Old 04-10-2020, 01:53 PM   #1
clepsydrae
Human being with feelings
 
clepsydrae's Avatar
 
Join Date: Nov 2011
Posts: 2,785
Default technical question about SNR

Theory: when I record an ensemble with multiple mics in a room who are playing a song at average level X, there is a certain amount of direct sound in each mic, as well as room ambiance and bleed from the other musicians. Each channel experiences this. When the ensemble then plays a different song at average level 4X with no change in mic'ing, the same ratios maintain: that is, there is more ambiance bouncing around, more bleed, but also more direct signal, and given linear relationships and the inverse square law and blah blah blah it all balances out and the same mix is obtained. Obviously instrument timbre varies but the signal-to-bleed balance is basically the same for all mics, all else being equal.

Reality: it seems like whenever I do this, the second song has way more bleed, ambiance, etc, in proportion to the direct sound. I've never quite understood why.

Certainly with a single mic, if I speak in to it from a fixed distance, there is some direct and some reverberant pickup. If I then scream, it makes the room echo more, but the direct sound is proportionally louder, and it all balances. But a multi-mic setup seems to introduce a lot more room sound.

I could see compression being to blame: the ring time will be longer so if the gain pumps back up between bursts you'll have a perceptibly stronger and longer ambiance, but it seems true even in the raw recordings.

Anyone have a technical explanation for this? Is my "math" off, or maybe there are nonlinear dynamics I'm not clever enough to anticipate? Or maybe I'm imagining it. I haven't rigorously tested it. It just always seems to bite me in mixing.
clepsydrae is offline   Reply With Quote
Old 04-10-2020, 02:10 PM   #2
Fergler
Human being with feelings
 
Fergler's Avatar
 
Join Date: Jan 2014
Posts: 4,242
Default

Partly could be perception, as volume and perceived volume of frequencies is not linear (Fletcher-Munson curve for more details).

Can also be that when they are playing louder, they are creating more harmonics and higher mid-range tone. This is in general true for all orchestral instruments. Play soft and you don't get as much brightness in your sound.
__________________
Check out my hip-hop, funk-rock band Theatre Crisp
Fergler is offline   Reply With Quote
Old 04-10-2020, 02:10 PM   #3
karbomusic
Human being with feelings
 
karbomusic's Avatar
 
Join Date: May 2009
Posts: 26,548
Default

Do decay times of room resonances increase when louder? Seems like if gain were also adjusted this may not matter but 'exiting the room' or whatever the technical term is crossed my mind. I would think this more likely where modes are concerned since often the secondary issue with those is their increased decay aka ring time - which is why we often need a waterfall plot.

Maybe you could sweep it and check the waterfall. Maybe REQs tools would assist? If it were me, that's where I would go next.

I'm not sure Fletcher-Munson would matter since you are monitoring at the same levels correct?
__________________
Good health is merely the slowest possible rate at which one can die.
karbomusic is offline   Reply With Quote
Old 04-10-2020, 02:33 PM   #4
clepsydrae
Human being with feelings
 
clepsydrae's Avatar
 
Join Date: Nov 2011
Posts: 2,785
Default

Quote:
Originally Posted by Fergler View Post
Partly could be perception, as volume and perceived volume of frequencies is not linear (Fletcher-Munson curve for more details).
Yeah... seems a likely guess, but it really seems like it's not a timbre issue that I'm hearing... More generally, I'm aware that loudness perception is highly non-linear (i.e. the relationship between SPL, intensity, and loudness), but in the above-described scenario, I'm effectively normalizing by monitoring at the same perceived volume during mixing, so I'd expect things to fall back to the same proportions... unless there is an additional nonlinearity I'm not accounting for in the physics of the microphone's transduction.

Quote:
Originally Posted by karbomusic View Post
Do decay times of room resonances increase when louder? Seems like if gain were also adjusted this may not matter
Yeah it gets longer until it falls under whatever threshold but AFAIK I don't think it gets proportionally longer in any way (including longer modes)... that'd be good to know, if true, though.

