Snakeoil from Sony?

Anything to do with computer audio, hardware, software etc.
Clive
Posts: 205
Joined: Mon Oct 07, 2013 11:12 pm

Re: Snakeoil from Sony?

Post by Clive »

sbgk wrote:
Clive wrote:If they reduce noise as well as they do with their camera sensors then they have a winner. There are some areas of digital they excel at.
why aren't they saying it's good for pictures as well ? are audiophiles more gullible ? it's easier to prove you are seeing something different rather than proving you're hearing a difference.
The wars on the camera forums are every bit as twisted as the hifi ones. Differences people claim I sometimes struggle to perceive. Pro photogs vs amateurs is similar to engineers vs audiophiles. Huge parallels.
nige2000
Posts: 4253
Joined: Thu Feb 14, 2013 10:47 am
Location: meath

Re: Snakeoil from Sony?

Post by nige2000 »

sd card player, modded soekris dac, class a lifepo4 amp or gb class a/b amp, diy open baffle speakers based on project audio mundorf trio 10's
sbgk
Posts: 1950
Joined: Mon Oct 07, 2013 9:45 pm

Re: Snakeoil from Sony?

Post by sbgk »

Sony explain the audio SD card design

http://av.watch.impress.co.jp/docs/seri ... 91795.html

seems to make sense
sbgk
Posts: 1950
Joined: Mon Oct 07, 2013 9:45 pm

Re: Snakeoil from Sony?

Post by sbgk »

Long story short, they tried to reduce RF noise radiation of memory card by decreasing bus clock of NAND controller

thanks to yamamoto2002

Following is the translation

Digital Audio Laboratory Issue 626

Interview with the developer of Sony premium sound microSD card

Some details abount the plot without number and stability/quality control

Recently SR-64HXA premium sound microSDXC card is released on March 5 from Sony.
AV Watch reports this news article and this news is propagated by Twitter and Facebook.
Most people treats this news as usual audio cult, something like
"Sound cannot be changed because it is digital media"
"I'm concerned about Sony for releasing such a snake oil product"

I recognized this news rather late, about the half days later in the midnight,
by Twitter message something like "Please investigate about the product" from several people.
Testing audio performance of such product is typically difficult but I thought
Sony does not release the product that has no effect so I asked about the interview with the
developer of Sony via editorial depertment of AV Watch. This is how this interview materialized

Why they develop such a product? Is there firm ground for improvement of sound quality?
I asked question to four development staffs, Singo Asou, project leader, Youzou Goto, product planning,
Kenichi Satori, design leader and Yosirou Sato, he is in charge of sound quality.
former 3 people work for device Solution business unit storage media department Memory media product division
and their primary task is no relation to the sound quality, they are focused on "larger capacity, faster speed,
and higher reliability"

Interviewer: I'd like to know how this development project started.

Goto: My motto as a planner is "to propose unique memory card only Sony can"
something different memory card other than simply increase capacity.
This product is the result of such way of thinking.
Quality of some aspect of read access of memory card makes difference of sound quality.
I thought that is only Sony can, other competitors is difficult to create such a product.

Asou: Every year we got propose from other business units in Sony about "premium sound storage media"
and on the summer of the last year, Credibility of sound improvement rises and the project is started.

(Figure: SQ improvement approach from the memory card. electronic noise stands for EMI,
noise frequency is higher than audible frequency.)

Interviwer: We knows SQ of analog media such as cassette tape tend to be different by the magnetic material
and everyone can hear the difference. But how flash media which contains only digital media influence the SQ?
It seems weird.

Asou: Certainly. I've developed about DRM technology before and thought "SQ must be the same because it is digital"
there was no conception about premium sound memory card. we thought the same as Twitter people at first.

Satori: All engineers on this business unit had the same opinion at
that time. therefore it is normal many people feels strangeness on this product.
I cannot believe the story about the sound difference from memory cards from Sato in the walkman development team.

Interviewer: Hmm. I understand this project is not proposed from the memory media division.

Asou: Our division produces storage media and in the past we released premium sound CDR so
this is not the first time to release the media focused on sound quality.
CD and DVD has caracteristics of some kind of analog-ish behavior and SQ can be improved.
but we did not think flash media can improve SQ at that time.

