Topic: Sony PCM-D100

[yates7592]

I have been researching and pondering over the relative merits of 2.8 MHz DSD and 24/192 recording with the D100.

Main conclusions so far:

* 24/192 has more information, is more 'detailed', but 2.8 MHz DSD generally sounds 'better' or more 'life-like' than 24/192 when used in a live music recording environment, dynamics and sense of depth/space are more analog-like
* 24/192 can be readily edited in a DAW as you would a 24/44 or 24/96 file
* 24/192 can be burnt to DVD-A (dual layer disc if needed)
* DSD cannot be edited in anything other than fairly expensive software programs (e.g. Pyramix $1,000 - edit)
* You can create your own DSD discs of recordings using the free software 'Audiogate' from Korg. This will let you trim, split, adjust levels up or down, join etc, and then  create the necessary DSD file structure for DVD burning (but not effects/plug-ins/EQ etc)
* Native DSD files can only be easily played on the recorder itself, hooked up to hi-fi (or in Foobar if you have DSD-capable DAC - edit)
* DSD discs created from Audiiogate can be played very nicely on a Playstation PS3 (using the analog out, not digi....)
* Most universal disc players that tout DSD capability will convert the DSD signal to PCM if using digi out
* Most experts seem agreed that 5.6MHz DSD and 24/384 are the future and are as near-analog as currently possible in sound quality

[H₂O]

I have read that you can buy Merging Technologies Pyramix software for under $1000 now - so you can edit DSD for much less nowadays

It's also important to note that most AD converters convert into a bitstream (ala DSD stream) before converting into PCM - so that 24-192 recording you made on ur D100 was a 2.8Mhz DSD stream first - this allows for noise shaping to help offset quantization error

AD's have worked this way for over 20 years

[yates7592]

^

Re the Pyramix software, thanks for that, I found it and it seems quite capable.

Re the conversion, I note in the D100 info released thus far from Sony says there are 2 separate A-D converters for LPCM and DSD, so it sounds like a signal may not be processed in the way you described, I don't know (which is better??, I presume so?) 

[H₂O]

It's unlikely that there is two seperate AD converters as most if not all DSD capible AD's also output 24 bit/192khz or higher pcm

The additional chip and power overhead for DSD would be to get the DSD serial output into a hardware format the main SOC can except -  the nice thing about PCM is that these same AD's support I2S bus which only supports PCM data and is supported by most SOCs (System On a Chip - the CPU in the recorder)

So to interconnect the AD > SOC is fairly trivial for PCM whereas to connect the DSD outputs of the AD to the SOC requires some support chips such as an FPGA, etc

SOC maybe a traditional CPU based solution such as ARM based or a DSP one

[2manyrocks]

What happens if you record video with a separate audio track in DSD and then you try to use the DSD audio track in the video?

If DSD is not 100% compatible with video editing software, then that would also be something else to consider. 

[H₂O]

I don't think DSD is considered to be used in conjunction with Video - it's more for audiofile or audio mastering purposes.

If I where recording video (which doesn't interest me personally for concerts) I would want to use PCM w/ a timecode solution.  If I was not using timecode and mixing in the audio post then it really doesn't matter what it was recording in (either DSD or PCM) for video - there is an extra step of converting DSD to PCM as the end result on the video is PCM.  If I was using a D100 I would record in DSD regardless of video or not as the audio quality would be very important to me (meaning I would want a DSD version of the Audio for personal use or archival purposes).

My main recorder is PCM only but I do like the sound of DSD over PCM there just isn't a more than 2 track portable DSD recorder at present and there may never be as there is no interest in the industry that truly drives the high end market (film/TV).

DSD seems to be a niche audiophile format and other technologies such as 32bit 384+Khz may replace it or maybe not.  Film/Video seems to be locked at 24bit 48Khz for now.

[voltronic]

* Native DSD files can only be easily played on the recorder itself, hooked up to hi-fi

You can play DSD or DFF files with Foobar, and convert them to PCM if you wish:
http://www.exasound.com/Blog/tabid/74/articleType/ArticleView/articleId/61/Configuring-Foobar2000-for-ASIO-DSD-DXD-Playback-with-exaSound-DACs.aspx
http://www.audiostream.com/content/how-play-dsd-file-using-foobar2000

[yates7592]

* Native DSD files can only be easily played on the recorder itself, hooked up to hi-fi

You can play DSD or DFF files with Foobar, and convert them to PCM if you wish:
http://www.exasound.com/Blog/tabid/74/articleType/ArticleView/articleId/61/Configuring-Foobar2000-for-ASIO-DSD-DXD-Playback-with-exaSound-DACs.aspx
http://www.audiostream.com/content/how-play-dsd-file-using-foobar2000

DSD goes in, but is Foobar outputting DSD or LPCM audio?

