PCM to DSD conversion

Subjects that don't have their own home
User avatar
Ray P
No idea why I do this anymore
Posts: 6294
Joined: Thu Nov 22, 2007 5:18 pm
Location: Somerset

#16 Re: PCM to DSD conversion

Post by Ray P »

Nick wrote: Fri Sep 04, 2020 5:52 pm I would be very surprised if there was not some DSD to PCM and back involved in that...
Isn't the converse true regarding some DACs; anyone using a delta-sigma DAC, such as the ESS range, is experiencing a modicum of 'DSD' because they convert incoming PCM to SDM? That fits your earlier comment about things just happening in different places.

My feet are starting to loose contact with the bottom about now...
Sorry, I couldn't resist!
Neal
Shed dweller
Posts: 2299
Joined: Fri Aug 03, 2007 10:57 am
Location: From the land of the Bodgers

#17 Re: PCM to DSD conversion

Post by Neal »

Some googling shows DSD was indeed an archival format for old analogue tapes, mastering requires a conversion to PCM at some point.

Edit, ok this article implies that there can be a conversion to analogue then back again to avoid going to PCM during mastering....still don’t get the point :D

https://dsd-guide.com/what-difference-b ... d-and-sacd
Only the Sith deal in absolutes.
User avatar
Nick
Site Admin
Posts: 15706
Joined: Sun May 06, 2007 10:20 am
Location: West Yorkshire

#18 Re: PCM to DSD conversion

Post by Nick »

Ray P wrote: Fri Sep 04, 2020 6:49 pm
Nick wrote: Fri Sep 04, 2020 5:52 pm I would be very surprised if there was not some DSD to PCM and back involved in that...
Isn't the converse true regarding some DACs; anyone using a delta-sigma DAC, such as the ESS range, is experiencing a modicum of 'DSD' because they convert incoming PCM to SDM? That fits your earlier comment about things just happening in different places.

My feet are starting to loose contact with the bottom about now...
Not sure what your point is, I said that DSD was similar to Delta Sigma. There are multi bit DS DAC's, it doesn't have to be single bit, but I agree that generally DS involves noise shaping, but as the process is internal I assume it can achieve a higher data rate. Not sure what "a modicum of 'DSD' " means, but its not a race I have a dog in. I generally believe that if two DAC's sound different at least one of them is broken in some way, and the problem probably lies in the analogue domain, either output stage or power supply.In the interest of symmetry I also think r2r dacs are equally broken in a different way, or at least limited by production tolerances.

After much messing with different DAC's I found the solution that worked for me was an off the shelf DAC chip put together with a simple valve balanced analgue stage combined with taking care about isolating stages and using lot of regulators on the chip (one per supply line). I lost interest after that

But what do I know, I much prefer vinyl.

My comment about manipulating DSD was just a observation that the format was not designed with such in mind and that to do so would involve conversion to and from DSD. I didn't consider conversion to analogue, but as Neal points out, that’s a possible method.
Whenever an honest man discovers that he's mistaken, he will either cease to be mistaken or he will cease to be honest.
User avatar
Ray P
No idea why I do this anymore
Posts: 6294
Joined: Thu Nov 22, 2007 5:18 pm
Location: Somerset

#19 Re: PCM to DSD conversion

Post by Ray P »

Nick wrote: Fri Sep 04, 2020 9:27 pm Not sure what your point is...
I don't think I was making any particular point beyond the observation that there can be multiple data conversions in the chain.
Sorry, I couldn't resist!
User avatar
Nick
Site Admin
Posts: 15706
Joined: Sun May 06, 2007 10:20 am
Location: West Yorkshire

#20 Re: PCM to DSD conversion

Post by Nick »

Ray P wrote: Sat Sep 05, 2020 7:44 am
Nick wrote: Fri Sep 04, 2020 9:27 pm Not sure what your point is...
I don't think I was making any particular point beyond the observation that there can be multiple data conversions in the chain.
Yep, though not all conversions are equal. Some are more lossy than others. Some are lossless, such as PCM -> USB -> PCM.
Whenever an honest man discovers that he's mistaken, he will either cease to be mistaken or he will cease to be honest.
User avatar
Ray P
No idea why I do this anymore
Posts: 6294
Joined: Thu Nov 22, 2007 5:18 pm
Location: Somerset

#21 Re: PCM to DSD conversion

Post by Ray P »

Prompted by a mail about the NativeDSD website being updated, I had another look for the DSD info I couldn't locate before and found it straight away :scratch:

Posted just for info/interest.

https://dsd-guide.com/

https://dsd-guide.com/faq/what-dxd-it-dsd
Sorry, I couldn't resist!
User avatar
Nick
Site Admin
Posts: 15706
Joined: Sun May 06, 2007 10:20 am
Location: West Yorkshire

#22 Re: PCM to DSD conversion

Post by Nick »

