Dismiss Notice
We are making some updates and reconfigurations to our server. Apologies for any downtime or slow forum loading now or within the next week or so. Thanks!

Audio Fidelity and HDCD

Discussion in 'Audio Hardware' started by PeaceableKingdom, Feb 24, 2009.

Thread Status:
Not open for further replies.
  1. fathom

    fathom Forum Resident

    The decoded HDCD WAV file will actually have 20-bit resolution, but the decoder saves it as a 24-bit file (the 4 extra bits are just padding and don't take up any appreciable space).
  2. I've always felt there was an improvement with HDCD mastered CDs.

    Was looking forward to purchasing some of Steve's discs with HDCD as I just acquired an OPPO player that has the function. I'd been without HDCD for about 2 years, since my old Toshiba HDCD 5 disc DVD player stopped working and was sent the recyclers, as it was too expensive to repair.

    My question for Steve would be, why did you use the HDCD process if you mastered the albums in such a way that there is no real difference between HDCD and no HDCD? I thought HDCD encoding was supposed to improve the audio quality of a standard redbook CD by giving it a 20 to 24 bit type signal within the 16 bit format.
    With my old Toshiba DVD player, I could change to a feature that would knock-out the HDCD encoding while a song was playing. It was some sort of "wide stereo" function or something. That function would not work with HDCD at the same time.
    I had the good fortune several years ago to be involved in a shoot-out in the main, very high-end sound room at Audio Ark in Edmonton. Got to hear a test with an HDCD encoded disc vs a different CD mastering of the same title without HDCD and a U.K. vinyl LP, of again, the same title. The HDCD knocked me on my a**. It was amazing.
  3. CraigVC

    CraigVC Senior Member

    Portland, OR
    I tracked down the thread in which Steve Hoffman says they tweaked their HDCD equipment so that the end-result was "neutral":


    The specific quotes that I'm referring to are these:


    The entire thread is interesting because initially Steve's opinion of HDCD was not good, but after tweaking the equipment he hit upon a solution that sounds fine to his ears.

  4. Dansk

    Dansk Forum Resident

    Ontario, Canada
    Very simple, actually, I just had the syntax wrong.

    Rip a CD to WAV using EAC or whatever program you like. Then put the hdcd.exe file in the same folder as the WAV files.

    Open command prompt and navigate to the folder with the WAV files in it. (Using the usual DOS commands, cd.., dir, etc)

    Then once you're in the right folder, type this:

    hdcd.exe output.wav input.wav

    You can change output to whatever you like, but input.wav has to be the name of one of the wav files in the folder.

    My problem is that I had the two backwards, input first, and then output. The order MUST be output - input.
  5. 5-String

    5-String μηδὲν ἄγαν

    Sunshine State
    Thanks Dansk,
    makes sense now.

    Also interesting quotes. Like I said, unless I did something wrong, I could not hear any major differences between the HDCD files and the CD played in a non capable HDCD system.
  6. Shakey

    Shakey New Member

    Chicago, Illinois
    I still think HDCD improvements over Redbook are subtle and virtually impossible to compare the two. I owned a Sonic Frontiers SFCD-1 that was HDCD compatible and since I could not switch it off, on HDCD discs, I wasn't able to A/B it.

    And unlike a lot here, I think too much emphasis is placed on the chips that are only part of the equation, with respect to overall sound of players and DACS.

    But YMMV
  7. beckfan2

    beckfan2 New Member

    any newer computer is equipped to payback in hdcd?

    I thought cd had full dynamics dynamics and was indistinguishable from an old master when done properly. I'm pretty sure that hdcd only increased peak levels. Seems kind of pointless to do it in hdcd.
  8. There is a big difference between playing back HDCD encoded CD's on a regular CD player and a HDCD-capable CD player IF the HDCD encoded CD uses the peak extension feature of the HDCD format.

    For example, the Neil Young HDCD's use that feature, and when you play back the respective CD's on a regular CD player, they sound compressed and have reduced dynamic range. The full dynamic range is restored when properly decoded with HDCD.

