The plot is indeed confusing and does NOT show the ability to localize sounds at the respective frequency.
Rather:
- it describes the ability to localize (mono!) sound sources in the median plane (before, behind, above, below), where neither interaural time differences nor level differences are present.
- under this conditions, the ear/brain can only rely on spectral information, i.e., a specific spectral band being boosted
- and this is what the plot is showing: the ability to localize a sound source when a (broadband!) spectral peak is present at different specific frequencies. Indeed, each of the spectral bands has exactly one of the localization probability function peaking onto it, but some positions (like "before") can be triggered by more than one frequency band.
As such, it does not have much to do with phase audibility or even stereo imaging, though it might influence how we perceive broadband peaks in a (poor) loudspeaker's response
Last edited:
----------------------
These are really not accuracy scores for localization. If you are familiar with psychophysical techniques, it is best to think of them as estimates of bias. I don't have time to address your other comments.
I agree that it is probably not relevant to setting up a crossover
I mentioned my original objection since someone claimed that you can't localize signals around 1 kHz. That is simply wrong. The discussion then degenearted into a misunderstanding of "directional bands" and somone accusing me of ego problems etc. Enough with the negativity. If someone in the field tries to get you back on track, then it is best to listen rather than to criticize.
Last edited:
The graph plots the chance of a test person correctly detecting the direction of a sound versus the frequency of the sound. Green is frontal sound, red is from behind, blue from above. The explanation given in the article sums up to: Ears are resonant systems, where the resonance is dependant on the angle of incidence, resulting in a angle dependant filter function and thus the phenomenon which is described in the graph.
The wiki article references a english publication: Jens Blauert: Spatial Hearing. The Psychophysics of Human Sound Localization. Revised edition. The MIT Press, Cambridge MA u. a. 1997, ISBN 0-262-02413-6.
No, that is not correct, and exactly the point of dr. Geddes. Please read again.
"Wird bei einer vorn positionierten Schallquelle in einem dieser Bänder der Schallpegel gegenüber anderen Frequenzbereichen angehoben, so kann daraus das Gehör die Information entnehmen, ob der Schall virtuell von vorn, von oben oder von hinten kommt."
It's all about the ear/brain working as a spectral analyzer and interpreting a boost in the relative level of a certain spectral band as being a front/back/above sound in a particular condition (the absence of the typical ITD/ILD cues, that is, in the median plane)
Attachments
it all depends
I used to have stereo subwoofer, I have better bass right now with my 8 inch 2 -way speaker. go figure.
Last edited:
Having a stereo sub woofer doesn't say much about how any of it was implemented or what the mains were at the time. Still that is nowhere close to being what a multi-sub implementation is about.
I literally translated the explanation on the wiki. Being correct is not my problem, but the wikis, so please take it to them instead.
You're welcome! I will have to get back to you as time permits. The short answer is I don't know. At the time I saw it I took it to be a graphical way of showing why/how the pioneers chose XO points based on then known hearing acuity data, but here it is many decades later and I still haven't gotten around to studying the DESIGNING FOR SPEECH INTELLIGIBILITY section of Don and Carolyn Davis's SOUND SYSTEM ENGINEERING simply because choosing XO points based on my [right/wrong] interpretation of it has worked well for me.
GM
I've seen similiar graphs as well, all peaking at 2khz.......which immediately reminds me of the all the 6" two ways crossed at 2khz. Maybe 4khz isn't even low enough?
The pioneers didn't think so, they chose 5 kHz, which makes sense based on the early hearing curves and really should be up around 7-8 kHz for the more recent ones. I've traditionally used 7.5 kHz once I found out that this was an easy way to deal with many driver's excessive sibilance.
GM
GM
So we also have some of these new large, 'mid' tweeters around. Are these the ideal candidates?
I mentioned earlier I'm using a 2" dome which works right through the 1-4k range. Sounds just fine 🙂
Quantifying tradeoffs as potential problems isn't as easy as you make it out to be and from a subjective standpoint becomes near impossible. The very existence of Bose speakers is a great example.
rather You don't read it 😉
it's not about "a three way with 2" compression driver", it's about:
- Status
- Not open for further replies.