Right. It would be easiest with an old-school analog tone generator where you could just turn the knob up and down but you could still do it with a software generator.
SL didn't think the Shure earbuds he tried needed any EQ and they are pretty inexpensive. But again that's his ears and it might not be true for others.
fb said:
Yes, thats my intent, but the fact is that I don;t have many "other speakers". I have an older set of Boston Acoustics that I will record as a baseline, but other than that I don;t know how I would get "other speakers" to record.
The music is choosen to be widely know, like Paul Simon Graceland (who doesn't have that?) and Diana Krall, again a widely known source. ITs not going to be easy to make this work especially with people who have no experience with binaural simple because binaural recordings take some getting used to. But I know many people who use the Etymotic Insert phones as a reference because they are that clean. The headphones will not add a lot of the kinds of errors so common in loudspeakers and rooms.
Will this work in a wide marketplace, I don't know. But if it does, think about it. You would never have to leave your computer to buy a pair of speakers. Currently distribution eats up 1/2 or more of the cost of the speakers at the consumer. This seesm to me to be excessive, and its all because people believe that you have to audition loudspeakers. The sales point knows this and they use it to their advantage. Markup on speakers is very high to exhorbitant for just this reason.
gedlee said:
The Sony MDR-V6 or MDR-7506 over-the-ear phones are somewhat a 'standard' in mixing studios. Not too expensive either. The two models are supposedly the same thing, one marketed to consumers and one to pros. It would still be a good idea to do tone sweeps to see if they need EQ.
Edit: Earl, you have a real advantage that you can quickly A/B the speakers and your recording of them to see if they sound the same.
So you are trying a budget version of the Realizer headphone?
An interesting way to sell speakers, if it works.
An interesting way to sell speakers, if it works.
Thats system works fine and will create a very good impression of any room and speaker combo that you have the impulse responses for. The advantage is that you can listen to any music that you want to. Maybe I'll sell the impulse responses for my speakers in my room!!
But its been know for a long time that one could easily build in the EQ for auralization into the recording itself rather than do it "on-the-fly" as that product does. In other words, I can, and will get the impulse responses from each source in my room to each ear and from this I can construct a filter that I can pass any wave file through in non-real-time. Then if I put these files on my Creative Player (like an I-Pod, but better) I can listen to my favorite songs as if I were at home, anywhere in the world. Once you get the headphone EQ worked out the rest is easy.
But its been know for a long time that one could easily build in the EQ for auralization into the recording itself rather than do it "on-the-fly" as that product does. In other words, I can, and will get the impulse responses from each source in my room to each ear and from this I can construct a filter that I can pass any wave file through in non-real-time. Then if I put these files on my Creative Player (like an I-Pod, but better) I can listen to my favorite songs as if I were at home, anywhere in the world. Once you get the headphone EQ worked out the rest is easy.
catapult said:
I have those headphones, I prefer the Etymotics. In reaserch, the Etymotics are accepted as reference sources, but the Sonys are not.
The auditioning will be a bit tricky since I only have singles of each one built.😀 Mostly I'll only be able to say whether there is some potential or not. The two little ones that have already been built definitly sound different. Very interesting to experience how the flare rates effect response and sound.mige0 said:
I've spent too much time under the Sony MDRs. Not crazy about them. Maybe they just need to be EQ'd.
Will be interesting to see if Earl can mimic the sound of his speakers (and room?) without the custom HRTF. Seems allmost all the reports of the Smyth A8 say the custom HRTF really makes the diference. Tho reports say the generic one is "OK", but not great.
Being able to get away from the custom HRTF would be a big step forward.
Will be interesting to see if Earl can mimic the sound of his speakers (and room?) without the custom HRTF. Seems allmost all the reports of the Smyth A8 say the custom HRTF really makes the diference. Tho reports say the generic one is "OK", but not great.
Being able to get away from the custom HRTF would be a big step forward.
panomaniac said:
As I said before this is exactly what the situation is. The generic ones are "OK", thats about it. The question is would the rank ordering of the loudspeaker systems change or would all loudspeakers suffer equal degradation. Lidia and I plan to do this study, but Genelec already did a similar study and found that with specific HRTF EQ even the absolute ratings did not change. They alluded to the rank order not changing even under generic HRTFs.
Whats important is not that the absolute sound quality be identical, but that the relative sound quality between different models and different manufacturers be the invariant. If the later holds then the concept works once people get used to the differences in binaural listening. Greisinger is correct in pointing out that these differences get adapted to, but it takes time.
What cannot be done, of course, is to use the binaural recordings in any comparison with a real audition. That just isn't going to work. It won't be possible to listen to the recordings and say. for example, that you do or don't like the speakers, because you have no real reference. At some point, when a sufficient library has been built of of competitive products, then one could come to some concrete conclusions, but not initially.
