mentioned this has a great impact on what may be deemed closer to the original master files, and a technical headache to decide which filter/co-efficients are closer to the music.
I understand Playback use an apodizing filter for the PCM?
If so it does have some strong advantages along with a few cons, technically it is still not the perfect impulse response and that is where a few of the cons come in (but can be said about nearly every PCM reconstruction filter).

It is a pain to do but a worthwhile test; try also creating original PCM source file to be 96kHz and also 88.2Khz and compare again on both MSB and the Playback - it is worth doing as some hardware chipset or clocking architecture may work better with these native rates.

Tailspn good question regarding DAD AX32 and ESS Sabre but only know about the DA.
Cheers
Orb

The above quote was from Winston. You'll have to take it in the context that the mastering was done with the PBD DAC.

 

[microstrip]

The above quote was from Winston. You'll have to take it in the context that the mastering was done with the PBD DAC.

Bruce,
What was the sequence of AD and DD conversions used during the mastering? BTW, can we conclude from your previous comments that the more significant gains of using DSD are at the AD stage in the recording phase, not in the playback?

 

[LL21]

What Winston and I heard from his CD test pressings is that the Playback was more true to the original mastered files and the MSB was more musical or more enjoyable to listen to.

Interesting comment given how many audiophiles equate true to mastered original files as the ultimate in 'real music'/more enjoyable...whereas you have distinguished the two...plus you are in the recording industry, not a pure hobbyist. Can you be more specific about what made the MSB 'more musical' or 'more enjoyable to listen to' for you? That is something I am very interested in reading.

 

[jeromelang]

Needle drops are not suitable for this type of tests.
It is always guaranteed on second plays, the sound will be different.
Why wasn't tapes used?