Exact sound level for quality comparisons....... assessing sound quality

In review, the Towshend Allegri Reference, which has 0.5dB volume steps................

While making precise subjective comparisons at a very high quality level I judged a moderate sound quality difference between the test examples, for several parameters, including clarity, timbre and image depth , and these were noted.

On rechecking we found there was a calibrated 0.5dB error in level for the comparisons and when this was corrected those audible quality differences fell away.

However the overall quality difference assesed by comparisons with other references remained so the tests themselves were suffiently discriminating.

How many published comparisons are potentially at fault simply due to a lack of attention to precisely matching source replay levels?

Martin Colloms

Hans, I dispute your use of the term ‘any quality differences' ...!

Perceived sound quality is dependent on numerous sensitive aspects, of which sound level is just one.
Great systems and installations may be played louder than inferior ones.

The ear has to put up with many deficiencies in reproduced sound and when excessive tend to increase fatigue, both with extended listening and with increasing loudness.

The exact level matching issue is commonly found with digital audio sources with a nominal ‘2V’ output, of more usually 2.1V , here to try and gain a demo advantage, and we always guard against this possible error in level .


Martin

Hans : Fair enough and ’any quality differences’ was overstating the point. I was reacting to your statement in the previous post ’
... I judged a moderate sound quality difference between the test examples, for several parameters, including clarity, timbre and image depth , and these were noted. On rechecking we found there was a calibrated 0.5dB error in level for the comparisons and when this was corrected those audible quality differences fell away’.
I find A/B-ing quite difficult for the reasons you mention, amongst others. The fatigue issue for instance is not always easy to notice in short comparisons. Another complicating factor, in my limited experience, is that different systems seem to have different optimum volume levels.

MC replies:
This is a wonderfully complex subject which has occupied many minds for more than 70 years.
Imperfections in the recordings, sound system and room together control tolerable dynamic range.
Tolerable because response irregularities, distortion, room standing waves and program quality work together to determine and even limit subjective loudness, on grounds of fatigue and even annoyance.
For almost every programme and reproducing system there is a ‘natural’ sound level, volume setting, for the best experience, the optimal degree of musical communication. Too quiet and the low level sections lose detail. Too loud and the ear is overloaded by the acoustically driven room modes.

In a given room the better the system and the set up, this for all components including the listener position, the clearer and cleaner it sounds and the louder it will play.

That is why headphone listening allows much greater sound levels, the room is eliminated though with other compromises to the sense of realism.


Martin Colloms

Edited by user 09 February 2020 10:02:29(UTC)  | Reason: Not specified