Thanks! I was actually looking for that on the HDTT site.
1 Like
Either I read it wrong or I just donât understand what is written, but the definitions there donât really explain what makes DSD potentially better.
I get different. Just not why (it can be) better.
But, itâs okay. Not looking for an explanation, here and now.
Iâll figure it out over time. Iâll be watching and reading the forums here and in other places.
I know Paul has some YT vids covering this topic.
Unless someone has the link to,
âHow to Explain DSD Compared to PCM And Why DSD Can Be Better, To Your Five Year Oldâ? 
To simplify a little (maybe a lot):
What makes pure DSD better is a much higher sample rate (64, 128, or 256 times as many samples per second than PCM) and itâs a continuous flow of information with no breaks as a natural music stream would be.
PCM is words describing the natural stream of music that have to be put back together into sentences to be played back so the natural flow is lost.
Thatâs about it.
2 Likes
probably a bit like Video Resolution, too, maybe?
At some point (different for each person), our eyes (ears), and our TV (stereo) canât discern a difference.
2 Likes
Thanks!
I get the concept. Mostly.
I think where I get lost is the difference in bit depth.
I am so used to a higher bit depth being better.
Yes. that makes sense for what we see (hear).
Without going overboard on technical details, I want to understand why DSD can (should?) be better than PCM.I think Baldy was able to clear that for me - âcontinuous flowâ.
I understand frequency of the samples.
Trying to wrap my head around the single bit thingâŚ
Donât stress. Iâll figure it out.
Thanks!
i think the idea for DSD is that the chunks are smaller and there are more of them.
Think fine sand vs aquarium gravel.
1 Like
I think where I get lost is that I am stuck thinking in image terms.
The more bits, the more color variance.
24-bit = true color.
And if I understand bits, they are simply, 0 or 1 - off or on.
Is it just that since DSD sample rates are so high, only 1 bit is needed in that moment of capture?
If so, how is off or on the specific sound frequency at that moment of measurement?
I will watch some YT vidsâŚ
You can find more technically accurate vids and pages if you search for âSigma Delta Modulationâ instead of anything audio related like SACD or even DSD. I donât know if that much detail is what you are looking for or not, but the keys to DSD are noise shaping and Sigma Delta Modulation. Even with only one bit the noise shaping allows getting the signal to noise ratio as low as youâd like over the audio band if your sample rate is high enough. With SACDâs the noise floor over the audio band is -120dB, about like 20bit PCM. The noise floor rises as the frequency goes up past the audio band but that can be filtered out with a very gentle analog low pass filter. The other features of one bit audio like linearity also give it a leg up over PCM.
4 Likes
I have always âvisually analogizedâ the advantage offered by âDSDâ as the ability to offer a smoother (and quieter) frequency response, so to speak/picturing a chunky sine wave vs. a near perfectly smooth sine wave. An imperfect (and technically wrong) visualization; but it allows me to set such wonders aside and fall to sleep at night.
Looking forward to any and all technical corrections I deserve to receiveâŚ
1 Like
Ironically the problem with well executed PCM is that it is too smooth - itâs missing the higher frequencies which add the detail.
5 Likes
Got a picture (or two) of that illustrating the difference�
Donât mean to task you but that might be something even I could understand.
1 Like
Does this help?
The top trace is a (very small) section of âThe Tennessee Waltzâ. The bottom is the same thing after lowering the sample rate (i.e. low pass filtering.)
Basically the bandwidth (1/2 the sample rate) determines how much detail you get. Lower sample rate PCM doesnât get more blocky, it gets smoother. The whole digital is jaggy thing is simply wrong.
8 Likes
Perhaps this will make more sense. This a section of the above zoomed in so you can see the samples and what the output of a DAC with those samples would produce:
Hereâs the same thing with 1/4 the samples:
6 Likes
Keeping this short is hard. But the key concept is actually simple.
Start with whatâs absolutely the same about PCM and DSD: what youâre given is a series of relative amplitude values for a waveform at regular time intervals. The waveform by definition is not allowed to have any frequency components greater than half of the sampling rate.
PCM makes more intuitive sense because the visual plot of those points âlooks likeâ the audio waveform we expect to end up listening to. But the way the points translate to sound is essentially the same for DSD⌠itâs just that the set of points given contains not just the audio we want to hear but also a massive amount of random-seeming ultrasonic noise. We canât âseeâ the audio in the points like we can in PCM.
