SBIR

[DanDan]

I would like to gather some input on SBIR.

We focus on modes but is this other phenomenon just as significant, or maybe more so?

This other phenomenon is in reality a collection.
There are many SBIR's. Much thanks to Andreas for this sketch.


These SBIR's can combine dramatically if the reflection path lengths coincide.
Note well, having the woofer an equal distance from the front and side wall will not cause equal reflection path lengths. Thanks to Jens for that insight.

It is suggested that a 4 inch panel behind the speaker may diminish SBIR.
My tests have never shown a significant improvement. Lupo concurs. I would love to see some successful attempts please. I believe Rod Gervais did achieve this but interestingly he placed the trap close to the speaker rather than the wall.

Over to you.

DD

[Jens Eklund]

Before discussing treatment, it’s necessary to clarify what the problem is. Depending on the position of the source in relation to the boundary, the fist null(s) (usually the biggest issue(s) with SBIR) will appear at different frequencies. If the speaker is close to a wall, this might be quite high up and therefore not to hard do fix with relatively shallow absorbers but if a lower frequency, it will require deeper construction or a dedicated (more or less "tuned") panel.

Behind monitors--absorption, diffusion?

[johndykstra]

Hey Dan,

Good thread. I would think a link to that nifty wall bounce calculator you have would be a good reference early on in this discussion.

[Hannes_F]

Originally Posted by DanDan

Over to you.

How to tell a mode from a SBIR?

Look at the bottom of the waterfall. If a branch grows out there we have a mode (long time phenomenon).

If we have a dip in the FR that is not sitting on a branch ...

then it is either

• next to a mode (modes can steal energy from neigbour frequencies)
• or it is finally a SBIR (one time occurance).

If it is a peak in the FR not sitting on a branch it must be a SBIR, too.

In bad luck you have both a mode and a SBIR at the same frequency

A SBIR should show in the ETC, too (right? always?).

All of the above is only my opinion attained from my measurements, so anything could be wrong. Take it with a grain of salt.

Next one, please.

[Rod Gervais]

Let us also stipulate that what we should be examining is the 2nd illustration in the sketch provided.

The 1st illustration is a situation easily handled - deals with first reflections, is well documented and (relatively) easy to deal with. It is also generally not a low frequency issue.

I always differentiate between simple early reflections and what is the much more complex issue being illustrated in the 2nd sketch provided.

Which is why I do not use the same term to describe 2 completely different "animals".

Situation 1 is a simple early reflection that may well cause comb filtering........

Situation 2 is a condition that is much more similar to the problems caused by room modes - and will never cause comb filtering.

Situation 2 is what I refer to if you hear me use the term SBIR.

[Jeffrey Hedback]

I keep this little Genelec marketing swag SBIR tape measure in hand (or nearby) at all times. It shows physical distance, the freq of that distance, & 1/4 wavelength freq which becomes the resulting null due to phase cancellation. Nifty little deal.

Of some dispute is where you measure (just covering points on topic): front of speaker to surface is correct as this is where the return energy is going to combine with the direct energy.

[Hannes_F]

Situation 2:

It should be noted that as long as we are below the Schroeder frequency it is not important that the mirroring surface is perpendicular to the wave propagation, nor do only surfaces count that are located on certain axes in regards to the loudspeakers. Any beveled hard surface near the speakers will do (but it is a different question whether much frequency content gets there that can create interferences).

[DanDan]

Thanks for joining me here.

Thomas Barefoots Excel Wall Bounce Calculator 2D download

Good point Jens. Ethan and I often recommend extreme closeness to front wall. This often measures better than the various out from the wall compromises, particularly with smaller speakers which like the LF boost.
It simply hadn't occurred to me to try SBIR treatment inserted between the wall and speaker. May be the icing on the cake. Thanks.

Rod, I have a memory of some studio you did. There was a trap, a sheet of 703 behind the speakers. Do you have links and/or tests on that?
I am assuming that the thinnish panel worked at that considerable distance because it was at a velocity max?

DD

[Jens Eklund]

Originally Posted by DanDan

Thanks for joining me here.

Thomas Barefoots Excel Wall Bounce Calculator 2D download

Good point Jens. Ethan and I often recommend extreme closeness to front wall. This often measures better than the various out from the wall compromises, particularly with smaller speakers which like the LF boost.
It simply hadn't occurred to me to try SBIR treatment inserted between the wall and speaker. May be the icing on the cake. Thanks.

