• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required as is 20 years of participation in forums (not all true). Come here to have fun, be ready to be teased and not take online life too seriously. We now measure and review equipment for free! Click here for details.

Does DSD sound better than PCM?

Status
Not open for further replies.

pkane

Addicted to Fun and Learning
Patreon Donor
Joined
Aug 18, 2017
Messages
514
Likes
444
#42
So converting from PCM to DSD (or DSD to PCM) in software introduce noise..... ic, but audiophiles do claim that they hear a better quality sound though....
DSD is not a wonderful format that most on the other forums think. It's a capable format that can work just as well as PCM, and can reproduce at least the same signal, with the same accuracy, as PCM, maybe at the cost of significant file size increase. It's not limited by the same Nyquist frequency as PCM, but that's of little value when comparing it to say 24/96KHz. I don't need all those extra frequencies, as I can't hear ultrasound and they are ultimately filtered out anyway before output.

The quantization noise generated by DSD is normally dealt with in a proper implementation. It's shifted out of audio band, and then filtered out.

What DSD does allow is a simpler hardware relative to a PCM. You don't even need a proper DAC to convert DSD to analog, an output filter will work if fed the DSD 1bit stream. Maybe in some DAC instances DSD can result in a higher fidelity playback. Then it might make sense to convert. Maybe, but not IME.
 
Last edited:
Joined
Dec 7, 2018
Messages
2
Likes
1
#43
My answer would be, compare yourself and draw your own conclusions. It was what I did, and I heard the differences, the clearest for me was the bass much more present and defined, with more texture, I also noticed an improvement in the position of the instruments, and most important for me, that made me adopt the DSD as the main format, in most songs, sounded more natural and refined, with a timbre closer to reality.

So I felt the difference and I say that it is better than PCM, I see that many disagree, so you'd better compare to whether or not you like format. I compared DSD and PCM using my LH Geek Out v1, Sennheiser HD580 and Hifiman EF5 amp.
 

Kal Rubinson

Addicted to Fun and Learning
Joined
Mar 23, 2016
Messages
867
Likes
650
Location
NYC/CT
#44
SACD is MultiChannel and for those who care, this is great. still MCH is easily accomplished with PCM, e.g. the various Dolby and DTS surround formats
None of these have had any significant impact on music recordings. Almost all multichannel music recordings released on disc are SACD. Multichannel PCM (not Dolby/dts) has recently become a factor in file downloads.
.... meanwhile DVD-A is DEAD.
Unfortunate but true. So?
If there were enough demand you could see streaming MCH music. I recently clocked 125 Mb/s on my North Miami Internet connection ... Enough to pump out any sane number of (PCM :D) channels/ MCH with 4 K video on top of it ...
Sure. Mebbe it will happen, mebbe not. Has nothing to do with DSD.
 

graz_lag

Addicted to Fun and Learning
Joined
Oct 13, 2018
Messages
845
Likes
643
Location
Le Mans, France
#46
Care to let us know what DSD does better than PCM?
I am not an expert in recording technologies but a consumer of music as many of us in here ... And I happily consume it in both DSD * and PCM formats, hoping they can both live in parallel as complementary recording methodologies for many years to come ... :)
* for that reason, I purposely bought a dedicated DSD DAC, the DS-DAC 100 from Korg

Reading here and there, experts in recording say for example :

#1 - "DSD is the best way to archive analog recordings if one needs maximum flexibility to release in different digital formats. DSD can be converted easily, without decimation errors into any PCM sample and bit rate. PCM is not so flexible."
#2 - "If a project requires multi-tracking, dubbing, and EQ corrections, PCM is a far better recording choice than DSD. Many bands use PCM, click-tracks, and multi-tracking because, frankly, they're not good enough to do an entire song in one take perfectly. Other reasons for using PCM include recording situations where instruments can't be recorded all in one take such as when a drum track is needed behind an acoustic guitar. Drums are so much louder that they bleed into the guitar's microphones, which requires either isolating the drums (sometimes in another room in the studio) or forcing the drummer to play at unnaturally low levels."
#3 - "DSD is best for live recordings of acoustic concerts. If one only needs two tracks, plan to mix down (if needed) real-time and don't need to use any overdubs or additional tracks, DSD is the way to go. With DSD, the recording engineer can easily generate a 44.1/16 version as well as 96/24 and 192/24 files. With even the simple the editing program AudioGate from Korg one can do edits, fade-ins and fade-outs, and all his PCM conversions, if needed."
 
