[Occam]

Feature
Power this: testing audio ICs
SOCs for set-top boxes, television monitors, disk players, and mobile media players have or soon will have HD capability. But another aspect of this evolution—one that could prove even more challenging to SOC designers and test engineers—is that, along with HD video comes a significant increase in the quality of the accompanying audio.

The PDF
http://www.edn.com/contents/images/6418209.pdf
save the above as the following HTML link will break in due time -
http://www.edn.com/article/CA6418209.html?nid=2431&rid=243880419#Can%20you%20even%20measure%20that?

Very interesting read.
FWIW

[PaulHilgeman]

Very Very interesting.

I liked it!!! Bravo, good post!

[mgalusha]

Great post Paul, that was a very good article.

[rollo]

Paul,
        Great find. Imagine using audiophiles methods to make judgements on design. Man, the fun I'm going to have sending this to my engineering buddies is priceless.
       rollo

[Scotty]

Way to go Paul! This article is the irresistible riposte to the Flat Earth viewpoint
that the human hearing mechanism is too unreliable to base an important decision about audio quality on. We are actually not all delusional. I think it is a very good thing that the nature of the market itself has forced chip manufacturers to try to understand what makes a chip sound good or bad when listened to and has made them use listening tests to improve the sound quality of their chips. The future actually looks pretty good for continued improvements in the sound quality of IC's in every product we might purchase. Somebody IS actually listening out there!
Scotty

[Fiji5555]

But looking further, he found that the investigators had conducted a blind test in which the listeners didn’t know which system they were hearing on which passages, and the results were consistent. So, engineers dug further and found that measurable differences existed between the rise and fall times of the optical transducers in the S/PDIF optical link. This difference caused data-dependent jitter, which, on the other side of the DAC, had an audible component. “I believe in blind tests,” Lave says, “but not always in the explanations.”


Ahhhhhhh so don't overlook the importance of measureable differences and also blind listening tests hmmmmm?  =]

[Scott F.]

But looking further, he found that the investigators had conducted a blind test in which the listeners didn’t know which system they were hearing on which passages, and the results were consistent.

I think you might have mis-read that passage....

Some listener observations that sound totally irrational turn out to be not only reproducible in blind tests, but also traceable back to something that actually does show up in measurements.

and then...

Listeners reported that the analog connection produced better sound. The report on the test concluded that this finding was yet another proof of the superiority of good old analog over digital. Lave initially ignored the result.

But looking further, he found that the investigators had conducted a blind test in which the listeners didn’t know which system they were hearing on which passages, and the results were consistent. So, engineers dug further and found that measurable differences existed between the rise and fall times of the optical transducers in the S/PDIF optical link.


That as an aside, there is defiantly a place for blind testing when performed properly but this article does seem to confirm that our ears are more sensitive to those subtle changes than some 'specs only' and 'if you can't measure it I don't believe it' engineers seem to rest their laurels on.

Great find Paul

[Scotty]

I kind of wondered if the audible performance of some chips was due to more than dumb luck and happenstance. The LM4780 based chip amp I am using is probably a result of this new design paradigm at National Semiconductor.
The fact that TI uses "Golden Eared" listening panels to insure that their chips have the consistent dry character that TI considers to be their "house" sound was a graphic illustration of how seriously a multinational corporation takes
the subject of how their product actually sounds. They are not guilty of delusional thinking and sprinkling floobiedust on their chips to satisfy the demands
of obsessive/compulsive audiophiles. I guess they are not convinced that the differences they observe in how a product sounds can be reduced to the effects of comb-filtering and response curve variations.
Scotty

[aerius]

I've wondered why TI class D chips sounded so different from the Tripath ones, now I know.

The most important point of the article to me is that the engineers have to measure the right things, flat frequency response and low THD aren't enough, there's a heck of a lot more going on that has to be explored and measured to get the big picture and start to understand things.  It's good to know that guys at TI and some other places are really trying to explore & add to our knowledge of how hearing works and how to better correlate measurements with good sound, which will help make everything sound better when the knowledge is applied.

Unfortunately, far too many of the "objectivists" stop at frequency response, distortion, and channel separation, then jump on the soapbox and harp on DBT's until the cows come home.  Instead of thinking "hmmm...this doesn't make sense, we need to research this more" they go "what we measured is all that matters and if you don't think it is you're delusional".  And that's a big problem, it leads to endless feces-flinging and nobody really learns anything from it.

[Fiji5555]

But looking further, he found that the investigators had conducted a blind test in which the listeners didn’t know which system they were hearing on which passages, and the results were consistent.

I think you might have mis-read that passage....

Some listener observations that sound totally irrational turn out to be not only reproducible in blind tests, but also traceable back to something that actually does show up in measurements.

and then...

Listeners reported that the analog connection produced better sound. The report on the test concluded that this finding was yet another proof of the superiority of good old analog over digital. Lave initially ignored the result.

But looking further, he found that the investigators had conducted a blind test in which the listeners didn’t know which system they were hearing on which passages, and the results were consistent. So, engineers dug further and found that measurable differences existed between the rise and fall times of the optical transducers in the S/PDIF optical link.


That as an aside, there is defiantly a place for blind testing when performed properly but this article does seem to confirm that our ears are more sensitive to those subtle changes than some 'specs only' and 'if you can't measure it I don't believe it' engineers seem to rest their laurels on.

Great find Paul

No, I didn't misread what he wrote. People heard a difference in the blind listening test and it turned out to be measureable after further investigation. You are trying to imply that people heard something wrong that wasn't measureable? I am confused by what you are trying to say here.

[mjosef]

You are both standing on the same side of the line. 

[rollo]

It just amazes me that what we have been hearing can NOW be measured. No scientist can tell me I do not hear or see what I hear or see. Don't get me wrong, measurements are very important in the design of audio components. However the final test is how it sounds and it does not have to measure well to sound good. Can anyone say SET.
    rollo

[Double Ugly]

Just saw this, but can't get the .pdf to open, thus I can't save it. 

If anyone has it, I'd be grateful if you would send it to doubleugly*at*cableone*dot*net.

Thanks.

-Jim

[Scotty]

Double Ugly,you may have to upgrade your AdobeReader program. A trick I use if the page will load far enough to give me the Adobe toolbar is to click on the save icon on the upper left hand side and save to My Documents. The page will usually be saved to that location and I can open it when it has fully downloaded, when I cannot open it in a browser window.
Scotty

[Double Ugly]

Thanks Scotty, but someone has already sent the file, and it opens just fine. 

FWIW, the link didn't even begin to load the file, and instead gave me -

File error.

Couldn't open the file.

I got the same when I selected the option to download the file instead of opening it.  I've never had a problem loading and occasionally saving .pdf files, so it's odd that this one won't work.

Thanks again.

-Jim

[Scotty]

Double Ugly,What did you think about the articles content?
Scotty