Post by Iain August 19, 2014 (321 of 352)
|
|
rammiepie said:
An excellent article, Wilhelm.
But painting the audiophile community as dorks, dweebs, whackos and 'drunken troglodytes' is so PCM, wouldn't you say?
Article is from 2006 and is old and irrelevant to current environment.
|
|
|
|
|
|
tailspn said:
Because unlike PCM, there are no samples in DSD (1-bit two level PDM). It's the density modulation of a repetitive waveform which is bounded by two voltage/current levels. Since that waveform (square wave)(bit clock)(whatever) is such a high frequency/repetition rate, compared to the signal modulating it, the waveform looked at within the bounds of the modulating signal passband frequency spectrum appears continuous. Even to the naked eye on an oscilloscope. It looks analog, acts analog...you know, quacks like a duck.
Lets agree to disagree Carl. We're viewing this from different perspectives.
Tom - no sampling rate? Every single piece of literature known to man about PDM or DSD defines them with a sampling rate using 1-bit samples. Density? It is the density of successive digital bit values - 1 or zero - that pulse density modulation refers to. After 30 seconds into the Wiki articles, you will get the idea.
You are saying those bits only appear to be digital bits, because it is really analog. As I said in my previous digression, all digital transmission involves special 1/0 pulse encoding on an analog electronic signal. So, digital transmission is a separate and distinct, special case of analog transmission, and analog and digital transmission are not to be confused. Receiving devices will not like it.
But, how do you know for sure that what you see on your 'scope only appears to be analog, when it is, in fact, digital? Sure, if you filter any digital stream of pulses enough, DSD or not, it can be turned into a continuous looking, apparently analog signal. The result of that may or may not be garbage, however. But, you seem to be having serious trouble seeing the difference between digital and analog. There is no such difficulty in interpreting the pits on a DSD disc. They are clearly 1/0 bits and nothing else. But, for some strange reason once inside the DAC, you refuse to call them digital bits. Once you filter the bitstream, it is not DSD or PDM anymore. Only then after filtering and other processing does it become analog, but it has lost its identity as DSD.
My computer, unlike your oscilloscope, thinks DSD is digital. Fooled again? I cannot plug a DSD bitstream into to one of the analog sound card inputs. Sony/Philips also thought it was digital, which is why they named it Direct Stream Digital. If you can find a DAC or player manufacturer who thinks DSD is analog, we sure would all like to know. What does your friend Ed Meitner say?
When it comes to matters of opinion, I am fine with agreeing to disagree. I respect you, and normally I respect your opinions. But, when it comes to matters of elementary fact, I have to ask why are you in total denial of all the established evidence, spinning and spreading distorted misinformation? If you were not normally a very credible and influential guy, I could let it go. But, here we have a mystery. It is just not like you.
|
|
|
Post by tailspn August 19, 2014 (323 of 352)
|
|
Iain said:
Article is from 2006 and is old and irrelevant to current environment.
So Iain, other than Sony no longer giving Sonoma workstations to labels to kick-start DSD recording, what's "old and irrelevant to current environment" in that article?
|
|
|
Post by jackan August 19, 2014 (324 of 352)
|
|
Fitzcaraldo215 said:
When it comes to matters of opinion, I am fine with agreeing to disagree. I respect you, and normally I respect your opinions. But, when it comes to matters of elementary fact, I have to ask why are you in total denial of all the established evidence, spinning and spreading distorted misinformation?
This
(I would "this" the whole post, but the software will not allow the length.)
DSD is very different from PCM, single bit.
But it is still bits. Still digital. Better digital, but still digital.
Until it is decoded, then it is analog, not DSD.
If it were analog, one would not need HDMI to plug it into something.
If it were analog, one would not need DACs, but could plug straight into an amplifier with an rca cord.
Both are digital.
One is change over a set period of time.
One is time over a set amount of change.
The set amount of change is the "single bit".
|
|
|
Post by jackan August 19, 2014 (325 of 352)
|
|
tailspn said:
So Iain, other than Sony no longer giving Sonoma workstations to labels to kick-start DSD recording, what's "old and irrelevant to current environment" in that article?
You are correct about this. But, other than that, there is nothing in this article to support your position.
Yes, it can be argued that DSD is a superior format to PCM.
No, it can not be argued that DSD is an analog format.
It is not. DSD is digital. In every way. Including the name.