Interviewer: Then you are heard from Sato that SQ can be improved by memory card

Sato: February this year we released NW-ZX2 walkman and the end of last year NW-A10 high resolution walkman.
These devices equip microSD slot and this enables the new feature, playback from memory card.
on the course of the development of walkman, we tested sound quality and found sound is heard
differently from internal flash and microSD. From listening test, sound difference is obvious so
I thought it must be visualized by some measurement. I measured frequency response, SNR, dynamic range and THD+N
and other characteristics but difference was not able to be recorded.
But difference is there and it can be heard by listening so I contacted to memory media product division.

Satori: I got the request from Sato so I provided several sample memory card with different parts,
different firmware etc. among this set, there are the same internals only different paint is contained.
I provide about 30 different samples.

Interviewer: Is SQ different from these memory card?

Sato: Yes, There is sound difference. Many people in my division listens to and feel the difference.
Surprisingly, it seems different sound is heard from the two cards its only difference is the paint.

Interviewer: I cannot believe this story that different sound is heard from the two cards its only difference is the paint.

Sato: Measuring result does not tell the difference. So I cannot explain the reason confidently.
Metal ingredient is different from the paints, so I guess this causes the difference.

Satori: I feel doubtful from the first time. But several people say the impression on the same trend
and I'm now convinced. I can now feel the difference of sound so this card may be materialized

Interviwer: What part causes SQ difference? This may be caused by unevenness of each copy.

Sato: I thought this difference is from the unevenness of each copy and
I ordered several samples with the same configuration and listened to the difference.
result was, feels the same on the same configuration, feels different from the different configuration.
I started to select good sound one from these.

Interviewer: Is there the definition of good sound on this case?

Sato: Measured result was the same. but from the listening, I feels somewhat hard,solid or light,gentle among those
cards. I think light,gentle one is more originally intended sound.
I remember the sound of early days of Compact Disks, that produces sound feels harsh, this problem is
caused by unmatured technology. Sound quality of CD is gradually improved by the advance of technology.
On memory card, sound difference is miniscule and has been overlooked but now high resolution is introduced
and subtle difference can be distinguished and memory card difference is heard

Interviewer: I'd like to ask the question about the plot. You said measurements was not able to distinguish memory
card difference and why this prot exists? and why the plot lacks number. This plot increases doubtfullness of this product.

Asou: Several complication prevents to release more complete plot. This plot may causes misunderstanding about
this product. This plot is measurement result from a quality test equipment of Sony EMCS. We considered
not release this prot but decided to release to public with some obfuscating. We recorded more higher frequency
but this region was cut because it is radio frequency of cellular phones

Interviewer: Hmm, horizontal axis was logarithmic scale and, oh I understand right end is 800MHz.
How about vertical scale?

Satori: Spectral intensity means amount of energy from point radiation source, which transmit energy to some direction per unit frequency per unit time.
We tested with real walkman PCB, inserting microSD card and while playing high-res sound.
More specifically please refer following figure. We attached the probe on the microSD slot and measured electromagnetic field created around microSD slot.
We decied not to disclose vertical axis number because this is trade secret.

Interviewer: This measurement result is about the high frequency, non hearable range. This difference does not affect SQ ?

Sato: Sure. On the audio device, power curcuit and analog signal curcuit is sensitive and in some case
incoming noise affects the resulted sound. Noise on this plot cannot be heard directly. I suppose this noise affects
to adjacent curcuit work and alter the sound.

Interviewer: I agree to that. I heared the SQ difference from the original CD and copied CD-R 14 years ago.
And at that time, I cannot measure the difference of these two but I certainly hear the difference.
I guessed the reason of the difference as follows: pick up servo motor causes electromagnetic noise and this affects
adjacent analog curcuits.