    I don't think that the Audio Fidelity HDCD's use the peak extension feature, so the playback on a regular CD player should not be compromised.
    artfromtex likes this.
  9. Key

    Key New Member

    , USA
    Yep rjstauber is correct. You will basically get a clipped/compressed CD if peak extension was used.

    I think when you cut through the hype around HDCD there is really nothing to be gained by it. Peak extension for example is throwing an unneccessary weak link into the recording chain - A compression before playback and an expansion during playback. Have you ever heard of more than one engineer agreeing 100% on what compressor to use or what expander is the "best"? Then they supposedly print the control signal for the expander on the least signifigant bit - has anyone ever been able to extract this info with a computer?

    You get better results with a regular old maximized CD and Waves C4 uncompressor imho.
  10. emmodad

    emmodad Forum Resident

    bay area, ca
    1/ regarding the various comments about HDCD decoding via use of WMP and hdcd.exe: note that hdcd.exe does not implement full HDCD decode functionality (no conjugate decode of the realtime dynamic decimation filter switching, hence a portion of the HDCD sonic benefit is not realized); WMP also has issues and may be phasing out HDCD compatibility (private communication) if it hasn't already disappeared from the latest release.

    you wrote "IF the HDCD encoded CD uses the peak extension feature" (which logically means any and all use of peak extension); this is simply not true. Peak Extension (which is optional) is applied only for very brief peaks; and for levels below onset of the soft limit, there is nothing but a gain factor.

    so, there could perhaps be a problem if Peak Extension is applied very poorly, ie to an inappropriate extreme... however this is certainly not typical in projects mastered by engineers experienced in using HDCD techniques.

    as above, the declarative statement of "if peak extension was used" is rather inflammatory.

    from that and the statement "they supposedly print the control signal... with a computer" it seems that you may not understand the technologies and functionalities used in the HDCD encode/decode process. here on SHF you can stroll through this thread for some useful info: http://www.stevehoffman.tv/forums/showthread.php?t=124020&highlight=hdcd

    really valuable is to read the Keith Johnson and Flash Pflaumer paper introducing HDCD (and a lot of very good info on the psychoacoustic research behind the technical processes) "Compatible Resolution Enhancement in Digital Audio Systems" presented at AES 101st Convention (Los Angeles 1996). Can be found using the Wayback Machine directed at the old hdcd.com website.
  11. Key

    Key New Member

    , USA
    I've read enough on HDCD and I guess I just don't buy it. Some of the techniques they use I believe are bottlenecked by the 16-bit format. I think that going from 16-bit>24-bit is still 16-bit - no matter what expander or carrier you utilize with it.

    I think the problem is that the more features you implement onto a CD the more subjective the results will be on the user end - too much processing that can yield different tonal results. And I don't think it's worth the supposed benefits. Now I have technically more dynamics but on HDCD player A it sounds thin and on HDCD player B it sounds fat.

    Anyway regardless of my philosophy on the process why not just try some simple listening tests. Get a CD that is maximized (limited) and run it through an expander like Waves C4 (UNcompressor preset with -12dB) and compare it to one of the supposedly good HDCDs with peak expanding and all that jazz. See which one recovers more dynamics and which sounds the most "transparent".
  12. PMC7027

    PMC7027 Forum Hall Of Fame

    I don't believe that the "Peak Unlimiting" feature of HDCD is its main selling point. The Pacific Microsonics digital filter provides multiple decoding algorithms. One of the things that HDCD encoding does is to embed a code that tells the digital filter which algorithm to use at any instant. I don't remember all of the details, and I no longer seem to have my electronic copy of the AES paper, but the properties of the filters differ in regard to impulse response, phase response, frequency response, etc. As part of the HDCD encoding process the musical information is analyzed and a determination is made as to which filter algorithm should be used at each moment. The sonic advantage of proper HDCD decoded playback is that the D/A process if more optimized by appropriate "filter switching" in comparison to using a single filter.
  13. PeaceableKingdom

    PeaceableKingdom Member Thread Starter

    Thanks for the response Shakey, rjstauber, key, CraigVC and everyone else!

