I am here on and off to question
But yes some of my terminology could do with being a bit more formal, so my apologies. As for being hard to follow, I should probably think a lot more before I type.Diffusion I use according to: having a discrete standing wave region and a diffuse sound field region being separated by the Schroeder frequency (in theory at least and a little more blurred in reality); and where the diffuse region is assumed to be that in which the two frequency responses at the ears are essentially decorrelated. So the use of diffusion is not correct, although not unrelated to diffusion at course surfaces at room boundaries, say, that can aid decorrelation at the ears.
The sensation of spaciousness decreases with increasing correlation between the ears, although I comment again that the spaciousness perceived in reproduction over loudspeakers is somewhat different to that in, say, a concert hall listening to a live performance. And it is for this reason that I originally brought up the effects of conventional stereo - and in particular the low frequency lateral energy attenuation below the audible comb filter notches (rather than the notches themselves).
This lack of low frequency energy can be compensated, so obviating any need for a more-live-than-desirable listening room acoustic. With such compensation, then high DI loudspeakers can possibly be employed to also obviate many of the early listening room issues that effect imaging, without comprising those more associated with late energy. But I do massively support the idea of analysing time-frequency responses in a manner that replicates our perceptive capabilities. Heisenberg anyone?
BTW I tried angling full range drivers at the room corners and they sounded noticeably skewed (off axis). I assume this is because of the changes in directivity.
How does the mind tell whether a driver is off axis, I assume its the narrowing fr
This paragraph makes no sense whatsoever, including the Heisenberg reference.
How does the mind tell whether a driver is off axis, I assume its the narrowing fr
This paragraph makes no sense whatsoever, including the Heisenberg reference.
Last edited:
My apologies, I will try again...
With the low frequency lateral information in a stereo recording restored to what it should be (assuming a near-coincident recording), stereo reproduction can give ample spaciousness without having a need for an overly-lively listening environment. One is then free to adopt high DI speakers to optimize imaging that may be compromised by early reflections, without simultaneously compromising the lateral energy that the room is having to sustain for a sufficient sensation of "spaciousness".
Heisenberg's Uncertainty Principle underpins analysis in the time-frequency plane and, for want of a better way to explain the trade-off, it determines our capability to better resolve either temporal information or frequency information, but not both simultaneously.
You see, this is not the acoustical definition. A diffuse sound field is one in which all angles of sound incidence on a point are equally probable.I am here on and off to question
Diffusion I use according to: having a discrete standing wave region and a diffuse sound field region being separated by the Schroeder frequency (in theory at least and a little more blurred in reality); and where the diffuse region is assumed to be that in which the two frequency responses at the ears are essentially decorrelated. So the use of diffusion is not correct, although not unrelated to diffusion at course surfaces at room boundaries, say, that can aid decorrelation at the ears.
What you are describing is the "reverberant field", that just beyond the direct field. The reverberant field can be more or less diffuse depending on the room characteristics.
I would agree with your definition of spaciousness, but the later effect is not one that I have found to be, for music made in a studio, significant. That's why I asked Mitch's opinion on that topic.
I have no real problem with what you are saying here, but I see that others may not read it like I do. It's not clear.
Last edited:
Yes, I saw a thread where you were discussing the maths of wave propagation........
I'll probably stick to attempting to understand Toole, more relevant to me anyway?
Let me see if I can clear up what's been said here.
"This lack of low frequency energy can be compensated, " - OK, no problem.
"so obviating any need for a more-live-than-desirable listening room acoustic. " - Well here you lost me.
" With such compensation, " - OK.
"then high DI loudspeakers can possibly be employed to also obviate many of the early listening room issues that effect imaging, without comprising those more associated with late energy."
- Which is exactly what I have been saying for a long time.
"I do massively support the idea of analysing time-frequency responses in a manner that replicates our perceptive capabilities." - That's good
"Heisenberg anyone?" He doesn't enter into this problem.
