Reference recording in multi formats released by Japan Association of Professional Recording Studios(JAPRS)

[Bunpei]

The latest issue of Japanese magazine for professional recording engineers, "PROSOUND"

Stereo Sound STORE / PROSOUND 2015

(this web page is in Japanese language)

includes a DVD that contains six audio source files in six different formats, PCM 48kHz/16bit, 96kHz/24bit, 192kHz/24bit, 384kHz/32bit(float), DSD 5.6MHz and DSD 11.2MHz for a single play of "Orfeu Negro" by a Japanese trio, sax.: Y. Tomakomachi, piano: M. Ishizuka & Percussion: S. Ishikawa. A recording engineer is Mr. H. Takada of Victor Entertainment, a chairman of technical committee of Japan Association of Professional Recording Studios (JAPRS). Those files were prepared by capturing an identical analog output signal of API console of CR-506 studio of NHK(Nippon Hoso Kyokai: Japan Broadcasting Corporation) with six DAWs, Avid ProTools for PCM 48kHz, 96kHz, 192 kHz and Merging Pyramix/Horus for 384kHz/32bit, DSD 5.6MHz, DSD 11.2MHz simultaneously.

 

They say their purpose of the release of the sources is to provide "a kind of reference standard sources" for professional engineers to recognize actual differences on sound characteristic depending on formats.　The recording was done in a joint technical seminar of NHK and JAPRS.

 

I think the six source files reflect the differences well. In other words, the highest resolution files of 384kHz/32bit(float) and DSD 11.2MHz make us recognize well enough an advantage of high-resolution recording.

If you are interested in plays of high-resolution sources, I'd like to recommend the sources. The publisher of the magazine offers an oversea shipping. You can order a single issue from the web page shown above.

(If you are very eager to listen to any of them, please send a private message to bunpei.)

 

Recording devices used are,

Mixing console

API

Microphones

Saxophone: Neumann M-149 x 2

Piano: Schoeps CMC-521 x 2 (main) , DPA 4011 x 2 (bass), DPA 4006 x 2 (mid range)

Percussion:<conga> Sanken CU-41 x 2 <cymbal, finger bell, etc.> AKG C-451 x 2 <Cajón> Audio Technica ATM25

Reverberators

Lexicon 960, Lexicon 480

Compressor

Tube Tech LCA2B (for sax only)

[Hiro]

Wonder what kind of ADC they have used, as Sony's professional ADCs max out at 5.6448MHz / 128FS DSD...

 

Ah, just noticed that you mentioned the Horus. It would be nice if they compared the SQ of the Horus vs the Sony...

[Bunpei]

ADC for Pro Tools was HD I/O and for Pyramix was Horus, they say.

I guess SONY DSD recording device is not so popular enough in recent studios in Japan.

 

*** Error Correction ***

The name of sax player is Y. Tomabechi.

[Hiro]

ADC for Pro Tools was HD I/O and for Pyramix was Horus, they say.

I guess SONY DSD recording device is not so popular enough in recent studios in Japan.

 

The Sony Double DSD front-end is an ADC so I presume it can be used with any recorder such as Tascam DA-3000 and perhaps also with the Pyramix system.

[Bunpei]

Hi, Hiro!

 

Would you tell me any web link for the DSD128 ADC unit manufactured by SONY?

 

Bunpei

[Hiro]

Hi, Hiro!

 

Would you tell me any web link for the DSD128 ADC unit manufactured by SONY?

 

Bunpei

 

I have only recently learned of its existence, as it has reportedly been used on the new Cat Stevens DSD128 release.

 

Cat Stevens-Tea For The Tillerman-DSD Double Rate 56MHz128fs Download|Acoustic Sounds

 

The ADC is discussed in this thread, but not much is known about it at this point. (The pic of the Sony DSD ADC comes from Phileweb.com so it's definitely available in Japan - if it's indeed the same SONY ADC Gus Skinas at Super Audio Center used when transferring Cat Stevens master tapes to 5.6MHz/128Fs DSD).

 

I hope that the future DSD tests conducted by Japan Association of Professional Recording Studios will be done with the 5.6MHz DSD SONY unit...