From my point of view individual HRTFs in the sound reproduction chain are an absolute must. With Beyerdynamics Headzone I experience some kind of externalization effect but not a realistic and stable localization. With some binaural recordings it's slightly better but there's always heavy front–back reversal. As the Headzone uses head tracking, the only possible explanation left is the use of non-individual HRTF. I'll find out when my order of the Smyth Realiser ships.
Best, Markus
Best, Markus
ZilchLab said:
Would it be much trouble to do a couple off axis FR shots of that mid too? Say 20º and 40º?
Its an interesting mid solution. One way, possibly, to mate a standard pro 15" to a 1" CD using available waveguides.
soongsc said:
??? "whether there is SOME POTENTIAL or not" ???
You must be kidding!
No clue what you are after, but if this baby performs half as good in your app as it does in my dipole horn - form measuements *and* first auditioning - you are king, man!
Michael
It's hard to describe what one is after in words, and probably the target shifts once one has reached one level. I guess one always wonder how much better it get get.
HRTF gets really tricky since we can only design for one point set of head. I have listened to a system at ITSEC many years ago, the idea seemed sound, but the sensation was not there. Don't know whether they were just using inadequate headphones or not.
gedlee said:
Praxis will do binaural convolution of a WAV file with room impulse responses. It's $1K or so for the full version but the free demo version will do all the post-processing if you get the impulse responses elsewhere.
http://www.libinst.com/praxis_downloads.htm
http://www.libinst.com/Room Sound Convolution.pdf
Very interesting... CARA also does the room simulation thing, but I don't know if it's for headphones. Might be. Of course it's for room sims, not speakers.
hello Earl,
You should see at the acoustic impedance curve of the human ear:
http://audilab.bmed.mcgill.ca/AudiLab/teach/circuit/meng24.gif
A usefil lumped model for the ear is described at:
http://audilab.bmed.mcgill.ca/AudiLab/teach/circuit/circuit.html
Best regards from Paris, France
Jean-Michel Le Cléac'h
You should see at the acoustic impedance curve of the human ear:
http://audilab.bmed.mcgill.ca/AudiLab/teach/circuit/meng24.gif
A usefil lumped model for the ear is described at:
http://audilab.bmed.mcgill.ca/AudiLab/teach/circuit/circuit.html
Best regards from Paris, France
Jean-Michel Le Cléac'h
gedlee said:
>> if you get the impulse responses elsewhere.
I once tried the rather old KEMAR HRTF data for a HRTF-based HP-crossfeed but it didn't work well.
Better we take real, individualized HRTFs...
... by placing WM61-A's (or the like) in the ear canal entrance, measure and correct the HP used for a perfect dirac response, then measure LS responses the same way (preferably the LS should be also corrected to best possible dirca response unless you want your real response with LS coloration etc, like Earl does), and finally acount for the ear canal resonance which might involve some guesswork and iterations.
For all of this you have to spend only zero $$$ for the SW, using tools like DRC
http://drc-fir.sourceforge.net/
- Klaus
I once tried the rather old KEMAR HRTF data for a HRTF-based HP-crossfeed but it didn't work well.
Better we take real, individualized HRTFs...
... by placing WM61-A's (or the like) in the ear canal entrance, measure and correct the HP used for a perfect dirac response, then measure LS responses the same way (preferably the LS should be also corrected to best possible dirca response unless you want your real response with LS coloration etc, like Earl does), and finally acount for the ear canal resonance which might involve some guesswork and iterations.
For all of this you have to spend only zero $$$ for the SW, using tools like DRC
http://drc-fir.sourceforge.net/
- Klaus
Jmmlc said:
Hi ...
I am very familiar with that data - having been head of Research for Knowles who made KEMAR and the Zwizlocki couplers, but I don;t see what they have to do with the problem.
Hello Earl,
It is interesting to see that Linkwitz optimisation (for his own ears) resulted in a electrical reponse to feed the earplugs which is the inverse or so of the human middle ear impedance.
See also:
fig. 6 in Bruel & Kjaer's document:
http://www.bksv.com/doc/bp0974.pdf
or fig 5 in
http://www.bksv.com/doc/bp0974.pdf
A more constant frequency response is probably better when the external ear is also used.
Best regards from Paris, France
Jean-Michel Le Cléac'h
It is interesting to see that Linkwitz optimisation (for his own ears) resulted in a electrical reponse to feed the earplugs which is the inverse or so of the human middle ear impedance.
See also:
fig. 6 in Bruel & Kjaer's document:
http://www.bksv.com/doc/bp0974.pdf
or fig 5 in
http://www.bksv.com/doc/bp0974.pdf
A more constant frequency response is probably better when the external ear is also used.
Best regards from Paris, France
Jean-Michel Le Cléac'h
gedlee said: