|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
Non-oversampling DAC concept was first proposed by a Japanese independent engineer, Mr. Ryohei Kusunoki, in 1997. He published his paper on MJ (Musen to Jikken) magazine, Nov. 1997 issue. It created a stir in the industry in Japan at the time, and today, non-oversampling concept is associated with his name as a common sence in Japan. His paper was followed by actual products from a couple of adventurous manufacturers in Japan in following years. 47 Laboratory's Progression DAC (1998) which doesn't have either digital or analog filter, Illungo Audio DAC (1998) with no digital filter but uses an analog filter at high frequency, Zanden DAC (2000) also with an analog filter. I suppose 1 x sampling by Audio Note is one and the same.Mr. Kusunoki published his concept instead of applying for a patent so that everybody can benefit from his original concept. I translated his papers into English with his permission and it's included in SAKURA SYSTEMS' website, "Reviews and Articles" section. Please check on it if anybody's interested in.
Follow Ups:
hi Yoshi -thanks for the info! BTW, any news or ETA on those new 47 Lab DAC's you were expecting? (the new lower-priced one)
thanks,
-Ed Sawyer
Dear Ed:The new "Shigaraki" DAC from 47Lab is about to land in US anyday now. It's the same concept and almost the same circuitry as their Progression DAC, but encased in a shigaraki ceramic casing. Shigaraki clay is a totally de-electronic, non-conductive ceramic material which has been used to make tea cups and vases in Japan since almost one thausand years ago.
The very good news is that this material is rather cheap in Japan, and the retail price in US will be $1250 including a seperate power supply, also in shigaraki casing.
A matching transport is expected to come out sometime this summer.
The detailed information will be on SAKURA SYSTEMS' website, www.sakurasystems.com, shortly.
Best,
Yoshi/SAKURA SYSTEMS
Yoshi,Actually, your dates are in error. As far as I know, the real background of the Kusunoki DAC is the DIY DAC contest which M&J held in 1996. The date of the contest final was July 14. Due to magazine publishing deadlines, the first photos from the contest were published in the October 1996 issue of M&J (you can see a photo of Kusunoki and his DAC on page 22), while the actual design and construction article was published in the following issue. I don't know Kusunoki personally, and I therefore have no idea when he started his design, but it must have predated the contest by a good while.
I remember my good friend Isao Shibazaki (he writes for M&J and if I am not mistaken was one of the contest panelists) talking about this "primitive DAC" which measured like hell but actually sounded pretty good. I think Shibazaki mentioned that most of the M&J panelists wanted to flunk the Kusunoki DAC from the contest at the schematic judging stage - but luckily they decided to listen to it before passing judgement on it.
BTW, wasn't Junji Kimura the industrial designer for the mechanical structure and cabinet of the Illungo DAC?
regards
jonathan carr
You are right. I just realised it and was about to make the correction though. Thanks for the notice.Yoshi/SAKURA SYSTEMS
Dear Yoshi,Just for the record.
Andy Grove and I started developing the 1xoversampling DAC concept in 1995, and the first functional prototype was tested in early 1996, my previous engineer Guy Adams (now with HP) started work on the principle already in 1993, but the project was abandoned in 1994.
Audio Note UK built and shipped the first 1xoversampling DAC5 in August 1997, a full three months before Mr. Kusunoki's article was published, by which time we had already sold several DAC5's.
Mr. Kusunoki's article is very interesting, but he was not the first as you claim and neither were we, because if you want to be absolutely historically correct then SONY's first CD-players launched in 1982 also had no oversampling or digital filters, so it is they who can lay claim to the original idea, which may be even older from all I know.
I don't want to take anything away from Mr. Kusunoki with this posting, he was and is undoubtedly one of the pioneers of the no oversampling filterless DAC theory and unlike what you are suggesting about Audio Note UK, I would never claim that he knew about our work either and therefore was inspired by it, but I want to state clearly that since neither Andy Grove or I read Japanese we only came to know about the article in MJ sometime last year by reading it on your web site, by which time Audio Note UK had already had a full product line with the technology on the market for over two years.
As it is, patenting an idea like this would have been impossible, because by 1997 there was already too much prior art in the public domain.
I do not understand your continued insistence that Audio Note UK has somehow plagiarised Mr. Kusunoki's ideas because this is patently not the case and I think it would serve the no oversampling filterless DAC concept better if we worked together bringing it to the front of the commercial queue, so that as many music lovers can benefit from its unique properties, rather than to in-fight about who "got the idea first" as if that somehow gives a "right" to the best product.
So would you mind ??
Sincerely,
Peter Qvortrup
I'm wondering if you guys realize that a non oversampling dac was a toy that many people played with in the early 90s.This was via 2 ways. First was wired up to the parallel port of an IBM compatible and the data lines were feeding the resistor ladder.
2nd was just modulating the PC speaker. This second technique is more of a strech but if you study it the idea can be seen.Granted neither of these was "hifi" by any stretch ( the LPT DAC is only 8bit) but the concept existed and was implemented.
-CAL
Dear Peter:If your statement about the timing is correct, I owe you an apology.
There were rather embarrasing harrasments against Mr. Kusunoki by major Japanese corporations when his first article was published, but he stood on his ground, published more articles, and continued his research which eventually produced those products I mentioned.