[Bunpei]

A person who knows very well about SONY DSD products for a recording industry told me that the device in the picture was a 8 channel AD/DA unit used for Genex or SONOMA. It covers only 64 fs. Its customer was SME-J.

[Hiro]

A person who knows very well about SONY DSD products for a recording industry told me that the device in the picture was a 8 channel AD/DA unit used for Genex or SONOMA. It covers only 64 fs. Its customer was SME-J.

 

The device is an ADC (not an AD/DA all-in-one), that much I know, because there's a separate SONY DAC unit to go with it.

 

That being said, it may not be the DSD128-capable SONY ADC Gus Skinas used for the Cat Stevens release.

 

Does the person you consulted with know about any 5.6MHz (128fs) DSD capable SONY ADC?

[bmoura]

Since the Recording Engineer on the JAPRS DSD 256fs recording was from Victor Entertainment in Japan, I'd guess that it is the Victor ADC that supports DSD 256fs/11.2 MHz recording that is in use here.

[Hiro]

Since the Recording Engineer on the JAPRS DSD 256fs recording was from Victor Entertainment in Japan, I'd guess that it is the Victor ADC that supports DSD 256fs/11.2 MHz recording that is in use here.

 

They used the all-in-one Horus Ad/DA, as Bunpei reported.

 

As for the 11.2MHz DSD ADC I tweeted about some time ago, I'm afraid, I misatributed the device to Victor company. I've recently learned that the 11.2MHz DSD ADC is used by Prof. Tsutomu Ōhashi, whom you may know from his research on hypersonic effect.

 

[Bunpei]

Oh, Hiro! You're great! I wanted to look at Professor Ohashi's proprietary recording device capable for DSD256. You showed it to me!

I have talked with engineers involved in the development of the system, separately. Its ADC/DAC part was manufactured by a small technology-oriented company in Tokyo. The company inherits DSD technology of SONY lab. Its recorder part was supplied by a famous DSD recording device manufacturing company in Japan.

[Bunpei]

We can purchase and download some albums in DSD256 format created by Professor Ohashi whose stage name as a conductor or a leader of his performance group is "Shoji Yamashiro" via e-onkyo online. My top recommendation is a recording of Java or Bali ethnic music, "Gamelan". It is really high-fi. You may feel something different and original in the sound. My interpretation is that the taste comes from an original delta-sigma modulator designed for Professor Ohashi.

[Hiro]

Oh, Hiro! You're great! I wanted to look at Professor Ohashi's proprietary recording device capable for DSD256. You showed it to me!

I have talked with engineers involved in the development of the system, separately. Its ADC/DAC part was manufactured by a small technology-oriented company in Tokyo. The company inherits DSD technology of SONY lab. Its recorder part was supplied by a famous DSD recording device manufacturing company in Japan.

 

You're welcome. And thanks for the additional information about the development of this most intriguing quad-rate DSD recording equipment. It's a pity that this DSD recorder, as well as the Sony 5.6MHz DSD ADC aren't used more often...

[bmoura]

ADC for Pro Tools was HD I/O and for Pyramix was Horus, they say.

I guess SONY DSD recording device is not so popular enough in recent studios in Japan.

 

*** Error Correction ***

The name of sax player is Y. Tomabechi.

 

Not a surprise. DSD 256 recordings these days generally use the Horus or HAPI converters from Merging Technologies.

[Hiro]

Not a surprise. DSD 256 recordings these days generally use the Horus or HAPI converters from Merging Technologies.

 

Yes, given that at the moment only ONE commercially available ADC can do DSD 256, that isn't surprising.

 

The above reflects the fact that we are currently in the very early days of DSD 256 recording.

 

The other two Double and Quad DSD ADCs, first one from SONY, the latter custom-designed for Professor Ohashi, are unobtanium for the wider market unfortunately. But we should hope that this market will eventually mature, with new generation of DSD 128 & 256 ADCs eventually being introduced, leading to maturing of the technology itself.

[bmoura]

Yes, given that at the moment only ONE commercially available ADC can do DSD 256, that isn't surprising.

