There have been previous threads about the distinction between upsampling and oversampling. I think if you stick to players that claim ONLY oversampling, and not upsampling, you will have dramatically narrowed the field to the kinds of players you (Mbonn) seem to be looking for. I would very much like to know what you find out in your searching, and hopefully listening, as I have become skeptical of upsampling (interpolation) based on my experiences with the Perp. Tech. stuff. When companies like ARC, Meridian and Muse decide to skip the oversampling/interpolation game, it makes me wonder whether this is not a major industry band-aid to improve lack-luster digital performance until something better comes along (maybe SACD)....... |
Hi Z'man,
I appreciate your threads. Your comparison on upsampling vs. non-upsampling was a fine piece of work. Didn't mean to come across as defensive.
The Audio Note DAC's do not use digital filters but do use custom analogue filters. In fact, the quality of these filters improves as you move up the line.
They all use the Analogue Devices AD1865 18 bit chip, they have no optical link and use these custom tailored analogue filters to roll off enough of the carrier frequency at 44.1 KHz to render it inaudible.
In addition, from the DAC 2 up they use the patented interface matching transformer coupling curcuit in varying quality as you move up DAC levels. Many actually using transformer coupling to the output stage instead of capacitor coupling. |
Mbonn, you just anwser your own question !
"Regarding 4x oversampling: Maybe the companies that claim "No oversampling" really mean "no more than 4x oversampling"".
But does it matter ? It's all come down to the music.
Regards |
Whoa, JC, maybe you should go back and reread my post - without reading so much *into* it! I made no "claims" and offered no "opinions", bad or good, about AN DACs' performance or their partnering characteristics with other gear (neither did I claim to have heard one, which I haven't). I was simply reporting an item I remember reading from Stereophile's Industry Update section a couple of years back, which described an AN player that eschewed all filtering (and overesampling), sending the unfiltered signal on down the chain; granted, this might well have applied only to one model, and not the entire AN line. Obviously, AN didn't think that this approach would harm partnering equipment, but as I recall the article writer wasn't so sure when it came to some rare designs that might not be entirely stable to begin with.
(I fail to understand the relevance of your Goldmund amp/speaker-response analogy - an amp's stated operating bandwidth doesn't mean that it is actually generating ultrasonics in its output; besides, as I imply in my post, any speaker [and the human ear] already functions as a high frequency filter. As I understand it, any questions having to do with the wisdom of the unfiltered approach revolve around the possibility of causing a connected amp to go into oscillation, and not the mere presence of ultrasonics at the speaker - although in theory I suppose some speakers are designed so that their tweeter resonances could be over-excited if unanticipated high-level ultrasonics were present, maybe to the detriment of the sound, though probably not the speaker itself. I don't know if in fact either of these is ever really a practical problem anyone encounters, and am as sure as you are that most properly designed amps would not create a destructive mismatch in such a case anyway.)
I do not follow AN technology closely, and deliberately prefaced my short remark with the phrase "I believe" due to that. Rather than excoriating me for intentions that I do not possess and things I didn't write, you could add something educational to the thread, if you have any additional info on the subject of DAC output filtering. It is an interesting area, and I respect your knowledge and opinions on digital, which I come to with no particular preconceptions or agendas (or even an extraordinary amount of personal concern, I might add, seeing as how my own collection is mostly vinyl by far). Happy listening, whatever you listen to, Z'man. :-) |
Here is the 47 labs website: http://www.geocities.com/rjm003.geo/rjmaudio/diy_dac.html |
There's more: looking at the web sites implies that 47 Labs Progression does NO filtering AT ALL, while Zanden does no digital filtering but DOES DO Analog filtering. |
6chac - Regarding 4x oversampling: Maybe the companies that claim "No oversampling" really mean "no more than 4x oversampling". |
6chac - Regarding 4x oversampling: Maybe the companies that claim "No oversampling" really mean "no more than 4x oversampling". |