Quote:
I'm not sure Fletcher-Munson would matter since you are monitoring at the same levels correct?
Beat me to it, yep, but the curves certainly would matter if what I'm really hearing is just a timbre illusion.

Does anyone else hear this in their recordings? Our band just recorded an album; ~14 tracks with the mics in the same place. The quiet parts all sound fantastic, and the loud parts mostly sound terrible (due to room ambiance and bleed.) The instruments don't sound terribly different in terms of timbre. When I solo individual tracks there is tons of bleed, proportionally speaking, on the louder tracks. It's confusing.

I wonder if some of our instruments radiate proportionally more sound out their sides when played loud.

It's also possible that some of us are dramatically increasing our volume on the loud songs where the others aren't, and when things are balanced in the mix that brings out the ambiance on the quieter instruments. I bet that's basically it, plus timbre illusions.

If that is it, and my "math" is right, then it's really easy to see how engineers get the impression that louder = more proportional bleed/ambiance, because it sure as heck seems true.
clepsydrae is offline   Reply With Quote
Old 04-10-2020, 02:41 PM   #5
Fergler
Human being with feelings
 
Fergler's Avatar
 
Join Date: Jan 2014
Posts: 4,242
Default

I dunno I've just sort of learned to surrender to the irrational processes our brain does with audio. For example this might be a problem with multiple mics only because you're hearing it in stereo and it sounds real and your ear can target those room sounds. But in mono with one mic maybe that whole process just falls apart.
__________________
Check out my hip-hop, funk-rock band Theatre Crisp
Fergler is offline   Reply With Quote
Old 04-10-2020, 02:54 PM   #6
karbomusic
Human being with feelings
 
karbomusic's Avatar
 
Join Date: May 2009
Posts: 26,548
Default

Quote:
Originally Posted by clepsydrae View Post

Does anyone else hear this in their recordings?
I can say from 30 years of recording experience, lower SPL recordings always sound better than higher SPL (live or studio) the vast majority of the time.

I didn't mention before because you were discussing bleed and didn't make the connection as I've always connected it to the gear and instruments themselves (but not 100% sure of the cause), because them working louder is simply not the same as more nominal levels.

This is audible even playing a clean electric guitar and picking louder where I assume that the PU compresses/chokes for a split second then recovers - where softly picked, even if supposed to sound heavier will often actually sound heavier with a lighter touch. That's a guess but that type of thing could exist electronically and physically all over the place?

I just don't really think ^it's related to what you are explaining.

Are you sure the SNR or DR is actually the same? Could something be compressing like my PU example but in the instruments/gear itself including the mics? I'm just spit balling here.

Or perception as Fergler said but, seems like you could rule that out.
__________________
Good health is merely the slowest possible rate at which one can die.
karbomusic is offline   Reply With Quote
Old 04-10-2020, 03:19 PM   #7
clepsydrae
Human being with feelings
 
clepsydrae's Avatar
 
Join Date: Nov 2011
Posts: 2,785
Default

Quote:
Originally Posted by karbomusic View Post
Could something be compressing like my PU example but in the instruments/gear itself including the mics? I'm just spit balling here.
Yeah that's what I'm wondering. That or like I'm not visualizing the math right.

I should maybe ask this: for a given sound source at perceived loudness X with mic A 1 foot away and mic B 10 feet away, let's say mic A hears it at such a level as to generate a resultant loudness level Y, and mic B hears it to generate resultant loudness level Z (through a hypothetical reproduction chain that doesn't change).

If I increase X by factor k, do I hear kY at mic A, and kZ at mic B?

I've been assuming so, all these years, but I'm not super-solid in that belief.

