You can start first with Michael Lavorgna's article at Audiostream.com http://www.audiostream.com/content/ayres-pcm-dsd-comparison then move on to the actual posting at Ayre.com. http://www.ayre.com/insights_dsdvspcm.htm
Thanks for posting. It's good to read some love for PCM. Here's another interesting comparison of DSD and PCM (this one gives the edge to DSD): PCM v DSD Comparison: 16/44.1, 24/96, 24/192, 64x DSD, 128x DSD
It's troubling that there is a 9db difference between the dsf and wav samples. That is going to cause problems.
Slightly off-topic, but with regards to the following statement in the Ayre article: "Standard DSD has low noise levels in the audio band. But at 20 kHz the noise rises sharply."…….."These high-frequency noises can lead to damage of downstream equipment including amplifiers and loudspeakers. Therefore the official Sony specification (the “Scarlet Book”) specifies a third-order low-pass filter starting at 50 kHz, and the actual usable frequency response of SACD doesn’t extend much beyond 30 kHz"…... I'm assuming that when DSD audio is packaged in the format of an SACD, the SACD player is equipped with the appropriate built-in filters to remove this high frequency noise on playback. Am I correct? But what happens when DSD audio is played back on a computer instead of a "standalone" SACD player? Is the HF noise still filtered out? And/or does it depend on the method of computer playback? E.g. - 1) Playing DSD natively through DSD-compatible software (e.g. Audirvana) and a DSD DAC (e.g. Schiit Loki). 2) Playing DSD with on-the-fly DSD-to-PCM conversion through appropriate software (e.g. Audirvana) and a conventional DAC. 3) Converting DSD files to PCM with appropriate software (e.g. Foobar).
I highlighted in bold (leaving the rest to read in context) two aspects of cd sound that appeal to me. Being that this is a DSD vs PCM thread, I'm free to express my opinion. I'm not a fan of DSD, because IMO it can soften leading edge transients. I also think that there is a frequency imbalance, favoring a highly resolved midrange depth yet lacking definition at the frequency extremes. In regards to hardness/harshness often associated with PCM. I think much of that has to do with the use of op-amp based output stages instead of discrete circuitry.
Here is my simple take on this: If it is THAT HARD and THAT CONTROVERSIAL to tell the difference this is not a battle worth fighting (for me, anyway). I'm sitting out the whole 'DSD' debacle. Sorry Sony, not going to buy all my records in yet another format, especially one that is obviously not destined to have widespread acceptance. If in three years DSD downloads are a commercial success and obviously offer better sound, than maybe I'll try some. You can tell who the Pioneers are by the arrows in their back. I am NOT being a Pioneer this time.
That is what the Audiostream article said. I'm a fan of DSD for a number of reasons but the clincher for me was that we did some live mic feed tests and DSD sounded closer to the source. DSD seems to have a flow that is more analog like and reb I find DSD actually does better on transients than PCM when I have done various comparisons. I have also found that DSD files on DSD DACs generally sound better than PCM files on PCM DACs. Your mileage may vary. The whole ultra-HF noise thing is misdirection as no one can hear it. One guy I greatly admire on recording quality is Todd Garfinkle and he was all PCM for the longest time but then he started using Korg MR1000 and later MR2000 machines eventually doing double DSD. He feels this is the best he has heard on his recording in terms of naturally capturing what is going on in the studio. But hirez PCM is so damn good these days it really doesn't matter and I'm not sure why Charlie Hanson is stirring up another religious debate. Can't we all just get along? Best strategy is to be format neutral and enjoy more music.
But doesn't it have the potential to damage your ancillary equipment (especially tweeters) if it is not filtered out?
Depends on the gear but many players filter this out at some point. My 777 has a switch on the back to do this.
Lee, I don't have access to all the high-end gear that you do. And have never heard well mastered SACD played back on the latest equipment as you have. I have some deeply rooted opinions that Universal players are compromised and are unable to playback DSD or PCM to their full potential. Perhaps that has much to do with my preference for cd playback on a dedicated cd only machine. Through vacuum tube output stages to give cd some needed spaciousness .
