Mark Waldrep and More Silliness...

[Paul R]

All Recordings Are DSD…NOT!

 

I just ran across this, and while I admit he has some justification, this is one of the most ridiculous and vindictive articles I have ever read. It takes recycled bull corn to a whole new level of odiferous unpleasantness.

 

Make your own decision of course, but I suggest this article is just crazy talk from someone that should know better. It would be interesting to "follow the money" here and see what is really driving him. If he is not just crazy, then he has to have a financial interest somewhere in this.

 

I will personally not provide any further income to this person. I would urge other people to consider doing the same.

 

The above is my personal opinion only. It does not reflect the opinion or policies of the management.

[One and a half]

When I looked just now at the article it had 808 visits in a little over 12 months.

That's 2 visits a day, perhaps a little more with this link.

 

Storm in a (doll's) teacup.

[alfe]

All Recordings Are DSD…NOT!

 

I just ran across this, and while I admit he has some justification, this is one of the most ridiculous and vindictive articles I have ever read. It takes recycled bull corn to a whole new level of odiferous unpleasantness.

 

Make your own decision of course, but I suggest this article is just crazy talk from someone that should know better. It would be interesting to "follow the money" here and see what is really driving him. If he is not just crazy, then he has to have a financial interest somewhere in this.

 

I will personally not provide any further income to this person. I would urge other people to consider doing the same.

 

The above is my personal opinion only. It does not reflect the opinion or policies of the management.

 

Calling for boycott, I think you are over reacting Paul.

[firedog]

Again, he's being a stickler for the term "DSD" meaning 1 bit and being something separate from multi-bit DSD.

 

Another person who doesn't understand that in audio vernacular, "DSD" refers to both 1-bit and multi-bit DSD.

 

If you aren't a stickler for that distinction, his whole argument falls apart.

[Paul R]

When I looked just now at the article it had 808 visits in a little over 12 months.

That's 2 visits a day, perhaps a little more with this link.

 

Storm in a (doll's) teacup.

 

Wow - that's pretty dismissive.

 

Things get out on the internet and never go away.

 

Someone here will be using that as total justification for some oddball belief sooner or later.

[Paul R]

Calling for boycott, I think you are over reacting Paul.

 

Perhaps.

[mansr]

Again, he's being a stickler for the term "DSD" meaning 1 bit and being something separate from multi-bit DSD.

 

Another person who doesn't understand that in audio vernacular, "DSD" refers to both 1-bit and multi-bit DSD.

 

If you aren't a stickler for that distinction, his whole argument falls apart.

Show me a single release in multi-bit "DSD". Then show me a DAC capable of playing it. Until you can do that, equating DSD with 1-bit is perfectly logical.

[Teresa]

I have a question.

 

If I understand correctly in Delta-Sigma Modulation (DSM) the change in the signal is encoded rather than as an absolute value as in PCM.

 

DSD is a trademark and the actual process is DSM but many here use DSD interchangeably with DSM.

 

So my question is: Does multi-bit DSM encode the same way as single-bit DSM? If so is it incorrect to call any type of DSM (change in signal) PCM (absolute value)?

[mansr]

I have a question.

 

If I understand correctly in Delta-Sigma Modulation (DSM) the change in the signal is encoded rather than as an absolute value as in PCM.

 

DSD is a trademark and the actual process is DSM but many here use DSD interchangeably with DSM.

 

So my question is: Does multi-bit DSM encode the same way as single-bit DSM? If so is it incorrect to call any type of DSM (change in signal) PCM (absolute value)?

 

That's an oversimplification at best.

 

Delta-sigma modulation (or sigma-delta if you prefer, same thing) is a method for applying spectral noise shaping when quantising a signal. It can be used for PCM, differential PCM (DPCM, which indeed encodes the sample to sample difference), or 1-bit "DSD" which can be seen as a degenerate form of PCM, DPCM, or PDM, whichever is convenient.

[firedog]

Show me a single release in multi-bit "DSD". Then show me a DAC capable of playing it. Until you can do that, equating DSD with 1-bit is perfectly logical.

 

Logical? Except to all the designers and industry professionals who regularly use the term multi bit DSD, and relate to it as a subset of DSD - and all the hobbyists who do the same. Your pedantic side doesn't like it, but the reality is that common usage doesn't make the distinction that you insist on between 1 bit and multi bit "DSD". Waldrep makes the same mistake in his blog, leading to a false argument/ retort.

[mansr]

Logical? Except to all the designers and industry professionals who regularly use the term multi bit DSD, and relate to it as a subset of DSD - and all the hobbyists who do the same. Your pedantic side doesn't like it, but the reality is that common usage doesn't make the distinction that you insist on between 1 bit and multi bit "DSD". Waldrep makes the same mistake in his blog, leading to a false argument/ retort.

 

Your reasoning appears to go something like this:

 

1. 1-bit DSD is produced by an SDM process.

