"No, I don't think I'm confused about precedence effect, unless you can explain how you think I might be ?
Typical wall/floor/ceiling reflections to most listening positions easily exceed the minimum 2ms required for precedence effect to work, which is what allows the direct signal's frequency response to be the dominant factor in perceived spectral balance in the first place......"
Well DBM / Simon:
I'll put it nicely ... You are confused and it looks like you did not take my suggestion about looking up the definition. I'll try to define it in a nutshell.
I'll create a short duration stimulus of say two transients. These will be spearated in time by 2 to 10 ms (let's not quibble about exact numbers). The binuaral cue in the first will be set to be different than the binaural cue in the second.
The precedence effect is the phenomenon that you will hear a single event and the lateral location will be (mostly) determined by the binaural value contained in the first transient. IOW, you are relatively insenstive to the binaural value of the second transient when it arrives shortly after the first.
Importantly, nothing is mentioned about a relative insensitivity to other aspects of the second transient in terms of its posible spectral or timberal contribution.
I could go on since this is something near and dear to me, but I'll stop.
Hopefully this clears up some of your (and others) confusion.
Typical wall/floor/ceiling reflections to most listening positions easily exceed the minimum 2ms required for precedence effect to work, which is what allows the direct signal's frequency response to be the dominant factor in perceived spectral balance in the first place......"
Well DBM / Simon:
I'll put it nicely ... You are confused and it looks like you did not take my suggestion about looking up the definition. I'll try to define it in a nutshell.
I'll create a short duration stimulus of say two transients. These will be spearated in time by 2 to 10 ms (let's not quibble about exact numbers). The binuaral cue in the first will be set to be different than the binaural cue in the second.
The precedence effect is the phenomenon that you will hear a single event and the lateral location will be (mostly) determined by the binaural value contained in the first transient. IOW, you are relatively insenstive to the binaural value of the second transient when it arrives shortly after the first.
Importantly, nothing is mentioned about a relative insensitivity to other aspects of the second transient in terms of its posible spectral or timberal contribution.
I could go on since this is something near and dear to me, but I'll stop.
Hopefully this clears up some of your (and others) confusion.
Last edited:
Very interesting link, thanks.
I think, for domestic listening rooms, large stone step staircases are out!
David S.
I don't see how it could be floor reflections at these frequencies, especially since they would be primarily from the tweeter, the one that would have the floor bounce frequencies well into its stop-band.
I initially dismissed the idea that it could be diffraction, but other than the crossover (unlikely IMO), an increasing "fizz" sound would most closely correlate to the dimensions that affect diffraction, both from the baffle and from a driver adjacent to the tweeter. Moving farther away (very) slowly reduces the effective delta involved, so that small geometry change would cause me to still question that.
The one other possibility could be reflections, not diffraction, off of the top of a base cabinet in a setup with an M/T box on a woofer. That probably would affect a wide dispersion midrange more than the tweeter, due to proximity of the midrange unit.
What possibility exists for it being related to HRTF? Distance from the source, be it mono or stereo, is going to affect the response.
Dave
Last edited:
... and probably more important, its spatial contribution.
By the way, here's a good paper about the precedence effect.
26_Scales with repetition pitch
Very cool demonstration of pitch effect due to reflection. 3rd example is best. Uses 15ms to 0.95 ms echo delay.
David S.
Very cool demonstration of pitch effect due to reflection. 3rd example is best. Uses 15ms to 0.95 ms echo delay.
David S.
The "problem" with repetition pitch is that it has no practical meaning in sound reproduction because we normally don't walk around while listening nor do we listen to pink noise 🙂
You are never "moved" by the music? Our motion makes the pitch much more obvious since it calls attention to itself by moving. But even if we are stationary the pitch of reflections is real and is an inaccuracy that should be dealt with. Effects audible on pink noise are generally audible on music, you may have to search a little to hear them but once they are noticed they become everpresent.
Listen to the various examples, they are not subtle. By the way, the locomotive examples are very close to my recollection of the sound of the loudspeaker bounce. I notice both the fundamental pitch and a modfied treble character.
David S.
What you call "inaccuracy" is also part of the spatial signatur of a room. I'm not sure removing it will result in better sound reproduction.
Play white noise from a single speaker and move around in the room. Now play decorrelated white noise from two speakers and move around in the room. What do you perceive in each case?
There is another interesting phenoma that can be heard with a point source of white noise. The shape of the pinnea is such, that the same noise coming from above (listening to the source facing down) contains more high tones than the noise coming from below (listening to the source facing up). This might also explain some of the phenomena mentioned above.
vac
vac
Com'on now. Flutter echo and standing waves are part of the room signature. They aren't good things. I do want to remove all reflections that result in perceived response aberations.
Uncorrelated between the two channels? Not sure how that is different than the single speaker case. For sure if you send mono noise to two speakers you get comb filtering as you move across the center line between the two (as an interesting aside the comb filtering virtually goes away when you are directly between, on a line between, the two speakers. The angular separation is enough that the interference drops.)
Reflections in the "wrong" time range cause audible coloration. This ranges from baffle edge reflections for mid and hig frequencies, floor bounces at low to middle frequencies, to the standing waves at low frequencies from multiple wall reflections. Reducing them is part and parcel to increasing fidelity. If reflections come from the sides and are sufficiently late in time (20ms or so), they can increase a sense of space and be helpful. Earlier than that, and especially coming from the direction of the source is very undesirable.
