I should also point out that this thread and my quote from Griesinger "above" (prior page) is essentially in relation to *source* location.. (even though it concerns reproduction).
This is different than reproduction of *sources* - aka "images" (..unless of course the source was "mixed" totally to either L OR R).
This is different than reproduction of *sources* - aka "images" (..unless of course the source was "mixed" totally to either L OR R).
Last edited:
That we can localize impulse sounds is fine, that's not the discussion. impulsive means that HFs are present and we localize on those. At LFs in a small room only a steady state signal could be said to be non-impulsive.
The LF standing waves in a small room are established in < 100 ms (a sound wave can move about a small room a few times in only a few periods of a 100 Hz tone.) And you can only average over many modes ("enough of them") when there are many modes present - this is hardly the case in a small room < 100 Hz.
I can buy this argument > 100 Hz, but not really below.
The LF standing waves in a small room are established in < 100 ms (a sound wave can move about a small room a few times in only a few periods of a 100 Hz tone.) And you can only average over many modes ("enough of them") when there are many modes present - this is hardly the case in a small room < 100 Hz.
I can buy this argument > 100 Hz, but not really below.
Email him.
As for "results" data.. he probably has that. Harmann's not shy about testing listening.
I doubt though that he can successfully prove that results and theory "match-up". (..it would require some thoughtful limits on the test.)
..and note: he changes his mind all the time (and has updated some of his publications to reflect it: showing the changes in his thinking) - it is after all not a particularly well-studied area.
Sadly, most thoughtfully limited tests/research seem to start and stop with headphones.
But why? How many speakers or subs are free of those artifacts? While it may be nice to know about purity, it's not a real world case. We have to deal with what is real.
BTW, thanks for the 1 story allowance and Merry Christmas to you!
Some of the reasons I build sealed subs in massively strong boxes and use steep crossovers. Of course, good drivers help too. I also think the steep crossover helps in not having as many integration issues with cancellations due to phase/difference. I imagine if the subs were sitting under the mains ( I don't have that luxury) it would allow higher crossovers of lower slopes. .
Earl brings up an intriguing idea of basically reverb isolated to the bass region. My movie room is very small, and it sounds small in spite of all my equalization and room treatments. Imaging is fantastic, but small. Not sure how one would do this without some DSP. Bucket brigade chips are long gone.
Email him.
As for "results" data.. he probably has that. Harmann's not shy about testing listening.
I actually have E-mailed him a couple of times, but I never get a response.
He has not worked at Harman in a very long time. Sean Olive is just about the only person left at Harman from "the good old days". Even Doug Button is gone. And then there was the "Apple Raid" that cleared out anyone that they could get their hands on. I don't know what Apple is doing with all the engineers, but hopefully something that pans out or all those guys will be back on the streets.
You could use a reverb unit for each sub, but that might get expensive.
I keep looking for a DSP unit that is really programmable in something other than assembly. Something like the Analog Devices platform, but more readily suited to DIY and not commercial product development like the AD stuff.
We did the simple studies using the AD VisualDSP platform (it was barely doable there), but that was in automotive and we only had one sub. It would be difficult to translate that system into something usable in a HT.
I think that's great. Shows he is learning. I hate to be held to something I may have said a long time ago. I used to believe that nonlinear distortion was a big deal - it had to be minimized. I don't believe that any more.
Last edited:
That is completely beside the point. One needs to know if its the HFs that are being localized or the artifacts or what it is. Otherwise we are just guessing and making statements about how we can localize LFs when in fact that is not what we are localizing. We're just doing poor experiments that don't tell us anything.
Cathedral
not the first octaves, but ears are decieved by eyes here : we look at towards Grand Orgue. And distance is important too as you have the place to experiment here... good toys, first disco rooms in the History... big bass in the stomach and no electricity ! If you don't see God you can hear him... BTW that's the goal here !
If ours rooms decieved us with bass localization, most of the times designers want not to waste the beginning of the voices with xo and phase trap ? Most of the times, drivers dictate their own choice (high frequencies of the bottom drivers are needed in stereo because often the XO are above 200 hz and 4 or 5 ways are fews?)
I experiment with one of my speakers without bass : XO is 125 hz LR3 (LR2 acoustical + LR1 electrical) : I can hear a difference at 3,5 meters at listening position between the two drivers. Just 2 cents experiment.
not the first octaves, but ears are decieved by eyes here : we look at towards Grand Orgue. And distance is important too as you have the place to experiment here... good toys, first disco rooms in the History... big bass in the stomach and no electricity ! If you don't see God you can hear him... BTW that's the goal here !