    After reading through all of these posts, I think I've decided that it probably isn't worth it to me to buy a new machine just to get the HDCD capability.

    I love the player I have now! It is a much modified unit and the technology it incorporates isn't something that is being duplicated at this time by any of the HDCD capable units that are available. If I were to switch to a new machine just to get HDCD compatability I would, for the most part, have to abandon my ideology and approach to building, what I believe, is a properly functioning audio system.

    I had considered buying an HDCD compatible unit in the past but decided against it for the reasons above and because I only had a handful of HDCD cd's. Knowing that I was going to be buying a lot of the new Audio Fidelity discs got me to thinking again about buying a new player to get the HDCD function. But, as Shakey has stated, I don't know that the change to HDCD decoding is more important than the overall design of the player itself. The real question is: Will a poorly designed HDCD player outperform a well design non-HDCD player?

    So, now that I've decided not to buy a new cd player, I've got some money burnin' a hole in my pocket! So, how about it!?! When are the next Audio Fidelity discs due out? What's taking so long?!?!?:laugh:
  14. reeler

    reeler Forum Resident

    Neil Young and the Grateful Dead seem to be on the HDCD train. I have a lot of HDCD discs by them, and some columbia jazz discs that are not marked as such, but the light comes on when played on my denon 2910. It was hyped up for so long before it actually came along that it sounded like too little too late. There's a little more space and dimension is about all I hear, nothing dramatic. On some discs there are weird tonal artifacts, maybe this was just the engineers learning how to use it. Since Steve seems adept at getting even his vinyl and cd/sacd close, I have no doubt his HDCD's will be unscathed however you play them.
  15. Rose River Bear

    Rose River Bear Senior Member


    Don't hesitate to buy an HDCD player if you have the money. I have a Rotel and have compared the sound and there is no comparison-T the HDCD player with an HDCD encoded disc has much more presence and realism. I have loads of HDCD discs and hope they are around for years to come.
  16. Gary

    Gary Nauga Gort! Staff

    I've "bought it". Literally. :D

    But I never did try to understand the way it works or anything like that. I had to get a new CDP, it had HDCD capability - good enough!

    Agreed! It's a definite level up from normal CD playback!
  17. Key

    Key New Member

    , USA
    Don't get me wrong I am sure they can sound good. I just wouldn't do it with my own music so it's really low on a priority list. If there was a CD player that sounded good to my ears and had it on there well that would be a cool bonus I guess. But would I buy a player because of it or, have that be a selling point in anyway? - no.

    And I just don't trust it, what can I say. It's like pre-emphasis. Just throws more complications into an already complicated chain of events. And makes yet another weak point in the chain for engineers to go back and forth over. I would rather get rid of as many of those as I can.
  18. Taurus

    Taurus Senior Member

    Houston, Texas
    Don't know if this applies to the AF discs, but years ago when HDCD debuted I remember reading that part of the (potential) improvement one hears with an HDCD-encoded disc could be attributed to the extremely high-quality analog-to-digital convertor (which includes the HDCD encoder) the studio used, which was manufactured by Pacific Microsonics.

    Pacific Microsonics was bought by Microsoft :( IIRC in the early 2000s so am not sure if they are still sold under that name (why Microsoft bought an ADC manufacturer is still a mystery to me).
  19. Jamie Tate

    Jamie Tate New Member

    They don't make that converter anymore. Those things are treasured among the mastering engineers I know. I remember when Denny Purcell died people were asking when they were going to sell his many pair. Newer converters come on the market yet they still use their Pacific Microsonic boxes. You have to believe there was something special about them.
  20. emmodad

    emmodad Forum Resident

    bay area, ca
    HDCD - various reference papers re 24/192 et al

    yes, the filtering is widely-regarded a key attribute of the HDCD process; this is one of the reasons that decoding with hdcd.exe is unfortunately only a partial solution.

    some people are of the opinion that the dynamic range functionalities are secondary, but there is no doubt that their use can yield wonderful results when skillfully applied by the mastering engineer. sounds like Steve has hit a really good balance of using these functions subtly, so that he's satisfied with what he calls "neutral" results on non-HDCD playback equipment! It would be great if he could chime in with general comment on this (without giving away any of his "secret sauce" ingredients, of course.....)

