Peer reviewed or not, there is never enough in published hearing research we argue about here to allow for proper replication. It means there aren't enough details, period.
Mark I have a real technical question for you, how do the 2nd and 3rd harmonic registers in the ESS DAC's account for phase (or do they)?
How about this for a cable test:
You have a common source
You run two cables, one being a ‘reference’ from the common source to subtraction amp
You decide on the audible threshold that determines if there’s an audible difference in sound - let’s say For argument it’s -100 dB
You then look at the output level (if any) of the difference amp to determine whether it is audible.
The signal source must be BW limited to 20 kHz And the diff amp must be flat to at least 200 kHz.
You have a common source
You run two cables, one being a ‘reference’ from the common source to subtraction amp
You decide on the audible threshold that determines if there’s an audible difference in sound - let’s say For argument it’s -100 dB
You then look at the output level (if any) of the difference amp to determine whether it is audible.
The signal source must be BW limited to 20 kHz And the diff amp must be flat to at least 200 kHz.
Only by sign, the numbers are 32-bit signed ints.
EDIT: They may be intended to compensate for resistor mismatches in the I/V and differential summing stages, in which case sign may be close enough.
Last edited:
Ah, yeah, I guess. I get grumpy about all the engineering/materials science journals I *used to* have access to, so the joke flew right over my head.
Or you just pass a signal from a source to a load and look at the signal across the cable. Try a source impedance of 100 ohms and a load between 10,000 to 1,000,000 ohms. Better yet use a dual tone signal that the difference tone is at a most perceptable range around 3,000 to 6,000 hertz. You might also want the test tones to be a few octaves above the difference frequency to make filtering easier...
I agree that the price is too high for an individual. But it may be fine for an interested company or university with mega-bucks to spend. On the other hand, an interested company or university with mega-bucks to spend is probably a member or doesn't care about the cost.
In either case, for $8 - $12 each, I'd be downloading papers left and right. There are at least a dozen over the past 10 years that I'd like to have/read. But I'm not paying the price of a book for a paper!
Last edited:
Some would have us believe it costs tens of thousands of dollars to conduct such a test, so maybe it's not a rip off, who knows?
After reading the content and grammar of the abstract, I would wonder who would be interested in the content of the paper.
It's a tough one to swallow and there is a mix of academic publishing companies abusing their position of near-monopoly and a reasonable need to pay these folks to do their job. A lot of funding organizations are now requiring open access as as result of these access issue.
I'd be sure nice to have better editorial standards, but then again, I might never have published. (along with 75% of all grad students/postdocs/etc)
I was part of the UC system for grad school, and we did not have an AES account, much less where I am now. Then again, UC did not renew their agreement with Elsevier, so we're seeing some major changes in this world.
Ed, I was more being facetious about the need for >300 dB of signal amplification to get a meaningful signal. 😉
I just looked at the paper. He is looking at change in time constants in cables. Basically it seems to point at a voltage coeffient in the cable dielectric that will affect the sound (in an high impedance circuit). Interesting and I'll attempt to read it (Japanese origin so not the easiest of reads). I would have thought the voltage effects would need to be huge to show in an audio link. Especially to headphones. Still something to look at and maybe improve.