Edit: and I should say that my intuition sometimes points the other way; besides the above-described scenario with our album recording, it has always seemed the case that close-mic'ed vocals seem to have way more dynamic range than moderate-distance-mic'd vocals, for example (setting aside issues of proximity effect and plosives, etc).
clepsydrae is offline   Reply With Quote
Old 04-10-2020, 03:42 PM   #8
Geoff Waddington
Human being with feelings
 
Geoff Waddington's Avatar
 
Join Date: Mar 2009
Location: Dartmouth, Nova Scotia
Posts: 5,057
Default

Isn't it as simple as, with louder playing you are putting more energy into the system, thereby increasing reverb tails (e.g. longer time to -40db) to the point where they are adding up more and more and more the louder you play, thereby altering the wet/dry/bleed level relationships fairly dramatically, as we all have experienced
__________________
CSI - You can donate here: geoffwaddington.ca
Beta software: https://stash.reaper.fm/v/38349/CSI%20beta.zip
installation and documentation here: https://github.com/malcolmgroves/reaper_csi/wiki
Geoff Waddington is offline   Reply With Quote
Old 04-10-2020, 03:42 PM   #9
karbomusic
Human being with feelings
 
karbomusic's Avatar
 
Join Date: May 2009
Posts: 26,548
Default

Quote:
Originally Posted by clepsydrae View Post
If I increase X by factor k, do I hear kY at mic A, and kZ at mic B?

I've been assuming so, all these years, but I'm not super-solid in that belief.
Yea, I don't know, I'd have to try to test it too.

Quote:
Edit: and I should say that my intuition sometimes points the other way; besides the above-described scenario with our album recording, it has always seemed the case that close-mic'ed vocals seem to have way more dynamic range than moderate-distance-mic'd vocals, for example (setting aside issues of proximity effect and plosives, etc).
I assumed it would be a given since it's traveling through that extra air - I don't have math to back it up just IME. Seems to me there would be a lot more going on there physically than just more gain but again spit-balling based on experience.

It's why I tend to rarely do vocals as close as most do - come to think of it, I do most of my micing further than what the internet shows as the norm (not always though) - to me it just sounds better with some distance or even slightly off-axis. But I have a bit of a dead room on purpose, just so I can do that.
__________________
Good health is merely the slowest possible rate at which one can die.
karbomusic is offline   Reply With Quote
Old 04-10-2020, 04:07 PM   #10
clepsydrae
Human being with feelings
 
clepsydrae's Avatar
 
Join Date: Nov 2011
Posts: 2,785
Default

Quote:
Originally Posted by Geoff Waddington View Post
Isn't it as simple as, with louder playing you are putting more energy into the system, thereby increasing reverb tails (e.g. longer time to -40db) to the point where they are adding up more and more and more the louder you play, thereby altering the wet/dry/bleed level relationships fairly dramatically, as we all have experienced
Ah, this seems like the explanation, and which Karbo was also saying way back in the third post: even if the tail is in the same proportion to the inciting sound, the threshold of hearing hasn't changed in the two scenarios, so the louder tail will appear to linger longer. So with louder instruments in a room, the tails all linger longer and the collected sum is therefore stronger. That makes sense to me... maybe that's all it is (plus timbre, etc.)

So the room isn't excited proportionally more by loud sounds (which is what I was stuck on, because the proportionality seemed to contradict that), it's excited longer, due to the fact that the human threshold of hearing doesn't scale with the increased loudness (well, much; the acoustic reflex obviously throws this off a bit, but we can probably ignore it.)

This is pretty obvious in retrospect, but that's why I posted: to try to figure out what I was overlooking. Thanks for the brainstorming and help.

Last edited by clepsydrae; 04-10-2020 at 04:30 PM.
clepsydrae is offline   Reply With Quote
Old 04-10-2020, 04:33 PM   #11
not-relevant
Human being with feelings
 
Join Date: Jun 2019
Posts: 92
Default

But isn't the reverb tail the same level below the main signal, in the same timeframe? And assuming you turn the resulting mix down to get the same headroom, haven't you also pushed the room reverb volume down as well, such that it decays below the threshold of hearing at the same point in time?
not-relevant is offline   Reply With Quote
Old 04-10-2020, 04:40 PM   #12
clepsydrae
Human being with feelings
 
clepsydrae's Avatar
 
Join Date: Nov 2011
Posts: 2,785
Default

Oh shoot, right, which is why that explanation didn't satisfy before... so maybe we're back to the other explanations...

