Skip to main content ↑ ↓

Notice

Please note that most of the software linked on this forum is likely to be safe to use. If you are unsure, feel free to ask in the relevant topics, or send a private message to an administrator or moderator. To help curb the problems of false positives, or in the event that you do find actual malware, you can contribute through the article linked here.
Topic: ABX of Nature/evo-adapted cues (i.e. non-music) (Read 5048 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

ABX of Nature/evo-adapted cues (i.e. non-music)

Abstract:
A classic subjectivist argument against ABX testing is that music happens over (extended) time.
Maybe music is not the ideal source medium for ABX. The human ear-brain system probably does better with evolutionarily-important (survival-benefit) Nature-based acoustical cues: lion roar, human scream, baby cry (esp. YOUR baby), thunderclap, rattlesnake rattle, insect buzz, bird chirp, etc. A lot of these are quickies, so it may better fit with the ABX test design.

Not sure this issue has been covered before. I non-exhaustively data-mined the archives – as well as the std. Google search – and came up with little. Throw some links this way if this was covered before.

ABX of Nature/evo-adapted cues (i.e. non-music)

Reply #1 –
Looks like utter bollocks to me.

Why would anyone care about objectively comparing such sounds? I mean, really, I’d like to construct a longer argument against that, but it just seems like complete nonsense.

And this is an abstract of what, exactly? Of your own invention, or of a study published in some silly place? If the latter, would you care to provide a link?

ABX of Nature/evo-adapted cues (i.e. non-music)

Reply #3 –
Yes, because I often hear disgruntled cinemagoers complaining about the deleterious effects of lossy audio compression.

I anticipate another thread full of question-dodging and tangent-hopping. Good luck with that!

ABX of Nature/evo-adapted cues (i.e. non-music)

Reply #4 –
Quote from: hollowman on
Abstract:
A classic subjectivist argument against ABX testing is that music happens over (extended) time.


Of course there's nothing inherent in an double blind test (ABX being one example)  that dictates a short time frame, unless it is set up to require people to be around to switch connections, guard against the testee peeking behind a curtain, etc.  If a subjectivist feels that he can better discriminate by listening to one sample for a week, then have at it!

If the test was suggested to verify someone's claim of obvious differences,  or differences in a quick sighted test, then one would think a quick blind test would be sufficient.

WRT ABX'ing with survival sounds, while evolutionary pressure may promote being able to hear a cheetah stalking you, until recent decades there's been no survival advantage to telling whether it's live or Memorex 

ABX of Nature/evo-adapted cues (i.e. non-music)

Reply #5 –
Quote from: DonP on
Exactly, and even if humans were better able to distinguish, say, the sound of their infant there is no bearing such a fact would have on the validity or usefulness of ABX.  ABX is a procedure and if anything science (if it existed) demonstrating humans are more sensitive to differences in "evo-adapted cues" would point towards new avenues to explore with ABX, and hardly point towards any sort of failing.
Creature of habit.

ABX of Nature/evo-adapted cues (i.e. non-music)

Reply #6 –
Quote from: DonP on
WRT ABX'ing with survival sounds, while evolutionary pressure may promote being able to hear a cheetah stalking you, until recent decades there's been no immediate survival advantage to telling whether it's live or Memorex 
There is also no survival advantage to being able to hear most of what we today call music (& especially complex music like Mozart, Mahler, Miles and Metallica) -- the creation/exposure to which are an evol.-recent phenomenon. We know, e.g, that for real important (emergency) stuff, a lot of ckts (between senses and amygdala) get bypassed (no cortex processing). Relaxed music listening isn't all that important -- so ear/brain handles it differently.
Might this give subjectivists an added argument? Not necessarily.
This is why I'm curious as to how (and how differently) the Nature sounds test. Especially given the same time chunks. E.g. a 5sec thunderclap vs. 5sec. music seg.

ABX of Nature/evo-adapted cues (i.e. non-music)

Reply #7 –
Quote from: Soap on
...
...and hardly point towards any sort of failing.
It may fail if the ABX test worked for Nature sounds but failed for pure music.
Lotsa things may have to be held equal (of course) like the aforementioned time interval. (See prev. response)

ABX of Nature/evo-adapted cues (i.e. non-music)

Reply #8 –
Quote from: hollowman on
Quote from: Soap on
...
...and hardly point towards any sort of failing.
It may fail if the ABX test worked for Nature sounds but failed for pure music.
Lotsa things may have to be held equal (of course) like the aforementioned time interval. (See prev. response)

That's not a failure of ABX you are describing.  That's a failure of human ability to differentiate music as well as "natural" sounds.  (in this artificial scenario you are describing)

The ABX test worked both times.
Creature of habit.

ABX of Nature/evo-adapted cues (i.e. non-music)

Reply #10 –
Quote from: Soap on
Quote from: hollowman on
Quote from: Soap on
...
...and hardly point towards any sort of failing.
It may fail if the ABX test worked for Nature sounds but failed for pure music.
Lotsa things may have to be held equal (of course) like the aforementioned time interval. (See prev. response)

That's not a failure of ABX you are describing.  That's a failure of human ability to differentiate music as well as "natural" sounds.  (in this artificial scenario you are describing)

The ABX test worked both times.
Yeah, hypothetical scenario. I thought that was clear from the OP. Anyway ...
This is what I meant (scenario, of course):

20 music samples (5sec each): 320 vs. WAV --> 65% correctly ident.
20 Nature samples (5ec each): 320 vs. WAV --> 95% correctly ident.

I don't know how realistic the music-sample odds are. Just that there is (hypothetically) a statistically-sig. diff. between the two sample types.

I'm not saying anything is legit. Just that there are no Nature-sound-only ABX tests in the "fossil record". Dunno about anyone else, but I know I'm listening to Nature most of the time.

ABX of Nature/evo-adapted cues (i.e. non-music)

Reply #11 –
As was explained earlier, that hardly invalidates ABX testing.

http://www.hydrogenaudio.org/forums/index....showtopic=29558
http://www.hydrogenaudio.org/forums/index....showtopic=38041
http://www.hydrogenaudio.org/forums/index....showtopic=73686
http://www.aes.org/sections/pnw/ppt.htm

ABX of Nature/evo-adapted cues (i.e. non-music)

Reply #12 –
I’m still waiting!
Quote from: db1989 on
And this is an abstract of what, exactly? Of your own invention, or of a study published in some silly place? If the latter, would you care to provide a link?

ABX of Nature/evo-adapted cues (i.e. non-music)

Reply #13 –
Quote from: hollowman on
Abstract:
A classic subjectivist argument against ABX testing is that music happens over (extended) time.
Maybe music is not the ideal source medium for ABX. The human ear-brain system probably does better with evolutionarily-important (survival-benefit) Nature-based acoustical cues: lion roar, human scream, baby cry (esp. YOUR baby), thunderclap, rattlesnake rattle, insect buzz, bird chirp, etc. A lot of these are quickies, so it may better fit with the ABX test design.

Not sure this issue has been covered before. I non-exhaustively data-mined the archives – as well as the std. Google search – and came up with little. Throw some links this way if this was covered before.


Doing tests like this doesn't seem to be all that hard. Seems like something best done by the advocate for doing them. Be my guest!
 

ABX of Nature/evo-adapted cues (i.e. non-music)

Reply #14 –
This isn't so much about the tests, but rather the erroneous conclusions that might be drawn from the results.

E.G.:
I can more easily ABX two rattlesnakes than I can 16-bit vs.24-bit, therefore ABX testing is flawed.

<rhetorical question>
Didn't someone mention something about jumping the gun in another thread with the same participants?
</rhetorical question>