Joined
Nov 24, 2018
Messages
25
Likes
67
#47
DSD is not a wonderful format that most on the other forums think. It's a capable format that can work just as well as PCM, and can reproduce at least the same signal, with the same accuracy, as PCM, maybe at the cost of significant file size increase. It's not limited by the same Nyquist frequency as PCM, but that's of little value when comparing it to say 24/96KHz. I don't need all those extra frequencies, as I can't hear ultrasound and they are ultimately filtered out anyway before output.

The quantization noise generated by DSD is normally dealt with in a proper implementation. It's shifted out of audio band, and then filtered out.

What DSD does allow is a simpler hardware relative to a PCM. You don't even need a proper DAC to convert DSD to analog, an output filter will work if fed the DSD 1bit stream. Maybe in some DAC instances DSD can result in a higher fidelity playback. Then it might make sense to convert. Maybe, but not IME.
There is nothing wrong in this post, but DSD can't produce same quality as PCM does in REAL world because of ultranoise. Actually, even PCM is not converted on its own bit and sampling rate. In modern delta-sigma DAC chips, PCM files are converted to about 5bit, oversampled sample rate file which is 5bit version of dsd. The reason that chip manufacturers use 5bit architecture is to reduce the amount of ultrasonic noise. It means simpler filter stage on amplifier and better performance.

The analog filter is far from ideal filter and is very hard and expensive to improve the performance without additional noise and distortion. So most popular dsd format, dsd64 is a very hard load to dac system. Even dac chip manufacturers recommend different output filter design for dsd and show lower spec on dsd playback condition.

Of course, the actual difference of pcm and dsd on properly designed dac is not audible. I just wanted to say that modern filter is not perfect yet and ultranoise of dsd file causes additional distortion on audible frequency in practical condition.
 

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
18,973
Likes
19,236
Location
Seattle Area
#48

Blumlein 88

Major Contributor
Joined
Feb 23, 2016
Messages
5,774
Likes
4,844
#49
I am not an expert in recording technologies but a consumer of music as many of us in here ... And I happily consume it in both DSD * and PCM formats, hoping they can both live in parallel as complementary recording methodologies for many years to come ... :)
* for that reason, I purposely bought a dedicated DSD DAC, the DS-DAC 100 from Korg

Reading here and there, experts in recording say for example :

#1 - "DSD is the best way to archive analog recordings if one needs maximum flexibility to release in different digital formats. DSD can be converted easily, without decimation errors into any PCM sample and bit rate. PCM is not so flexible."
#2 - "If a project requires multi-tracking, dubbing, and EQ corrections, PCM is a far better recording choice than DSD. Many bands use PCM, click-tracks, and multi-tracking because, frankly, they're not good enough to do an entire song in one take perfectly. Other reasons for using PCM include recording situations where instruments can't be recorded all in one take such as when a drum track is needed behind an acoustic guitar. Drums are so much louder that they bleed into the guitar's microphones, which requires either isolating the drums (sometimes in another room in the studio) or forcing the drummer to play at unnaturally low levels."
#3 - "DSD is best for live recordings of acoustic concerts. If one only needs two tracks, plan to mix down (if needed) real-time and don't need to use any overdubs or additional tracks, DSD is the way to go. With DSD, the recording engineer can easily generate a 44.1/16 version as well as 96/24 and 192/24 files. With even the simple the editing program AudioGate from Korg one can do edits, fade-ins and fade-outs, and all his PCM conversions, if needed."
So when are you getting to the advantages of DSD. #2 isn't an advantage. #3 is rarely going to be the situation for commercial recordings. Rarely as in not even 1/10th of 1% of the time. Plus I see no problem making PCM into another format. Easy to do and is done all the time. Which also means #1 isn't really a particular advantage of DSD.
 

cjfrbw

Active Member
Joined
Mar 5, 2018
Messages
283
Likes
238
#50
From my own memory archives, DSD was a recording format developed by Sony because they purchased archives of analog master tapes and wanted a digital recording medium that would best conserve the analog impulse response and waveform. Thus, DSD was a RECORDING medium, not necessarily a playback medium. The original intent did not take editing into account nor did they at first regard it as a commercial medium. The SACD was a playback method for DSD files that Sony decided to commercialize, because it fit in with their one bit digital stream philosophy and desire to re-vivify digital marketing because standard PCM and CDs had lost a bit of their credibility and luster as audiophiles began to 're-discover' and resurrect traditional analog media.