If thats in response to
Nick, I've had a quick look but can't find it at the moment but
I didn't see any of that info about DSD encoding, just a lot of articles. Ifs there one I should look for?
Whenever an honest man discovers that he's mistaken, he will either cease to be mistaken or he will cease to be honest.
User avatar
Ray P
No idea why I do this anymore
Posts: 6294
Joined: Thu Nov 22, 2007 5:18 pm
Location: Somerset

#23 Re: PCM to DSD conversion

Post by Ray P »

Nick wrote: Sat Sep 12, 2020 10:43 am If thats in response to
Nick, I've had a quick look but can't find it at the moment but
I didn't see any of that info about DSD encoding, just a lot of articles. Ifs there one I should look for?
I've not had time to drill down yet Nick - supposed to be working - later.
Sorry, I couldn't resist!
User avatar
ed
retired
Posts: 5384
Joined: Thu Jun 21, 2007 4:01 pm
Location: yorkshire
Contact:

#24 Re: PCM to DSD conversion

Post by ed »

alert...... skeptic mode on

Thought I'd have a fiddle this morning(I think I mentioned it in the dac update thread):

downloaded the AUI ConverteR(haven't got round to direct mastering with the studio setup yet)
converted one of my tunes from wav to dsf D64(2.8mhz)
updated picoreplayer to v6.1.0(not sure if it was necessary)
continued playing the wav version and the dsd version alternately for some time

So far I can't detect any difference except the dsd is significantly down on volume without touching any controls.
down on volume means some immediate loss of detail
increasing volume on dsd version brings back the detail, but no more than was there with the wav version.

the E30(dac) momentarily displays 356 on the display then resumes displaying the SPL(0) and the PCM indicator(not sure what this means at the mo)
I obviously need to do more testing and listening, but I'm disappointed in not hearing any differences, however slight.


objectively the file sizes: wav=34.3mB, dsf(D64/2.8mhz)=137mB
considering that's the lowest resolution dsd I'm struggling to remain positive.
end of skeptic mode
There's nowhere you can be that isn't where you're meant to be
Wolfgang
Old Hand
Posts: 229
Joined: Mon May 07, 2018 3:08 pm

#25 Re: PCM to DSD conversion

Post by Wolfgang »

This article gives a nice overview without being too complicated.

https://headfonics.com/dsd-vs-pcm-real-competitors/

Here are some results from this article that make most sense IMO:

If reconsider some of the common myths or opinions of DSD and PCM we have 3 we would like to focus on:
1. DSD can be considered as format superior in audio quality than PCM.

For DSD noise is pushed to inaudible (high) part of the spectrum. For pushing (noise shaping) significant energy of noise out of the audible range you need a reserve of the band. I.e. higher sample rate, than for PCM, is needed.
PCM quantization noise correlates with useful signals: no signal equals no noise. DSD noise does not depend on signal and is present in silence also. DSD DACs eliminate this noise.

Applying a DSD-DAC allow maximally simplify scheme and adjusting of DAC. DSD DAC is a simple low-frequency filter (that passes low-frequency audio only). It is a higher sample rate than for PCM and simplifies the design of an analog filter. No need for steep transients to the suppression area as you need for PCM. No need for so many precise components.
Almost all modern DAC use internal PCM to DSD conversion for digital-analog conversion. If you use DSD as an end-user format you need 1 precise reference voltage and simple analog filtering only. I.e. same result with less effort than “native” PCM.

I have yet to compare digitizing/restoring to the analog of the square wave for PCM and DSD. There is an advantage in DSD due to its steeper front and less ringing in front/end sides of square impulse. Let consider how to ideally digitize/restore the square wave.
A Square wave has an infinite spectrum. I.e. for ideal digitizing/restoring you need infinite sample rate. DSD has a significantly higher sample rate than that of PCM. It is the reason for a steeper front/end of the square impulse.
Lower ringing for DSD is the result of a shallower filter than that used for PCM signals that have lower sample rates. On the flip side, using wider (more 20…24 kHz) bands for DSD generates more noise energy particularly beyond 24 kHz. I.e. price of a better form of a square wave is a higher noise level.
Lesser ringing due to a shallower DSD DAC filter (less ringing) leads to worse filtration. Thus leading to a higher noise level. By increasing the PCM sample rate it is possible to achieve a steeper front/end of the square impulse also. I.e. no difference between DSD and PCM in the approach to restoring the square wave. There are results of sample rate and filter steepness only.
Digital audio format quality is the degree of loss of an original analog waveform that is restored from digital forms.
1. It is technically impossible to compare DSD and PCM as pure digital formats. You need to compare released systems, that use these formats.
User avatar
ed
retired
Posts: 5384
Joined: Thu Jun 21, 2007 4:01 pm
Location: yorkshire
Contact:

#26 Re: PCM to DSD conversion

Post by ed »