Last edited:
I have qualified my observations many times by stating near-coincident microphones and real acoustic events. With music made in a studio, what comes out is anybody's guess. I have heard vocals that "hang in the air" in front of me, but I have no idea what was supposed to be heard. On someone else's speakers and/or in someone else's room, something quite different could be generated.
This is why I'm reading what Toole has to say. There are many aspects to the reflections he talks about, things that would be very difficult for most of us to work out for ourselves. Much has to do with finding a balance and one we personally prefer. Perhaps the simplest way to get an initial idea, is to use our "normal living room environment" and move ourselves and our speakers around in an attempt to experience the variations and find our preference. Then if one finds the actual set up to be impractical, deduce what room treatments might help. Just an idea.
Just to the throw some more wood on the fire - one thing I rarely see mentioned here is the change of tonal balance with the change of the amount of reflected sound. It's well known in cinema sound that the less reverberate the room, the more high frequency is needed for a good tonal balance.
In the world of spaces much bigger than domestic rooms, the "house curve" changes with the size of the room. Larger rooms with more and longer reverberation need less high frequency energy to sound properly balanced than do smaller, drier rooms. This is manifested in the SMPTE X-curve established in the 1970s. Top end roll-off should vary depending on the size and reverberance of the room.
Most of us are familiar with the classic B&K domestic listening room curve that is a roll-off of ~1dB/octave from 160Hz-20kHz. Or the more recent Harmon curve that is usually stated as ~1dB/octave roll-off from 20Hz-20kHz. Those two curves were found to sound balanced in good or critical listening rooms of domestic size. The SPMTE X-curve goes further, specifying a top end roll-off that is room size dependent.
It's argued that the reason for these curves is an artifact of steady state measurements that accumulate late reflected sound along with the direct sound, skewing the measurement depending on the mix of direct vs reflected energy. That's certainly part of it, but some people (me included) argue that the ear also hears in this way. It's been my experience that drier rooms need more treble than reverberatent rooms to sound balanced. Engineers who are used to mixing in small spaces with near field monitors will mix overly bright in large rooms.
Why does that matter here? Because the ratio of direct vs reflected sound can change the subjective tonal balance. In the extreme; a highly directive system in a heavily treated room will not sound as bright as a less directive system in a live space. They need a different EQ of the direct sound. Of course there is almost infinite variation directivity and room reflection. How we hear those differences can lead to arguments and frustration over what are the best practices.
In the world of spaces much bigger than domestic rooms, the "house curve" changes with the size of the room. Larger rooms with more and longer reverberation need less high frequency energy to sound properly balanced than do smaller, drier rooms. This is manifested in the SMPTE X-curve established in the 1970s. Top end roll-off should vary depending on the size and reverberance of the room.
Most of us are familiar with the classic B&K domestic listening room curve that is a roll-off of ~1dB/octave from 160Hz-20kHz. Or the more recent Harmon curve that is usually stated as ~1dB/octave roll-off from 20Hz-20kHz. Those two curves were found to sound balanced in good or critical listening rooms of domestic size. The SPMTE X-curve goes further, specifying a top end roll-off that is room size dependent.
It's argued that the reason for these curves is an artifact of steady state measurements that accumulate late reflected sound along with the direct sound, skewing the measurement depending on the mix of direct vs reflected energy. That's certainly part of it, but some people (me included) argue that the ear also hears in this way. It's been my experience that drier rooms need more treble than reverberatent rooms to sound balanced. Engineers who are used to mixing in small spaces with near field monitors will mix overly bright in large rooms.
Why does that matter here? Because the ratio of direct vs reflected sound can change the subjective tonal balance. In the extreme; a highly directive system in a heavily treated room will not sound as bright as a less directive system in a live space. They need a different EQ of the direct sound. Of course there is almost infinite variation directivity and room reflection. How we hear those differences can lead to arguments and frustration over what are the best practices.
