Post by ARQuint February 6, 2011 (1 of 50)
|
|
Audio online forums are not for the faint-of-heart. This is a fact of life I was reminded of once again when I came across a posting on Audio Asylum from Charles Hansen of Ayre that "ripped me a new one" for an aspect of my recent feature on HDtracks in TAS (Issue 208, December 2010.) My crime was to make comparisons between the sound of the stereo high-resolution program on a number of SACDs with that of the same material as heard as an HDtracks download—using for the former my trusty Oppo BDP-83 via its HDMI input into an Anthem processor. Here's the link: http://www.audioasylum.com/cgi/vt.mpl?f=critics&m=53687&VT=C
The issue for Hansen was the HDMI interface. To quote him: "HDMI is without a doubt the worst sounding digital connection ever devised, with incredibly high levels of jitter, especially in a mass-market product like the Oppo." In fact, my always thoroughly grounded editor, Robert Harley agrees that HDMI, objectively, manifests measurably more jitter than, say, an S/PDIF digital connection.
And yet. Many of us have used HDMI for multichannel high-resolution playback quite happily. When I wrote a basic level introduction to multichannel for Fanfare a few years back ('Multichannel Without Fear - Issue 32:1 Sept/Oct 2008), Kal Rubinson—an indefatigable advocate for multichannel music for whom I have the highest regard—wrote in to say that he agreed with most of what I had to say, except that, instead of my suggestion that those getting started with multichannel utilize analog outputs from player to receiver/pre-amp, they instead employ....HDMI. My friend, neighbor, and thoughtful listener (and frequent SA-CD.net contributor} Carl Weber, continues to use an Oppo outputting via HDMI as his source; he's the one I got the idea from in the first place.
Leaving behind the unpleasant tone of Charles Hansen's posting—I'll accept that as part of the package because I write for an audiophile magazine—I'd like to get a sense of how many critical listeners are actually employing HDMI as the interface from an SACD/Universal machine. What's driving your choice? Is it convenience? Is a coaxial or optical digital connection even an option for SACD with your system? Especially if you've been able to make comparison's, is this level of jitter audibly important? The denizens of SACD.net have an incredible wealth of experience, and I would love to know.
Andy Quint
|
|
|
Post by Fugue February 6, 2011 (2 of 50)
|
|
HDMI is a necessity with the Sony 5400ES if you want multi-channel playback since there are no m-c analog outputs. The HATS system is supposed to reduce/eliminate jitter. I'm very pleased with its performance. (faster Fast Forward/Reverse would be nice, though.)
|
|
|
Post by diw February 6, 2011 (3 of 50)
|
|
At one point, I think Kal R had an earlier Oppo SACD player modified with 3 coax outputs for MC, and ran them into his Meridian. I don't know if he ever did a direct comparison of that unit to an Oppo 83 with HDMI into the same prepro.
I used to use the Oppo 83, now I have purchased a Sony 5400 and use the HDMI output into my Anthem. I have not tried doing a direct comparison of the 2 units into the same prepro, but maybe I will someday. I bought the Sony mostly because I have intermittent transport issues with my Oppo on SACDs (but not any other media).
As with Kal, I think the HDMI issue is a difficult one, because a big part of using it for me is the ability to perform room correction.
|
|
|
|
|
|
diw said:
At one point, I think Kal R had an earlier Oppo SACD player modified with 3 coax outputs for MC, and ran them into his Meridian. I don't know if he ever did a direct comparison of that unit to an Oppo 83 with HDMI into the same prepro.
I used to use the Oppo 83, now I have purchased a Sony 5400 and use the HDMI output into my Anthem. I have not tried doing a direct comparison of the 2 units into the same prepro, but maybe I will someday. I bought the Sony mostly because I have intermittent transport issues with my Oppo on SACDs (but not any other media).
As with Kal, I think the HDMI issue is a difficult one, because a big part of using it for me is the ability to perform room correction.
Nothing difficult with HDMI on the Sony Combo either for 2 channel SACD/RB CD/FM and only way for MCH SACD, with HATS enabled jitter is virtually eliminated.Analogue cable only required from DA5400ES pre-out to stereo amplifier if latter is used for front speakers.
|
|
|
|
|
|
ARQuint said:
My crime was to make comparisons between the sound of the stereo high-resolution program on a number of SACDs with that of the same material as heard as an HDtracks download—
1. Which SACDs? and
2. Would it not have been far more sensible to use something like a Marantz KI-Lite Pearl (USA Marantz SA8004) which can play both SACDs and handle download files ... and then you have the same architecture for both sources.
|
|
|
Post by ARQuint February 6, 2011 (6 of 50)
|
|
I've written two pieces about HDtracks, the one for TAS and a second for Fanfare. The latter is in the March-April issue. That isn't in print yet, though it's available online to subscribers now.
In the TAS piece, I compared Kreizberg Dvorak Symphony No. 6 (Kreizberg/PentaTone), Chesky's Piano Concerto (Chesky) and the Blu-ray version of Divertimento (Trondheim Soloists/2L). For the Fanfare article, I also used Holst: The Golden Goose (Hickox/Chandos.)
