Hi!
I now have done some impulse measurements with a microphone. All the measurements shown here have been done with the microphone at the listening position.
The picture numbered 1 corresponds to the LS standing in a "conventional" way. The distance between the LS and the mic is 2,3m (7,5 feet). The impulse is not that great but still quite near what is “commonly” accepted for loudspeakers. The large peak at 0,8 ms correspond to a reflexion inside the box where the damping material is placed. Needle felt is placed in the middle of the depth of the speakers, which are 30 cm deep (12 inches). I will try other damping materials and maybe other positioning of it in the future.
The picture numbered 4 corresponds to the LS lying along the side walls (as shown here). The distance between the mic and the LS is 2,5 meter (8,2 feet) in this case. I had to enter this distance manually since the measurement program considered the large peak at around 5 ms as the direct sound! This peak corresponds to the roof reflexion, meaning that I was listening mostly to the roof… This might explain my impressions of spaciousness and 3D feelings.
After a while, the euphoria due to the different sound went down and I moved my speaker back to the conventional way where I have a much better tonal balance.
Regards,
Etienne
I now have done some impulse measurements with a microphone. All the measurements shown here have been done with the microphone at the listening position.
The picture numbered 1 corresponds to the LS standing in a "conventional" way. The distance between the LS and the mic is 2,3m (7,5 feet). The impulse is not that great but still quite near what is “commonly” accepted for loudspeakers. The large peak at 0,8 ms correspond to a reflexion inside the box where the damping material is placed. Needle felt is placed in the middle of the depth of the speakers, which are 30 cm deep (12 inches). I will try other damping materials and maybe other positioning of it in the future.
The picture numbered 4 corresponds to the LS lying along the side walls (as shown here). The distance between the mic and the LS is 2,5 meter (8,2 feet) in this case. I had to enter this distance manually since the measurement program considered the large peak at around 5 ms as the direct sound! This peak corresponds to the roof reflexion, meaning that I was listening mostly to the roof… This might explain my impressions of spaciousness and 3D feelings.
After a while, the euphoria due to the different sound went down and I moved my speaker back to the conventional way where I have a much better tonal balance.
Regards,
Etienne
Attachments
great!
the measurements clearly confirm that the loudspeaker cannot be heard in such a positioning as a distinct source of sound
which is a good thing – a key to realistic sound
and this is all the measurement really shows
Well, strictly speaking You were certainly not "listening to the ceiling"!
You were listening to the music instead of listening to the speakers
tell me - had you an impression that "sound comes out of the ceiling"?
reflection off the ceiling and all other reflections in such a positioning are not perceived as distinct sound sources, just like the speakers
it is just a sequence of reflections that occur in the room and that carry the left-right signal encoded in a stereo recording
much better that is…?
can You give more details?
tonality can be or in case of a fullranger positioned in such a way even should be corrected by means of equalization, passive or active
and what about other things You wrote about?
what about the bass? feeling of space? 3D? deeper silences between the notes? "sweet area"?
by returning to standard positioning without first trying frequency equalisation didn't You "throw out the baby with the bath water"?
best regards,
graaf
Not directly, it comes from higher than with the speakers in a conventional way... Unfortunately we are not very good at locating thing in the vertical plan!graaf said:
I will try!graaf said:
If you look at the attached chart, it shows the psychoacoustic response calculated from the impulse response I posted earlier. (I don't know about the real value of the psychoacoustic response but you might find more info here). curves 1 and 2 are from the LS positioned in a conventional way. Curves 3 and 4 are from the LS lying along the side walls with the correct delay. Curves 5 and 6 are from the LS lying along the side walls with the automatic delay.
You can see that with the LS lying the response drops regularly from 500 Hz upwards which I cannot correct with the tone controls of my current amplifier. At the end, the correction would have "linearized" the frequency response but it would have done nothing to the time domain.
I already explained the bass, the bass output is increased by the fact that the speakers are closer to the floor and side walls. Between a little deeper bass + missing highs and less deep bass with acceptable highs, I chose the second one!graaf said:
The feeling of space and 3D might come from the roof reflexion...
The deeper silences between the notes most probably come from the excitement or euphoria!
The sweet spot is now smaller but I have more depth.
As you can see, it is not all black or all white! Both positioning methods have advantages and drawbacks. The "conventional" positioning is to my opinion the best compromise for my loudspeaker-room combination.
Regards,
Etienne
Attachments
Etienne88 said:
well, OBJECTIVELY it is much better balanced than in case of conventional positioning
the black and green curves (1 and 2) are clearly the worst - thin and bright sound
the orange curve (6) although whith attenuated highs is still clearly the best balanced of all
not with tone controls but even simple passive equalization of the BSC (baffle step correction) type could do the jobEtienne88 said:
and what is wrong with time domain? Did You hear anything wrong which could be attributed to the time domain distortions? What was it?Etienne88 said:
I think that You are really suggested by the impulse response measurement which You misinterpret
with simple equalization You would have those desired highsEtienne88 said:
BTW Your "psychoacoustic response" of the conventional setup (curves 1-2) shows bright, excessive highs and higher midrange (typical for such a Fostex and usually corrected with apropriate RLC) plus lower midrange suck out
both things typical for typical audiophile sound and for me quite unacceptable, highly unrealistic, unnatural
but, well, this is all of course very subjective
well, "might"...Etienne88 said:
this doesn't mean that it is an artificial "ceiling effect"
the decisive test is whether different recordings sound more similar one to another with conventional positioning or with the unconventional
the question is whether the "3D" You experience is uniform and the same for different recordings (="it’s an added ceiling effect") or is it different for different recordings (="the system reveals 3D that is in the recording and that is lost in conventional stereo setup")
You mean the effect is gone? You cannot hear those silences anymore?Etienne88 said:
best regards!
graaf
Hi Graaf,
I choose to have a wide band speaker for simplicity reasons. I wanted first to avoid filter since I have almost no idea how they work, but I would like to learn! So unless you teach me or you serve me the appropriate filter on a silver plate I will live with my speakers as they are!
I did not hear anything wrong, no. But I trust better my microphone than my brain!
How do you interpret the impulse response from the lying speakers?
regarding the rest, I come back to your later...
Regards,
Etienne
I choose to have a wide band speaker for simplicity reasons. I wanted first to avoid filter since I have almost no idea how they work, but I would like to learn! So unless you teach me or you serve me the appropriate filter on a silver plate I will live with my speakers as they are!
I did not hear anything wrong, no. But I trust better my microphone than my brain!
How do you interpret the impulse response from the lying speakers?
regarding the rest, I come back to your later...
Regards,
Etienne
Etienne88 said:
I choose to have a wide band speaker for simplicity reasons. I wanted first to avoid filter since I have almost no idea how they work, but I would like to learn! So unless you teach me or you serve me the appropriate filter on a silver plate I will live with my speakers as they are!
I can understand that
Etienne88 said:
the problem is that microphone has VERY little in common with our hearing mechanism which is very selective, it integrates the data it receives, it is "active" (as opposed to "passive" reception of sound by a microphone)
read this to find out what is sound from the point of view of psychoacoustics: http://www.moultonlabs.com/more/nick_batzdorf_interview/
Etienne88 said:
well - I can see NATURAL SOUND that is the first transient and a volley of early (but nor too early) reflections
IMPORTANT is that we don't hear those reflections as separate sound events
for our hearing mechanism it is all ONE SOUND coming from the direction of the first wavefront
the strongest reflection comes over 5 ms after the first wavefront and as You mentioned it comes from the ceiling and it means that it comes in lateral plane from exactly the same direction as the first wavefront, no problem at all
the second in strength is delayed by over 10 ms, no problem at all
best,
graaf
Graaf,
I think you are mistaking sound production and sound REproduction when you speak about natural sound. Early reflections are more than desirable when playing a music instrument. Then, reflections during the 10 to 20 ms after arrival of the direct sound are to be avoided when reproducing sound… The early reflections should be on the record, that's why you don't want them to interfere or to be masked by the reflections from your room.
El’Ol,
Try Espacenet.com to find some patents from Stig Carlsson. Many are in Swedish, but the drawings are international!
I found some that might be of interest for you which are in English: 4,558,762 and 4,475,620.
Regards,
Etienne
I think you are mistaking sound production and sound REproduction when you speak about natural sound. Early reflections are more than desirable when playing a music instrument. Then, reflections during the 10 to 20 ms after arrival of the direct sound are to be avoided when reproducing sound… The early reflections should be on the record, that's why you don't want them to interfere or to be masked by the reflections from your room.
El’Ol,
Try Espacenet.com to find some patents from Stig Carlsson. Many are in Swedish, but the drawings are international!
I found some that might be of interest for you which are in English: 4,558,762 and 4,475,620.
Regards,
Etienne
Etienne88 said:
no, I am not mistaking anything
read the Batzdorf interview with Moulton carefully
I must admit that an analogy He makes with musical instruments is unfortunate as a little bit misleading but there is more than this analogy in His reasoning.
when early reflections are on the record it is practically impossible for them to be masked by room reflections, it is rather that the room reflections carry the information about reflections in the recording making them better audible, INCREASING (NOT DECREASING) the spatial resolution of the system
Read again our (and earlier mine) discussions with Dr Geddes who is clearly of different opinion than Moulton BUT from what He explains follows that it is all very complicated
one must distinguish between reflections <5 ms, reflections 5<10 ms, reflections 10<20 ms and finally reflections >20 ms
it all has been discussed recently on diyaudio between me and Dr Geddes
best,
graaf
graaf said:
post scriptum
in fact there is nothing in reality like "masking of early reflections by other early reflections"
it is all perceived as ONE SOUND coming from the direction indicated by the first wavefront (Haas effect, sound fusion)
only late reflections, echoes, can mask the reproduced sound
Graaf,
I am not at all convinced by the interview you linked to: he tells his point of view without backing it up with tangible data…
Reread post #95 and #97, I hope it will convince you!
El’Ol
If I remember well, Stig Carlsson said in his paper that the drivers should be pointed so that they cross behind the listener. It is somewhere in this paper... From that you can construct speakers adapted to your own listening room!
Regards,
Etienne
I am not at all convinced by the interview you linked to: he tells his point of view without backing it up with tangible data…
Reread post #95 and #97, I hope it will convince you!
El’Ol
If I remember well, Stig Carlsson said in his paper that the drivers should be pointed so that they cross behind the listener. It is somewhere in this paper... From that you can construct speakers adapted to your own listening room!
Regards,
Etienne
Etienne88 said:
view? data? what data? this is not "His view". This is simply RUDIMENTARY psychoacoustics - something that is not questioned outside of HiFi discussions where it is questioned because rudimentary knowledge is unfortunately ignored in audio industry - engineers are rather interested in what they were educated that is "signal processing in electronic circuits" etc. than in physiology of human hearing
and Moulton is very credible, HiFi loudspeaker business is marginal in His activities, it is His hobby, he has no "marketing agenda"
browse His website
anyway, let me repeat my question:
Did You hear anything wrong which could be attributed to the time domain distortions? What was it?
Of course You have already answered. Nothing was wrong.
The problem is that You simply believe more the apparently "bad look" of impulse response measurement which You misinterpret.