Then we are clear that we have drastically different expectations. Yours may best fit coincident micing in a real venue, while mine more studio work. I'm fine with that as long as we are all clear on where we are coming from so readers can weigh the objectives for themselves.
I bought Toole’s book when it came out in 2008 and have read it from beginning to end several times. It is a good compilation in easy to read language of other peoples reseach mostly. Good I would say but not that many answers to ones questions if one wants to plan a listening room including choice of speakers / absorbers / diffusers /deflectors and their necessary sizes or material properties. I haven’t read Toole's 2nd or 3rd editions, there might be more info in those (?).
A book I find better in that case would be ”Acoustics of Small Rooms” by Mendel Kleiner and Jiri Tichy: https://www.amazon.com/Mendel-Kleiner/e/B003NP9OSI/ref=ntt_dp_epwbk_0 Reviews: https://www.amazon.com/gp/product/B00L2EBN7C/ref=dbs_a_def_rwt_hsch_vapi_taft_p1_i1#customerReviews Most research and findings are from large rooms / auditoriums / concert halls which in many parts are not applicable in small rooms, so it was nice to come across this one. The focus is on how to get an accurate reproduction of stereo recordings in small rooms. I find Toole’s pros and cons advises with early reflections are more into enjoyable reproduction in stereo and / or multichannel sound. (Early reflections, in my personal view, gives a false soundstage width which some find fine, others not so fine.) To me it seems Dr Geddes and Kleiner share many opinions.
A book I find better in that case would be ”Acoustics of Small Rooms” by Mendel Kleiner and Jiri Tichy: https://www.amazon.com/Mendel-Kleiner/e/B003NP9OSI/ref=ntt_dp_epwbk_0 Reviews: https://www.amazon.com/gp/product/B00L2EBN7C/ref=dbs_a_def_rwt_hsch_vapi_taft_p1_i1#customerReviews Most research and findings are from large rooms / auditoriums / concert halls which in many parts are not applicable in small rooms, so it was nice to come across this one. The focus is on how to get an accurate reproduction of stereo recordings in small rooms. I find Toole’s pros and cons advises with early reflections are more into enjoyable reproduction in stereo and / or multichannel sound. (Early reflections, in my personal view, gives a false soundstage width which some find fine, others not so fine.) To me it seems Dr Geddes and Kleiner share many opinions.
Last edited:
I know that Tichy and I do - he was my thesis advisor. Somehow I had not heard of this book! I will have to get it as Mendel was a good friend as well.
That sounds like good advice, I do think most people learn more from experimenting and trying things than anything else.
Its my understanding that all things equal direct sound will sound louder than
reverberant because its hits you all at once, so its more intense. This is why in typical speakers when the woofer begins to beam it can enhance the mids and often something like a bbc dip can help.
Many British speakers seem to use techniques like the dip to soften the sound.
Also wondering about using off axis speakers that beam into a certain direction to give spaciousness and more late reflection in a small room?
Last edited:
Unfortunately, bias in the human mind makes that approach almost useless. I have seen it shown to be invalid in so many cases that I simply rule out this kind of "experimentation".
It seems logical to me. As a fan of classical music and multichannel recordings, why would one have qualms about doing anything to enhance the listening experience with recordings that are neither?
Last edited:
You talk about things like spaciousness and imaging and that's all perceptual and has never even been shown to exist.
It's whatever your preference is the point... The industry research shows peoples preference is for smooth on and off axis frequency response.
In the industry guideline I referenced earlier, a spec range for directivity index is given. The JBL's I listen begin to have constant directivity starting around 400 Hz. I have reviewed the Kii THREE and Dutch and Dutch 8c which have constant directvity starting around 100 Hz using cardioid tricks.
The graphic above shows the measured directivity over frequency for a double 15” and 90 x 50 CD horn, which are the JBL 4722’s (what I listen to). The 15” woofer and 120 x 100 CD are the JBL M2’s and the Domestic cone/dome system is the Revel Salon 2. The KEF LS50 is a cone and dome system and likely similar to the Salon 2 from a directivity perspective.
