At this point I don’t think it really matters what I say or what sources I provide..
Still, I’ll try again..
The Blauert paper Frank references:
The abstract (at the beginning) expressly references recognition and segregation of concurrent sound sources in acoustically adverse conditions.
Yes, that’s concurrent sound sources – NOT a source and a reflection.
Ex. Loudspeaker A replays a trumpet while loudspeaker B replays exactly the same trumpet slightly delayed – potentially while both are playing in a reflective environment. Basically an often encountered sound reinforcement problem with setup and delay.
What’s being examined is only the directional effect – effects of loudness, timbre coloration, or an added spatial attribute are NOT being discussed except for the purpose of noting their existence.
HOWEVER, while those attributes do exist (loudness, timbre coloration, added spatial property) – it’s not being observed as a reflection. Moreover, timbre coloration and added spatial properties are the result of inter-channel combing in this example. This is in 9.1.1 of Toole’s book where Toole specifically references Clark’s work and that the delay of the contra-lateral loudspeaker sources had a “greatly degrading effect”.
Moreover, to do the “breaking” - the delayed direct sound is contra-lateral. (..see the left side of the picture in fig. 1). It’s basically a worst case scenario designed to push the precedence effect to its limits and beyond – to “snap” the directional behavior to two identifiable sources instead of one. It also happens to be a worst case scenario with respect to inter-channel combing effects.
-so no, I don’t really think this paper is “right on the money”.
Earl, I honestly don’t think you are that “far off” from Toole with respect to very early reflections. I mean – every ms only represents about a foot of distance. From what I’ve seen with most of his experiments the loudspeakers were always at least that far away from walls – even the nearest wall. I’ve also repeatedly stated that there is basically a 3 foot minimum to avoid most detrimental effects (..you can still have a channel balance problem of course if one loudspeaker is 3 feet away from its near side-wall and the other is like 5 feet from its near side-wall.)
While Toole does “extrapolate” in his book – I’ve not found an instance where he hasn’t provided a caveat. (..which isn’t to say that it doesn’t occur, just that I haven’t found any – and I wasn’t really looking for any.) For the most part he is using core material from the research of others, and “dips” into his (and Olive’s + Toole's) research for smaller percentage (if any) depending on the topic.
Nor do I think Blauert and Greisinger are really in disagreement with Toole.
Nothing in the paper Frank provides is in disagreement with what’s stated in Toole’s book that I can find, and in fact 11.1 basically reiterates the quote that Frank referenced. In fact, Toole specifically uses core material from Blauert (..with cites) and references Greisinger.
As far as sources – I’m not “stuck” on Toole because of Toole. Rather I’m referencing his book because it is in fact a reference – referencing the work of OTHERS and material that I’ve read from OTHERS.
Moreover my opinion isn’t at all biased toward Toole’s (or Olive’s) research, I personally like Greisinger’s research better (even though the lion’s share is specifically concert-hall in nature). Hell, I was reading Ando and Baron long before I even heard of Toole (..what can I say, my library skills weren’t great in undergrad – and it was a hobby-interest
). Additionally, most of my opinion isn’t formulated by reading – but rather through experiences. So yes, I look at more sources than Toole – and I look to Toole’s book because it’s based on far more sources than Toole himself.As to what Blauert agrees on?
Please ask him in the correct context (his email is on the paper) – because the paper really is not representative to what we are discussing.
I’d specifically like to know his “take” on the timing, amplitude level (relative to the source), and expected similarity/dissimilarity to the source that reflections will have with regard to listener significance with stereo reproduction.
My guess is (based on Toole’s cite of his work – 11.1), that very near reflections are so similar (..for most loudspeakers – though perhaps not a loudspeaker like yours with its enhanced directivity well off-axis at higher freq.s), that there will be very little timbre coloration (if any) – that it just “sums” to source, and likely little in the way of enhanced spatial property beyond the minor shift in angle for pressure due to the enhanced pressure provided by the very near wall. Again though, guess..
Last edited:
The cite is:
1. Biased to a control room setting
2. Stale (or old)
3. Is not written by "reference" author(s) with respect to psychoacoustics
..So I respectfully disagree with the "authority" you have provided.
I have talked with Blauert on similar topics and one of my key principles comes directly from him. Namely: That very early reflections ill be a tradeoff between enhanced spaciousness and improved imaging.
Take what you want from this, but to me it implies that very early reflections degrade imaging.
I have no questions for Prof. Blauert, I think that I understand the issues. If you have questions then you write him.
I do think that you misunderstand that in Blauert's paper that the second speaker can be a surrogate of a reflection. That's the way it is typically done.
As I have said before, I don't think that there is a high level of agreement among all parties on these issues. It remains a wide open field for investigation. Unfortunately there is no interest in doing studies of loudspeakers anywhere.
Last edited:
No, I understand that it's designed as a "simulated" reflection - but that the simulated reflection is not a reflection (or group of reflections) and is in fact an added source.
I also understand that with respect to the precedence effect and source detection that the result is virtually identical for real reflections - provided the "lag" speaker is properly limited in spl corresponding with the delay and relative to the "lead" speaker. (..and his "figure 1" specifically has the precedence effect in-action for broad-band signals starting at just 1 ms.)
It's NOT however a particularly good simulation of real reflections in other aspects - like timbre coloration and spatial properties, because it's (the "lag" speaker) an actual sound source (.."concurrent sound source").
Again, look to the Clark material in 9.1.1.
Last edited:
It won't. But above 12Khz the difference is so minimal as to no big deal. String also gain warmth for the higher harmonics.
