R/E/P Community

Please login or register.

Login with username, password and session length
Advanced search  

Pages: [1]   Go Down

Author Topic: Yet more snake oil?  (Read 4292 times)

kraster

  • Full Member
  • ***
  • Offline Offline
  • Posts: 199
Yet more snake oil?
« on: August 18, 2006, 11:25:30 PM »

http://www.anagramtech.com/technology/q5-upsampling/


"Q5™ Upsampling allows digital audio signals from virtually any audio source to be resampled, resynchronized and retimed to extraordinary levels of detail. In effect, upsampling the digital data stream (up to 384kHz) allows existing “standard” CDs to have the enhanced audio clarity, richness and dynamic range that would normally be associated with SACD or DVD-A disks “upgrading” the tremendous investment that has already been made in high quality CD recordings. You may be astonished once you begin to hear what you have been missing!"


It goes on to say that they use "adaptive time filtering" to correct minute deviations from the master clock resulting in pristine audio etc. etc. How can something remove jitter from a signal that has acquired jitter in the original AD process?

Hmmmm....

Best,

Karl Odlum
Logged

chrissugar

  • Newbie
  • *
  • Offline Offline
  • Posts: 19
Re: Yet more snake oil?
« Reply #1 on: August 19, 2006, 03:08:52 AM »

kraster wrote on Sat, 19 August 2006 06:25

How can something remove jitter from a signal that has acquired jitter in the original AD process?



Where did you read that? There is nothing mentioned about removing jitter in the AD process.
And there is no snake oil, it is just a way to say that their DA eliminate/reduce jitter induced in the process of digital signal transmission. You can have a perfect source with zero jitter and you will still have jitter at the receiver side in the DA. It is because of the digital transmitter, cable, receiver that induce jitter whatever you do. So the only right way is to reclock the DA to a precise clocking system (inside the DA) and not to improve the jitter at the transmitter side or to buy a 500$ digital cable.
ANAGRAM's method is probably something based on SRC like in the Benchmark DAC1 or the Lavry in wide lock mode. No snake oil.

chrissugar
Logged
Christian Mike Sugar
CMS-LAB

Jon Hodgson

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1854
Re: Yet more snake oil?
« Reply #2 on: August 19, 2006, 05:51:28 AM »

Looking at their white paper, not exactly snake oil but a lot of marketing hype around something that may be useful in certain situations.

The first bit to look at is the claim that upsampling improves the sound of your system. Some people might infer from this that it somehow brings back elements which were missing at the lower sample rate, it can't. There's only two ways in which this might be the case.
1) The DAC used in the system works better when driven by a higher sample rate, for example because it avoids an internal interpolation filter which is inferior to the SRC.
2) The SRC process is imperfect and generates distortions in the audio band, they just happen to sound nice.

The second part is the whole reclocking thing. It makes sense, however you would have to be able to measure the jitter very accurately, and then calculate coefficients to compensate it very accurately, so I have my doubts that it would work better than a decent phase locked loop.

Looking at their numbers I would have to say they really aren't impressive. 114dB THD+Noise on a Sharc?

In a cost constrained consumer system where you might have an excess of cheap DSP power available but don't want to spend money on high quality phase locked loops and DACs (component count is hugely important when trying to make cheap gear) perhaps it makes sense, but for higher end gear it looks like it's going to introduce more problems than a decent PLL and DAC
Logged

Sin x/x

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 72
Re: Yet more snake oil?
« Reply #3 on: August 20, 2006, 06:42:26 AM »

Up sampling does not alter the amount of information in the signal.

So no extra detail.

All descent converters use a phase locked loop, witch removes all jitter.


Snake oil.
Logged

Jon Hodgson

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1854
Re: Yet more snake oil?
« Reply #4 on: August 20, 2006, 07:34:53 AM »

Sin x/x wrote on Sun, 20 August 2006 11:42

Up sampling does not alter the amount of information in the signal.

So no extra detail.
.

No extra detail, however in most cases of D/A conversion there are upsampling stages in the digital domain to reduce to requirements of the analogue reconstruction filter.

This upsampling process will itself introduce errors, though with proper design they will be insignificant. It may be that in the case of cheaper DAC chips the built in SRC is of limited quality, and if you have some DSP power to spare in your system it may make sense to use that and bypass the SRC on the DAC, rather than upgrading to a better quality DAC (remember that in mass market audio consumer products they count pennies when it comes to production cost).

Sin x/x wrote on Sun, 20 August 2006 11:42


All descent converters use a phase locked loop, witch removes all jitter.
.

Well that would depend on the quality of your PLL, they are an exercise in compromise... which is why Dan came up with CrystalLock.

However the more I think about it, the less useful I can imagine this functionality being, because it doesn't actually make sense, it would only work if you were removing jitter in the original sample clock (which I would expect to be smaller than you could measure and compensate for)  and not jitter in the transmission line. Otherwise you're actually going to be introducing distortion, not removing it.

Actually reading their white paper a bit more I think I may have misunderstood what they were suggesting, mostly because the repeated use of the word "jitter" is a red herring.
Sin x/x wrote on Sun, 20 August 2006 11:42


Snake oil.

Well I's say more like moisturerizer being advertised as the secret to eternal youthful looks. It appears to be a sample rate converter, aimed very much at consumer playback equipment. It MAY be a good solution to meeting achieving a good cost/performance ratio in some cases.
Logged
Pages: [1]   Go Up
 

Site Hosted By Ashdown Technologies, Inc.

Page created in 0.066 seconds with 17 queries.