For harsh music, which unfortunately has a lot to do with the loudness war with extra midrange distortion due to too much dynamic range compression, a narrower directivity speaker may sound less harsh than a wider one. This is because the wider directivity speaker is adding more room sound at those same midrange frequencies and just accentuate the harshness (assuming both speakers have similar frequency responses). You can hear the comparison on the binaural recordings I referenced earlier between the JBL 4722 (very narrow directvity) and the KEF LS50 with much wider directivity.
Just to be clear, there is a lot of talk about high vs low directivity.
If you are in a space with terrible resonances like a big pipe then one wants to absorb the reflected sound or increase directivity.
However assuming the room isn't all that bad then there are other factors that might make one seek more directivity or less.
Last edited:
The "exactly" seems to be the crucial point, because in most cases we don´t know either what that is.
Obviously the only way "to hear exactly what´s on the released recording" is to listen to the said recording in the _same_ environment that was used during the production.
Everything else is an approximation - maybe based on a some assumption about linearity for example - that might be (in a statistical sense) be often correct overall (means if using a large number of different recordings from different production facilities, but can be quite often wrong for some recordings.
Even under the same conditions it is not ensured that it works for you the same way it works for the people who did the productions, as the usual reproduction via a restricted number of discrete channels relies strongly on the listener´s ability to fill in (based on the given cues) what in reality is missing in the recording.
One exception from the rule can be binaural recording (and reproduction with headphones/in ears) as it realy tries to recreate the original soundfields.
Most of what I listen to is created in a studio environment and expectations or such divisions are not my point.
A near-coincident miked recording can deliver amazing image depth and width. I do not think that should raise any contention. But, when replayed in an anechoic environment, the anomalies of conventional stereo reproduction become evident. It is clear that the listening room acoustics act in a mannner to obviate some of these effects. There are measures one can take to compensate for these deleterious effects: I have mentioned treating the M and S channels separately, narrow-angle reproduction and shuffling. But each of these is "off-topic" and probably warrant threads of their own.
So aside from these topics, my point of discussion has been that the effects of the room in compensating for conventional stereo reproduction are common to whether you have a near-coincident acoustic recording, an acoustic recording using spaced mics or studio created material (that may exploit a variety of recording techniques). The reference to near-coincident recorded material is simply that it has a valid reference to identify what is occurring. Spaced mic arrays distort the reference irrecovarably, whilst studio created material just doesn't have one to begin with (for a listener sat at home, that is).
Early reflections are well-known to effect image perception (depth, width etc). Early reflections in a concert hall are well-known to create spaciousness, but the time delays mean that such delays could render them more likely late in a typical listening room. Stereo reproduction also confines lateral reflections to the frontal plane. In a well executed acoustic recording, significant amounts of this information can be retained, however.
By contrast, a studio created recording will likely have a number of different acoustic footprints and these may be "enhanced" by early reflection generators and reverb units. I have yet to hear a good anechoic recording. The notion that somehow a multitude of such tracks can be assembled to recreate a whole is not realistic. What is left is an "effect" that is subject to a number of different factors and the notion of a 'best' solution appears flawed.
In my previous post, you rightly took issue with my use of the word diffuse. I used the term to describe a perfect reverberant field devoid of regular temporal elements. My contention is that this reverberant energy is what makes conventional stereo reproduction possible. This is common to conventional stereo reproduction how ever it is recorded. You also rightly point out that removing temporal aberations in the room response would be a good idea - that is improving the diffuseness. I merely point out that fixing the inherent cause of the problem with conventional stereo might be a better first fix, and one that when implemented leaves (IMHO) the effects of poor room acoustics downgraded somewhat.
To reiterate... Stereo recording and reproduction has a reference point that enables the effects of any room treatement or loudspeaker directivity to be ascertained. Stereo reproduction of studio created material is compromised for this task - at least for the listener sat at home.
I think that probably does a much better job of explaining what I have been trying to say!
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.