The Hopelessly Derailed ODAC/Objective DAC Anticipation/Discussion Thread

[kiteki]

Originally Posted by nullstring /img/forum/go_quote.gif
 
Can we measure every single thing?

 
There is zero evidence to support that we can measure everything in audio, ask for it and see what happens, you'll be presented with all kinds of assertations (see: belief systems), all lacking in evidence too.  So the only true answer is "No".
 
There is an attitude that audio is an advanced field like mathematics or chemistry, so we should rely on formulas and forensic lab equipment in audio evaluation.  The counter attitude is that the formulas and lab equipment in audio lack evidence and possibly just suck, since they != what we hear.
 
I use the word attitude, because I see a common pattern that both active parties usually share an emotional incentive with marketing.
 
There is a lot of dissonance and inanity in this divide, I think there should be a neutral spirit accepting the continual vices and virtues of science and human evaluation at the same time in an open-minded pursuit of audio, immune to advertising, and stop cherry-picking at experimental science. 
 
 
Irrespective to all this, the ODAC is still a positive product in these areas...
 
- It has a lot more documented specifications than any other DAC I know of
- There is no advertising, and no profit involved
- It can serve as a reference point in audio evaluation / reviews
- It sounds very good (at least, my AKD-23S-HF does)
 
kiteki

 

[frenchbat]

- It has a lot more documented specifications than any other DAC I know of
kiteki

I'd say the Benchmark also enters that category

 

[estreeter]

Can mathematics or chemistry explain the genius of Michaelangelo or Shakespeare ? Audio is part science, part black art, with neophytes endlessly seeking their 'precious', and we all know how that ended ..... 

 

[bellsprout]

sound and hearing are 2 separate concepts that are used interchangeably when referring to audio. this is the source of most confusion and controversy.
 
sound is understood and completely measurable. hearing is not.
 
edit: ODAC knick knack this and that

 

[maverickronin]

Quote:

sound and hearing are 2 separate concepts that are used interchangeably when referring to audio. this is the source of most confusion and controversy.

 
This.
 
If you've got the cash it's not to hard to measure any parameter of a device's electrical output to insanely precise accuracy   It's just a voltage that changes level over time.  That's all there is too it.
 
Relating that data to perception is quite a bit more difficult but the experimental evidence indicates that the standard measurements are good predictors of audibility.

 

[Sapientiam]

Quote:

Relating that data to perception is quite a bit more difficult but the experimental evidence indicates that the standard measurements are good predictors of audibility.

 
Really? Got any links that detail research that relates THD to sound quality?

 

[bellsprout]

Quote:

 
Really? Got any links that detail research that relates THD to sound quality?

take your favorite track, create copies of it that are 1,2,3...n octaves higher and lower, add them together and listen to it and see if it sounds any good to u? 

 

[Sapientiam]

So that would give me in excess of 100% THD, right?

 

[maverickronin]

Quote:

Really? Got any links that detail research that relates THD to sound quality?

 
If you want information about specific thresholds of audibility most of the relevant primary sources are unfortunately behind the AES paywall.  It's not as simple as just THD.  Some kinds of distortion are more audible and/or objectionable so different kinds of distortion have different thresholds.  Just a low THD number isn't some kind of guarantee though it can be a good indicator because of the way such circuits operate.  You need more comprehensive measurements that just a THD percentage to be sure.
 
The proof of concept that such numbers strongly correlate with audible differences is much simpler but we're not allowed to talk about it here...
 
Quote:

So that would give me in excess of 100% THD, right?

 
You can attenuate the harmonics to varying degrees.  -20dB, -40dB, -60dB, -80dB, etc and see where you stop hearing anything extra.  I haven't done that specifically but I've heard similar demonstrations as shown in this video.

 

[kiteki]

 
This.
 
If you've got the cash it's not to hard to measure any parameter of a device's electrical output to insanely precise accuracy   It's just a voltage that changes level over time.  That's all there is too it. [/]

 
Sounds familiar.
 

 

[Sapientiam]

Quote:

 
If you want information about specific thresholds of audibility most of the relevant primary sources are unfortunately behind the AES paywall.

No, I'm not interested in specific thresholds of audibility of various amounts of THD. Rather support for claims that THD (or even another measurement number) is a good predictor of perceived sound quality. I take it that there is none.
 
This is really a philosophical issue - can any quality be expressed in a sequence of numbers? If anyone's seen the movie 'Dead Poets Society' then the scene where Mr Keating gets students to rip out pages from the book 'Understanding Poetry' by Dr J Evans Pritchard is instructive.
 
http://arcadioblog.wordpress.com/2009/03/29/understanding-poetry-by-dr-j-evans-pritchard-phd-fictional-character/

 

[bellsprout]

Quote:

So that would give me in excess of 100% THD, right?

well yea...the trend being the more noise there is the worse it sounds, and a lot of people who prefer colored sound forget how bad something with a lot of noise sounds.
 
small amounts of noise is tolerable, but i don't have any reason to believe there's any form of noise where our response is innately positive. eg, there's nothing innately good sounding about cracks and pops of a vinyl - but it's nostalgic and reminds people of archaic technologies

 

[Sapientiam]

Yep, I agree - the more noise the worse it sounds. So then the problem becomes - how to measure noise? Doing it by taking an FFT (very popular method) sucks rather, its rather like trying to estimate traffic density at night by means of time exposure photography.

 

[maverickronin]

Quote:

 
Sounds familiar.

 
Not really...
 
Quote:

No, I'm not interested in specific thresholds of audibility of various amounts of THD. Rather support for claims that THD (or even another measurement number) is a good predictor of perceived sound quality. I take it that there is none.
 
This is really a philosophical issue - can any quality be expressed in a sequence of numbers? If anyone's seen the movie 'Dead Poets Society' then the scene where Mr Keating gets students to rip out pages from the book 'Understanding Poetry' by Dr J Evans Pritchard is instructive.
 
http://arcadioblog.wordpress.com/2009/03/29/understanding-poetry-by-dr-j-evans-pritchard-phd-fictional-character/

 
If you're talking about that kind of "quality" then it's just whatever you happen to like.  You can still use measurements in comparison to something you already know you like.
 
We can measure fidelity without any reference to a preference and that's what most people mean when they say "sound quality".  That's what the "Fi" in Head-Fi and Hi-Fi is short for.
 
If you want to talk philosophy...
 

 

[Sapientiam]

Quote:

If you're talking about that kind of "quality" then it's just whatever you happen to like.

My use of the word 'quality' doesn't have that meaning. Its OK that yours does, but we're obviously arguing about different things. Which is pointless, so I'll stop.