If ours rooms decieved us with bass localization, most of the times designers want not to waste the beginning of the voices with xo and phase trap ? Most of the times, drivers dictate their own choice (high frequencies of the bottom drivers are needed in stereo because often the XO are above 200 hz and 4 or 5 ways are fews?)
I experiment with one of my speakers without bass : XO is 125 hz LR3 (LR2 acoustical + LR1 electrical) : I can hear a difference at 3,5 meters at listening position between the two drivers. Just 2 cents experiment.
That's a bummer.
Apple's sitting on a pile of cash.. and can't seem to figure out where to spend it to see future profits - sort of a "spend and hope it sticks" formula. Patents, Talent, and Real-Estate seem to be on their shopping list. (..the last element is particularly fiscally troubling with cost/benefit, but no doubt their new "campus" has some good tax perks to go along with it.. still though, it reminds me of the titanic-that-was Compaq.)
Scott - I agree - Apple is in risky territory. If they can't generate something "new and great" every year then things will go south in a hurry. They certainly cannot compete in a price based environment. It is very difficult when you are on top. They seem to be doing alright thus far, but anything can happen.
Hmmmm...
My suspicion is that, even with the best drivers in the world, we're fighting equal loudness contours here: I'd expect that even relatively small amounts of THD at VLF (say, 20Hz) would be clearly audible, and aid in determining the source of the sound.
How can we test our low-frequency localisation if we're unable to produce pure low frequency tones?
Maybe I just need a better subwoofer (currently 2x JBL GTO1214 in a sealed cabinet, around a kilowatt between them)...
Chris
My suspicion is that, even with the best drivers in the world, we're fighting equal loudness contours here: I'd expect that even relatively small amounts of THD at VLF (say, 20Hz) would be clearly audible, and aid in determining the source of the sound.
How can we test our low-frequency localisation if we're unable to produce pure low frequency tones?
Maybe I just need a better subwoofer (currently 2x JBL GTO1214 in a sealed cabinet, around a kilowatt between them)...
Chris
No. That completely IS the point. While it may be fun to sit in ivory towers and chase fairies with a telescope - most of us live in the real world. Just how clean is a real subwoofer? Do its very real faults allow it to be located?
Once we move away from academic tests of pure tones over headphones and into the real world of dirty old speakers - the landscape changes. That is the point.
Of course can, but it essentially depends on the signal type. In a small room it requires signal having a modulated envelope before localisation takes place. Constant sinusoidal signals are harder or impossible to localise. Fortunately real music is no sinusoid but has pseudorandom like envelope.
This comes down to speaker's ability to reproduce the modulation of the signal. Believe or not there can be very interesting modulations at low frequencies in music
It's not all about 'steady state', not at all..
Confusion as to the source is a different issue. (reflections, etc. can cause confusion in a room)
There is a constant failure to differentiate between *localization* (detection of postion/location) and *omni directional radiation*.
Often the idea of omni-directional radiation is taken to mean that one can not at all detect the source of LF energy. It's not the same as saying that LF energy disperses equally in all directions.
The simple test is to take a sub and put it out with the speakers, then move it next to you... (change the levels suitably). I'd guess you notice that the sub is off to one side, or behind you.
There is a constant failure to differentiate between *localization* (detection of postion/location) and *omni directional radiation*.
Often the idea of omni-directional radiation is taken to mean that one can not at all detect the source of LF energy. It's not the same as saying that LF energy disperses equally in all directions.
The simple test is to take a sub and put it out with the speakers, then move it next to you... (change the levels suitably). I'd guess you notice that the sub is off to one side, or behind you.
Could you provide references?
Assuming that humans can localize low frequencies, there's still the question what to do when localization cues become heavily distorted by placing speakers in a room. Like Art already mentioned:
Of course can, but it essentially depends on the signal type. In a small room it requires signal having a modulated envelope before localisation takes place. Constant sinusoidal signals are harder or impossible to localise. Fortunately real music is no sinusoid but has pseudorandom like envelope.
This comes down to speaker's ability to reproduce the modulation of the signal. Believe or not there can be very interesting modulations at low frequencies in music
.
I'd think that bass instruments produce rather sinusoidal than percussive sounds.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.