Русский      Audiophile Inventory on Facebook Audiophile Inventory on Google+ Audiophile Inventory on Twitter Audiophile Inventory on YouTube Audiophile Inventory on Instagramm Audiophile Inventory on Pinterest     

What is DSD Audio [Quality. DSD PCM Comparison. Players. Converters]


Audio Basis - educational articles

DSD is format to store audio ♬ in high resolution. Read about audio quality ► players ► converters ► file formats ► editing ► DSD versus PCM ► other issues. Brief variant of the article here.

1. DSD features2. Bit depth and noise level3. Sample rate4. Noise, Maximal Level and Stability Issues5. DSD in figures6. DSD vs PCM7. File formats and compression method8. DSD players9. DSD converters10. DSD editing softwareDSD audio samplesConclusions


• Input DSF, DFF, ISO (1-bit audio) is supported in Audiophil-E, PROduce-R and configurable Modula-R
• Output DSF, DFF are supported in maximal PROduce-RD and configurable Modula-R
• For ISO tracks, DSF, DFF with length more 3 minutes FREE demo mute 2 second silence in the output middle
• DVD ISO is NOT supported

What is DSD audio?


Watch educational video: DSD vs FLAC comparison


1. DSD features


DSD features
Abbreviation of:  Direct Stream Digital
Audio data coding method: sigma-delta modulation (details and video)
Bit depth: 1 bit or more (details about sound quality)
Sample rate: DSD64 (2.8 MHz), DSD128 (5.6 MHz, double), DSD256 (11.2 MHz, quad), etc. (details)
Channel number: Stereo or multichannel
Medium: SACD optical disk, including hybrid SACD (with CD layer), computer files [SACD ISO (ripped albums from SACD) and DSF, DFF, CUE+DSF/DFF] (details)
Specification: "Scarlet book" (1999)
Applications: Music production, home hifi/hiend audio

Direct Stream Digital is one of high resolution audio formats to improve CD-audio dynamic range in the audible band.

Read below about sound quality issues (noise, bit depth, band, sample rate, DSD versus PCM).



2. Bit depth and noise level


As rule, this format have bit depth: 1-bit. And quantization-error (noise) energy have significant level.

However, there is, so-called noise shaping of spectrum is used. And quantization-error energy is re-distributed.


Noise shaping (NS) of 1-bit signal (spectrum)

Noise shaping of 1-bit signal


The quantization noise spectrum (at the left of the picture) have level comparable with musical signal. Sigma-delta modulator push significant part of the energy from low to high frequency range, out of audible band (0 ... 20 kHz).

When the 1-bit record is played back, low frequency filter into sigma-delta demodulator cut the noise.


DSD decoder (demodulator)

DSD decoder


Therefore, at spectrum error level into the audible range is comparable with multibit pulse code modulation's (PCM) one.

Also sigma-delta modulation may have multibit resolution. Read details below and watch video at this page.



3. Sample rate


The format use standard range of sample rates based on 44100 Hz:

  • DSD64     = 44100 *   64   =   2 822 400 Hz =   2.8 MHz
  • DSD128   = 44100 *  128  =   5 644 800 Hz =   5.6 MHz
  • DSD256   = 44100 *  256  = 11 289 600 Hz = 11.3 MHz
  • DSD512   = 44100 *  512  = 22 579 200 Hz = 22.6 MHz
  • DSD1024 = 44100 * 1024 = 45 158 400 Hz = 45.2 MHz
  • Etc.

Also 48000 kHz base is possible. There are not technical limitations for the sample rate value. But compatibility issues are very probable.



4. Noise, Maximal Level and Stability Issues


When sigma-delta modulator is designed, the engineers pay attention to:

  • noise level in audible range and
  • tolerance to overload.

To solve these issues the engineers have:

  • bit-depth,
  • sample rate,
  • noise shaping.


These parameters should be considered in complex.


Bit depth

Bit depth reduce quantization error level as itself.


Noise shaping

NS "push" the quantization error energy out of audible range.

We can suggest that more energy may be pushed out of audible range in sigma-delta modulator. But it demands more steep noise-shaper.

Steeper NS can cause higher probability of glitch (broken stability) of sigma-delta modulator when input overload happens.
After glitch, silence or some oscillations are generated at output of sigma-delta modulator.
After broken stability the modulator should be forcibly reset.

Overload tolerance of sigma-delta modulator

Overload tolerance of sigma-delta modulator


Sample rate

Higher sample rate reduce error spectrum level due distribution of its energy into wider band. Also it give ability to use sloper NS.

Energy is square of figure, concluded between spectrum line and horizontal axis into band 0 ... [sample rate]/2.


Sample rate and quantization-error-spectrum level

Sample rate and quantization noise level


At the left and right pictures squares of the noise energy figures are same. But figure, that more expanded in horizontal axis, give lesser noise level.

Higher sample rate allow tio reduce noise level in audible frequency range. It allow to reduce noise shaping steepness, that increase overload tolerance of the modulator.



We can see that lower noise and higher stability to overload can be achieved different ways.

As example, better quality is noise shaping implementation matter for same bit depth and sample rate. But , on the other hand, we can increase sample rate and/or bith depth to decrease error level for same NS method.



5. DSD in figures


Pro audio modulators have noise level in audible frequency range for sample rates (read details):

  • D64 about -125 ... -145 dB (comparable with PCM 24 bit)
  • D128 about -165 dB (better than pulse code modulation 24 bit)
  • D256 and higher about -170 ... -200 dB (comparable with PCM 32 bit)

Noise level in the audible frequency band almost don't depend on demodulator. But the noise should be maximally suppressed out of the band. Ultrasound noise can cause audible intermodulations distortions.