Etienne88 said:
ok. let's have a look at them:
gedlee said:
Did You hear any "image blur"? Did You hear any "coloration" (apart from attenuated highs)?
notice that Dr Geddes wrote "next 8-10 ms", not 20 ms
look at Your measurement - most of reflections occur after the first 10 ms
"image blur" or "localization distortion" can occur but under specific conditions, absent in the case You have measured, the level of reflected sound would have to be much higher for these to occur
and all those effects were studied by Haas and others specifically for the case of reflections in the lateral plane (side reflections) and not in the vertical plane
gedlee said:
Dr Geddes expected "a whole lot of VER (< 10 ms.) coming from all over the place"
well, Your measurement proves that He was WRONG
I can see only three VER <10 ms - NOT "a WHOLE LOT from all over the place" - the first is probably off the floor, the dominating second off the ceiling and the third off the opposite wall or the front wall (?)
just as I had expected
in my proposed positioning it would be even better – reflection off the floor would be practically neutralized and ceiling and front wall reflections would arrive later in time
it looks that Your measurement supports my claims
best,
graaf
Graaf,
About the interview, let me quote a sentence in the answer to the first question: "We have a long history of knowing that we prefer to hear musical instruments in reverberant spaces with reflections; there's no reason we shouldn't hear loudspeakers the same way."
I agree with the first part, I have much more problems with the second part. To me, this is a highly subjective position not supported by any scientific argument...
And it goes on like that for the rest of the interview!
Back to the impulse response.
I mentioned earlier that the program I used with the microphone placed the start of the impulse at the big peak after 4 ms on graph 4. I am wondering if the brain would do the same? I don't know the answer to that question, but:
1) if yes, then you will agree with me that the "impulse" right after 4 ms is followed by plenty of reflections. Which is not what we want!
2) if the brain "feels" the impulse as shown in graph 4, then I agree with you that there little reflections at first, but only for little less than 5 ms. This is not either what we want (which is: as little reflections as possible during the 8 to 10 ms)
My guess would be that the brain put the start of the impulse at the big peak right after 4 ms. It is only a guess, maybe somebody with more knowledge than me could confirm or infirm my guess...
Regards,
Etienne
About the interview, let me quote a sentence in the answer to the first question: "We have a long history of knowing that we prefer to hear musical instruments in reverberant spaces with reflections; there's no reason we shouldn't hear loudspeakers the same way."
I agree with the first part, I have much more problems with the second part. To me, this is a highly subjective position not supported by any scientific argument...
And it goes on like that for the rest of the interview!
Back to the impulse response.
I mentioned earlier that the program I used with the microphone placed the start of the impulse at the big peak after 4 ms on graph 4. I am wondering if the brain would do the same? I don't know the answer to that question, but:
1) if yes, then you will agree with me that the "impulse" right after 4 ms is followed by plenty of reflections. Which is not what we want!
2) if the brain "feels" the impulse as shown in graph 4, then I agree with you that there little reflections at first, but only for little less than 5 ms. This is not either what we want (which is: as little reflections as possible during the 8 to 10 ms)
My guess would be that the brain put the start of the impulse at the big peak right after 4 ms. It is only a guess, maybe somebody with more knowledge than me could confirm or infirm my guess...
Regards,
Etienne
Etienne88 said:
You are right as far as You refer to informal character of an interview, it is INFRORMAL - hence questionable analogies as rethoric means
but the matter of the argument is something else - it is pure science, check some handbooks on psychoacoustics
it is all simple and it works
look - I told You that such positioning would work, Dr Geddes told You that it "cannot work"
I based my proposal on scientific grounds, I knew that it would work - neither the soundstage nor imaging is corrupted/destroyed in such positioning because no physical mechanism for such corruption/destruction exists in nature
You have tried it yourself and now You KNOW that it works
amazingly You still question Moulton position (which is also mine) because "is is not supported by any scientific argument"
and at the same time You quote with approval Dr Geddes who gave no scientific argument for support of His position that "it cannot work"
and You know that He was wrong because You have tested it Yourself
Your attitude amazes me
Etienne88 said:
our hearing works in time domain - "signal processing" starts with the FIRST wavefront, not with the loudest
Etienne88 said:
"little less than 5 ms"? I can see 5 ms or perhaps a little more for the ceiling reflection
(and those time limits of 5 ms and so on were experimentally established for lateral reflections, not vertical)
Can You decribe what You see on the graph?
I am not sure I can read it correctly.
anyway, I ask again: Did you hear anything wrong in terms of spatial reproduction?
best,
graaf