Edit: post 10 explains why it would sound like the room was more excited during a live music experience, but when recorded it doesn't seem like it should matter.
clepsydrae is offline   Reply With Quote
Old 04-10-2020, 04:47 PM   #13
not-relevant
Human being with feelings
 
Join Date: Jun 2019
Posts: 92
Default

My money is on it being psycho-acoustics, and the same kind of mechanisms that are work in loudness-war super compressed stuff that sounds louder than material that's notionally the same level but not been subject to the same processing: more 'density', whatever that is!
not-relevant is offline   Reply With Quote
Old 04-10-2020, 04:55 PM   #14
clepsydrae
Human being with feelings
 
clepsydrae's Avatar
 
Join Date: Nov 2011
Posts: 2,785
Default

Quote:
Originally Posted by not-relevant View Post
My money is on it being psycho-acoustics, and the same kind of mechanisms that are work in loudness-war super compressed stuff that sounds louder than material that's notionally the same level but not been subject to the same processing: more 'density', whatever that is!
Sounds like you're implying that the nonlinearity is a kind of compression, i.e. my little formula in post 7 is not accurate, which seems plausible except for the point you just made: since we're normalizing when monitoring the results, any such nonlinearity/compression would have to happen in the hardware and not be psychoacoustic in nature... which feels harder to explain, to me.
clepsydrae is offline   Reply With Quote
Old 04-10-2020, 05:38 PM   #15
Geoff Waddington
Human being with feelings
 
Geoff Waddington's Avatar
 
Join Date: Mar 2009
Location: Dartmouth, Nova Scotia
Posts: 5,057
Default

Quote:
Originally Posted by not-relevant View Post
But isn't the reverb tail the same level below the main signal, in the same timeframe? And assuming you turn the resulting mix down to get the same headroom, haven't you also pushed the room reverb volume down as well, such that it decays below the threshold of hearing at the same point in time?
Yes, but there are many previous reverb tails that still haven't fallen below that threshold. There is more energy in the room with the louder playing and it takes longer overall time to decay.
__________________
CSI - You can donate here: geoffwaddington.ca
Beta software: https://stash.reaper.fm/v/38349/CSI%20beta.zip
installation and documentation here: https://github.com/malcolmgroves/reaper_csi/wiki
Geoff Waddington is offline   Reply With Quote
Old 04-10-2020, 05:39 PM   #16
karbomusic
Human being with feelings
 
karbomusic's Avatar
 
Join Date: May 2009
Posts: 26,548
Default

Quote:
Originally Posted by Geoff Waddington View Post
Yes, but there are many previous reverb tails that still haven't fallen below that threshold. There is more energy in the room with the louder playing and it takes longer overall time to decay.
That's what crossed my mind but wasn't sure - I bet testing this would be easier in an empty gymnasium.
__________________
Good health is merely the slowest possible rate at which one can die.
karbomusic is offline   Reply With Quote
Old 04-10-2020, 05:57 PM   #17
clepsydrae
Human being with feelings
 
clepsydrae's Avatar
 
Join Date: Nov 2011
Posts: 2,785
Default

Quote:
Originally Posted by Geoff Waddington View Post
Yes, but there are many previous reverb tails that still haven't fallen below that threshold. There is more energy in the room with the louder playing and it takes longer overall time to decay.
Well but there aren't more tails that haven't fallen below... the point is that if you are turning the level down in post to match the direct signal level, you are turning all the tails down, so they fall below the threshold faster, meaning they are around less time. Basically, we're back to linear proportionality.

E.g. I clap and record it. The reverb tail lives at some proportion to the direct clap. The tail lives some amount of time in the room before it's inaudible.

I clap twice as loud and record it. In the room, the clap is twice as loud, and the tail is twice as loud in proportion, but the tail lasts some amount longer as well, which means your explanation works (AFAIK) in the room: the tails build up and are proportionally louder with louder claps.

But in the recording we take the second clap, normalize it to match the level of the first clap: now the clap is the same, the tail is the same proportional level, and the tail lasts as long as with the first clap until it falls under audibility.