Well I don't have as good a player as many. But I have heard a lot in any event. Many universal players do compromise SACD such as the older Pioneers like the 47. But again all that matters is what you find sounds good. As long time readers here know I was quite enamored with Super Audio when I joined here back in 2001 or so. Over time, I heard better and better SACD playback but DVD-Audio playback got really good too and I got a cheap Oppo 980H and really enjoyed playing that. Then around three years ago the DACs got really excellent and it seems every year PCM playback gets better and better. Honestly at this point, I could live with either format. And the gap between CD and hirez has even narrowed though I prefer the latter. And output stages make a big difference, there are good ones of tube and solid state variety. I may be wrong but my view is this is a possible order of formats in decreasing sound quality: Reel to reel tapes LP SACD/DSD Files Closely followed by 24/192 and 24/176 24/96 and 24/88 16/44 AAC/MP3
I use a Korg MR2000 for my vinyl transfers @ 5.6 MHz. When I A/B between the analog source (LP) and the DSD file of the same LP, it's negligible if there is a noticeable difference. When I make the same recording to a PCM file @ 192kHz, it's a noticeable difference between the two.
Sounds like a problem with the Korg. It should sound the same. I'd also suggest using 24/96 which is more than adequate. Tim
Do I have this right? We are meant to judge DSD vs PCM via a dub of vinyl replay, with no knowledge of how the original actually sounds? Tim
There's that argument ...24/96 is more than adequate. I remember when I was looking for an ADC, and a vendor told me there was no reason to go with anything more than 24/96. I purchased his product coughbenchmarkcough, I really tried to get the most out of that product, but it just never made my jaw drop. Once I purchased the Korg and made a transfer, my jaw hit the floor. Then I A/B'd the two and was convinced, for me, this was the way to go. The Korg will record at 24/96 or 24/192 and it does sound very, very good. So there is no problem with my Korg. I was referencing 24/96-24/192 recording I had made with other ADC's For me, the proof is in the pudding, at least to my ears. I prefer to go 5.6MHz, then dumb it down to a 192kHz PCM file for editing and then to CD for use in my auto. Trust me, I've gone the whole gamete of 24/96 to 24/192 PCM. For me, and only me cause I'm the one who get's to listen to what I'm transferring, the 5.6MHz DSD file is the way to go. I'm not going to try and convince anyone that my method is any better than theirs. It just works for me. My use is mainly with first press LP's. A1/B1 type stuff. I could care less about using this technique for new vinyl. I find that I can replicate the best sound and dynamics with my method that transfers just fine to my listening needs. Of course, YMMV. Carry On.
Thanks, and I am not doubting your report. The question though would be why? As a matter of interest, does your conversion from DSD to 192 kHz PCM change the sound at all? How does a recording done directly to 24/192 compare to the result of your LP->Korg->DSD->24/192? What about from 24/192->16/44, can you hear any difference? I'm fascinated by the discussion about the new Platinum SHM-CDs which follow the same mastering principles as the SHM-SACDs. People are discovering that they sound very similar. Leaving aside the platinum nonsense, and the SHM nonsense, it shows how good the CD format can sound when done right. Tim
You go through all this for something you listen to in your car? Interesting. I hope this is a Rolls Royce Ghost, where they claim the car is so quiet at 60MPH, the loudest thing inside the car is the ticking of the clock... Love the retractable hood ornament.
Thanks. I also found a helpful thread on CA that explains how to configure the on-the-fly DSD-to-PCM settings in Audirvana to filter out the HF noise. It is more of a problem if you are sending high sample rate PCM to your DAC as the noise increases with frequency (a rising slope above 30kHz). E.g. if you do on-the-fly DSD to 88.2kHz PCM then there will be an inherent low-pass filter around 44.1kHz, whereas if you are converting the DSD to 352.8kHz PCM then you are potentially letting through HF noise up to 176.4kHz, which is more likely to cause instability issues, but the customisable filters in Audirvana seemingly allow you to negate this problem. Regarding passing the DSD stream directly to a DSD DAC without PCM conversion, I emailed Schiit and they confirmed that the Loki has appropriate in-built filtering to remove the HF noise .