2. An SDM can also produce multi-bit output.

3. Therefore any output from an SDM is DSD.

4. All DACs include an SDM stage.

5. Therefore all DACs convert to DSD.

6. Therefore PCM playback cannot outperform 1-bit DSD.

 

If you can't see the flawed leaps of logic in this, there's probably no point trying to educate you.

[tailspn]

I have a question.

 

If I understand correctly in Delta-Sigma Modulation (DSM) the change in the signal is encoded rather than as an absolute value as in PCM.

 

DSD is a trademark and the actual process is DSM but many here use DSD interchangeably with DSM.

 

So my question is: Does multi-bit DSM encode the same way as single-bit DSM? If so is it incorrect to call any type of DSM (change in signal) PCM (absolute value)?

 

Hi Teresa,

 

The difficulty that everyone has in the beginning with understanding DSD/DSM/SDM/PDM (confusing already, too many terms describing the same thing) is attempting to define/understand "DSD" in terms of Pulse Code Modulation. They're completely different processes using completely different techniques to arrive at completely different results. Their only commonality (from our interest) is they're both used to encode and deliver music.

 

To your question; yes, it's incorrect to call ANY type of DSM Pulse Code Modulation (I prefer the term Pulse Density Modulation, for it describes the end result, not the process used to achieve it). The reason is simple, there's no ABSOLUTE VALUE present in "DSD". PCM, on the other hand, is only a series of 2's compliment binary words VALUES at some bit depth, and at some discrete sampling rate.

 

"DSD" (insert your favored term) is a very simple concept, implemented through a complex modulation process of modulating a bit clock with an audio signal to produce a varying density bit stream proportional to the audio signal's amplitude. The only thing "digital" about the RESULTING bit stream is that it's storable and retrievable in a digital media. For 1-bit two level (DSD), that's all there is. It's an analog signal expressed in the density of bits within a continuous bit stream.

 

To make "DSD" useable beyond just playing it back, several processes have been used to date, including conversion into PCM, and employing parallel multi-bit PDM techniques. The latter is simply used to create a digitally processable digital VALUE based system acceptable to a digital processing system.

 

Hope this helps.

[Jud]

there's probably no point trying to educate you.

 

There's really no point in trying to "educate" any of us. Having a pleasant conversation from which we might learn something is extremely welcome, however.

[Jud]

Hi Teresa,

 

The difficulty that everyone has in the beginning with understanding DSD/DSM/SDM/PDM (confusing already, too many terms describing the same thing) is attempting to define/understand "DSD" in terms of Pulse Code Modulation. They're completely different processes using completely different techniques to arrive at completely different results. Their only commonality (from our interest) is they're both used to encode and deliver music.

 

To your question; yes, it's incorrect to call ANY type of DSM Pulse Code Modulation (I prefer the term Pulse Density Modulation, for it describes the end result, not the process used to achieve it). The reason is simple, there's no ABSOLUTE VALUE present in "DSD". PCM, on the other hand, is only a series of 2's compliment binary words VALUES at some bit depth, and at some discrete sampling rate.

 

"DSD" (insert your favored term) is a very simple concept, implemented through a complex modulation process of modulating a bit clock with an audio signal to produce a varying density bit stream proportional to the audio signal's amplitude. The only thing "digital" about the RESULTING bit stream is that it's storable and retrievable in a digital media. For 1-bit two level (DSD), that's all there is. It's an analog signal expressed in the density of bits within a continuous bit stream.

 

To make "DSD" useable beyond just playing it back, several processes have been used to date, including conversion into PCM, and employing parallel multi-bit PDM techniques. The latter is simply used to create a digitally processable digital VALUE based system acceptable to a digital processing system.

 

Hope this helps.

 

Thanks Tom. I really like one of the Mallinson brothers' explanations using two ways to calculate an average price of gasoline as an example. It's in an RMAF video, I believe. I'll see if I can find it and link to it.

[Hiro]

Show me a single release in multi-bit "DSD". Then show me a DAC capable of playing it. Until you can do that, equating DSD with 1-bit is perfectly logical.

 

There are plenty of DSD DAC architectures in use today, and some use multiple 1-bit DACs. Such SDM architecture for DSD DACs was embraced by Philips long time ago.

 

You are right that some ADCs use 5-bit SDM front ends but as Miska said in another thread their noise performance isn't that much different.

[Hiro]

Your reasoning appears to go something like this:

 

1. 1-bit DSD is produced by an SDM process.

2. An SDM can also produce multi-bit output.

3. Therefore any output from an SDM is DSD.

4. All DACs include an SDM stage.

5. Therefore all DACs convert to DSD.

6. Therefore PCM playback cannot outperform 1-bit DSD.

 

If you can't see the flawed leaps of logic in this, there's probably no point trying to educate you.