[yates7592]

It's unlikely that there is two seperate AD converters as most if not all DSD capible AD's also output 24 bit/192khz or higher pcm

The additional chip and power overhead for DSD would be to get the DSD serial output into a hardware format the main SOC can except -  the nice thing about PCM is that these same AD's support I2S bus which only supports PCM data and is supported by most SOCs (System On a Chip - the CPU in the recorder)

So to interconnect the AD > SOC is fairly trivial for PCM whereas to connect the DSD outputs of the AD to the SOC requires some support chips such as an FPGA, etc

SOC maybe a traditional CPU based solution such as ARM based or a DSP one

I must admit I don't understand a lot of what you have said, but to quote Sony's press releases:

"The PCM-D100 is equipped with a dedicated AD converter for both LPCM and DSD, providing specialised functionality for respective modes"; and

"The PCM-D100 uses separate A to D converters for PCM and DSD recording"

Is there any significance in this, or are Sony having us on?

[H₂O]

It is possible that there are 2 seperate AD's but I still am leaning to this statement as being lost in translation between the engineers and Sony marketing

[F.O.Bean]

Count me IN! After 2014 gets here, I am going to buy a D100 and comp it against the 24bit of my m10s

[dnsacks]

* Native DSD files can only be easily played on the recorder itself, hooked up to hi-fi

You can play DSD or DFF files with Foobar, and convert them to PCM if you wish:
http://www.exasound.com/Blog/tabid/74/articleType/ArticleView/articleId/61/Configuring-Foobar2000-for-ASIO-DSD-DXD-Playback-with-exaSound-DACs.aspx
http://www.audiostream.com/content/how-play-dsd-file-using-foobar2000

DSD goes in, but is Foobar outputting DSD or LPCM audio?

if the d>a supports DSD, foobar2000 can support direct DSD (not LPCM) output -- see http://www.matrix-digi.com/pdf/fb2k_dsd_Setup_Guide.pdf for instructions

[Nickelodeon]

It's unlikely that there is two seperate AD converters as most if not all DSD capible AD's also output 24 bit/192khz or higher pcm

The additional chip and power overhead for DSD would be to get the DSD serial output into a hardware format the main SOC can except -  the nice thing about PCM is that these same AD's support I2S bus which only supports PCM data and is supported by most SOCs (System On a Chip - the CPU in the recorder)

So to interconnect the AD > SOC is fairly trivial for PCM whereas to connect the DSD outputs of the AD to the SOC requires some support chips such as an FPGA, etc

SOC maybe a traditional CPU based solution such as ARM based or a DSP one

Hi H2O,

I think I know what you are saying here, and it tends to resolve, in my mind, a conundrum I felt when I first saw the battery lifetimes listed for each of the recording formats supported by the D100 as given on Sony's Pro Audio website; i.e., "25 hrs @44.1kHz/16 bit; 18 hrs @192kHz/24bit or 12 hrs @ DSD2.8" (See - http://pro.sony.com/bbsc/ssr/cat-audio/cat-recorders/product-PCMD100/, and look under Specifications). If most (if not all) DSD capable AD's also output 192kHz/24bit LPCM or higher, then that signal can be sent off directly to the SOC (System On a Chip) without incurring any additional power burden on the battery. But if a straight out of the AD DSD signal also requires ancillary chip hardware, such as an FPGA (Field-Programmable Gate Array), be up and running in order to make that signal understandable to the SOC, then it makes sense that the stated battery lifetime would be shorter with the D100 operating in a DSD only mode. This tends to support your claim that the D100's ADC is of the conventional variety.

However, from the ProSoundWeb press release (see - http://www.prosoundweb.com/article/sony_unveils_pcm-d100_high_resolution_audio_portable_recorder_at_aes_2013/), its unclear as to what Sony really has going on in their ADC. Sony has been known to come up with some proprietary circuit arrangements in the past, so perhaps they have something new going on here as well. I guess we'll all have to wait for some info to come in from those who have taken a peek inside a D100, and/or have seen the schematics of one.

[Nickelodeon]

I forgot to add that the reason I though that the battery lifetimes, as list by Sony for each of the D100's recording formats, was a conundrum was due to my assumption that the conversion of a DSD signal from the D100's AD would require additional PCM conversion hardware to be up and running, which would imply a shorter battery lifetime for the D100's LPCM mode. But if the D100's DSD ADC can output a native PCM signal, and the DSD signal itself requires ancillary hardware to be up and running instead, then the battery lifetimes listed by Sony make a good deal more sense.

[Mike Davis]

Coming soon?

$799 at B&H

http://www.bhphotovideo.com/c/product/1008089-REG/sony_pcm_d100_portable_stereo_field.html