...unless there is some non-linearity to the decay time, so a test might be helpful.
clepsydrae is offline   Reply With Quote
Old 04-10-2020, 06:14 PM   #18
ashcat_lt
Human being with feelings
 
Join Date: Dec 2012
Posts: 5,023
Default

Are the amplifiers more compressed/overdriven/distorted? Drums and some other acoustic instruments actually have different volume envelopes when you hit the harder. What I’m getting at is that if the sources are less dynamic then so will be the reverb and bleed.

But also yes if one instrument turns up and the other doesn’t the yes the bleed in the second will (obviously?) be proportionately louder.
ashcat_lt is offline   Reply With Quote
Old 04-10-2020, 06:20 PM   #19
Geoff Waddington
Human being with feelings
 
Geoff Waddington's Avatar
 
Join Date: Mar 2009
Location: Dartmouth, Nova Scotia
Posts: 5,057
Default

Let's try it this way.

You are in a very live room.

You are on speaker phone and speak very quietly into the mic.

The person on the other end hears no discernible room.

Now, without moving your position you speak more loudly and soon the person on the other ends starts to sense you are in a room.

One might be tempted to say we are starting to "excite the room", but that is BS, even the slightest sound "excites" the room, it's just a matter of our perception. It's not like the room magically isn't excited below some threshold

It's like the ridiculous statements that if you keep the levels low you don't excite the bass modes in the room -- once again BS, they are excited at every level, you just don't perceive them as much at lower levels because of Fletcher-Munson.

So what is going on with the room when we speak more loudly ?

We are putting more energy in than before when we were speaking softly.

That energy decays because of the reflections and absorptions that occur over time.

More energy = longer overall decay - alters balance between wet/dry/bleed.

Maybe I'm missing something here, not sure what
__________________
CSI - You can donate here: geoffwaddington.ca
Beta software: https://stash.reaper.fm/v/38349/CSI%20beta.zip
installation and documentation here: https://github.com/malcolmgroves/reaper_csi/wiki
Geoff Waddington is offline   Reply With Quote
Old 04-10-2020, 06:32 PM   #20
clepsydrae
Human being with feelings
 
clepsydrae's Avatar
 
Join Date: Nov 2011
Posts: 2,785
Default

Quote:
Originally Posted by ashcat_lt View Post
Are the amplifiers more compressed/overdriven/distorted? Drums and some other acoustic instruments actually have different volume envelopes when you hit the harder. What I’m getting at is that if the sources are less dynamic then so will be the reverb and bleed.
Hmmm but once the sound is coming out of the instrument it's all the same to physics, right? Meaning the proportions and all that stuff are going to follow the same rules... or maybe you're saying this is a way an instrument can seem louder without actually being louder; i.e. band member plays "twice as loud" but it isn't, really. That makes sense.

Quote:
But also yes if one instrument turns up and the other doesn’t the yes the bleed in the second will (obviously?) be proportionately louder.
Yeah listening to the recordings a little more closely I think this might be the main culprit: biggest issue is with the cellist. When she plays louder, it's not really that much louder, whereas when the fiddle and guitar play louder, they are. Most of the effective extra bleed seems to be in the cello mic. And that makes sense. Unless all the instruments increase by the same proportion, the bleed proportion will shift.

Quote:
Originally Posted by Geoff Waddington View Post
Now, without moving your position you speak more loudly and soon the person on the other ends starts to sense you are in a room.
The problem with that scenario is that you haven't equalized the results. If you turn the second one down so that the voice loudness is the same as in the first, the reverb goes down with it, and I believe the person can no longer hear the room, as in the first case.

Quote:
One might be tempted to say we are starting to "excite the room", but that is BS, even the slightest sound "excites" the room, it's just a matter of our perception. It's not like the room magically isn't excited below some threshold

It's like the ridiculous statements that if you keep the levels low you don't excite the bass modes in the room -- once again BS, they are excited at every level, you just don't perceive them as much at lower levels because of Fletcher-Munson.
Agreed on all points, which is the source of my confusion: this all should be linearly proportional to energy input, AFAIK.

Quote:
More energy = longer decay - alters balance between wet/dry/bleed.
...the thing is that you measure decay length based on a perceptability threshold. Double the energy = longer time to decay below that threshold. But double the energy, then cut the energy in half in post when you normalize the recording level, and it should bring the decay time back to where it was.
clepsydrae is offline   Reply With Quote
Old 04-10-2020, 06:56 PM   #21
ashcat_lt
Human being with feelings
 
Join Date: Dec 2012
Posts: 5,023
Default

Quote:
Originally Posted by clepsydrae View Post
Hmmm but once the sound is coming out of the instrument it's all the same to physics, right?
I guess what I was saying is that the source has more average energy, so the reverb/bleed will too. Like if you hit a quickish note without much compression, the whole thing will (note and reverb) will die out faster than if there were more compression, so the more compressed source might make the reverb seem to last longer because the sustained portion actually is longer.

The more you post, though, the more I'm leaning toward just some instruments didn't get as much louder as some others, and so the S/N proportions are actually changing in a simple and predictable manner.

Then, too, unless the mics are really fixed distance from the instruments, it is at least possible that one might "play the mic" a bit more than some others - moving away from the mic when they get loud or something - which would also affect the relative bleed level.
ashcat_lt is offline   Reply With Quote
Old 04-10-2020, 07:05 PM   #22
karbomusic
Human being with feelings
 
karbomusic's Avatar
 
Join Date: May 2009
Posts: 26,548
Default

I think it's fairly easy to hit or enter into the slope of the compression/threshold on an instrument which changes it's DR and not really notice that's happening. I'm experience this on both my electric and acoustic guitars for example - acoustically, no amps involved.

Seems we could take all of that out of the picture though by playing a click or something and micing the monitor/speaker.
__________________
Good health is merely the slowest possible rate at which one can die.
karbomusic is offline   Reply With Quote
Old 04-10-2020, 07:32 PM   #23
clepsydrae
Human being with feelings
 
clepsydrae's Avatar
 
Join Date: Nov 2011
Posts: 2,785
Default

Quote:
Originally Posted by ashcat_lt View Post
I guess what I was saying is that the source has more average energy, so the reverb/bleed will too. Like if you hit a quickish note without much compression, the whole thing will (note and reverb) will die out faster than if there were more compression, so the more compressed source might make the reverb seem to last longer because the sustained portion actually is longer.
Ah OK, this is kinda what not-relevant was getting at... basically that "louder" instruments aren't necessarily louder in a simple sense, but that the "louder"ness can mean a different distribution of energy through time (i.e. compression) which could affect tail length... I could see that.

Quote:
The more you post, though, the more I'm leaning toward just some instruments didn't get as much louder as some others, and so the S/N proportions are actually changing in a simple and predictable manner.

Then, too, unless the mics are really fixed distance from the instruments, it is at least possible that one might "play the mic" a bit more than some others - moving away from the mic when they get loud or something - which would also affect the relative bleed level.
Yeah I think these issues are probably the main thing. My recent recording isn't enough to go on given the confounding variables, but I'm still unable to entirely square my intuition on this.

E.g. 10 people in a large room having 5 conversations in pairs. One mic in front of one of them at a fixed distance. They all talk quietly; record. Then they double in volume; record. Then normalize the result for the targeted speaker's voice. My intuition strongly suggests that there would be more ambiance and bleed in the second recording. But maybe that's wrong. Either my intuition needs to change or I need a better explanation of what's happening. :-)

Quote:
Originally Posted by karbomusic View Post
I think it's fairly easy to hit or enter into the slope of the compression/threshold on an instrument which changes it's DR and not really notice that's happening. I'm experience this on both my electric and acoustic guitars for example - acoustically, no amps involved.
Totally. And the opposite, too: some instruments suddenly get four times as loud and you don't really notice until you record them and try to work them in to a mix (pennywhistle, I'm looking at you...)

Quote:
Seems we could take all of that out of the picture though by playing a click or something and micing the monitor/speaker.
I'm usually the "I'll run some tests" guy but I'm a little occupied. :-) Maybe later on.
clepsydrae is offline   Reply With Quote
Old 04-11-2020, 02:10 PM   #24
clepsydrae
Human being with feelings
 
clepsydrae's Avatar
 
Join Date: Nov 2011
Posts: 2,785
Default

Couldn't help myself; ran some tests. I don't have a big room, unfortunately, so it may not be all that useful, but it has helped convince me...