I agree that who-did-it-first does not benefit customers, but the superior sound of non-oversampling (1 x sampling?) DACs do.
Best regards,
Yoshi Segoshi/SAKURA SYSTEMS
Dear Yoshi,Thanks, I appreciate your honesty.
Perhaps we meet in NYC for the show in May?
Sincerely,
Peter Qvortrup
Can't wait!George
Dear George,Certainly, where do you suggest, email me about that?
Sincerely,
Peter Qvortrup
Dear Peter Qvortrup,
Refer to your website, all Audio Note DACs are capable for 18 bit only. I am curious is the 1x oversampling technology restrict your DACs have to stick with 18 bits due to not many D to A chip with no filter in it. Would you please explain?Do all Audio Note DACs upgradeable to 24 bit/192 kHz or DSD capability?
Thanks!
Yours,
PPtriode
Dear PP triode,It is possible to make a 24Bit 1xoversampling DAC with the Burr Brown 1704 chip, we tried and I preferred the sound of the AD1865 18Bit Analogue Devices chip which is why we use that.
Our DACs are all Redbook protocol PCM converters, they would not work with DSD, they should work with 192 kHz signals, provided input chips are available that will read that.
Sincerely,
Peter Qvortrup
Dear Peter Qvortrup,
> > Our DACs are all Redbook protocol PCM converters, they would not work with DSD, they should work with 192 kHz signals, provided input chips are available that will read that. < <Even you change the input chips, however the AD1865 Analogue Devices chip still a 18 bit chip, and only can handle anything up to 18 bit; not 24 bit.
Thanks!
Yours,
PPtriode
Dear PPTriode,Quite true, however, if you compare the AD1865 running a 24Bit DVD-A at 18Bit resolution and the BB1704 doing the same but at 24Bit, the AD1865 sounds far better despite the fact that it truncates, the question is why it sounds better?
I am sure there is an explanation, but I have neither the time nor the resources to get to the bottom of this, so I have to make my decisions on what I hear, not what the theory or the measurements tell me.
There is an awful lot of high resolution theory out there at the moment, however, the practical applications more often than not struggle to support the resulting sonic performance.
I have to say that most of the theories have real problems when it comes to proving themselves in real life.
Sincerely,
Peter Qvortrup
Hello Peter,Your approach to building dacs is a breath of fresh air!
Now I am really curious as to what they sound like.Regards,
Kyle
Hi Peter,Your comment on BB1704 vs AD1865 is interesting. I'm
not sure of internal architecture of AD chips but
1704 has built in flaw which is Bipolar offset.
On PCM63 this feature is defeatable and quite
audibly degrades DAC performance when connected.Did you compare PCM63 with AD1865?
Regards,
Terry
Dear Terry,All our older oversampling DACs used the PCM63, which whilst better than the BB1704, did not sound as good as the AD1865N.
I remain unconvinced that higher sampling rate alone improves fidelity, I think that the design of the chip and many other factors also play an important part in how the end result sounds.
Sincerely,
Peter Qvortrup
Thanks Peter,I'm surprised the AD chip sounds better than the '63,
I'll have to try it.However there are a few considerations. '63 must be
run without bipolar offset as this has significant
degradation.Also Jitter performance of DAC will favour
some chips over others and affect critical
choices. Jitter must be removed first before
other design choices are made.An example, I auditioned AN DAC3 and overall thought
it was really nice. But there was no provision for jitter
reduction and lots of components inside pointed towards
a euphonic and warm musical sound.So I put the AN analog stages on my DAC which has PMD100
and extremely low jitter design.My surprise was that it didn't sound much better.
Yet when I put the AN's equivalent dig circuitry on
my DAC it sounded worse than AN. My DAC has 0 FB
tube SS hybrid but is much more linear than transformer
to CC tube stage. So it seems like a "warts and all"
scenario.Conclusion was that the AN's OP stages are beautifully
musical but cover up some of digital sections
shortcomings such as DF and high jitter. So when
jitter is removed, analog stages can be made less
euphonic but the result still sounds musical
and with greater resolution.It's a can of worms, and I will have to try
the AD1865 chips, but am also looking closely at newer
chips from BB (1738)and NPC (5865/6) which are
totally different architecture... hopefully
better.Has anyone here tried these new DAC chips?
Here we go again!
Regards,Terry
Dear Peter Qvortrup,
Did you guys tried the AKM 4393 (24/96) and AKM 4394 (24/192)? Those are very good sounding DAC chips, I would like to hear your comments on this one.Yours Sincerely,
PPtriode
Peter
Can your DAC be used with a room correction device such as the Tact?Thanks,
David Shapiro
Dear David,Using a digital room correction device really runs against the grain of removing as much of the processing from the signal as possible, as we do in the 1xoversampling topology.
Otherwise I think Werner has pretty much covered what I was going to say.
Sincerely,
Peter Qvortrup
Yes, one would expect so. Perhaps the out of band HF images
produced by the DAC could interfere with the RC device's
calibration measurements, but this effect would then be
limited to the treble and it is wise anyway to limit the
correction action of RCs in the treble.No, because in the light of a non-oversampling DAC a
room correction unit and its signal processing would
rather look non-PC :-)
This post is made possible by the generous support of people like you and our sponsors: