[Ardour-Users] Experience? Rework an external 44.1 KHz 16 bit recording

[Lamar Owen]

Ralf, you misunderstood my statement.  I'm simply cautioning that audio engineers are not immune to the placebo effect.  Now, when gross errors are caused by cabling issues that's different; but if two delivery techniques produce equal measured specifications and yet one method 'sounds' better, then it's time to start looking at why.  But do see my comments inline.

Ralf wrote:
> So you would use a 40-conductor IDE instead of a 80-conductor IDE, even
> if there would be issues with a PATA drive? 

If throughput of UDMA33 or less is acceptable, why not?  An 80 pin cable is not going to change anything if the speed is UDMA33 or less, and the length is within the ATA spec.

> You would use an audio cable for the antenna, but a 75 ohm antenna cable?

In a pinch, sure; the efficiency will be less, but it can work (I was a radio broadcast engineer full time for over ten years, specializing in AM directional phased arrays that present mutual impedance issues, and have been a consulting engineer now for twenty years; this is one topic that falls inside my area of professional expertise).  For short lengths a paperclip and a piece of chewing gum wrapper can work for S/PDIF; I actually did that one time with two Panasonic SV3200's back to back, and got a perfect DAT copy through S/PDIF. It was one of those situations where the tape was going south, it was very late at night, and I didn't even have an extra audio patch cable available. 

> S/PDIF coaxial 75 ohm and optical might cause issues to, this has
> nothing to do with audiophile placebo effects, resp. I don't know what
> would cause "failure" (to avoid the word "loss", even if such failure
> does cause loss), but there could be issues for S/PDIF. I don't know if
> they belong to the cable, it's just a guess.

Digital cabling tends to either just work or not; however, digital signals are still analog voltages, and capacitance in the coax and optical issues in the case of fiber can occur.  My experience is that there is a very narrow margin of error between working perfectly and not working at all; however, there is an area of hysteresis that exists right on the threshold, and there you can get bit errors and such that are distinctly audible (they tend to not be subtle; something like a dirty head or tape dropout noise on a DAT deck; very distinct pulse noise is the result, at least in my nearly twenty years of experience with DAT decks).

> You don't need a double-blind test, if something is "really" audible.

Certainly; I'm talking about subtle things. From the medical side of things, where the term placebo originates, the placebo effect has always been in subtle cases; no placebo ever (to my knowledge) cured terminal cancer, for instance.

> Audiophile subtleties aren't important for audio engineers and
> musicians, they are just important for idiots. And I wasn't talking
> about this.

No, I've seen enough talk on gearslutz to say that you're not correct on this point; plenty of gearheads, engineers, and musicians out there are fully impacted by the placebo effect when differences are subtle.  Things like which year of the Martin D35 is the best and gives the fuller sound, for instance.  Or certain subtleties in different manufacturer's tubes (which measure identically, but 'sound' different).

> Anyway, theory has less to do with the area of practice.

In practice engineer's ears are just as easily fooled as audiophile's ears are, as both are human ears.

Hmm, I wonder how difficult it would be to code an A/B double-blind plugin switcher?