I think any analysis of jitter audibility without measurements to back what we are hearing is fraught with uncertainties.
Agreed, but few people have the necessary expensive equipment to measure jitter. In the absence of this we have to use best guesses as to what is underlying the sonic differences we hear (I'll address the bias & placebo below).At the end of the day, it's almost immaterial if the improvement is because of reduced jitter or some other change - as long as we can consistently achieve the improvement. My view is that it may take time to research & analyse exactly what is the underlying cause of the improvement.
Let me give you an example which is easy to test. Try using an RF attenuator (these
ones -10dB or -15dB from Minicircuits) on your SPDIF cable & report what you hear. Use DBT if you need to. This is a simple change in a system which I believe effects the jitter arriving at the DAC. Knowing exactly what mechanism is behind this change in sound is interesting but doesn't change it's effect.
I just downloaded a copy of an AES paper on audibility of distortion between op-amps (little amplifiers used in countless ways in our electronics). They tested two op-amps one of which has a lot of high frequency harmonic distortion and one that had the same low harmonics but no higher ones. The listeners in blind testing could not distinguish between the two. Masking theory tells us that these distortions should have mattered more than the lower ones but it did not work out that way in this test. Such disagreement makes us then wonder about our understanding and testing. Same with jitter. I like to have the measurements and see if I know for sure I lowered it before I say X is better because it sounded like it had lower jitter.
I don't know the paper to comment but often the devil is in the details of the test set-up
The other bit is that it is *exceedingly* simple to mistake placebo for effects of jitter. I have tested this myself. I can easily "imagine" more edginess and proceed to hear it that way! I take a *single* audio file and play it on my PC. I can literally play it again and make myself believe it is harsher and hear it that way! Sounds crazy but it is very possible. If the mind can imagine things on itself, it can also be told on demand to do the same (with some discipline).
So while like you I find Bob's words familiar to effects I hear, I have to put out the variability that our ear+brain bring to this space. Measurements don't have this issue so we need to combine the two to get high confidence learning.
Again, it's wise words you speak but there are a couple of points to be made:
- just because a measurement doesn't show any change does not necessarily mean that we are hearing a placebo effect. It could also mean that our measurement is not appropriate or sensitive enough for what is being heard.
- What it behoves us to do is to keep an open mind on the above & basically keep it an open question i.e which is correct, our hearing or our measurements.The sample size begins to matter here - if it's just a sample size of one that hear the effect then yes low confidence level; if a large number of people (on different systems) hear the effect (bearing in mind expectation bias) then we can increase the confidence level. If trained listeners can consistently pick out the improvement (blind) then we can increase the confidence level again.
- Finally, negative bias can have just as strong an effect as positive bias i.e. a person doesn't expect (or want ) to hear a difference & doesn't hear one - even though it exists. Some people even measure before they listen