I could probably live with an 8K limit on a lot of music, but jazz with its heavy use of the ride cymbal and violins in classical would lose a lot of life. The question isn't always "How high can I hear?" but rather "What frequency limit makes a noticeable difference."
I'm very glad can't hear that high. There must be a lot of nasty environmental sounds up there. Maybe you'd get used to it.
Yeah, you'd think we are making this sh!t up as we go along. Of course so must Toole - he must have been making sh!t up all along..
Monaural emphasis and depth of field.. Bah, what other fool my have said such a ludicrous thing?
Last edited:
Of course it could be "Data confuses me, what I hear is all I have." When a person spends long enough chasing data and finding no reasonable correlation with what they hear... I don't need to say, frustrating at the least.
I've always believed that data was more reliable than hearing. There was some time where I thought that data alone was not enough, that much could be true. An understanding of the hearing mechanism is needed to make sense of the data. This information may or may not be complete, for sure it seems to be scattered around. A person certainly wouldn't stumble across it all by accident.
Last edited:
Funny, I just reread both and I don't see much similarity at all. We believe what we want to believe I suppose (confirmation bias?).
Allen - if you are saying that one needs to actually understand the data in order to see a correlation then I strongly agree. To me (and Toole/Olive) the correlation is extremely strong. But of course getting and interpreting data is a LOT of work. Listening is far easier. If only listening weren't so biased then life would be so much simpler.
+ several.
It's too late here(
Somewhere
I have a paper on absolute ear interarrival times....Meanwhile, I have more pressing matters...
* yes we also use spectral content.
Scott - it appears to me that what we have here is a situation where there is no data that directly supports or denies what is being discussed. We each see the others references as not really applying or not a complete validation, and of course we see our own as a "close" confirmation of our positions.
I have wanted to do a subjective test of the audibility and effects of the near reflection versus the contra-lateral one. To my knowledge this has never been done and I once even suggested this to Toole. I have a hypothesis about how this would come out based on my own experiences and understand of hearing, but I do not have the data to validate this hypothesis (and none exists TMK.)
I also think that what we get wrapped up in around here is the difference of standpoint. For example, it is vastly different to make claims for "common" speakers in "typical" rooms versus a specific speaker design in highly optimized rooms. I always speak from the later because that is my area of interest. However, the former is most likely far more common.
I have wanted to do a subjective test of the audibility and effects of the near reflection versus the contra-lateral one. To my knowledge this has never been done and I once even suggested this to Toole. I have a hypothesis about how this would come out based on my own experiences and understand of hearing, but I do not have the data to validate this hypothesis (and none exists TMK.)
I also think that what we get wrapped up in around here is the difference of standpoint. For example, it is vastly different to make claims for "common" speakers in "typical" rooms versus a specific speaker design in highly optimized rooms. I always speak from the later because that is my area of interest. However, the former is most likely far more common.
The cite is:
1. Biased to a control room setting
2. Stale (or old)
3. Is not written by "reference" author(s) with respect to psychoacoustics
..So I respectfully disagree with the "authority" you have provided.
Oh yea
Well my reference is better than your reference
so there.
Anyway, in a control room setting u want accuracy of the sound.
Since when does old matter?
I did not provide the paper as an "authority" but rather to share my position.
I do not plan to argue that " my authority is better than your authority"
with you.
I have stated my disagreement and provided information as to why.
Not so bad these days.
But years ago...
In the tube days TV flybacks were really a terror ( 15.734KHz)
Earl,
I saw you choosed a heavy cone on the New Suma for the 15" mid-bass (80 hz -700 hz iirc) ! Is it necessary to have a heavy cone aginst energy storage or "non cone life" after the voice coil stops (between two impulse I mean)!
Is is just about choosing a huge BL and low Qes with such Mms above 100g for those FR mid-bass register (or did you also choose a heavy cone to fight the back rear wave in the sealed cabinet !)
Seems OT but in fact I ask myself questions about a musical snappy mid bass register like can be some ALtec in A VOTT load or W Onken ! But here a little like you need : a dedicated driver for mid-bass... It's new to me that such FR can have a heavy 15" instead a light 8" or 10" (I'm more worry of the come back of the cone between two impulses for a 15"... How to choose a good driver for 80-100 hz to 700-800 hz ???)
I saw you choosed a heavy cone on the New Suma for the 15" mid-bass (80 hz -700 hz iirc) ! Is it necessary to have a heavy cone aginst energy storage or "non cone life" after the voice coil stops (between two impulse I mean)!
Is is just about choosing a huge BL and low Qes with such Mms above 100g for those FR mid-bass register (or did you also choose a heavy cone to fight the back rear wave in the sealed cabinet !)
Seems OT but in fact I ask myself questions about a musical snappy mid bass register like can be some ALtec in A VOTT load or W Onken ! But here a little like you need : a dedicated driver for mid-bass... It's new to me that such FR can have a heavy 15" instead a light 8" or 10" (I'm more worry of the come back of the cone between two impulses for a 15"... How to choose a good driver for 80-100 hz to 700-800 hz ???)
Last edited:
Thanks Earl for the clear answer,
Looking at the datasheet of the 15" driver you choosed, I saw a peak in the far high end (present in many brands btw) and far above the XO you choosed, peak here is 2Khz in the manufacter datasheet).
No worries about this peak: because it's far enough above of the XO (XO around 700 Hz in your design; peak aound 2Khz in the B&C datasheet)
Looking at the datasheet of the 15" driver you choosed, I saw a peak in the far high end (present in many brands btw) and far above the XO you choosed, peak here is 2Khz in the manufacter datasheet).
No worries about this peak: because it's far enough above of the XO (XO around 700 Hz in your design; peak aound 2Khz in the B&C datasheet)
Last edited: