Thursday 27 June 2013

Two Reasons Why Audio Comparisons Are Contentious

How often is it you hear claims made that A sounds better than B, or conversely that C sounds indistinguishable from D.  And no sooner is the claim made than somebody steps in to vehemently assert the opposite position.  Opinions on high-end audio-related issues are often held with what amounts to a religious fervour, and very seldom do they resolve into anything better than a broad consensus.  On some issues, even a consensus fails to emerge.  Either way, blood-pressure-raising emotions can be counted on to rule the day.  Why is this?  What is so difficult about high-end audio?  Here are a couple of discussion points that attempt to shed some light on this.

1.  Apples and Oranges
Paul McGowan, President of PS Audio, recently took some heat for a post which promulgated the notion that a good digital rip of a vinyl LP should be indistinguishable from the original.  Many objections to this were raised, of which the most important (in my view) was that Paul's analog playback system was not of the same sonic quality as his digital playback system.  To be fair, Paul accepted the validity of this argument, but he still sticks by his position.  It is a great illustration of the Apples vs Oranges aspects which can be validly raised in regard to pretty much any comparative exercise in the high-end audio domain.  While it may be that a perfectly valid conclusion was reached in Paul's listening room, using the equipment Paul had available, and the source material Paul used, it would be a mistake to suggest that his experiment - taken in isolation - properly supported the conclusions Paul attributed to it.  The point is that Paul's viewpoint may ultimately turn out to have been right or wrong, but either way his experiment does not on its own prove anything.  In order to prove that A is indeed different to B, every single aspect of the comparison which is not inherently part of either A or B in the totality of their manifestations must be identical (or separately proven to be effectively identical).  In Paul's case, he was unable to eliminate the separate contributions of the analog and digital playback hardware.  Sometimes complete separation can never be achieved.  For example, there is a great deal of contemporary interest in DSD and a consensus is emerging that DSD sounds better than PCM.  However, when making an assertion on this issue either way, it is impossible as things stand to eliminate the separate aspect of any given DAC's design and fitness for purpose from the inherent qualities of the format.  Any comparison may simply favour the better-designed DAC.


2.  Is Something Inaudible if I Can't Hear It?
The supposedly gold standard of audio testing is the double-blind test with a panel of listeners, although many criticisms are directed towards what its proponents claim as the "only indisputable scientifically based" methodology.  But how can one be assured that the listeners will end up noticing any differences that in reality do exist?  When it comes to the ultra high-end, there are very few truly "golden-eared" listeners.  If a listener repeatedly and reproducibly cannot hear any difference between A and B under controlled conditions, must we conclude that there are no differences?  I think not, and I want to illustrate this with an episode from my own recent past.  I have never been able to hear differences in absolute phase.  Try as I might, I have never been able to reliably detect anything.  Recently, Peter McGrath, a highly respected recording engineer who also serves as Sales Director for Wilson Audio Specialties (which designs and manufactures loudspeakers that most people's retirement funds couldn't buy), was setting up a demonstration for me using some of his own recordings.  The last thing he did was to establish if the absolute phase was correct.  After swapping the phase a couple of times he pronounced himself satisfied.  I was floored.  At each stage of the setup process I could follow what he was doing, and heard everything he was listening for.  But I heard no differences as he switched between absolute phases.  I won't go through it all, but Peter was kind enough to show me the things to listen for (and why he chooses a particular recording to listen for it), and to train me to pick up on them.  Within ten minutes I could clearly identify when the absolute phase was correct, on those recordings where it makes a difference (and not all do).  Self-evidently I was "hearing" nothing that I wasn't hearing ten minutes before, but I had clearly become "aware of" things, of which I previously was not.  That is but one small example of the fact that when a random panel of listeners unanimously report that they cannot hear something, that cannot be taken as proof that there was nothing there to be heard in the first place.  Admission to the panel of listeners needs to be based on validated listening skills.

Wednesday 5 June 2013

iTunes 11.0.4 working fine !!

We have been using iTunes 11.0.4 all day now and have encountered no problems.  It looks like it should present no major problems for BitPerfect users.

OS/X 10.8.4 working fine !!

We have been using OS/X 10.8.4 all day now (and all last night) and have encountered no problems.  It looks like it should present no major problems for BitPerfect users.

iTunes 11.0.4 !!

Wow.  iTunes 11.0.4 is now being pushed out.  I hope they have not "fixed" the problem with 11.0.3 that caused us to push out BitPerfect 1.0.7!  If so, 1.0.7 will stop working.  Please do NOT download this latest version of iTunes until we have checked it out...

Tuesday 4 June 2013

OS/X 10.8.4 Update

OS/X 10.8.4 has just been released.  I will spend tomorrow checking it out for any incompatibility issues and will report here in due course.

iTunes Shuffle Mode

We have discovered another problem with iTunes 11.0.3.  It seems that when we use Apple Scripting Bridge to ask iTunes whether or not it is in Shuffle mode, the answer now is always NO.  Therefore, BitPerfect now has no way of knowing whether or not iTunes is in Shuffle mode!

This is only really a concern from the perspective of Gapless Playback.  When BitPerfect detects that iTunes is in Shuffle mode, it temporarily disables Gapless Playback.  But if it thinks Shuffle mode is still off, it won't do that.  The net result is that it will pre-load the wrong track into memory.  When the current track finishes, the (wrong) pre-loaded track starts playing for a brief moment before BtPerfect realizes that iTunes wants to play a different track.

The only work-around I know of at this time is to temporarily check "Disable Gapless Playback" in BitPerfect's Preferences menu every time you enable iTunes' Shuffle mode.

This problem may also affect iTunes 11.0.2 as well, but I can no longer test that.