Originally Posted by Jens Eklund

Usually (but naturally there’s exceptions), it’s preferable to move up against the front wall unless you can move back a considerable amount (> approx. 1,5 - 2 meters) and still sit in front of the center in the room. A close distance to the wall will push the SBIR effects up in frequency and is therefore easier to deal with (absorb).

[Rod Gervais]

Originally Posted by Jeffrey Hedback

I keep this little Genelec marketing swag SBIR tape measure in hand (or nearby) at all times. It shows physical distance, the freq of that distance, & 1/4 wavelength freq which becomes the resulting null due to phase cancellation. Nifty little deal.

Of some dispute is where you measure (just covering points on topic): front of speaker to surface is correct as this is where the return energy is going to combine with the direct energy.

On a speaker with the port in the rear (like the Genelec speakers recently specified by a client for his 5.1 mixing room) It would make sense to measure to the port......

[Rod Gervais]

Originally Posted by DanDan

Thanks for joining me here.

Thomas Barefoots Excel Wall Bounce Calculator 2D download

I would point out that the calculator Thomas wrote was designed specifically for near field speakers - and must be viewed with that in mind......

Rod, I have a memory of some studio you did. There was a trap, a sheet of 703 behind the speakers. Do you have links and/or tests on that?

Dan,

That was Hit Production's Studio "H" in Manila, using Barefoot Speakers (Micromain 27s) in the room. I never made it to that project and was (therefore) never able to perform testing on that studio.

Everything was done at a distance.....

I am currently designing some AV rooms for them including a 5.1 mixing room (with a voice over room attached), and, a couple of Foley Rooms, both with their own mixing rooms.

You can check Hit Productions out at:

Hit Productions

I am assuming that the thinnish panel worked at that considerable distance because it was at a velocity max?

Dan,

I just went back and reviewed all of my communication on this between Dennis Cham and myself regarding the issue relating to the speakers in this room (which centered on the reason for the treatment directly behind the speakers you refer to).

This was actually an issue related to an early reflection that existed off the window looking into the tracking room - and not a coupling of a signal at the speaker itself.

The issue was easily solved with some 703 - the placement close to the speaker was due to the fact that I did not want to close of any part of the window view from the person sitting at the mixing desk.

I took a lot of pains with that design to make certain that whoever was at the desk had an excellent view of the room - was not going to let anything screw that up.

There are 2 huge base trap located in the frot corners of that room - which helped tremendously - there is also a huge cloud directly over the speaker (actually the whole room - 4" of 703 which runs to within 2' of the walls on all side), and a monster bass trap in the rear of the room.......

None of the rooms I have designed have ever had any real issues with low frequencies. Studio H needed a 3dB cut at 125 and a 1.5dB boost at 80Hz at the mixing desk.

They also had a 3 - 4 dB boost at the lowest frequencies all the way in the back of the room at the producer's couch.

Not perfect - but close enough to suit my client.

By the way, although almost ashamed to admit it - I never knew who Dan Dan was before today - other than the fact that you were knowledgeable.

I took the time to check out your website while i was there - read your biography in the process, and I must admit that I am (rightfully so) impressed.

Rod

[DanDan]

Thanks for recalling that very specific information Rod. My eye gets caught by little anomalies or curiousities. Now the mystery of that back speaker panel is resolved for me.
Further thanks for your kind words. The decline in work for us freelance recordists plus the unfortunate 'paused' state of this little island means I don't get as much opportunity to do my thing with great talent in great studios these days. It is becoming apparent that I may have to join the 50,000 leaving here yearly.
All good here in our little cyberworkshop though, DD

[JohnTheRevelator]

I wrote something not too long ago on the topic and thought it might help someone new to acoustics grasp the topic easily. Check it out and don't make fun of my artwork.tutt

audio blog: A little more into boundary conditions

My screen name is a pseudonym for which I apologize. Some of you may know me around the web as DanTheMan. That name was taken.

Dan

[Jeffrey Hedback]

Following up on SBIR and where to measure, I discussed this with a speaker designer friend of mine and the result:
- sealed box: the face of the woofer dust cap
- bass reflex: has a condition variable answer. The port tuning typically affects (general statement) an octave above the tuning frequency and down. Therefore depending on the distance of the speaker to surface & the specific speaker design, you would measure from woofer dust cap if the distance places SBIR above port effects OR from the port if SBIR is within the range of port tuning effects.

