Digital in the signal path is bad? Really? A look at something that happened.

Discussion in 'AV Receivers' started by Chu Gai, Jul 7, 2003.

  1. Chu Gai

    Chu Gai Lead Actor

    Joined:
    Jun 29, 2001
    Messages:
    7,270
    Likes Received:
    1
    Trophy Points:
    0
    A lot of people don’t like digital and prefer to run straight analog even going so far as to use vinyl. That is, of course, their prerogative to make that choice. Others are quite certain that any sort of AD-DA conversion is to be avoided because it leads to audible degradation.

    But how much merit is there to this? Many of us have performed our own tests on equipment of our choosing and haven’t been particularly careful with regards to either level matching or paying attention to the elimination of biases. As a result, our findings can be skewed one way or the other, perhaps affected by our own predisposition to one philosophy or another.

    The article below, which occurred about 20 years ago recounts some careful work that was designed to test this. The test was conducted in what would have to be considered most favorable for anyone who wants the ‘skeptic’ to be made a fool of or proven wrong. The person who set out to prove the inadequacy of digital was Ivor Tiefenbaum, who is presently (?) the managing director at Linn. He is, as is often said, a respected man and designer. The test was conducted at a place of their choosing using their equipment.

    What’s of particular interest to me are the findings where it strongly suggests (I won’t say proves) that this test had a bit of cheating in it. Read it and you’ll see what I’m talking about.

    Perhaps this will make some rethink their positions or rethink the pains that are taken in order to arrive at meaningful results. I certainly hope it’ll make you think that often times, the results that you see or hear when going to demonstrations are not necessarily because a product is better but because it’s been arranged that the chips fall a particular way. Sometimes you know, the Pepsi Challenge, Monster demonstrations, audible cable differences, happen in a particular way because of something pulled out of Houdini’s book.

    Things are seldom what they seem, skim milk masquerades as cream…and all that.

     
  2. John Garcia

    John Garcia Executive Producer

    Joined:
    Jun 24, 1999
    Messages:
    11,571
    Likes Received:
    25
    Trophy Points:
    0
    Location:
    NorCal
    Real Name:
    John
    I read the whole thing. Very interesting, to say the least.
     
  3. Michael R Price

    Michael R Price Screenwriter

    Joined:
    Jul 22, 2001
    Messages:
    1,591
    Likes Received:
    0
    Trophy Points:
    0
    That's a great article, although it won't stop me from improving my CD player (perhaps not audibly, though).

    Has anyone blind tested the audibility of changing the passive components in speaker crossovers? I hear a lot of people claiming that putting in good components (costing up to $50 for a single capacitor, for example) made the sound a lot smoother, clearer, etc. Chu, what do you think of this one?
     
  4. Chu Gai

    Chu Gai Lead Actor

    Joined:
    Jun 29, 2001
    Messages:
    7,270
    Likes Received:
    1
    Trophy Points:
    0
    I'll confine my reply to things like resistors, capacitors, inductors.

    I think it's entirely possible to replace one component with another and change the nature of the sound. Whether this is good or bad is difficult to answer but one must factor in something called 'builder's ear' which is the tendency for people to be enormously satisfied with something they've created. If you recall the wonderful movie, Unforgiven you'll see an example of that in the house that Gene Hackman built.

    When people replace things there are a couple of things that often aren't done.
    1) They don't measure the component that's being replaced to determine what it's value was. Nor do they measure the value of the component being put in its place. Not all components are stable for an infinite amount of time. Sometimes their values change in but a few years and being that human auditory memory is not a rock, we don't even notice it happening until one day something just isn't right. Depending upon the environment the component is in such as the temperature and even vibration (think about the insides of a speaker cabinet) this deterioration can be accelerated. Also a particular manufacturer, say Motorola for instance, occasionally comes out with a batch of whatever that doesn't age gracefully. As Elvis Costello said, Accidents will happen, it's only hit and run.
    Electrolytic capacitors are a known culprit here because they can dry out and it's entirely possible to find that the capacitance has dropped by a factor of 10 or 100. Plus or minus 5% is one thing. Minus 100% or 1000% and you can be pretty sure there's a profound difference in what you bought and what you now have. You just might've lost the top end of your speaker response or lost the low end on your CD player. Of course, in the latter case, you could always fall back to the digital.
    So if one knows what type of components are likely to age unfavorably, where they're located, and has some skill and chutzpah to delve inside with snips and a soldering iron, they can often restore a component like a CD player to meet it's original specs. Confirmation that there may be a problem can always be done by burning test tones on a CDR and playing it while measureing the output voltage. Of course this takes the fun out of upgrading and putting a new look on our equipment that now looks a bit dated.

