Simon
Subjectively,if you adjust treble, then you have also adjusted the bass. That also could be a contribution to the effect.
Subjectively,if you adjust treble, then you have also adjusted the bass. That also could be a contribution to the effect.
My 2 inch cotton rope along five corners of my listening room made a huge difference in room ringing, in the midrange frequencies. In order to significantly affect lower midrange and bass, an "acoustic absorber" needs to have mass. Bass energy is more physical, as opposed to treble being like a beam of light from a flashlight. Bass loses its energy when it tries to move or vibrate something heavy. Corners seem to be the place in a typical listening room where absorbtion is by far the most effective. I think it's more about ringing than comb filter effects, but this kind of thing varies with frequency. A bean bag seems like a great idea for any two or three surface corner. It would be likely to be an effective acoustic arrestor at most frequencies. Although listening room reflections can enhance the sound (especially if there's a lot of them and they're very random in nature), I think any kind of ringing is pretty much always a bad thing. It elongates certain notes (spectra) thereby creating a bad psycho-acoustic effect.
Perception, in the words of Helmholtz, is an "unbewusste schluss" (please correct my German) meaning "unconscious inference" a creative process. It is based on a great jumble of sometimes conflicting cues. These cues are initially processed by more or less hard-wired physiology peripheral to the central processor.
Some of the great profusion of experiments mentioned in this thread kind of trip gaily among the peripheral mechanisms. Yet these special-effects of these mechanisms may have only slight influence on the final perception, even if they look kind of important in an isolated experiment.
As a useful exercise, think Helmholtz-like. Given a bunch of cues, what is the best posit perception for the brain to glom onto? Things get really hairy if you include the cues arising from the music room; so start with direct-to-eardrum cues even if delivered by speakers. What cues - if any - are around to help the person hear depth, height, the shape of the concept hall, etc. from a recording?*
While we can easily picture in our minds Disney Hall, can anybody suggest what cues are present to cause that perception or is it just something we are imagining without much basis on cues?
*Toole's ambience echos are a cue that helps the mind "think big hall" and we like that. Does it do any more than that?
Some of the great profusion of experiments mentioned in this thread kind of trip gaily among the peripheral mechanisms. Yet these special-effects of these mechanisms may have only slight influence on the final perception, even if they look kind of important in an isolated experiment.
As a useful exercise, think Helmholtz-like. Given a bunch of cues, what is the best posit perception for the brain to glom onto? Things get really hairy if you include the cues arising from the music room; so start with direct-to-eardrum cues even if delivered by speakers. What cues - if any - are around to help the person hear depth, height, the shape of the concept hall, etc. from a recording?*
While we can easily picture in our minds Disney Hall, can anybody suggest what cues are present to cause that perception or is it just something we are imagining without much basis on cues?
*Toole's ambience echos are a cue that helps the mind "think big hall" and we like that. Does it do any more than that?
Last edited:
Yes.
If, in your listening room these ambiance echoes are coming correctly from the sides (and perhaps from back of listener), they contribute greatly to perceived dynamic range. They particularly make the really loud stuff sound really loud.
This is quite noticeable with singers because of the range of their formants (2500 Hz and up) and the pulsating nature of their sound - this is extremely noticeable with big chorus moments.
In the hall the direct sound can only be maybe 50% of the total SPLs getting to the ears of someone in the balcony. Add in the rest and you're still not getting a lot of dBs - and it's what 75 -85 dB's? - but it sounds loud.
Somehow, our hearing adds up a lot of closely repeated echoes as loudness and we seem to be more sensitive to this as the energy increases.
Well, that's my "phenomenogical" account. I'm sure someone's done the science but I've not time right now to find it.
Edit: Also, I suspect if these "ambient echoes" are coming too much from the front, our hearing might pick up on them as some sort of linear distortion in loud passages leading to speaker localisation.
Last edited:
Only a minor correction, it is "unbewusster Schluss"
because Schluss=inference is considered male in german.
When i think about the commonly used practice of letting
a fairly directional speaker point to the opposite side wall,
so that the left speaker's reflections reach the listener via
right side wall and vice versa, i would think about the
following effects contributing to "spaciousness":
- early reflections highly correlated with the direct sound
of the speakers are mitigated.
- later reflections reaching the listeners from the right,
originating from the left speaker (and vice versa) are
more decorrelated from the speaker on the same side of
the room, as they originate from a different stereo channel,
which again may contribute to a more stable image and
possibly image 'depth'.
- toeing in -a seamless transition to "RLX" is possible -
may enable trading between earlier/louder in the
direct sound from either side. Given the speaker has a
dispersion pattern compatible with that (CD like), the
sweet spot (stereo listening zone) may be widened
considerably.
---
To me achieving interaural decorrelation in the
reflected/reverberant sound is a major key.
In my view that often may be more important than the
relative level of the reflections compared to the level of
the direct sound.