[DanDan]

That blog piece from JohntheRevelator is well worth a look. It wobbles from metric to imperial, but we are all getting used to that.
In it there is a link to a Floor Bounce Calculator. This can also be used as a Ceiling Bounce or Side Wall Bounce Calculator.
Nice.

Worth noting that the 'classic' SBIR, the front wall one causes a fixed frequency null. This is because the difference between the direct and reflection path remains constant.
The null frequency for all the others changes with listener distance.

This combination of effects makes it difficult to determine how effective an SBIR panel behind the speaker is. One would have to diminish the other SBIR's to isolate that one to get a readable test result.

Double whammy. Can someone remind me why the Back Wall LF trapping is so prioritised? Surely there is more energy at the Front Wall? Should we redirect to massive LF front wall trapping, which would have the SBIR side benefit? Or distribute our LF trapping efforts equally front and back?

DD

[Hannes_F]

Originally Posted by DanDan

Can someone remind me why the Back Wall LF trapping is so prioritised?

I have come to the conclusion that The Front Half (all surfaces near the speakers except those that are very near) dominantly determine the FR. Which is underrated imo. The Back Half is a good place to dominantly dampen the modes. This is no either - or, just dominances.

[JohnTheRevelator]

DanDan, that's essentially what I recommend with my quick and dirty room treatment advice. So I'd agree with your assessment. The post is more like 'a place to start' when thinking of room treatments.

audio blog: Room treatment advice--quick and dirty

Sorry for the Imperial/Metric waffling. It was primarily done for Americans b/c the calculator is in metric and anyone could do their own easily. The tables would require a bit of work for Americans. Perhaps I should add some for the rest of the world. It would take long.

Dan

[SörenHjalmarsson]

Wazzup everybody!!

Finally got some time to read this thread through... For those of you who do not know, this thread was partly triggered by a discussion that started in another thread. In that thread i tried to give an answer as to why you would treat the front wall with absorbtion, While at the same time, describing the negative effect of SBIR - i would now like to continue that discussion here to get some clarification...

My post was contradicted by Mr Gervais - but i have to say that i am not yet completely satisfied with this new (for me) understanding of the effect.

Here is the post by Mr Gervais: http://www.gearslutz.com/board/6358770-post64.html

I think it would be good at this point to try and lay down/explain (in fairly small words) what SBIR actually is - and how it negatively effects the accurate listening.

My understanding:

When you apply yourself to the serious business of listening to your speakers indoors, you should know that the boundaries of your room will alter the LF response and accuracy of your listening (more or less depending on what your walls are made of and placement of speakers etc...). This is due to the fact that omnidirectional Low Frequencies will reflect of the nearby surfaces and bounce back to the listening position. On it's way back it will mingle with the direct signal, and by doing so, cause an interference in the accuracy of your listening - this is what you would call a boundary interference - and is by definition a comb filtering effect...

i.e. I fail to see how SBIR could NOT lead to comb filtering??

Oh! and just to avoid us talking about different/same thing here, lets try to not let terminology confuse us here - 'wave interference', 'acoustic interference', 'polar lobing', or whatever you want to call it. The end result is,after all, comb filtering.

If anyone has an explanation that can ad or correct, feel free to do so...


Cheers,

[RKrizman]

I would think that independent of any comb filtering effects you would also get amplitude changes in various frequency ranges, because the proximity of the wall affects the energy of the reflected signal. In other words, if you move your speakers closer to the front wall would you not typically get more low end, because that's what's hitting the wall from the speakers, and the closer they are to the wall the more gets reflected back.

Sorry if I'm not putting that in correct scientific terminology. My point is that there is more than just comb filtering in SBIR, right?

-R

[DanDan]

Dan (JohntheRevelator) I neglected to say welcome to GS.

I have come to the conclusion that The Front Half (all surfaces near the speakers except those that are very near) dominantly determine the FR. Which is underrated imo. The Back Half is a good place to dominantly dampen the modes. This is no either - or, just dominances.

Me too, probably at the same moment, and Eureka!

DD

[JohnTheRevelator]

Thanks DanDan--or would it be better to say DD?