SACD relies on noise shaping to work due to the signal to noise tracking. The noise shaping algorithms were proprietary and influenced the 'sound' a lot, which became kind of a pseudo analog type of thingie. Some listeners seemed to be negatively impacted by the noise shaping rather than 'fooled'. Standard DSD has high nominal frequency response, but the bit depths drop off above 8Khz in standard DSD, and thus, standard CD PCM actually has more bits in it's high frequency response within the audible range.

For some reason, I have always preferred PCM playback to SACD. SACD sounds kind of nice at first, but bothered me for some reason after listening to it for a while, for reasons I can't fathom except for preference. I used to get in trouble with the SACD fanatics on Audio Asylum when I questioned SACD sound quality.
 

andymok

Active Member
Joined
Sep 14, 2018
Messages
145
Likes
45
#51
One of the technical merits of DSD, is its high sampling rate/bandwidth allows it to better track the transient in the content. It can do what PCM ADC cannot do. It will still be better even after converting to PCM cause the transients / dynamics were there.

It is like shooting JPEG in-camera vs converting JPEG from a RAW.
8-bit vs 12/14-bit > 8-bit, everybody can tell the difference.

or

a film shoot at low fps and low shutter speed vs high shutter speed, high fps down-convert to lower fps
natively [email protected]/60s vs [email protected]/240s > 30fps (reminds me of "Billy Lynn's")
 
Last edited:

March Audio

Major Contributor
Manufacturer
Patreon Donor
Joined
Mar 1, 2016
Messages
3,507
Likes
3,015
Location
Perth Western Australia
#52

pkane

Addicted to Fun and Learning
Patreon Donor
Joined
Aug 18, 2017
Messages
514
Likes
444
#54
One of the technical merits of DSD, is its high sampling rate/bandwidth allows it to better track the transient in the content. It can do what PCM ADC cannot do. It will still be better even after converting to PCM cause the transients / dynamics were there.

It is like shooting JPEG in-camera vs converting JPEG from a RAW.
8-bit vs 12/14-bit > 8-bit, everybody can tell the difference.

or

a film shoot at low fps and low shutter speed vs high shutter speed, high fps down-convert to lower fps
natively [email protected]/60s vs [email protected]/240s > 30fps (reminds me of "Billy Lynn's")
I'd be curious to see any evidence of 'better transients / dynamics' in DSD compared to PCM.
 

graz_lag

Addicted to Fun and Learning
Joined
Oct 13, 2018
Messages
845
Likes
643
Location
Le Mans, France
#57
With some good dose of pragmatism, going back & forth around the point that either DSD or PCM are technically better than one another is a waste of intellectual resources, I would like to say ... :rolleyes:
I mean, what's wrong with the scenario in where both exist as complementary methodologies ? :oops:
More complicated hardware since it has to decode both formats ? Very well I would like to add, more R&D, more technical technological development, more demand of skilled people by the manufacturers, higher sales as hardware, software, music for the whole audio industry, so more business for every one involved at the end ... That's is pure capitalism, nothing less nothing more than that. :cool:
Do not like that ? If so, go to North Korea ! o_O
 

pkane

Addicted to Fun and Learning
Patreon Donor
Joined
Aug 18, 2017
Messages
514
Likes
444
#59
I mean, what's wrong with the scenario in where both exist as complementary methodologies ? :oops:
... That's is pure capitalism, nothing less nothing more than that. ...
Purely out of intellectual curiosity and the desire not to waste my own time and money on a format that doesn't provide any improvement, I'd like to know if there really is a difference and what that might be. Capitalist argument may work for a real marketplace, but I'm not running a marketplace in my listening room ;)
 
Status
Not open for further replies.
Top Bottom