Read more about DSD vs DSF vs DFF >



6. DSD vs PCM


Direct Stream Digital (sigma-delta modulation) is like to pulse code modulation, but quantization error spectrum is shaped for decreasing noise into audible range.

We can apply NS for usual PCM. But difference here is band reserve for pushed noise out of audible frequency range.


DSD versus PCM



PCM have lesser band reserve (above audible range) than sigma-delta modulation format, but pulse code modulation have higher bit depth.

Noise shaping for pulse-code modulation is implementation matter too, like sigma-delta modulation.

Therefore, no format advantages as itself. But implementation makes difference.

Sigma-delta-modulation decoder (demodulator) is 2-position (1 / -1) voltage generator and low frequency filter. It is simpler than pulse-code-modulation hardware demodulator. Because pulse-code demodulator contains multi-voltage matrix or 1-bit decoder(s). So we have more abilities to make cheaper 1-bit DAC better than multibit one.


Read details here >

Also look to infographic DSD versus FLAC >



7. File formats and compression method


  1. Sigma-delta modulation is applied on optical disks SACD (Super Audio CD).
  2. The modulation contains into files DSF, DFF, SACD ISO (SACD disk image).
  3. WAV, FLAC as DoP (DSD over PCM) container.

SACD disc may be losslessly converted to SACD ISO [1], [2], [3].

SACD ISO image may be extracted to DSF and DFF files. The lossless extraction may be applied.

Read more about DSF and DFF.

1-bit audio files (DSF, DFF, SACD ISO) and disks may be either uncompressed or compressed with DST method.

DoP is open protocol that allow pack 1-bit audio as multibit for compatibility with software and hardware [4]. DoP cannot be played as usual pulse-code modulation.


Also 1-bit audio may be streamed via network.

Uncompressed DSD64 demands capacity 2.7 Mbit/sec = 44100 Hz * 64 / 1024 / 1024.


Direct Stream Digital sources

DSD sources, mediums


Also index file CUE + DSF/DFF audio files may contains 1-bit album.



8. DSD players


DSD audio players (hardware and software) is able to playback all or some of 1-bit file formats.

1-bit files may be played back directly on DSD DAC/player or converted to PCM "on fly" to playback to PCM DAC/player. About SACD conversion read more here

1-bit playback may be performed via special ASIO-driver under Windows, including DoP (DSD over PCM) digital audio packing format (example).

SACD optical disks may be played back at hardware player. Author know nothing about available SACD-drives for consumer computers to playback/record SACD optical disks.

Stereo player may downmix multichannel to stereo. As alternative multichannel files may be pre-converted to stereo. It allow to save space on hard disk of audio player. Downmix is lossy audio processing. Its quality defined by implementation.

Read more about the players here > and here >


9. DSD converters


DSD converters perform:

  • conversion DSD to PCM,
  • conversion PCM to DSD,
  • extracting SACD ISO to DSD,
  • extracting SACD ISO to PCM,
  • resampling 1-bit audio,
  • altering volume,
  • volume normalization,
  • click suppressing on track borders,
  • other processing.

Read how to works the converters here >


Converter list (convert some or all of DSD files [DSF, DFF, SACD ISO])



10. DSD editing software


Read main article about DSD editing >

DSD editing is sophisticated issue due modulation noise. Non-linear processing can cause audible intermodulation distortions by ultrasound noise.

Currently no information, that is known for author, about "native" 1-bit processing (example: gain altering, resampling, etc.) without conversion 1-bit to multibit and back. Except, merging/dividing audio file.


Read article about DSD versus DSF versus DFF >


DSD editing

DSD editing


PCM here may be considered as "multibit DSD". Pulse code modulation is not obligatory mean "24 bit / 352 kHz" or so on. Author recommend use 32- or 64-bit float point bit depth. This PCM contains high frequency modulation noise. But for conversion this "multibit DSD" to 1-bit need re-modulation with NS.

Losses of editing with 1-bit/multibit conversion is comparable with resampling.

Recording studios may distribute 1-bit records without editing.

There is DXD format. It is PCM (as rule "24 bit / 352 kHz" or so on) with high sample rate and bit depth and legacy DSD noise. However the noise can cause audible intermodulation distortions. Before non-linear processing, the high frequency noise cutting is recommended.

Video: Looks to experiment with cutting ultrasound noise




Read the articles



DSD audio samples


Audio samples may be downloaded here:

  1. Demo tracks #1
  2. Demo tracks #2
  3. Test simple signals





DSD is 1-bit audio data (sigma-delta modulation) format for hi-fi audio.

For same sample rate and bit depth the format may have different quality, that depend on implementation.

DSD digital-to-analog converter is simpler than PCM one. It allow to get better quality simpler way.

DSD may be edited via 1-bit to multibit and back conversion. The edition is lossy, except merging/dividing.



Read more about 1-bit format

  1. DSD vs. PCM. Real competitors? >
  2. DSF vs PCM. What is common base? >
  3. DSD Decoder Audio >
  4. How work sigma delta modulation in audio >
  5. DSD Converter of Audio Files: What Inside? >




  1. SACD ISO ripping (way #1)
  2. SACD ISO ripping (way #2)
  3. SACD ISO ripping (way #3)
  4. DoP specification



Read the articles

Copyright © Yuri Korzunov
["Audiophile Inventory" (since 2011), "AudiVentory", "Audiophile Inventory by".
Также "Audiophile Inventory by" (прежде "Audiophile Inventory") охраняемое коммерческое обозначение],
2010-2018. All Rights Reserved.

Site map      Terms and Conditions