SörenHjalmarsson, the front wall reflection is not likely to lead to a comb d/t the radiation pattern of the speaker. Most become pretty directional above 1 kHz or so. Here's a number of gated polar graphs I've taken:
audio blog: Review of Polar graphs
Just so everyone is aware I'm not making this stuff up. With speakers that are more omni, and definitely dipoles will cause a comb filter. That directivity pattern is largely dictated by the size of the woofer and the shape/size(or lack there of) of the tweeter's waveguide. This pattern also occurs from the floor, ceiling, and side walls though they don't eliminate it. So narrow patterned speakers have less comb filtering in general. Then you add to the mix things like line arrays--especially of the shaded variety, that will remove the floor comb completely.

They real question though is how audible is comb filtering? There is a lot of evidence on this. For reading on this you can start here: audio blog: Psychoacoustics
This might be disturbing to the pro audio crowd.

That's certainly not to say room treatment is overrated or anything. Drs. Toole and Olive (2006,1995) once did an experiment where they played a loudspeaker in four rooms of various sizes and shapes to several listeners and made a binaural recording of it. In the room, the loudspeaker quality was the dominant factor of perceived SQ, in the binaural playback it was also the speaker. When the same speaker was rated against itself binaurally in the different rooms, the room was the dominant factor. P=0.05 for the first and p=0.001 for the second. IOW the results were highly significant. The speakers were measured in anechoic chamber. What does this tell us about room factors in general? Well, I know what it means to me. I'll let you decide for yourself. Think about this however, could you recognize a Gibson vs. a Fender in a gymnasium? But wouldn't they sound better in a good studio?

Dan

[Hannes_F]

Originally Posted by Hannes_F

I have come to the conclusion that The Front Half (all surfaces near the speakers except those that are very near) dominantly determine the FR. Which is underrated imo. The Back Half is a good place to dominantly dampen the modes. This is no either - or, just dominances.

Dan, in order to put some more beef to the table I calculated some constructive and destructive SBIRs, and with the string method I was able to identify some possibly corresponding surfaces.

Eureka indeed ...

The graph is for the left monitor. In a way if you look downwards your FR it is similar to a fish-eye perspective photo of your room (actually to pictures layered onto each other, one for constructive and one for destructive inferference.

[bwo]

Personally I don't think placing the speakers close to the frontwall works well for most speakers. For the midrange and treble, the best speaker placement is away from the wall. Too close to the wall, and the image is distorted.

I think a better option is to use separate units for the bass and place them randomly in the room.

[Hannes_F]

Originally Posted by bwo

Personally I don't think placing the speakers close to the frontwall works well for most speakers. For the midrange and treble, the best speaker placement is away from the wall.

bwo, you are right insofar as we need to control the reflections by geometry... of course. Just placing speakers close to a wall without further construction is not good. Even with construction it is critical as even little reflections to the wrong direction can cause big imaging mess, as you say. Hence the necessity to check that all with ETC measurements.

[PaulRain]

Originally Posted by Hannes_F

Dan, in order to put some more beef to the table I calculated some constructive and destructive SBIRs, and with the string method I was able to identify some possibly corresponding surfaces.

Eureka indeed ...

The graph is for the left monitor.

sorry...I must miss something....and of course I'm a great ignorant, could you explane a little better this measures?
P

[Hannes_F]

Originally Posted by PaulRain

sorry...I must miss something....and of course I'm a great ignorant, could you explane a little better this measures?
P

You are right to ask because if this is true it is a very powerful method. But I need to verify this a little before I get into details.

[avare]

Originally Posted by Hannes_F

Dan, in order to put some more beef to the table I calculated some constructive and destructive SBIRs, and with the string method I was able to identify some possibly corresponding surfaces.

Eureka indeed ...

Fantastic stuff!

Andre

[gullfo]

nice work Hannes! i guess if you had a spreadsheet with prominent object-in-room coordinates and the data from the FR you could automate this to some degree as well as see where overlapping boundaries and objects conspire.

[Hannes_F]

Originally Posted by gullfo

nice work Hannes! i guess if you had a spreadsheet with prominent object-in-room coordinates and the data from the FR you could automate this to some degree as well as see where overlapping boundaries and objects conspire.

Haha Glenn, I guess I am more a pencil on paper guy but it might be time for a spreadsheet.

BTW it is strange, I have a very slow way of learning. I must write things down, do the calculations myself step by step, do some real life experiment, wonder about it over and over and come to a conclusion then.

What happens if an infinite force hits on an unsurmountable resistance?

It ... takes ... time ...

[claend]

As usual, I'm astonished.

...



Thanks for the info... hopefully, one day I'll understand it!