    a good article regarding the art of using the HDCD process (in which Jon Marks also addresses so-called "unmarked HDCDs"..... why you may see discs identified by players and SW as "HDCD" even though they do not use all/most of the HDCD capabilities, ie perhaps only PM's dither function was used in creating the final 16/44.1 master): http://www.jmrcds.com/Techsample.html scroll down to "About HDCD"

    David, as noted previously, the AES 76 paper is (still) available for download from the (now defunct) Microsoft HDCD website by using the Wayback Machine (all: if you don't know what this is, check wikipedia entry for Wayback Machine; read the Rocky & Bullwinkle portion for context; then read about the Internet Archive Wayback Machine. aside from being useful to find this old HDCD info, if you have never used the Wayback Machine, you may end up spending as much time as when you started to use Google search, or discovered YouTube....)

    [ Wayback > HDCD.com > 20010618 (that's 2001 Jun 18) > HDCD.com home page > HDCD Technology Resources (Articles and Papers) ] is a page which still contains available public download of the complete referenced AES paper. Since these archived sites can be slow to load (very few accesses, or perhaps located very deep in the archive.org mirror caches...), here is a direct link to the page:


    along with the AES76 paper, there are useful documents specifically addressing Gain Scaling (the dynamic range processing incl Peak Extension); a Decoding FAQ; and a "quaint" dated old Mix article from 1999.... which is also available without pics here: http://mixonline.com/mag/audio_pacific_microsonics_hdcd/

    in addition to presenting significant detail about HDCD processes and architecture, the AES paper goes into some of psychoacoustic research background underlying HDCD (KOJ's work in this regard is often referenced) and issues about importance of hearing sensitivity and high frequencies -- basis for work in realizing the dynamic selection of optimal filters as David alluded to above.

    for a very complete read, get the Pacific Microsonics Model Two manual here (from when it was last being manufactured, by Euphonix): http://connect.euphonix.com/documents/model_two_op_man_301.pdf

    the PM boxes are not just A/D, but contain A/D, HDCD encode, HDCD decode, dithering, D/A, & other stuff... the D/A was a tweaked UltraAnalog module (Dan Lavry design) and is still highly-regarded, despite its age. I heard Keith speak recently and IIRC I think it was stated that the A/D was also an UltraAnalog design, but I am not fully certain of this. Whatever it was, the Pacific Microsonics Model One and Model Two are also sought after for their A/D.

    Although the full story is certainly not available outside of non-disclosure agreements, it has been noted publicly on several occasions that MS acquired Pacific Microsonics for IP (Intellectual Property) reasons: not for audio purposes, but specifically for the method used to transmit HDCD control information.

    HDCD inserts pseudo-random-noise-encrypted control signals into the audio data stream a small fraction of the time (this is the "coding inserted in the LSB" stuff, read above papers for both simple and about-as-detailed-as-you-can-find-publicly overviews). MS apparently uses this as part of a watermarking protection scheme for software distributions, to detect changes and potential modifications.

    This is also alluded to briefly in the article linked earlier above in post #2.

    yah, revered. just go onto gearslutz, PSW or other such webboards and see how folk swoon when they found that one was for sale and was snapped up. Paul Stubblebine in San Francisco is certainly regarded as a Mastering Engineer; to HDCD fans he's like the king of pRon for the pictures he has posted of his 12 racked PM units...old page http://www.paulstubblebine.com/mastering/ma_ad.php ; search gearslutz for an HDCD thread with lots of history and the aforementioned pics...

    There is word that the Berkeley Audio Design folk (several of the PM principals and designers, creators of the superb "alpha DAC" which has HDCD decoding) are contemplating design of a Model Three... this could be a serious wow.......

    hope this was useful for y'all
  21. PMC7027

    PMC7027 Forum Hall Of Fame

    Thanks for the webarchive link. I can't get to it from work but will visit it tonight from home.
Thread Status:
Not open for further replies.

Share This Page