Playback of both types of encoding follows the same overall process (ignoring all the permutations of oversampling and multi-bit SDM etc etc because they still boil down to this):
- For each sample, generate an output voltage of the appropriate level at the appropriate time
- Pass the signal generated in this way through an analog low-pass filter
You see with PCM that the samples occur much less often but individually have a voltage which is much closer to the expected output voltage at that time. DSD has samples at least 64x more often than CD audio but the only voltages it knows are âfully positiveâ and âfully negativeâ.
The trick with DSD is that the samples have been calculated to produce a specific effect after the filter. Because the sampling rate is so high compared to the frequencies we are interested in post-filter, each sample has the effect of ânudgingâ the output closer to where we want it. They literally get averaged out. So you donât look at the current sample of DSD to understand what the audio signal level should be right now. You look at the average value of the past 64 or more. (Thatâs an ELI5 version, not an actual technical instruction.)
And as I said in the other thread DSD vs PCM Hi Rez - your opinion? - #11 by dvorak the primary reason for the difference in sound quality is the freedom to use much gentler filters which do way less damage to the audio that they let through.
2 Likes
Itâs certainly a difficult concept to get your head around. Iâve come up with what I think is a good visual analogy. Take a look at this video.
The âmurmurationâ has always fascinated me but I can see parallels with the way DSD works. In a murmuration of starlings, if you zoom in on an individual bird (read âbitâ) itâs difficult to discern any relationship with the birds (bits) next to it. Itâs only when you zoom out that you start to see these marvelous patterns emerge. In a similar way, as far as I can tell, the mathematical algorithms used to interpret a DSD bit stream canât do much with 1 bit. They need to âzoom outâ and analize a collection of bits but they arenât 16-, 24- or 43-bit âwordsâ in the PCM sense. They make little sense until you âzoom outâ and look at the broader picture.
3 Likes
Oh boy! Seemingly not so simple as a I thought.
But I think that I am getting closer to understanding.
Thanks to everyone jumping into this conversation.
Here are some of my revelations(?).
-
Am I wrong to say that PCM and DSD are not so differently far apart as we might be led to believe? There is a generational relationship between them? The same âDNAâ, just at different levels of evolution? Or more to the point, different levels of processing?
DSD is like a RAW image file closer to what the sensor âseesâ before being processed as a jpeg file. The DSD file is like a preprocessed PCM file? -
Thank you @tedsmith. âSigma Delta Modulationâ is key.
What I have learned here is that only 1 bit is needed because the current sample voltage bit is comparing itself to the previous sample. There is a range, but is this sample higher or lower than the previous sample, yes or no, off or on. Something like this.
And yes, for this pursuit, more detail is needed. There is no way around it. -
Yes @dvorak! I was hung up in the âvisual intuitiveâ plot of PCM matching an audio waveform.
I think fixing my fixation, and the information about âSigma Delta Modulationâ is where my answers lie. -
Filtering is key. Both of you have mentioned this. This is where the âbetterâ comes from right? I mean, with all of the above and more, taken into consideration.
I will continue to dig and report back. Thanks again to everyone!
DSD for Dummies! 
Ooh! I like this. Thank you!
Just saw your response after I had posted my previous postâŚ
âThe murmuration of the starlings is still a mysteryâ⌠The more I dig the more I want to understand the âmurmuringsâ of a DSD file.
But having a visualization for the words is great!
Thank you, again!
1 Like
I donât know if this will help or not, but sigma delta modulators are only one way to get DSD. In the most simple terms possible: DSD is a single bit stream that when filtered gives you what you want. This is literally true. I wrote a program that converted PCM to DSD by navigating the tree of all possible bit streams, filtering them, and then comparing the filtered output to the PCM. Clearly doing this fully takes way too much time so I looked ahead about, say, 10,000 samples and only kept the closest, say, 1024 bit streams at any moment. This did give me a DSD stream that when filtered gave me the PCM I was looking for, literally. In some sense there is nothing magic going on at all.
8 Likes
ââŚNavigating the tree of all possible bit streams.â
Whoa! Deep! Life imitates art?
Zen and the Art of PCM to DSD File Maintenance.
Seriously, I think I get what you are trying to say. Itâs just another file type.Donât get too hung up on the detailsâŚ
I wonât. Just want to overcome the mental block.
And if I am being really honest about this, how can I rationalize for myself why paying more for a DSD file over a PCM file is worth it.
And,
âDonât knock rationalization; where would we be without it?â
âI donât know anyone who could get through the day without two or three juicy rationalizations.â âTheyâre more important than sex (the music?)â