I put a LDC up a couple feet from one monitor speaker angled out into my (small) room. The other monitor was farther away and pointed in its normal direction, which was sort of away from the mic.

I played three identical 50ms bursts of white noise, increasing in volume by 6dB each time.

I normalized the results. The amount room sound was identical in quality and length, AFAICT. (The ambient noise floor was of course less with the louder samples, e.g. computer fan, etc.)

I then put up two other LDC's: one near the other monitor, and one a few feet away and pointed away from the speakers towards a wall, and repeated the experiment.

Got the same results: the degree of room tone following the burst was indistinguishable for me, as one would expect if everything was nice and linear. Then did it again and turned down the first of the three bursts an additional 6dB just to maximize the difference between the first and last burst (now ~18dBFS). Same result.

I'd like to repeat this in a large reverberant space, though, with repeating sounds (to cause tails to overlap) and different simulated instruments in each monitor speaker, in case I'm just not hearing something. But crudely it seems to support the "linear" model that I think most of us are expecting.
clepsydrae is offline   Reply With Quote
Old 04-11-2020, 04:15 PM   #25
Geoff Waddington
Human being with feelings
 
Geoff Waddington's Avatar
 
Join Date: Mar 2009
Location: Dartmouth, Nova Scotia
Posts: 5,057
Default

Quote:
Originally Posted by clepsydrae View Post
Couldn't help myself; ran some tests. I don't have a big room, unfortunately, so it may not be all that useful, but it has helped convince me...

I put a LDC up a couple feet from one monitor speaker angled out into my (small) room. The other monitor was farther away and pointed in its normal direction, which was sort of away from the mic.

I played three identical 50ms bursts of white noise, increasing in volume by 6dB each time.

I normalized the results. The amount room sound was identical in quality and length, AFAICT. (The ambient noise floor was of course less with the louder samples, e.g. computer fan, etc.)

I then put up two other LDC's: one near the other monitor, and one a few feet away and pointed away from the speakers towards a wall, and repeated the experiment.

Got the same results: the degree of room tone following the burst was indistinguishable for me, as one would expect if everything was nice and linear. Then did it again and turned down the first of the three bursts an additional 6dB just to maximize the difference between the first and last burst (now ~18dBFS). Same result.

I'd like to repeat this in a large reverberant space, though, with repeating sounds (to cause tails to overlap) and different simulated instruments in each monitor speaker, in case I'm just not hearing something. But crudely it seems to support the "linear" model that I think most of us are expecting.
As you say tone bursts aren't musical phrases

I think the rhythmic repetition inherent in music is an important consideration here, especially WRT the reverb/energy/density/field/etc. stuff.
__________________
CSI - You can donate here: geoffwaddington.ca
Beta software: https://stash.reaper.fm/v/38349/CSI%20beta.zip
installation and documentation here: https://github.com/malcolmgroves/reaper_csi/wiki
Geoff Waddington is offline   Reply With Quote
Old 04-11-2020, 04:19 PM   #26
clepsydrae
Human being with feelings
 
clepsydrae's Avatar
 
Join Date: Nov 2011
Posts: 2,785
Default

Quote:
Originally Posted by Geoff Waddington View Post
As you say tone bursts aren't musical phrases

I think the rhythmic repetition inherent in music is an important consideration here, especially WRT the reverb/energy/density/field/etc. stuff.
My conviction is that any mechanism that would be revealed by what you describe should, in theory, be revealed by my test, but I acknowledge confusion on the point and agree that a better test would be more convincing in case I'm just not detecting something. Unfortunately I am not going to be able to do that until you-know-what goes away or is beaten by science, so if anyone has access to a large room and wants to do it, that'd be great. :-)
clepsydrae is offline   Reply With Quote
Reply

Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump


All times are GMT -7. The time now is 05:07 AM.


Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2020, vBulletin Solutions Inc.