Very different from PCM. Not analog.
|
|
|
Post by tailspn August 19, 2014 (326 of 352)
|
|
jackan said:
If it were analog, one would not need HDMI to plug it into something.
If it were analog, one would not need DACs, but could plug straight into an amplifier with an rca cord.
Jack, don't believe all you've read/heard about DSD in the past. It's wrong.
For starters, DSD is not transmitted over HDMI. DSD (1-bit two level PDM) is encoded into .dst file format at SACD authoring, and it is that format transmitted over HDMI. But more to the spirit of you point, DSD is a modulated density of pulses bounded by two levels. Being only two levels, it's therefore transmittable, and storable/retrievable over a digital medium. That doesn't qualify it as being digital (for it contains no quantifiable digital values, only varying density of pulses who's density represent a reference-able signal level). But it does allow for digital transmission and storage.
You don't need a Digital to Analog Converter to retrieve the signal content. All you need is a passive integrator analog circuit per DSD channel. Read back to the Ted Smith/Audio Asylum link for how his PS Audio Direct Stream "DAC" circuit operates.
|
|
|
|
|
|
Hi Tom,
Your explanation of DSD and it's characteristics relative to PCM and analog are probably the best summation of the technology I've read to date. The point that may be confusing our fellow sa-cd.net contributors is, and correct me if I am wrong, but DSD performance emulates the manner in which a analog signal unfolds with each bit's amplitude relative to the previous bit as opposed to having an assigned value per multi bit word length PCM. Essentially DSD is a digital encoding process that closely emulates the manner in which analog is encoded. Digital behaving and performing as would analog in simple semantics.
|
|
|
Post by AmonRa August 19, 2014 (328 of 352)
|
|
Idiotic discussion. DSD is digital, as a numerical value is assigned to the direction of the pressure change (voltage value) of the audio signal. 1 for one direction, 0 for the other. That alone makes it digital/numerical. Different from PCM, sure, but just as digital anyway. Just the method is different and in practice DAC is needed to convert it back to analog signal just as with PCM. If it was analog, no filtering etc would be needed.
|
|
|
|
|
|
|
jackan said:
http://www.positive-feedback.com/Issue26/dsd.htm to support your position.
"After some consideration, I have decided to weigh in on the DSD/PCM debate, not because I think I have any particular insight on the distribution side, but because I have the benefit of some unique experience on the front end…
…specifically, what DSD is capable of as a RECORDING medium."
|
|
|
Post by tailspn August 19, 2014 (330 of 352)
|
|
off the grid said:
....DSD performance emulates the manner in which a analog signal unfolds with each bit's amplitude relative to the previous bit as opposed to having an assigned value per multi bit word length PCM
Hi Neal,
Thank you for your comments. The amplitude/direction change from bit-to-bit idea of DSD has been probably the biggest confusion factor clouding its understanding since DSD introduction. It sure confused me.
It implys that each bit has a numerical weight, giving rise to the idea that it's a sample, like PCM. It's all much simpler than that.
Picture a picket fence, where its length instead of being distance, is time. Also, this is a special picket fence where pickets can neighbor one another without spaces, like a solid wall. And an individual picket can either be present, or absent in each of the picket positions on our fence. Lastly, every picket has the same height, as they would be in a real fence.
Lets define a rule that if all the pickets are present (solid surface), the signal level represented over the period of time (length of the fence) that all pickets are present is Maximum (Positive). Also, if there are no pickets present, over that length of time, the signal level represented is Minimum (negative). Lastly, if the pickets alternate being present, and absent, then we have half the number of pickets, and the signal level represented is half way between Maximum, and Minimum. We'll call that Zero level.
So, now if we have a way of placing pickets on our fence, as we're running down it as Time progresses, such that as our signal level goes higher than Zero, more pickets are correspondingly nailed to our fence filling in the empty spaces (Zero is represented by alternating pickets and no pickets), until the signal level reaches Maximum, and the all pickets are present (solid surface).
You get the idea if the signal level is decreasing from Zero, then we'll be removing pickets until we're at Minimum represented level, and there's no pickets.
That, in a simplified nutshell, is 1-bit two level PDM (DSD). It's the density, the INTEGRAL, of the pickets present at any time that represents the signal level. There's no numerical values represented, or expressed as with digital PCM. No individual picket has any expressed value, but the density of pickets integrated over time, determines the signal level represented. No digital, very analog.
I apologize if I insulted by this primitive depiction.
|
|
