Which is related to the 100 million dollar question "how exactly does the indirect sound field influence our perception of timbre and spatial characteristics"?
I did not undestand/interprete his statement like that. What I extract from it (and also what Watkinson claims) is that reflections that have the same frequency response as the direct sound are not detrimental (or to a smal degree) to localization.
But when there are large differences in frequency response between them timbre might be quite strongly affected because a big part of that sound used for timbre recognition and loudness perception comes from reflected sound usually.
Regards
Charles
Just because Mr. Watkinson claims such a thing, the precedence effect and all the panning "laws" stop being valid?
And why do wide dispersion speakers generally sound bright? Maybe common recordings aren't made for such a radiation pattern but rely on colored reverberation?
I forgot "loudness". For example, most home theater owners don't listen at reference level. It is commonly perceived as being too loud. This is obviously caused by a difference in soundfields of movie and home theaters.
9.2.1: Persuasive evidence points to several beneficial and few
negative effects of early reflections.
A room with abundant reflections is not likely to ex-hibit audible evidence of comb filtering from any single
reflection.
Multiple reflections improve the audibility of timbral
cues from resonances in the structure of musical and
vocal sounds.
Early reflections improve speech intelligibility.
Need I go on?
This is not what I say.
Its the direct sound that determines the localisation.
Bad reflections can alter localisation.
Good reflections help with loudness and detail.
As for the precedence effect: Precedence effect - Wikipedia, the free encyclopedia
If the same sound signal arrives time delayed at a listener from different directions, only the direction of the first arriving sound signal is perceived. The delayed sound signals are localized from the direction of the first arriving signal (the first wave front).
If the reflections are not frequency dependant iaw have peaks and dips across the frequency band, these dips and peaks are perceived as timbre changes.
Simple test: Add reverb to a signal, then eq only the reverb.
If the reflections do not have enough similarities with the direct sound we can not distinguish between direct sound and reflections.
The frequency characteristics of the reflections are added to the direct sound, if the match it will result in a louder perceived sound.
Your statements are oversimplifications and not directly applicable to sound reproduction in small rooms. Toole is very careful in his wording when it comes to draw conclusions for stereo and multichannel sound reproduction from his tests. By the way, those tests were mostly done with single reflections under unechoic conditions.
When it comes to recommendations he advocates absorption and diffusion - how does this go together with your claims?
You've said that bad reflections are uncorrelated reflections and good reflections are more like the original. It should be obvious that correlated signals are more likely to get considered by our hearing as an additional directional cue than uncorrelated signals.
Dirk, you're NOT talking to someone that has heard the word "precedence" for the first time. My question goes deeper. We would need to look at delay, level, spectrum and angle. Nobody ever did this.
Toole's recomendations on absorption/diffusion in home listening rooms are quoted verbatim above in this thread in one of my previous posts - they go perfectly together with Dirk's "claims"
That's not a valid test because you're changing the signal that is radiated. Toole describes exactly that in chapter 9.
What's the perceptual result in your opinion?
I frequently hear about the spectrum of reflections matching the spectrum of the direct sound and this being some essential requirement.
Remember that the direct sound and reflection may be exact duplicates but can add with a highly colored combined response due to comb filtering. If the mutual arrival directions are close that will be the sonic effect.
I don't know of any studies that give much importance to the spectrum of reflections. The Lipshitz and Vanderkooy study did show that response holes in the later arrivals were innocuous but peaks might be detected.
David S.
Remember that the direct sound and reflection may be exact duplicates but can add with a highly colored combined response due to comb filtering. If the mutual arrival directions are close that will be the sonic effect.
I don't know of any studies that give much importance to the spectrum of reflections. The Lipshitz and Vanderkooy study did show that response holes in the later arrivals were innocuous but peaks might be detected.
David S.
opinion? it is a question of fact not of opinion - such incoherent reflections constitute additional spurious sound sources creating kind of "noise" detrimental to realistic sound reproduction
isn't a comb filter about a holes in response pattern? Then there is a selfcontradiction in what You say
BTW is disregarding Toole's latest book a new fashion here?
If the direct response is flat and the later reflections are not flat but have holes rather than peaks (such as power response dips at crossover frequencies of a multiway system) this was found to be relatively innocuous by L&V.
If the direct response and reflections come from similar directions then the combined response is key. In that case comb filtering has much more to do with the perceved response than the particular spectrum of the reflections.
David S.
If the direct response and reflections come from similar directions then the combined response is key. In that case comb filtering has much more to do with the perceved response than the particular spectrum of the reflections.
David S.
Below the schroeder frequency of a room there are very few reflections. Few reflections are bad and you need to 1 dampen them or 2 diffuse them.
The first thing to remember is that in order to survive in the wild we need to know where the danger is coming from. Our hearing does that by locating the sound source. And its only the direct sound that determines that. This is called the precedence effect.
Reflections in the wild have a strong relation to the direct sound and we can differentiate between the direct sound and these reflections, if we couldn't we would not be able to survive.
It is a test to see if frequency dependant reflections alter the perceived timbre.
And the test shows they do.
Spreading of the sound source. Iaw you can not pin point the instruments any more.
If you have few reflections you get comb filter effects. Example being the speaker on the mixing desk.
If you have lots of reflections you do not get comb filter effects. Adding reverb to a sound is an example of this.
How does that an answer my question?
The problem at low frequencies is not that there are "very few reflections" (which is simply wrong) but the lack of modes.
Sorry but I don't understand how that relates to correlated/uncorrelated reflections and phantom source localization.
Last edited:
Please re-read chapter 9 of Tooles book. You're altering the direct signal itself. This is different from acoustical summation of reflections.
But if the reflection would be highly correlated, then there would be no spreading?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.