Most of our perception is "figure (image?) background"
recognition in my view (please correct my english).
As long as the background is random enough, we can
tolerate "a lot of background" without the image getting
seriously blurred.
Too much of correlation between direct and reflected
sound will be analysed as "smeared sources",
"near walls" or something like that, maybe just because
our perception is made/trained to search actively for
correlations rather than ignoring them.
A diffuse reverberant field (in the listening room) will
contribute just 'spaceousness' or -in the best case- does
not even 'superimpose' itself too much to the spatial cues
present on a well made recording.
Given this is true, i would like to suggest some minor
changes to the 'BOSE 901' ... 😉
Cheers
Last edited:
I would like to simulate the cross talk in my headphones and see if I "like" the sound... reflection free. Will it sound like it's coming from the stage in front of me or will it still sound somewhat "inside my head" or maybe even behind my head as did some of the VRM simulations I listened to.
I've never looked inside and it's a sealed bag, but I would say it's small styrofoam pellets.
I take your point that a reduction in apparent treble (especially low treble in the presence region) can cause an apparent increase in bass as well as an increase in warmth.
I'm quite familiar with this effect, and it was my first thought, however to clarify what was going on I did as mentioned take measurements that showed clear and significant changes in the bass region.
A 1dB increase in a broad hole and 1dB reduction in a peak in the bass is certainly an audible phenomenon. You're talking about a total improvement in flatness and balance of the bass of 2dB.
Readers here might be interested to see this information on sound absorbers. ASC ATTACK Wall for Control Room and Mix Environments
The writeup is certainly "traditional" in the sense that they strongly endorse the LEDE strategy.
That's what I'd always believed - that to have any significant effect at bass frequencies an absorber would need to have a lot of mass, like a wooden panel bass trap. I have tried small bean bags in the corners once before in a different room (just because I had some) - I put one in each corner behind the speakers, (it wasn't possible to put them in the far corners) and there was a small but just noticable change in the mid/upper bass region - mostly a reduction in cancellation notches in the upper bass, but nothing of huge significance.
It was a much larger room though, at 4m x 8m, with quite different acoustics.
One possible advantage of a bean bag over ropes as you describe is that it's much thicker, even this small bag is about 30cm in diameter, so that helps to get some of the absorbent material away from the corner so the pressure will be lower and the velocity higher than it is right at the corner. Even though the beans will be fairly transparent at bass frequencies by simply travelling through the thickness of the bag and back there should be significant frictional loss, so if thick enough maybe it could work better that I first thought.
More bean bags and experimentation needed 🙂
The RT30 was fairly flat from low midrange up to treble both before and after the bean bag was added, which suggests that it is probably quite effective across a reasonable frequency range, and the odd shape of it will probably diffuse what it doesn't absorb at higher frequencies.
I might take some RT30 measurements again tomorrow and do a more accurate comparison.
The bean bag experiment has got me thinking about broadband room treatments with a high WAF - something along the lines of a self standing wooden frame screen about 2m tall by 0.5m wide, designed to push into each corner, with some kind of aesthetically pleasing fabric tapestry or similar in the front which is visually opaque but acoustically largely transparent - and filling up the space behind it a stack of bean bags. The idea would be to absorb primarily midrange and treble in the corners, but any effect on the bass would be a nice bonus.
Yeah, ringing is always bad, especially slap echo type ringing. I also find that too much reverberation makes it uncomfortable to listen at high volumes - the sound field gets so cluttered and confused to the point where it can overwhelm you, even though it can sound ok at low volumes.Although listening room reflections can enhance the sound (especially if there's a lot of them and they're very random in nature), I think any kind of ringing is pretty much always a bad thing. It elongates certain notes (spectra) thereby creating a bad psycho-acoustic effect.
I think it's because you need to hear "space" between sounds in loud music to have "breathing room", and if the RT30 is high enough and/or you're sitting right near a live corner of the room the average volume is much higher (and with a lower crest factor) because the reverberation fills in any small time gaps between the instruments.
Possibly the same reason that heavily (dynamically) compressed music is fatiguing and uncomfortable in a similar way.
I tend to find that a better damped room (to a point) makes loud listening levels far more pleasant and enjoyable, not to mention more punchy and dynamic, and I appreciate the transient attack of a well damped room. I guess I'm just not a fan of a lot of room reverberation whether in the mid/treble, or the bass.
Last edited:
The cross-talk must conform to the natural cross-talk in time and frequency or our brains are not fooled. But if done correctly then the "inside the head" effect is surpressed. One of my current projects is to try and find out what the tradeoffs of this are. It seems from my early investigations that one must get pretty close to the real cross-talk before the brain gives in. Partial corrections just seem to do very little. There is a timbre shift and an increase in spaciousness, but its still inside of the head.
I would agree that decorrelation of the reverberant field is a good goal, but I would say that it is not so much the "levels" of the direct versus reflections that matters as much as it is the timing and direction. Certainly all three matter (timing, level and direction), but your statement excludes two of the most important ones.
A difuse reverberant field is the goal, agreed, and all reflections become difuse/decorrelated after a long enough delay and enough reflections. Its the very early ones that can be strongly correlated with the direct sound and hence cause the most problems with imaging versus spatiousness. If the reflections were uncorrelated with the direct sound then there wouldn't be any imaging problems at all, they would just be perceived as spatiousness, but the first few reflections are not uncorrelated.
Yes, very much so. And that's true in large and small rooms for both live and playback. As long as it doesn't get to loud, it can be OK - but turn it up a bit and it gets out of hand fast. Strange effect, but I'm sure it's documented and explained somewhere.
Amen to that. I've tried every headphone correction gimmick I can get my hands on. Still inside the head. Would love to hear something better.
Nippon-Gakki
From old AES paper: On the Room Acoustic Design of Listening Rooms (Paper Number:1524)
I don't have AES membership but this is the intro at the AES website:
Handbook for sound engineers - Google Books
Most interesting was this observation regarding the Nippon-Gakki experiments:
The google book also has this recommendation for "entertainment" rooms.
From old AES paper: On the Room Acoustic Design of Listening Rooms (Paper Number:1524)
I don't have AES membership but this is the intro at the AES website:
And I found this book with some comments: (see page 106)
Handbook for sound engineers - Google Books
Most interesting was this observation regarding the Nippon-Gakki experiments:
Obviously any system design will have some trade offs in spaciousness vs localization and the choices made will be based on subjective evaluation. So... is it possible that GedLee place higher preference on localization and SL on spaciousness?
The google book also has this recommendation for "entertainment" rooms.
You might be interested in trying Rockbox with the "crossfeed" plugin to quickly simulate and evaluate the tradeoffs of various configurations. I suspect that an effective crossfeed setup may require settings quite specific to the listener in order for it to achieve full potential.
Disclaimer: I haven't tried it yet but many rockbox users seem quite happy with crossfeed.
I'm going to try it to see it I can get the sound out from "inside my head" and perhaps it will help me evaluate the influence of "early reflections".
We did a large, beamed-ceiling living room with a whole wall covered in Tectum directly over the wall studs (no plaster or wallboard underneath the Tectum) and with heavy/acoustic fiberglass stuffed in all the spaces between the studs.
Few treatments eat sound down to low frequencies like Tectum, esp. used that way. And it can be acceptable to spouses.
Tectum
You might consider regular acoustic ceiling tiles.... to "give your ceiling some nice texture" and also for use in strategically placed "genuine canvas artwork in nice frames".
Perhaps better not to call them "acoustic treatment". 🙄
If you hook up a Bob Carver Holographic Generator in the signal path, you'll create the very "inter-aural crosstalk" that the generator attempts to undo with speakers. I would expect that to collapse the perceived soundfield in headphones, but who knows. 😎
One thing that was very clearly demonstrated at the Don Davis Sound System Engineering seminar I attended back in the '80's, was that acoustic damping material was vastly more effective when it was attached tightly to the surface or wall of the acoustic environment. When I nailed my cotton rope into the corners of my room, I angled the nail directly into the corner rather than a side wall. Putting absortive anything out from the corner is probably a waste of time and space. The frequency range of effectiveness will be minimal.
In the case of bass freqs., room boundaries (walls, heavy furniture, etc.) is the only thing that I would expect would make much difference. Putting large sealed cubes made of MDF in corners may help significantly. A large folded resonant cabinet (Helmhotz?) might be partially effective. Those are so obtrusive and non-frequency adjustable that I've never wanted to try one. Below 200 HZ I'm very in favor of using electronic EQ to pull down resonant areas, and am VERY hesitant to try and pull up any cancellation frequencies much at all, if at all, since they create several other problems (peaks elsewhere in the room, damaging excessive cone displacement, potential frequency modulation if crossed over very high due to excessive cone displacement).
I'd ignore any changes that measure only one or two dB. I've never known any test setup to be that consistent. You move the mic one inch and measurements change more than that.
In the case of bass freqs., room boundaries (walls, heavy furniture, etc.) is the only thing that I would expect would make much difference. Putting large sealed cubes made of MDF in corners may help significantly. A large folded resonant cabinet (Helmhotz?) might be partially effective. Those are so obtrusive and non-frequency adjustable that I've never wanted to try one. Below 200 HZ I'm very in favor of using electronic EQ to pull down resonant areas, and am VERY hesitant to try and pull up any cancellation frequencies much at all, if at all, since they create several other problems (peaks elsewhere in the room, damaging excessive cone displacement, potential frequency modulation if crossed over very high due to excessive cone displacement).
I'd ignore any changes that measure only one or two dB. I've never known any test setup to be that consistent. You move the mic one inch and measurements change more than that.
- Status
- Not open for further replies.