 

DSD chain is not subjected to brickwall filters, decimation filters and oversampling filters, so looking at the 1-bit vs 5-bit issue alone is moot. And it is doubly moot because DSD DACs can also have 5-bit delta-sigma conversion stages.

[Paul R]

That's an oversimplification at best.

 

Delta-sigma modulation (or sigma-delta if you prefer, same thing) is a method for applying spectral noise shaping when quantising a signal. It can be used for PCM, differential PCM (DPCM, which indeed encodes the sample to sample difference), or 1-bit "DSD" which can be seen as a degenerate form of PCM, DPCM, or PDM, whichever is convenient.

 

Indo not think PDM can be called a degenerate form if PCM. They qre not mathematically the sme thing at all...

[mansr]

Indo not think PDM can be called a degenerate form if PCM. They qre not mathematically the sme thing at all...

Then how is it that a standard low-pass filter turns DSD into PCM?

[Hiro]

Indo not think PDM can be called a degenerate form if PCM. They qre not mathematically the sme thing at all...

 

It's PCM that's a degenerate form of SDM, as the analog signal is first captured using SDM by delta-sigma ADCs, and only after that converted to a PCM form. 99% of PCM recordings are downsampled from 128FS SDM to 44.1kHz (1FS) PCM.

[MikeJazz]

There's really no point in trying to "educate" any of us. Having a pleasant conversation from which we might learn something is extremely welcome, however.

+1. Some folks are in a very bitter mode these days...

[tailspn]

Then how is it that a standard low-pass filter turns DSD into PCM?

 

It can't and doesn't. PCM is a series of 2's compliment binary words expressing actual signal amplitude values for the windows of time its sampling rate dictates. They're real binary encoded values, in 2's complement form to allow for positive and negative values with a mid scale "0" reference.

 

A Pulse Density Modulation (SDM/DSM/PWM/DSD whatever) bit stream has no actual expressed amplitude value(s), no data value, binary, or any other coded value. It's an audio signal modulated bit clock bit stream pure and simple. The integral of the bit stream is/are continuously varying analog levels. The bit stream itself is the percent of modulation, where 50% modulation has been defined as 0dB.

 

Low pass filtering (signal integration) turns DSD into continuous analog levels, which it is in its DSD form modulating a bit clock carrier.

[Paul R]

Then how is it that a standard low-pass filter turns DSD into PCM?

Ummm - it does not.

[jabbr]

I think if we remove the term "DSD" from this discussion entirely, and replace with SDM, and qualify SDM with the rate and depth in the same way that PCM is qualified by rate and depth, that much confusion would go away. Similarly people would not be able to so easily write stupid comparisons of SDM and PCM "one has more information than the other" (gag)

 

Take the following quote from Waldrep and apply my suggestion:

Modern PCM sigma-delta converters produce much lower error signals than 1-bit sigma-delta DSD converters. The errors in the DSD system are due to the 1-bit quantization that occurs in 1-bit sigma-delta DSD converters. Multi-bit PCM sigma-delta converters can be fully dithered and do not suffer from this un-dithered truncation. Every added bit reduces the noise signal by 6 dB. A 4-bit sigma delta converter is 24 dB quieter than a 1-bit sigma-delta DSD converter. Right from the start, 1-bit DSD signals have higher losses than multi-bit PCM signals.

 

Obviously this isn't true comparing 16/44 PCM with DSD512. Obviously. It is also obvious that multibit *at the same rate* reduces the noise signal. The trick here is that he is using "DSD" to refer to DSD64 and comparing this to arbitrarily high bit depth/rate PCM. If we stop using the term "DSD" and instead "SDM" then can make fair mathematical comparisons.

[jabbr]

Indo not think PDM can be called a degenerate form if PCM. They qre not mathematically the sme thing at all...

 

Dithered 1 bit PCM?

[mansr]

It can't and doesn't. PCM is a series of 2's compliment binary words expressing actual signal amplitude values for the windows of time its sampling rate dictates. They're real binary encoded values, in 2's complement form to allow for positive and negative values with a mid scale "0" reference.

 

A Pulse Density Modulation (SDM/DSM/PWM/DSD whatever) bit stream has no actual expressed amplitude value(s), no data value, binary, or any other coded value. It's an audio signal modulated bit clock bit stream pure and simple. The integral of the bit stream is/are continuously varying analog levels. The bit stream itself is the percent of modulation, where 50% modulation has been defined as 0dB.

 

Low pass filtering (signal integration) turns DSD into continuous analog levels, which it is in its DSD form modulating a bit clock carrier.

 

Ummm - it does not.

 

If that's what you believe, then try this:

 

1. Download my version of Sox with DSD support: https://github.com/mansr/sox

2. Pick a DSD file and process it thus: sox file.dsf -b24 file.wav rate -v 96k

 

The output consists of the input digitally filtered with exactly the same low-pass filter it uses for PCM inputs. The DSD signal is simply treated as PCM with each sample at one extreme or the other.