The experiment you suggest could be telling - I hope someone undertakes it. I am just a lowly music writer without rapid access to a lot of equipment so I used what I use every day for record and video reviews. In addition to the Halide Bridge, though, for a computer-to-Anthem interface I also have a Squeezebox Touch.
|
|
|
|
|
|
ARQuint said:
In the TAS piece, I compared Kreizberg Dvorak Symphony No. 6 (Kreizberg/PentaTone), Chesky's Piano Concerto (Chesky) and the Blu-ray version of Divertimento (Trondheim Soloists/2L).
For the Fanfare article, I also used Holst: The Golden Goose (Hickox/Chandos.)
Welllllll ... both the Kreizberg Dvorak 6 and the Hickox Holst are DSD recordings.
And I assume the download version of these files are 24bit/96kHz, which is:
a) not only lower resolution than DSD, but
b) also has to be artificially converted from the files natural format (i.e. which is an imperfect process)
So ... what you have claimed is basically an impossibility theorem: namely, that a lower-resolution file sounds better than the higher-resolution source file from which it is derived. Which is, of course, nonsense.
[ Unless, we are talking about better in the sense that some people prefer mp3s to CDs ... ]
> The experiment you suggest could be telling - I hope someone undertakes it.
> I am just a lowly music writer without rapid access to a lot of equipment so I used what
> I use every day for record and video reviews.
Well, sorry - but, based on what you have said, your piece appears to have very little to do with downloads vs SACD ... it is really just a comparison of your outboard DAC with your Oppo universal player.
If you want to compare the two, use the same hardware.
|
|
|
Post by ARQuint February 6, 2011 (8 of 50)
|
|
canonical said:
Welllllll ... both the Kreizberg Dvorak 6 and the Hickox Holst are DSD recordings.
And I assume the download version of these files are 24bit/96kHz, which is:
a) not only lower resolution than DSD, but
b) also has to be artificially converted from the files natural format (i.e. which is an imperfect process)
So ... what you have claimed is basically an impossibility theorem: namely, that a lower-resolution file sounds better than the higher-resolution source file from which it is derived. Which is, of course, nonsense.
[ Unless, we are talking about better in the sense that some people prefer mp3s to CDs ... ]
> The experiment you suggest could be telling - I hope someone undertakes it.
> I am just a lowly music writer without rapid access to a lot of equipment so I used what
> I use every day for record and video reviews.
Well, sorry - but, based on what you have said, your piece appears to have very little to do with downloads vs SACD ... it is really just a comparison of your outboard DAC with your Oppo universal player.
If you want to compare the two, use the same hardware.
Do have a look at the TAS piece if you get a chance. When HDtracks works from a DSD original, they use 24/88.2, precisely because the math works better.
And you're right, my article wasn't mainly about comparisons of downloads to discs, it was about the promising future of downloads as a source of audiophile quality recordings.
|
|
|
|
|
|
ARQuint said:
And you're right, my article wasn't mainly about comparisons of downloads to discs, it was about the promising future of downloads as a source of audiophile quality recordings.
> downloads as a source of audiophile quality recordings
Well, that's the funny thing. Because of this thread which you started, I popped over to HDTracks ... and I am frankly appalled at how these audiophile recordings end up as download files.
a) If a recording starts off life as a PCM file (e.g. a 96kHz/24 bit conversion of Kleiber's Beethoven 5th analog recording), then it should correcltly be sent over to HDTracks as that 96 kHz /24 bit file. But that is not what is happening. Instead of receiving the source PCM file, HDTracks appear to get sent the SACD of the file (which started out as a 96 kHz /24 bit file, then upsampled to DSD), then HDTracks send the SACD off to some conversion lab somewhere or other, the conversion lab then downsample the SACD to 88.2kHz ... and you end up with a bastardised file of the original ... with 2 unnecessary conversions ... which you pay money to download ... just plain ridiculous.
On the other hand:
b) If a recording starts off life as a DSD recording ... as per the examples you listened to in your test ... then that pure DSD original is on SACD in its native DSD format, the SACD is sent off to some conversion lab ... downsampled all the way down to 88.2kHz/ 24bit ... and then downloaders pay money for that ... again a bastardised version of the original. There IS NO WAY a downsampled 88.2kHz file can sound better than the DSD recording it is sourced from ... all that has happened is the lab has thrown out more than half of the original recorded data to fit it into 88.2kHz/24bit.
That's not a promising future for downloads. It's lossy at best, and scrambled eggs at worst.
|
|
|
Post by AmonRa February 6, 2011 (10 of 50)
|
|
DSD dynamic range is 120 dB at best, and frequency range in practice ends at 35 kHz because of the infamous noise shaping. 24/88.2 has 144 dB dynamic range and noiseless reproduction to well over 40 kHz. So where did half of the DSD data go to? Black hole of technical ignorance?
|
|