 

The above reflects the fact that we are currently in the very early days of DSD 256 recording.

 

The other two Double and Quad DSD ADCs, first one from SONY, the latter custom-designed for Professor Ohashi, are unobtanium for the wider market unfortunately. But we should hope that this market will eventually mature, with new generation of DSD 128 & 256 ADCs eventually being introduced, leading to maturing of the technology itself.

 

There are ADCs planned by Mytek that will support DSD 256 using the Manhattan platform. Some of the Mytek ADC owners out there may be tempted to upgrade when those come out. We will see.

 

For now, the reviews of the Horus and HAPI recorded DSD 256 albums are quite good. So the market for those titles will continue to grow. And indeed, more are on the way. Watch for additional DSD 256 album titles soon....

[Nikhil]

Bringing up another old discussion thanks to this interesting post on another thread. After watching the Youtube of those fascinating custom JBL speakers from Kenrick Audio of Japan, I was intrigued to ask Bunpei (the OP) for more info on the JAPRS recording of "Orfeu Negro". He was kind enough to Dropbox me the files of all six formats which gave me an occasion for an interesting listening session.

 

Before I proceed a small toast to the wonderful world we live in today. Within a matter of days of getting my curiosity aroused on these recordings (and the fantastic jazz scene in Japan), I had with me high quality files ready to go on my setup.

 

So taken from the original post: " ... files were prepared by capturing an identical analog output signal of API console of CR-506 studio of NHK(Nippon Hoso Kyokai: Japan Broadcasting Corporation) with six DAWs, Avid ProTools for PCM 48kHz, 96kHz, 192 kHz and Merging Pyramix/Horus for 384kHz/32bit, DSD 5.6MHz, DSD 11.2MHz simultaneously."

 

... purpose of the release of the sources is to provide "a kind of reference standard sources" for professional engineers to recognize actual differences on sound characteristic depending on formats.

-

 

After listening to these 6 different formats on my setup I penned down the following observations:

 

The PCM recordings are not really that bad at all. If one were to listen to them just by themselves they would provide perfectly good listening. However in comparison to the higher res formats their shortcomings can be observed making them sound a little less lively and limited. More peaks less valleys.

 

The 48kHz file has maybe 80% of the information in terms of the recording but in comparison lacks the decay and texture of the higher resolution files. Transient information is there but with increase in resolution percussive "thuds" become "thumps" and cymbals shimmer just a little longer in the higher resolution files.

 

The 96kHz and 192kHz files are much quieter in comparison with more low level detail. You sense the piece begin to move when Ishizuka and Tomakomachi get going at the 00:55 mark. You can sense another gear shift at the 01:50 mark with Ishikawa getting in on the action a little later. The music does get a little more involving with the additional cues one gets from the additional information. Piano notes linger a little longer and the sax gets a little livelier.

 

Note: I did notice that the 96kHz did not have the ping of the bell/cymbal at the opening which maybe hints that despite being fed a parallel signal some info can get lost.

 

The 384kHz file has much more low level information in comparison. This is after all DXD/DSD64 area. The percussion has a ring to it and texture. You sense the dynamics. You begin to appreciate the shift from mezzo-piano to mezzo-forte as Ishizuka coaxes the notes from the piano. You sense the use of the damping pedal as he moves along with the piece.

 

The DSD 128 and 256 files really bring on another level in realism. Percussion notes hang in the air. Piano notes and chords have more tone. The resonating piano strings add another layer of information. There is a lot more "presence" information in the sound. You can hear the air as Tomakomachi coaxes the notes from his sax. Percussive notes have a feeling of membrane being brushed. To me the difference between DSD128 and DSD256 are again the presence cues that give the sound a little more realism.

 

Note: I noticed that the DSD recording is at a slightly higher volume than the PCM.

 

Clearly with increase in resolution a lot more information is available that adds to the listening experience. I have to (qualify) say that my observations are limited to the extent of my audio chain and vocabulary. But anyone can listen to these recordings and observe for themselves the detail of information that is available as you go higher up in resolution/format.

