|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
125.237.109.135
In Reply to: RE: IF you want a DAC to play normal "redbook" cds then you shouldn't buy the latest and greatest... posted by morricab on April 15, 2014 at 02:44:46
To a certain extent I agree with you... but only to a point!
No matter how good the DAC and supporting analogue circuitry, the final result will be ruined by a poorly designed digital filter.
I still find that the PMD100 is still a remarkably good sounding digital filter compared to the internal filters bundled with devices like the PCM1792 that claim to have superior stop band rejection. The only problem with these older devices (and I don't think the PMD100 is the only one to suffer from this) is that the bitclock must be reclocked for optimum performance otherwise jitter is injected into bitclock driving the DAC.
I still think it is possible to get equivalently excellent sound with the modern devices - it is now understood that delta sigma DACs have odd idle tones and patents have been filed that incorporate a dither module in the feedback loop to eliminate these effects. Like anything, it has taken time to understand how to implement this thing we call Digital Audio properly, it's just that you have to work harder to solve the problems that get introduced with progress!
"Beauty is Truth, Truth Beauty.." Keats
Follow Ups:
FUnny enough the PMD100 and the BB DF1704 were found in one player, the Sim Audio Moon Eclipse cd player that used BB PCM1704K chips. Universally it was thought that the DF1704 was the superior digital filter and Sim Audio only added the PMD100 for the HDCD functionality. There is a German company that has a swap for the PMD100 (with the DF1704!) because frankly it is not the best of the older style oversampling filters out there.
I've not heard the DF1704 so can't comment on a direct comparison. However, you need to define whether you mean the reconstruction filter (as in the ACTUAL reconstruction filter impulse, phase and amplitude response or whether you mean the integrated system output from the analogue outputs of the DAC.
Without studying the circuit implementation, this still doesn't prove that the PMD100 reconstruction filter characteristic is necessarily inferior. It only shows that in the final system integration the DF1704 resulted in a subjective improvement and that could be down to inattention to device specific requirements in dealing with corrupted clock signals or suboptimal dither.
Note also, that clock jitter affects the performance of the reconstruction filter and can result in aliasing artifacts when it would seem unlikely based on the specification of the device.
The PMD100 requires the bitclock to be reclocked. Any design that doesn't do this important step will have a corrupted bitclock driving the DAC. This will cause a degraded sound from the DAC. The other bit of information you haven't provided is how the PMD100 was configured. You have control over the output bit depth and the dither amplitude and type of dither. An inappropriate combination for the chosen DAC will also cause an audible difference
If the original designer merely substituted one for the other without having specific circuitry to handle the clocking for the PMD100, then it is possible to hear a difference I'm sure, but I'm afraid that this is not sufficient proof until all of the information is known!
"Beauty is Truth, Truth Beauty.." Keats
All the DAC chips I am talking about required an external DF of some kind (PMD, DF1704 or similar). All I can tell you is that enough people were ultimately dissatisfied with the PMD100 to warrant replacement offers. To my knowledge no one wants to remove their DF1704s unless they are thinking of going NOS.
WHo knows?? Maybe you are right and the PMD100 is wonderful but misunderstood chip but the history suggests otherwise.
I'm not precious about the PMD100 at all. Leaving aside the DF from the equation, it has been my experience that the I-V conversion and output analogue filtering topology has a much more significant effect on the output sound...without changing anything on the digital side.
For example, I wouldn't use any other dielectric than polystyrene for my signal path caps, and I would implement a DC servo to avoid the need for NP electrolytics that many manufacturers pollute the sound with!
The other thing I think that is important is the op amp topology to implement the analogue filter.
If you look at the PCM63 and PCM67/69 application notes, they recommend a GIC configuration rather than the more common MFB topology. This is a much better topology I feel as it has the advantage that the passive component properties are what will dominate the sound quality. Next to PTFE, polystyrene is the next best thing in terms of low dielectric absorption and the result is a huge lift in realism and clarity (to me).
Secondly, this may explain why older DAC designs may sound better than new ones, simply because the designers followed the guidelines which just happened to have the better topology!
For what it's worth, I personally wouldn't choose to use a PMD100 now for any perceived benefit since the sample rate is limited to 55kHz and the DF1704 is a bit more versatile.
I'm actually working on a DAC design of my own now (and actually have a PMD100 from my X-DAC which now has a totally trashed PCB from all my reworks!), but I will implement the DF in software using a DSP. A separate DF IC is just too limiting!
"Beauty is Truth, Truth Beauty.." Keats
I don't disagree with you about the importance of the analog I/V conversion and filtering (or lack thereof); however, even the best implementations I have heard fall short of what I am hearing when a good ladder DAC is in place with a really good output stage.
I find it interesting that my Audio Note 2.1 dac using the analog device dac uses no digital or analog filtering and the Audio Note transport uses no jitter reduction at all. Members of my audio club think it is the best sounding redbook playback they have ever heard. Maybe the simpler the better is the clue here. I also have a Mytek dac for "HIREZ" and I almost never listen to it any more. The audio note is much better. I also love vinyl
Alan
Heard the 3.1x with 3.1 transport into a Jinro amp with higher model E speakers...very natural in some ways but a little too rounded for my liking (a bit more than I think I hear live as well). Maybe too much AUdio Note in one chain??
regarding their Zero times oversampling and no digital/analog filters is this review of the DAC 1.1 linked below."With Audio Note the return to non-oversampling ties in with their earlier efforts on transformer-based DAC current-to-voltage conversion: a transformer forms an interesting low-pass filter to be used with a non-filtered DAC."
It's still different type of process than other design types (although there are some copycats out there) - and on sound quality I agree with you. The SQ is the key thing not one internal part - DACs are dirt cheap and Audio Note you can be sure tried them ALL.
It should be noted that the below review was when Audio Note used an Analog filter and termed it that.
Edits: 04/15/14
I hate to break it to you, but there is analogue filtering...it's just that it has been disguised in the output transformer as you can see. Not only is a shunt capacitor part of the transformer block, but the transformer itself will limit the bandwidth of the signal getting through to the tube output stage.
Regards
Anthony
"Beauty is Truth, Truth Beauty.." Keats
nt
Observe, before you think. Think before you open your yap. Act on the basis of experience.
Post a Followup:
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: