An externally hosted image should be here but it was not working when we last tested it.
That picture you show kind of reminds me of Pano's lava cave
He has shown measurements from within that cave and they do look interesting.
This tends to be the consensus and feels intuitively right.
But a very simple test can verify or disprove this: Add lots of reverb to a sound, see when the stereo image gets messy. I've done this and you need a LOT more reverb than original sound to mess up the stereo image. I can make some soundfiles if people want.
If on the other hand the reverb is not diffuse and frequency dependent, then only a bit of reverb will screw up the stereo image. This is the case in 99.9% of all studio setups. Think big mixing desk between the speakers and soundengineer/speakers on mixing desk causing comb filtering. So the consensus of to much reverb/RT60 is bad is understandable, but reality could be different/a lot more complex.
In short:
Comb filtering BAD
For the rest, the jury is still out. We need more blind listening tests.
Music is generally more forgiving, does it sound as good on a track where there's just one person talking?
In general we seem to be able to ignore (a lot of) room effects when listening inside that room. Once you make yourself aware of that room, for instance by recording the in room sound and playing that back trough headphones, or even worse; play it back on the same speakers you just recorded, it's way less easy to not hear the room.
Once you treat the (0 to 20 ms) reflections and have a left and right balanced first sound wave it gives you more of a "you are there" kind of experience. This is further enhanced with late, laterally arriving, reflections (somewhere between 15 and 25 ms). That last part actually gives you back the "life" in the music without disrupting imaging.
In my opinion we, as humans, use the combing patterns within the signal arriving at our ears to give the room dimensions. Take away your room specific signals (or comb patterns, by taking away the reflections) and you'll hear more room queues from within the recording (if present, fake or not).
We could always use more tests in this area, but you'd need a good, controlled environment to test something like this.
The most interesting DIY room I've seen (but not heard) that sort of follows the outline given in the first post has to be jim1961's room: https://www.gearslutz.com/board/studio-building-acoustics/817205-my-listening-room.html
The whole thread is filled with interesting measurements and discussion about this subject. This has to be one of the best passive results I've ever seen. A huge combination of absorption redirection and diffusion. Everything that hits the listening spot is carefully planned.
To bring it back to my earlier point: measure what you've got! As interested as I was in Jim's awesome efforts in room treatment and beyond I asked Jim to run the APL_TDA demo software.
See how clean that plot looks compared to my humble living room? He does have first reflections down by an actual ~15 to 20 dB. (I don't know the exact number as he plays with that in this room). You also see the Haas kicker, which is a result of a multitude of panels that arrange that reflection to come back from the lateral angles delayed about 24 ms.
As I said, I don't think it is easy at all, and it certainly isn't what my significant other would allow me to do in our living room. I still managed to improve my listening experience just by following this theory, be it in a more DSP kind of way. I couldn't have done it without room treatment, that's for sure.
Last edited:
Milkshake i do see your point but it depend from a lot of things in my view: reverb used (brands: a Lexicon and a TcElectronics do sound differents from an AMS or Alesis or... and there is difference between ref in lines, and algorythm, and if you tweak settings... You see the point!).
If you make some 'modern' reverb kind (a la TCelectronics: quasi no tail and a lot of ER) what you say is true to me, try this with with long tail reverbs (Lexicon type) and if non high quality (read non Lexicon!
) this will mess everything very quickly. Especially when used in stereo. It is the same using IR (convolution plug or hardware units).
This is why i have caution about that.
And for my own case, i can make my own soundfile to try...
I don't see the comb filtering induced by desks as reverb. But you are right here too, this is a plague of big analog desks/control surface (Icon) Even if some brands tryed to reduce the issues in theyr time (SSL in the 80's).
In control room low RT is more for giving a more 'insight' view of what you are doing, and make reproductibility between room easier.
I do agree, i've heard great results in room with RT60 as long as 1.5s... But it was not the whole story: speakers was pristine quality, room designed around the speakers...
Room+Loudspeakers seen as a whole.
Last edited:
It seems to me that the stereo image is more fragile with music, because of the multiple sound sources, than speech, one sound source. But I can make files for both if people want to.
A duh...
If you record sound with reflections, you'll hear the room. Its supposed to sound like that. And I know from experience that the stereo image is NOT screwed up by this.
As we are talking about reproduction in listening rooms, why not use a IR from a real modern living room.Milkshake i do see your point but it depend from a lot of things in my view: reverb used (brands: a Lexicon and a TcElectronics do sound differents from an AMS or Alesis or... and there is difference between ref in lines, and algorythm, and if you tweak settings... You see the point!).
If you make some 'modern' reverb kind (a la TCelectronics: quasi no tail and a lot of ER) what you say is true to me, try this with with long tail reverbs (Lexicon type) and if non high quality (read non Lexicon!
It is the same using IR (convolution plug or hardware units).
This is why i have caution about that.
And for my own case, i can make my own soundfile to try...
Yet you don't perceive the room as much listening to your speakers at your favourite spot. Listen at that spot to a recording of those speakers made in that spot (in other words, double the room effect) and after that go back, listening to a normal recording. It will make you more aware of the room.
It took me quite a while to revert back to normal listening behaviour
. Which is ignoring room related queues. And this is in a treated room. "Supposed to sound like that" wouldn't be my idea. I want my room to 'disappear/get out of the way' as much as possible.
Test it and know for sure. It's among the many things I have tried already. Now everything will depend on how and what you test, with a single voice I could more easily hear added sound effects. It will stand out more.
Last edited:
The Lava cave was a fascinating study in playback acoustics. No parallel surfaces, no soft surfaces, rough chaotic walls and ceiling, and great depth. I do not know if the lava was absorptive or reflective, I suppose a combination of both.
What was remarkable about the sound was its depth. First reflections came late or highly diffused. That gave the effect of huge depth to recordings that had it, but little or no depth in close miked recordings. That was true for stereo or mono recordings. Stereo sounded fuller and maybe a little deeper, but mono recordings done in a big space could sound very deep. Some recordings were a mix of flat and deep, for example Sinatra in front of an orchestra. The sound did not come out in front of the speakers, it was always just at or behind the plane of the speakers.
What the lava cave did not have, was envelopment. No sense of sound beside or around you. Nothing. It cried out for surround sound. Also for the stereo image, it was precise side to side and front to back, but rarely ever did it go past the speakers. It rarely (1 or 2 recording) spread wider than speaker positions. I found that very interesting and a bit frustrating, considering the wonderful depth and layering.
I never had time to fully understand what parts of the room acoustics were causing what effects. The extreme depth but lack of width was intriguing.
I have 2 audio rooms, one is a fully acoustically treated, sound proofed music studio that has full of diffusers, OC 704, floating floor, double wall, etc. Very flat RT60 for the size. The other room is a zero treated low ceiling large living room. This one is very reflective.
Of course, the music is surprisingly clear and accurate in the studio, but I like my living system better in some ways. Much less microscopic, more relax, and actually more natural, I mean the music is much more naturally and seamlessly connected to my life. Instead, the studio envionment makes the music disconnect from my real life. There is nothing other than music in the studio.
I enjoy both rooms, and they are just 2 different music listning environment. I think the goal of Linkeitz is something between those 2, and nothing wrong.
Of course, the music is surprisingly clear and accurate in the studio, but I like my living system better in some ways. Much less microscopic, more relax, and actually more natural, I mean the music is much more naturally and seamlessly connected to my life. Instead, the studio envionment makes the music disconnect from my real life. There is nothing other than music in the studio.
I enjoy both rooms, and they are just 2 different music listning environment. I think the goal of Linkeitz is something between those 2, and nothing wrong.
Linkwitz has described many interesting experiments to enhance his listening experience.
Including: surround sound for a 2 channel system
And his Watson experiment
Experiments like these are a fun and educative experience. I've done my share as well and looked at his experiments for inspiration.
I will probably continue to do so, as long as I'm having fun and maybe learn a thing or two. It's good to know what you have though.
Including: surround sound for a 2 channel system
And his Watson experiment
Experiments like these are a fun and educative experience. I've done my share as well and looked at his experiments for inspiration.
I will probably continue to do so, as long as I'm having fun and maybe learn a thing or two. It's good to know what you have though.
In a room with diffuse and frequency independent reflections, you should only hear an increase in reflections.Yet you don't perceive the room as much listening to your speakers at your favorite spot. Listen at that spot to a recording of those speakers made in that spot (in other words, double the room effect) and after that go back, listening to a normal recording. It will make you more aware of the room.
It took me quite a while to revert back to normal listening behavior
If you record sound with reflections, it should sound like sound with reflections not sound like it has no reflections. That's what I mean by "Supposed to sound like that".
We are talking about different things. I'm talking about stereo image, iow location of sound sources. Your talking about hearing the listening room.
I think we agree on both.
It is not constant directivity per se but a monotonically falling power response with increasing frequency (which obviously has a limit at some point) that causes odd reflections to be easily removed from perception. The more directional a sound source and the more non-linear its power response so the easier it becomes to discern room artefacts.
Not at all. It depends on their direction, intensity, frequency content and their delay. Only a reasonably diffuse reflection (that is unlikely to occur from a naturally generated first reflection) outside the requisite time window will necessarily improve clarity. In most practical cases well designed power responses give the required results.
However, this thread seems to be awash with ill-defined terminology. Image width can be controlled by a shuffler (see Gerzon's work on the subject for example) with no changes to the nature of early reflections.
Sorry I'm a bit confused by your last sentence. By "easier to discern room artifacts", do you mean that as a good or bad thing?
As in... smoothly rising directivity creates room artifacts that are easier to discern and therefore perceptually ignore.
Or
Creates room artifacts that are more apparent and harder to ignore.
One goal (and a good one in my opinion) is to have the room power evenly distributed. Ie: over time, and so the reverberant response has the same tonal balance as the direct sound. This gives a spatial character related to the room, but nothing else. When one wall becomes inordinately involved there is the potential for that reflection to be more distinct.
I think the mono test can demonstrate success here. Eg: the image can seem to be coming from a small window between the speakers. If you stand up and move toward the speakers and left/right it still sounds as if it's coming from that window.
Reflections are not the only things that can betray the location of a speaker. The sound shouldn't be coming from the speakers, they should appear to be silent.
I think the mono test can demonstrate success here. Eg: the image can seem to be coming from a small window between the speakers. If you stand up and move toward the speakers and left/right it still sounds as if it's coming from that window.
Reflections are not the only things that can betray the location of a speaker. The sound shouldn't be coming from the speakers, they should appear to be silent.
For reproduction it is a bad thing
It is unlikely an individual wall can effect the reverberant response that is the sum of many reflections in the latter part of the room response.
This makes no sense to me. Directional cues emanate from the direct sounds from the loudspeakers and the listener's position relative to them.
Less likely maybe, but there are timing, angle relative to the direct, and level issues that will make them troublesome.
A good reverberant field IMO is something you can relegate to the background in your mind. Trying to ignore an early reflection can lead to fatigue.
I'd sooner say it has much to do with the speaker locations relative to each other.
Again, I think that when there is little to betray a speakers location, and that means secondary sources of sound such as diffraction, it behaves as a point source and doesn't have a distinct three dimensional signature to give it a 'front and back' for example so when you move about it doesn't readily change.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.