I did a couple more at D256
a 28mB wav went to 454mB
I still couldn't detect a difference
I will persevere with the listening to DSD

It occureed to me that a client(however unlikely it may be) asking for a master demo on CD may require upwards of 10 discs for an album.
There's nowhere you can be that isn't where you're meant to be
User avatar
Ray P
No idea why I do this anymore
Posts: 6294
Joined: Thu Nov 22, 2007 5:18 pm
Location: Somerset

#27 Re: PCM to DSD conversion

Post by Ray P »

As I experienced the rare occurence of earning money on a Saturday today I thought I would splash some of it on NativeDSD and I've just been listening to this;

https://www.nativedsd.com/catalogue/alb ... =dsd-256fs

recorded as DSD256. The hour long album unzips to about 9GB. It sounds spectacularly good.
Sorry, I couldn't resist!
User avatar
Nick
Site Admin
Posts: 15706
Joined: Sun May 06, 2007 10:20 am
Location: West Yorkshire

#28 Re: PCM to DSD conversion

Post by Nick »

Here are some results from this article that make most sense IMO:
IMHO its a bunch of waffle that seems to be all over the place. Any time you see someone talking about square waves having a infinite spectrum you know you have someone who either doesn't understand information theory or is trying push dogma. Yes, DSD (which the text seems to conflate when it wants with DS modulation) may have a higher sample rate than PCM, but PCM can change its output from - to + full scale in one of those steps, DSD can only change one bit's worth of output in its sample time.
Whenever an honest man discovers that he's mistaken, he will either cease to be mistaken or he will cease to be honest.
JamesD
Old Hand
Posts: 997
Joined: Tue Oct 09, 2007 4:26 pm
Location: North Yorkshire

#29 Re: PCM to DSD conversion

Post by JamesD »

Hi Nick,

With regard to the rate of change of information per quantisation step, shouldn't we consider the amplitude of the change per unit time? In other words during the single step change in the PCM signal multiple step changes occur in the DSD signal and hence the single step comparison is an apples and pears comparison?

Doesn't it come back to the point you made earlier about the amount of information carried by the two signals - with PCM carrying more info than DSD supports and therefore having a higher potential for fidelity and a faster change in information per unit time?

Of course that doesn't mean that a given person would necessarily prefer PCM to DSD, sound quality is much more complex that that but it does mean that DSD isn't technically better - even if it does sound better to some people...

We're back to horses for courses and supplier bullsh*t to help sell their wares... I made the point in a different thread that white papers are an attempt by people with vested interests to persuade others to support their vested interest and should be regarded with all due scepticism - much like forum posts :D :D :D

I do wonder what the real information rate is for unique information per unit time across the end to end audio chain is? That is from original ADC to final DAC - I suspect its a lot less than we all think it is or than we would like it to be... and that would help inform these debates - although I think this is more about euphonic digital signal processing than unique audio information processing...

ciao

James
User avatar
Nick
Site Admin
Posts: 15706
Joined: Sun May 06, 2007 10:20 am
Location: West Yorkshire

#30 Re: PCM to DSD conversion

Post by Nick »

With regard to the rate of change of information per quantisation step, shouldn't we consider the amplitude of the change per unit time? In other words during the single step change in the PCM signal multiple step changes occur in the DSD signal and hence the single step comparison is an apples and pears comparison?
Yes, of course, that was my point, I was comparing that with the claim that was made in the quoted text that DSD is better at encoding a transient (in the example the edge of a square wave) because the sample rate is higher, ignoring the amount of information than can be encoded in the single step.

It seems to be full of double speak. DSD relies on the fact that the information we are interested in is either at a low frequency so that it can be encoded by the bit stream with sufficient fidelity, or if at higher frequency is of low amplitude such that the rate of change available from the single bit stream is sufficient to encode the signal. And that's fine, and correlated to what we find in the real world of audio, But to then state that the higher bit stream frequency (as opposed to word steam frequency) allows it to better encode a square wave is contradictory IMHO.

In effect DSD has a built in slew rate. The fact that its built into the data stream is one of the reasons that filtering is simpler (because much of the filtering has been done in the digital domain).

One way I think you could look at DSD is a encoding method that allow the transfer of audio in a lower data rate carrier by selectively removing information that is decided plays no useful part in what we listen to.

I guess my problem with it is:

1. DSD is seen as better than PCM because it can be reconstructed by the use of a simple low pass filter, so is more analogue (in some way)
2. Single bit DAC's are seen as less musical than r2r DAC's in part because the single bit DAC has to do digital filtering which in some way removes fidelity because of bogy-men like transient response and even more evil pre ringing.

Those two views seem to me to be contradictory

Don't get me wrong, If it sounds better then it is better, its the justifications that are given for that which I think requires logical analysis.
Whenever an honest man discovers that he's mistaken, he will either cease to be mistaken or he will cease to be honest.
Post Reply