Sato: About 15 years ago, I tested a lot about SQ of CD. current caused by servo motor causes noise in RF range
and when it is reduced, sound quality is improved. We know that at that time. noise suppression and reinforcing power supply affects SQ.
this is orthodox way to improve sound. Without such experience, this product idea can not be materialized

Interviwer: I looked into CD sound quality before and I was convinced SQ difference exists if digital data
itself is the same. but at that time I thought following thing: If servo current is provided from the different power source from the analog curcuit and analog curcuit moves away from the servo motor, problem is solved.
Once upon a time, some audio critic was saying "this CD-R is good for high frequency but that CD-R is better vocals" about the CD-R media review. But I think this sound difference is specific for the equipment used by the audio critic
and cannot say it as the general characteristics of the CD-R.
I'd like to know, What equipment is used for listening comparison?

Sato: We use NW-ZX2 prototype and NW-A10 for testing. These two show the similar tendency of sound difference for the memory card difference.

Interview: I understand that. I think sound superiority of SR-64HXA is shown only on NW-ZX2 and NW-A10.
for instance, think about playback of memory card slot on the PC. on this case SQ difference is diminished ?
I'm still not convinced this memory card itself improves SQ generally.

Sato: I did not test intensively on other companies products so I cannot say generally this memory card improve SQ
but some smartphones shows the similar tendency. I do not know about the phylosophy of sound of all other products
so this is difficult matter. I know the audio device that this memory card causes negative sound impression.
About the separation of analog curcuits, this is one of the improvement method but if ground line is connected
this line transmit noise.

Interviewer: I understand. From the previous talk, Sato selected the best combination of parts and this becomes the product as premium sound memory card. This card has some characteristics from engineering stand point?

Goto: Typically we always try to improve speed of memory card. increase bus width from NAND flash to controller
and increase bus clock and then increase host-controller bus speed.
But on this project, we decreased bus clock and increased stability. For instance, When PC is overclocked
computer becomes unstale and noise is increased. but when PC is underclocked, stability increases and noise and performance are decreased.

Satori: But when clock speed is decreased too much, sound buffer underrun may happens so We created samples and tested many times to determine the design of final product.

Interviewer: So this is not off the shelf memory card. custom made for audio application.

Goto: Exactly. We listened to many times and found the best configuration.
Also we must retain the same quality for this memory card. Typically memory card experiences frequent part
change, NAND, controller, firmware and its parameter without changing stock number.
but on this premium sound model, if on the future, when some design change is necessary, sound listening check will be performed

Interviewer: I'd like to ask about the design of walkman, audio data is buffered to the main memory when
playback. I think if this buffer is huge, microSD card does not runs on playback at all and SQ becomes unaffected by microSD card.

Sato: Surely we use main memory to buffer but walkman is consumer product and buffer size is not large.
There is enough room not to cause sound dropout but it cannot contain whole track. Also microSD card does not shutdown
immediately when read access stops so EMI emits and this affects SQ.

Interviewer: I understand fully. Thanks.

after this interview, I listened to ZX2 and A10. I reflained to say detail of objective impression but I felt the difference. I listened to high resolution contents and CD quality contents, I felt the difference from both.
This difference is smaller than MP3 192kbps and CD difference but if you pay attention to by your favorite instrument
or vocal quality, you may hear the difference? Anyway, this test is performed on walkman, sound difference on PC or other devices will be inspected at the next opportunity.
nige2000
Posts: 4253
Joined: Thu Feb 14, 2013 10:47 am
Location: meath

Re: Snakeoil from Sony?

Post by nige2000 »

i dunno
not that i have no faith
this is one component its difficult to imagine a substantial benefit
sd card player, modded soekris dac, class a lifepo4 amp or gb class a/b amp, diy open baffle speakers based on project audio mundorf trio 10's
Sligolad
Posts: 1089
Joined: Tue Jul 19, 2011 9:52 pm

Re: Snakeoil from Sony?

Post by Sligolad »

I have been watching these guys in the Netherlands for over a year now with their SD based playback systems so it will be interesting to see what they have to say about it if anything at all, it made a lot of sense to me at the time reading about the technical side of the development.
Had considered this unit for some time but backed off when i went with the Lamp.
http://www.ecdesigns.nl/products/sd2/
___________________________________________
SD Card DAC, Gryphon Essence Mono's & Pre Amp, Wilson Alexia 2 Speakers,VPI Scout 2 & Supatrac arm, Studer A812 R2R.
Post Reply