David S.
I looked back a bit and it looks like we (same folks) all debated the audibility of reflections back in March.
http://www.diyaudio.com/forums/multi-way/172806-flat-not-correct-stereo-system-141.html
At that time I posted some simple simulations of pink noise with a single reflection, meant to roughly simulate what might come back from the back wall behind a dipole (don't need to debate that again). At full strength a reflection is very audible. At -5dB it is as audible and -5dB with filtering above 2k it is nearly as audible. It is true that binaural hearing and different arrival angles might make the reflection less audible. It is also true that it is more obvious due to turning it on and off repeatedly. Still, would you rather try to deal with this as an issue or spend all your money on expensive interconnects?
From March:
Here are 3 MP3 files. All are pink noise with a reflection added. The first is a delayed (11ms) reflection, inverted and added at full strength. The second is similar but the reflection is reduced 5 dB, approximately what the extra distance would reduce the reflection level (depends on your distance from the speakers, of course). The third is also with a -5dB reflection but with highs above 2k filtered out.
THESE ARE NOT ZIP FILES. CHANGE THE ZIP TO MP3 TO PLAY (had to cheat the system to get them to upload).
The reflection is turned on and off every 2 seconds (first 2 seconds without).
Regards,
David S.
http://www.diyaudio.com/forums/multi-way/172806-flat-not-correct-stereo-system-141.html
At that time I posted some simple simulations of pink noise with a single reflection, meant to roughly simulate what might come back from the back wall behind a dipole (don't need to debate that again). At full strength a reflection is very audible. At -5dB it is as audible and -5dB with filtering above 2k it is nearly as audible. It is true that binaural hearing and different arrival angles might make the reflection less audible. It is also true that it is more obvious due to turning it on and off repeatedly. Still, would you rather try to deal with this as an issue or spend all your money on expensive interconnects?
From March:
Here are 3 MP3 files. All are pink noise with a reflection added. The first is a delayed (11ms) reflection, inverted and added at full strength. The second is similar but the reflection is reduced 5 dB, approximately what the extra distance would reduce the reflection level (depends on your distance from the speakers, of course). The third is also with a -5dB reflection but with highs above 2k filtered out.
THESE ARE NOT ZIP FILES. CHANGE THE ZIP TO MP3 TO PLAY (had to cheat the system to get them to upload).
The reflection is turned on and off every 2 seconds (first 2 seconds without).
Regards,
David S.
What happened to acoustic reflectors with piston drivers? This turned into a repeat of a conversation we've had in a dozen different places over the last few years. Objectives of speakers in small rooms, Toole's book, Geddes designs, flat is not correct, omni vs dipole vs waveguide etc.
I was interested in the original topic of the thread. If nothing else lessons here could be useful to musicians who are using something like this: Weber Beam Blocker -- high frequency diffuser to try to make their 12" full range speakers more useable. They're stuck with the big fullrange speakers due to favourable contributions and colorations they add to the overall sound but the beaming is very unwelcome.
I was interested in the original topic of the thread. If nothing else lessons here could be useful to musicians who are using something like this: Weber Beam Blocker -- high frequency diffuser to try to make their 12" full range speakers more useable. They're stuck with the big fullrange speakers due to favourable contributions and colorations they add to the overall sound but the beaming is very unwelcome.
When you think this through then the anechoic chamber is the goal. Apart from not being practical, do we get the "best sound" there?
With uncorrelated noise from two speakers binaural decolorization is even stronger. Binaural decolorization is very real and is probably part of perceiving a sound source as "real". While combfilter effects are real and can be heard with special signals and under special circumstances (plug one of your ears and do the single speaker white noise test), this doesn't necessarily imply those effects would be bad and must be avoided at all costs.
20ms isn't achievable in probably 99% of all listening spaces.
"In small rooms, it is improbable for natural refl ections to initiate impressions of envelopment, but sensations of ASW (apparent source width), image broadening, and early spatial impression are very real and desirable."
Isn't a speaker concept that exploits the effects of "early spatial impression" worth exploring?
---
By the way, an interesting quote from Blauert: "There is no such thing as a "law of the first wavefront" for timbre. Timbre is determined by the incoming energy over up to 100 ms - like loudness."
I agree but we have to fill the gap until I get to measure a bigger cone.
Is the shape, size and distance just random or is there any math that could be applied?
🙂 but this IS the real topic of this thread 😉
it is the same topic again and again - objectives of speakers in small rooms
this IS the question, without answering it first the rest of discussion is pretty pointless, it is like straining out a gnat while swallowing a camel if I am allowed to use a biblical metaphor, actually it's rather hopelessly choking on a camel
I think there shouldn't be any need to quote Blauert, I think it should be allready obvious to all, there are many reasons for that and myself I have already posted hundreds posts about it on the forum
substantial parts of discussions in this thread look funny and sad at the same time
Hi markus, what this means regarding the narrow gating technique supposed to exclude the reflections from the measured Fr ? Sorry if I miss something, but a little help would be welcome.
It questions the usefulness of such a measuring technique for in-room measurements. For measuring the direct field, gating the impuls remains a poor man's anechoic chamber and is highly valuable.
- Status
- Not open for further replies.