 

Note: We could also just be observing the quality of the respecting ADC.

 

Thanks again to Bunpei for sharing these files so generously. This has been a very interesting learning experience for me.

 

[Bunpei]

What a detailed review of professional quality! I appreciated it very much.

 

Mr. Uchinuma, president of JAPRS, once stated in a seminar held last autumn, "96kHz is very reasonable or well-balanced when we consider a total value of various aspects." He also said, "Engineers attended the recording event agreed that a replay of DSD256 was the most realistic comparing to a live sound output from a monitoring system."

[Bunpei]

Mr. Uchinuma also added, "the most of participants said a conventional analog reel to reel tape recording is the best in sound quality, frankly speaking".

 

I fully agreed with his last comment.

[Nikhil]

Bunpei,

 

It was very interesting for me to do the comparison and hear the differences. Some differences are subtle and some not. Very frequently we talk about what is hires audio and it seems like stating the obvious when you say DSD is better. With these files one can clearly "hear" the comparison between the 384kHz and the DSD128 for example.

 

The main problem remains that of availability of true hires audio material. While new material is being created the majority of my choice of music is from the CD era and getting a true hires file is not possible. However, demos like these make a strong case for DSD. Listening to the stunning realism of the DSD files versus the PCM versions - one can actually sense the un-involvement of the PCM files.

 

Reel to reel is very interesting but the eye watering prices of tape material (not to mention availability) will keep it a little out of reach for sometime.

 

Regards.

[bmoura]

With the number of DACs that can play DSD 256 music files increasing, more listeners can now check out and enjoy these files at that DSD bit rate. And we're seeing more DSD 256 recorded albums coming available as well. Exciting times indeed!

 

https://www.nativedsd.com/homepage/quad_dsd_music

http://www.dsd.sonore.us

[Nikhil]

Musicscope analysis of the JAPRS files enclosed for those who may be interested.

 

 

PCM 48kHz/16bit:

 

 

PCM 96kHz/24bit:

 

 

PCM 192kHz/24bit:

 

 

PCM 384kHz/32bit(float):

 

 

DSD 5.6MHz:

 

 

DSD 11.2MHz:

 

[Bunpei]

Very valuable results!

 

The frequency spectrum of DSD256 one shows no increase of quantization noise in this frequency range!

[Nikhil]

Very valuable results!

 

The frequency spectrum of DSD256 one shows no increase of quantization noise in this frequency range!

 

Yes. I was also a little surprised to see how "quiet" DSD256 is in the ultra high frequency band. I suspect this might be a clue as to why DSD256 is where the magic starts for the DSD enthusiasts.

[sandalaudio]

I've used these files to do a group listening with a bunch of people at a recent headphone meeting.

 

Although everyone noticed the marked differences between the files, it was quite surprising that there was almost a 50/50 divide in the preference of PCM vs. DSD.

 

A lot of people said the high res PCM version sounded better because the bass feels more solid and direct. Others thought the DSD256 sounded better because it has so much more air and atmosphere information that creates the whole soundscape.

 

It reminded me of the ancient CD player war between multi-bit and one-bit DACs.

 

In any case, it's good to be able to demo these high quality recordings, because it definitely debunks the general myth that higher sample rate is a waste of time (which quite often is true in many albums...).

 

FYI we used Chord Mojo and iFi micro iDAC2 / micro iDSD for the demo, so the opinions may change if other DAC is used.

 

Yes. I was also a little surprised to see how "quiet" DSD256 is in the ultra high frequency band. I suspect this might be a clue as to why DSD256 is where the magic starts for the DSD enthusiasts.

 

I think the fact that DSD256 has no noise hump all the way down to 96kHz, -96dB is a great thing. There will be very little stress to the subsequent amp and it doesn't have to be too picky about the LPF. We can simply let the music do its own thing.

 

One big issue with DSD256 is that it's reaching the limit of USB2.0 DAC throughput. I've been having many issues with pop noise caused by bad PC USB bus or poor USB cable deteriorating the data. This is where high quality USB cable really helps.

 

When it works, it sounds great.