    2) When one looks in an electronic parts catalog like Mouser, one finds a ton of different flavors of a particular component. Resistors for example differ in things like rated power, maximum working voltage, operating temperature, etc. They also can have other electrical properties that cause them to differ. Inductance for example. A wire wound resistor has a non-significant amount of that and were you to replace a resistor with neglible inductance with a wire-wound one, you would be changing what that circuit was designed to do.

    It is a strange thing that compels people to replace good components with other good and even more expensive components. Most though would not tolerate this sort of behavior from their auto mechanic. Golly gee Mr. Good Wrench, thanks for taking my brake pads out that had 30,000 miles left on them and putting in these Raybestos ones. Yeah, that chick that advertises them sure looks hot on TV. $450 for the job...thanks!!! Can I have a calendar too?
     
  5. Michael R Price

    Michael R Price Screenwriter

    Joined:
    Jul 22, 2001
    Messages:
    1,591
    Likes Received:
    0
    Trophy Points:
    0
    "It is a strange thing that compels people to replace good components with other good and even more expensive components. "

    This is exactly the thing I was talking about... for example, upgrading the crossover in a nearly new GR AV-1, JBL S38 or Klipsch RF-7... that would mean putting in inductors with fancy wire, polypropylene instead of mylar capacitors, etc. See this thread: http://www.hometheaterforum.com/htfo...hreadid=140533

    The point I'm driving at is that any distortion caused by passive components such as those in speakers is known to be less than that of any active component, like a transistor... and then there are also dozens of capacitors and such in all our equipment. So if there really are differences in crossover components, there sure as heck are differences to be found in digital equipment.
     
  6. RobertR

    RobertR Lead Actor

    Joined:
    Dec 19, 1998
    Messages:
    9,767
    Likes Received:
    197
    Trophy Points:
    9,110
    Good article, Chu. [​IMG] I've read of a similar experiment done by the late Gabe Weiner (recording engineer), who had people listen to a direct mike feed of music being performed vs listening to it through a real time A/D D/A chain.

    They couldn't tell which was which.
     
  7. Chu Gai

    Chu Gai Lead Actor

    Joined:
    Jun 29, 2001
    Messages:
    7,270
    Likes Received:
    1
    Trophy Points:
    0
    Hey if you want to replace components, knock yourself out but do yourself a favor and measure what's being replaced. That seems like a prudent thing to do. Take a look at many of the threads of people who replace things and claim improvements. All DBT issues aside, just how often do you see this being done? For some fun, try logging in and asking that question.
     
  8. Jack Briggs

    Jack Briggs Executive Producer

    Joined:
    Jun 3, 1999
    Messages:
    16,738
    Likes Received:
    129
    Trophy Points:
    0
    Stanley P. Lipshitz, by the way, is an enormously respected audio theorist and engineer. His informed commentary has often graced the pages of Peter Aczel's irregularly published The Audio Critic (a magazine which began, in 1977, as yet another golden-eared venue and morphed, over the years, into a scientific, subjectivist publication).
     
  9. Michael R Price

    Michael R Price Screenwriter

    Joined:
    Jul 22, 2001
    Messages:
    1,591
    Likes Received:
    0
    Trophy Points:
    0
    I agree no one measures the value of the components, which is sort of silly. It's almost as scary that some expensive "audiophile" capacitors used in crossovers have tolerances of 10%. No one bothers to match them. It's another reason I will use cheap, matched 2% capacitors and 1% resistors in my crossover. [​IMG]
     
  10. Kevin C Brown

    Kevin C Brown Producer

    Joined:
    Aug 3, 2000
    Messages:
    5,712
    Likes Received:
    0
    Trophy Points:
    0
    Chu- Good article. I only skimmed it (suspected I knew the results anyway... [​IMG] ), but I didn't see anywhere what the bit length and rate was? Kind of neat too: 1984, almost 20 years ago!
     
  11. RobertR

    RobertR Lead Actor

    Joined:
    Dec 19, 1998
    Messages:
    9,767
    Likes Received:
    197
    Trophy Points:
    9,110
     

Share This Page