yes, absolutely, thank You once again
reflections with delays of the order of 700 microseconds - exactly!
only very rarely we have wall reflections of the order of 700 microseconds, even in very small room, can we agree on that?
This is only true for natural sound sources. Stereo uses and needs to use delays of up to 1.5ms. See http://www.sengpielaudio.com/InterchannelLevelDifferenceTimeDifference1.pdf
Markus comment is completely out of context - we are talking about what reflections do not about how much delay is needed for ITD stereo to work
Watkinson makes some good pints in this article but his argument about localization errors due to a speaker's individual "footprint" doesn't make much sense to me. If the footprint of each speaker is the same then the transient might be inaccurate but the same error is present in each speaker, so there can't be any detrimental effect an localization.
thank You Markus
now if we take a look at p. 204 in Blauert's "Spatial Hearning":
Spatial hearing: the psychophysics ... - Google Ksi??ki
where we can find "about 1 ms" as a limit to summing localization "in most cases", we can ask again the question more forcefully:
is 25 cm in case of the setup that Markus linked to large enough to avoid "large summing localization reflections from the nearby walls"?
then Dr Geddes - how can You say that it is seldom the option ???
He means speaker enclosure diffraction and if You have any doubts whether speaker enclosure diffraction is detrimental to imaging then ask Dr Geddes
both experts agree on that, I can quote Dr Geddes' old post if necessary
Last edited:
Very early reflections up to as mush as 10 ms. can be detrimental. And don't quote me the "precidence" effect because it doesn't apply. This effect does NOT state that the image is not colored by the reflection only that its apparent location is that of the first arrival. I regard any reflection less than 10 ms as detrimental with a weighting of about 1/t(ms), then all lateral reflections greater than this time as highly positive. One must minimize < 10 ms while maximizing > 10 ms. This can only be accomplished with speakers that have a narrow directiivty.
so precisely speaking it is not a question of "summing localization" but of "coloring of the image by the reflection"?
Is this "coloring of the image" about localisation/imaging or about tonality?
Use of the word "coloring" suggests it is about tonality, am I right?
It's a question of "it all", it all matters. "Coloration" is a common term which means a change in "tone color", completely analogous to the spectral shifts in optics that cause an appearance of a color change. Different delay times, levels and directions can cause either or both a change in apparant image location and/or its color (tonality if you like). A clear demarcation of one to another does not exist IMO. The earliest tend towards image and the later (up to about 10 ms.) tend towards coloration.
Very true but often ignored.
...or with a room size that is >6x5m. Linkwitz's limit is 6ms which equals to a room size of 5x4m.
Last edited:
... if you have freedom for placement of the speakers.
Usual living rooms are multi purpose rooms and placement
may often be compromised, even though the room may
exceed say 5x6 meters or any other "magical limit".
Often a german living room e.g. has 22square meters.
That is quite exactly at the limits where "placing speakers freely enough"
is possible if i take the measures from your post.
A view into germany`s "most frequent" living room as constructed
by "Spiegel Online":
Deutschlands häufigstes Wohnzimmer - SPIEGEL ONLINE - Nachrichten
For an estimation of freedom in speaker placement please choose
menu item "Männer Ecke" (Men`s Corner) ... OK, kidding aside.
So i would say, some "narrower than wide" directivity won`t be
too bad in most cases.
Usual living rooms are multi purpose rooms and placement
may often be compromised, even though the room may
exceed say 5x6 meters or any other "magical limit".
Often a german living room e.g. has 22square meters.
That is quite exactly at the limits where "placing speakers freely enough"
is possible if i take the measures from your post.
A view into germany`s "most frequent" living room as constructed
by "Spiegel Online":
Deutschlands häufigstes Wohnzimmer - SPIEGEL ONLINE - Nachrichten
For an estimation of freedom in speaker placement please choose
menu item "Männer Ecke" (Men`s Corner) ... OK, kidding aside.
So i would say, some "narrower than wide" directivity won`t be
too bad in most cases.
Last edited:
Take a listen to the various simulations of reflections from the previous thread on acoustic reflectors.
homepageFAB
http://fabilsen.home.xs4all.nl/locanimation.mov
http://fabilsen.home.xs4all.nl/loc_RP.mov
http://www.santafevisions.com/csf/demos/audio/426_scales_repetition_pitch.htm
Link broken?
http://www.diyaudio.com/forums/mult...reflectors-piston-drivers-14.html#post2677631
Perceived pitch is a function of reflection delay. Delays significantly longer than 10ms still give a discernable pitch. The 4th link (it was working last week) demonstrated a 2 Octave musical scale.
Neat stuff.
David S.
Remember that the dimensions of the room may impact the delay time to the first reflections but many 2 and 3 bounce reflections will also arrive with significant level. The key is to reduce all of the early reflections to a level of minimal impact. Reflections of -20dB (or lower) for the first 20msec would be an admirable goal. Absorption, diffusion, speaker directivity and speaker aiming can all help reduce the reflection strengths.
The Bech papers go into this in great detail.
David S.
The Bech papers go into this in great detail.
David S.
David, I have some doubts with respect to the usefulness of those sound samples with a delayed copy of the original. How well do they translate to what one actually hears in a room? In particular I'm thinking about the samples that supposedly proved that dipoles degraded sound quality (backwave added to signal). Very similarly edited sound samples could be used to 'prove' that any reflection colors the sound, but in real rooms reflections are somehow perceived differently.
If you have a single reflection on the median plane without a lot of reflections around it (time wise), then I think they would be very relevant. Laterally arriving in the midst of other reflections, I'm not so sure. Certainly my experience is that strong reflections in the median plane are audible and have an associated pitch. I've cited other examples I've heard before: waves at the sea shore, picket fences at the tennis courts, rustling paper above a hard surface. Note that the links refer to audible real world phenomena: trains going past or handclaps in front of the steps of a monument, not just to electronic or anechoic simulations.
Are they perceived differently in a real room or have you just gotten used to them? Would they not be as dramatic if you could turn them on and off?
David S.
Last edited:
Heyser ETC (energy time curve). Might be good to try it with various bandwidths.
Note that most auditorium acoustics programs of the ray tracing type will create an impulse response that is just the sequence of calculated reflections. An ETC is the log magnitude plot of energy (pressure squared) so it can come from the room impulse response. The same simulators will typically give a binaural simulated output (a convolution of whatever you want to feed in). CATT Acoustics let you define the frequency dependent source polar curves, so you could try any potential speaker in your ideal room.
Anybody have the programming time?
David S.
but how can this "tendency towards image" be explained except for "summing localisation" which as we are told by Blauert and others does not apply in case of secondary signals delayed more then about 1 ms?
I can understand that reflections <1 ms can be detrimental to imaging, it is explained by "summing licalisation" but what about those >1 ms?
Is the precedence effect not real at all, even with regard to localisation?
I am talking about localisation distortions - smear, shifts etc. exclusively
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.