Jc, Zikesman - If you read the Audio Note web site carefully (the section on DACs), it says clearly that no DIGITAL filtering is used. But if you look at the comparison tables, it does make mention of ANALOG filters. It just does not describe the characteristics of the analog filter. My difficulty in understanding all of this is as follows: no oversampling means 44.1 kHz, and that means aliasing components down to 22.05kHz. Filtering this out EFFECTIVELY implies a BRICK WALL filter. Audio Note claims some proprietary filtering. If it's not brick wall, then it doesn't filter out (effectively)the aliasing conponents that are JUST ABOVE the audio band. |
Zaikesman. There have been hundreds if not thousands of Audio Note Dac's sold world wide. I have never heard of one person who had a problem because of what you have described. Many are not used with Audio Note amps. A Goldmund amplifier produces a bandwidth from dc. to over 2 mghz. Does that imply it can't be used with a speaker that has a frequency response of 20Hz-20kHz? What are you basing your opinions on? Are you familiar with the technology behind a Audio note DAC? Ever heard one? Eliminating a brickwall filter should not hurt any amplifier or speaker on the planet. You should do some research before making such claims. |
Every Compact Disc player on the market use oversampling. See the "Compact Disc" logo on every compact disc player ? Do you think those logo was put on all cd player by accident ? They pay to have it there. They pay to have the formula to build and sell compact disc player. Think of the Intel Pentium CPU's. The new Intel CPU's are basically X86 code, made faster, for example XT, AT, 386, 486, Petium, II, III, IV etc... The basic instruction set has to be there, to be backward compatible with the old softwares. Back to the compact disc, the samething applied. Some company does better oversampling, some do worst, some just bought OEM parts from Sony or Philips then build their own power supply (tube or solid state). TEAC develope their own transport (Wadia) but still the basic oversampling is there plus more (Wadia use 32X, 64X etc...)
Let say you build a brand new CD player, and try to sell it on the market. Sony /Philips will sue you until you take it off the market. You cannot call it a "cd player", neither. Because it is copyright. The same apply for media, not the content but the "disc" itself, also copyright. That mean your new build "what ever it is that you build" cannot (not suppose to) read the standard Compact Disc. Like a "pure" SACD player, which might not need oversampling, Sony own the "SA"CD player"" so somebody need to pay for the copyright or pattern anyway to sell "SACD player and/or media" see the circling ?
I was an adult that own a cd player in 1985 :-)
Regards |
Actually, I believe Audio Note is depending on the speakers and the listener's ears to accomplish the ultrasonic filtering, letting one's amplification fend for itself with an unfiltered signal from the DAC. |
D/A convertors do not need to have oversampling to work, it's just that they are almost always implemented to smooth out the filter curve. A pure analog "Brick Wall" Filter can be used somewhere past 20,000 hz with no over sampling. This was done in some very early CD players from the early 1980's. However, this didn't work very well because it caused distortions and artifacts in the audiable bandwidth. To say that the "Basic sample rate... is 4x oversampling" is completely incorrect. I had a 2x oversampling CD player back in 1985. At that time 4x oversampling was a totally new thing. I'm sure a company like Audio Note has found a way to implement an analog filter without causing distortion in the audiable bands. |
If it is 4 times oversampling it is not the same as Audio Note. In addition, The ARC does not use the same chipset as Audio Note or the 47. The Audio Note uses no up or oversampling and no digital filter. The results are astonishing. |
Why would you want a cdp that did not oversample? I too am confused....I thought the cd redbook standard called for some form of oversampling. Also, Wadia cdp's DO oversample. |
Frank Van Alstine claims that his "OmegaStar" DAC uses 4x oversampling. Does this mean he is in the same camp as ARC, Audio Note, 47 Labs etc (no oversampling)? this is very confusing. thanks |
6chac - thanks for your response. Audio Note claims "1x-oversampling" (and NO digital filtering). Is their claim not correct ? thanks - |
6chac - thanks for your response. Audio Note claims "1x-oversampling" (and NO digital filtering). Is their claim not correct ? thanks - |
I believe Wadia doesn't use oversampling neither. |
At least it used to be the case that the 47Labs transport had a dac in it. Although it did not sound as good as using the outboard dac, it probably was not oversampling as was the outboard dac. |
|
The basic sampling rate for Compact Disc format is 4X oversampling. Any company claims no oversampling is lying. |
