DSD debunked?


#1

Berkeley Audio is announcing a new DAC:



http://cybwiz.blogspot.co.uk/2013/12/berkeley-audio-design-alpha-dac.html



It ships with an offline DSD to PCM converter



From the announcement:



"Careful consideration was given to providing the highest possible reproduction of DSD files by the Alpha DAC Reference Series. 99% of modern DAC’s, including the Alpha Reference Series use mult-bit D/A converters because they provide better performance than 1-bit converters – even those who advertise “native” DSD compatibility. So, at some point, the 1-bit DSD stream must be converted to multi-bit for all of those DAC’s.



We could, like many other manufacturers, convert 1-bit DSD to multi-bit within the Alpha DAC Reference Series and show “DSD” in the front panel display. That would be the easiest approach from a marketing perspective. But that would also mean increasing the amount of processing in the DAC during playback which would degrade audio quality, and audio quality is the reason the Alpha Reference Series exists."



MY TAKE ON THIS:



They are basically saying, we don’t do true 1-bit DSD, never have and never will, and 99% of manufacturers claiming they do, don’t really do native DSD because they convert to multi-bit, which sounds better anyway. I guess the “1%” That does true native 1-bit DSD would be likes of EMM Labs. This must sound like a bit of heresy in the rapidly growing church of DSD converts…



It also appears they are suggesting for the 99% of multi bit DACs parading as DSD dacs (which would include my MSB), you are better off converting offline to High Rez PCM, because it offloads processing duties from the DAC. This would make eminent sense to me, and would be consistent with my experience that (on my MSB DAC anyway) DSD buys you absolutely nothing high rez PCM cannot deliver.



ANY THOUGHTS?





#2

I tend to agree with this concept.

The “bits are bits” argument is “logical”. so I suspect that it is the delivery system that adds to what we hear. Something like my newspaper is the same as my neighbor’s but if it gets wet from the rain, my experience will be different from his.


#3

I may be wrong, I think while we are in digital domain we can convert between data representation format without loss of data, however D/A and A/D conversion may be impacted significantly by the source/target digital data format.


#4

May I ask what model msb DAC ??


#5

Yes Master, bits are only logical. A bit is or is not - there is no maybe. :slight_smile:



Mikhail, transcoding in the digital domain can change the data. Going from DSD to PCM or PCM to DSD involves calculating waveforms and the result of going from one to the other and back is almost certain to create a different result, even if only slightly so. Within the PCM domain when changing sample rates the same is also true. About the only time the data will remain intact would be in going from one bit depth such as 16 bit to a higher bit depth such as 20 or 24 bit while staying at the same sample rate.



On the other hand, when transcoding lossless storage formats such as wav, flac, alac and such without changing the sample rate or bit depth will result in storing the exact same audio data.



J.P.


#6
May I ask what model msb DAC ??


Signature IV plus


#7

Cool how do you like it?



I have a platinum 4 plus coming . Do you use the universal player and music server function??



Thanks for replying.



Al D


#8

The beauty of the binary principle is that ‘either you understand it - or you don"t’ ;))


#9

Well I do not claim to know it, but I do have this question. I use software to run devices not in the audio related industry . And if there is errors the software will not function . Now beyond this is simple computers we all use to perform our everyday tasks , now here is where thy plot thickens . You can perform functions even if some part of the system is missing or corrupted. So I am assuming here that audio could be altered and in effect change the sound and still perform. Again I am just guessing here .


#10
The beauty of the binary principle is that 'either you understand it - or you don"t' ;))


Read "fuzzy logic" by Bart Kostko. Brilliant book.


#11

Ok thanks is it available for downloads. Amazon or kindle?


#12
Cool how do you like it?

I have a platinum 4 plus coming . Do you use the universal player and music server function??

Thanks for replying.

Al D


I was on the wait list to get the demo UMT Plus unit to compare as a streamer against my CAPS 3.0 USB server. However, based on observations of a fellow MSB DAC owner I decide to pass on it (he found the two signal paths to be very close, so no point in spending $7K (UMT Plus + I2S Pro Card) for a lateral move).

However, a very interesting third alternative just presented itself. Blue Smoke audio will release a USB to I2S Pro converter that (according to them!) smokes the USB input on the MSB DAC. Worth looking into in a cost no object system.

http://www.bluesmokesystems.com/

#13

The umt plus is the one I picked up. The system is used about 6 months old. It retailed for about 27500.

I paid 15000 us . So to me it’s worth the try. I am not big on transports , what I really want is a network server solution . It is for my office. I just want to walk in and play music, and pause when I need to to take phone calls. But if does not do it simple and reliable I a just going to use transport .



Thanks for replying.



Al D


#14
edorr said: They are basically saying, we don't do true 1-bit DSD, never have and never will, and 99% of manufacturers claiming they do, don't really do native DSD because they convert to multi-bit . . .

All of this is true. I have posted quite extensively here on the topic of DSD v. PCM.

Almost all DACs convert and decimate both PCM and DSD to multi-bit PDM. To add to the irony, very little DSD is not routinely converted to PCM for editing and processing, and converted back to DSD prior to sale.

It is nearly impossible to listen to a pure DSD file decoded in native format. It is only a bit easier to listen to pure PCM decoded in native format.

J.P. is also correct that transcoding back and forth from PCM to DSD to PCM is not transparent and perfect.

It makes the arguments of DSD v. PCM pretty amusing.

#15

Wow so I should just stop reading about this and not get mad . This whole hi def music industry is just not cool.


#16

Given that most dsd downloads are bull. What about sacd,s ? Especially ones from old recordings?


#17
Yes Master, bits are only logical. A bit is or is not - there is no maybe. :)

Mikhail, transcoding in the digital domain can change the data. Going from DSD to PCM or PCM to DSD involves calculating waveforms and the result of going from one to the other and back is almost certain to create a different result, even if only slightly so. Within the PCM domain when changing sample rates the same is also true. About the only time the data will remain intact would be in going from one bit depth such as 16 bit to a higher bit depth such as 20 or 24 bit while staying at the same sample rate.

J.P.


J.P., Thank you for correcting me! I have never seriously looked into this and it is really interesting. During digital format conversion algorithm has to virtually calculate 'analogue' signal. Theoretically, analogue signal can be completely recovered (limited by noise and frequency). As such, if I convert from 24/44 to 24/96 and back, I should come up with exactly same data. Is my logic wrong?

@Elk Can you please point me to info on why PCM<->DSD is not transparent/perfect?

#18

@edorr Debunk…reminds me of Scully’s assignment in the X-Files…to debunk Maulder’s work.


#19
mikhail said: Can you please point me to info on why PCM<->DSD is not transparent/perfect?

I'm sorry; I do not have a ready source to point you to.

Both are very high resolution formats, and capture higher frequencies and greater dynamic range than we can hear. The errors upon transcoding are minute.