Monitor & Mixing Desk Placement - Reflection Issues

[hillzz22]

So, thanks to inspiration from similar projects on this forum, I have designed and built a mixing desk (not quite finished yet, but close). I have Tannoy near field monitors (501a) and ideally would like to keep them off the desk, if possible. Once I lock in on the height I plan to build some stands. I recently gave the room the GIK treatment and before I got the desk in, the sound was great (with a temporary setup). I am now trying to figure out the best location for the monitors, around the desk. The desk surface is pretty large so it seems like my options are fairly limited if I am truly going to avoid reflections from the desk. I plan to rotate the 242 panel between the monitors so that it is horizontal and mount the two 242s up higher once I know exactly where the early reflection points will be. I also have two Monster traps on the back wall and have four Gridfusor panels that I can mount where needed. Am I better off putting the speakers on the bridge of the desk and building some type of absorber to place on the sides of the desk surface? Or are the speakers best as-is? Any suggestions would be much appreciated!

[localhost127]

uou can utilize the Envelope time Curve (ETC) response (part of the free measurement suite: Room EQ Wizard) to measure how specular energy impedes the listening position (gain with respect to time) - of which you can then work backwards to trace which boundary or surface the early, high-gain indirect energy is indicent off of.

the ETC needs to be measured one speaker at a time.

you're going to have early-early (1-3ms) specular reflections from the desk as well as edge diffraction as high priority issues to address with the current layout.

[Glenn Kuras]

Originally Posted by localhost127

uou can utilize the Envelope time Curve (ETC) response (part of the free measurement suite: Room EQ Wizard) to measure how specular energy impedes the listening position (gain with respect to time) - of which you can then work backwards to trace which boundary or surface the early, high-gain indirect energy is indicent off of.

the ETC needs to be measured one speaker at a time.

you're going to have early-early (1-3ms) specular reflections from the desk as well as edge diffraction as high priority issues to address with the current layout.

+1
Hill,
If you would like to send us your measurements I can have Bryan work with you to try to figure out the best set up.

[DanDan]

The current height looks way low to me.
The ETC is specifically for reflection work. And REW is brilliant, as is FuzzMeasure, which is a bit more Mac -ish.
ETC can do with a bit of a hand.
If you see the suspected spike on the time line, insert one of your traps on the desk. Spike gone, that's the one.
A tape measure and piece of string will also be invaluable.

Comb filtering can show up the effect of these desk reflections very vividly.
Desktop Reflection Revealed

Not to forget the humble mirror. Lay a mirror on the desktop, place your speakers on the bridge. Move your head around in the area you will be actually listening. If you can't see the tweeter/speaker, a good sign.
If you are clear for HF, check out he LF response also. Sometimes the extra baffle extension caused by a desk can be welcome, but a null obviously not.

DD

[hillzz22]

Thanks all; my full featured trial of FuzzMeasure expired so I'm now working with a limited feature set but I think I can still get the full spectrum room response and ETC response. Unfortunately, I'm have a MacBook Pro and FireWire interface so REW is really not an option for me unless there is a new workaround with that combination...I'll try to take some measurements with FuzzMeasure and upload for your review.

I actually did use a large mirror and moved it around the desk - both on the lower working surface and the bridge - as I moved the speakers around in an attempt to find the "best" position. The reason I have the speakers where they are now is because I could not see most of either speaker from the listening position (in the mirror). I agree that the height of the speakers does need to come up a bit, so I'll try to raise them and measure the response.

Glenn - Thanks, I'll definitely forward the measurements on to Bryan; he and I have emailed back and forth a bit and he has been very helpful.

I am starting to wonder if I should modify the shape of the desk? Without having to completely start over, I could taper the large working surface so that it is more trapezoid/triangular, with the shortest part towards the listening position. I would lose some working space but would also potentially eliminate some reflection. Any suggestions here?

Thanks again for all the help.

[Sirmikeywest]

I honestly think that the monitors don't need to be spaced that far apart. They should be closer together so they are not out beyond the corners of the desk. Then they should be placed a little higher up, maybe another 8-9 inches or so. But that's just my opinion, and I'm not as well-versed in acoustical treatment techniques as some of the other guys on here......Best of luck!

[DanDan]

They do look wide to me too, that is one of the reasons I suggested on the Meter Bridge.
You could buy FM!
In any case REW works just fine on modern Macs.
Just use the internal sound I/O. It is of decent quality and certainly good enough for purpose.
If you don't have a stand alone mic pre, there may be insert points on your interface. The 'send' is usually a Direct Out from the pre.

If you don't have that, then you could try SoundFlower. I am told it works. I haven't tested it yet, and another layer of complexity when starting with the feature rich REW is unwanted. Thus I recommend the onboard.
Works absolutely fine on my iMac and MacPro.
Don't use the trackpad or scrollwheel zooming. Wild. Use the Graph Axis Limits, which REW helpfully remembers.

That linked thread shows how a small angle on the desktop can diminish the reflection enough.


DD

[localhost127]

Originally Posted by hillzz22

I actually did use a large mirror and moved it around the desk - both on the lower working surface and the bridge - as I moved the speakers around in an attempt to find the "best" position. The reason I have the speakers where they are now is because I could not see most of either speaker from the listening position (in the mirror). I agree that the height of the speakers does need to come up a bit, so I'll try to raise them and measure the response.

1 - sound has size (it is not a laser beam)
2 - a mirror is not a measurement tool
3 - edge diffraction generators

[hillzz22]

I have a FireStudio Mobile, so I may be able to use that with REW and SoundFlower. I'll have to give that a shot, especially since it seems that REW is more feature-rich and FuzzMeasure costs $150!

I moved the monitors up onto the bridge and oriented them to form the typical equilateral triangle position. The attached measurements were all taken with the microphone in the mixing position, roughly where my head would be. I'm actually surprised to see that the room response seemingly hasn't changed all that drastically from when it was untreated. Maybe the desk has significantly impacted the room? Or I need to move some panels around??

[localhost127]

zoom in on the ETC 0-50ms

[hillzz22]

Ahh, I was wondering if that might be helpful; the time scale did seem pretty long. I'll change the to 0-50ms and post a new file when I get home this evening. I'm reading as much as I can about how to read and use ETC response plots, but my understanding is still pretty limited so I could certainly use help in interpretation of the results!

[hillzz22]

OK, so attached are the L and R speaker ETCs to 50ms. I shifted the curve to the left by 2.63ms to cover the delay and hopefully make the rest of the delay times more absolute in reference. What does this tell me, exactly?

[localhost127]

the room looks quite dead.

you can identify boundaries incident of early, high-gain indirect specular energy as measured via the ETC by utilizing the simple yet elegant "string test" (thanks, SAC!)

http://www.gearslutz.com/board/6133764-post8.html

http://www.gearslutz.com/board/6397239-post80.html

[hillzz22]

Dead as in free of reflections (a good thing) or too dead (bad)? For some reason, I feel like the low frequencies are not as present and clear at the mixing position now that I have the desk in the room. Could it have changed the room response so significantly? Is there something that I can do with the panels that I have to address the nulls at ~65 Hz and ~200 Hz or do I simply need more traps?

[localhost127]

Originally Posted by hillzz22

Dead as in free of reflections (a good thing) or too dead (bad)? For some reason, I feel like the low frequencies are not as present and clear at the mixing position now that I have the desk in the room. Could it have changed the room response so significantly? Is there something that I can do with the panels that I have to address the nulls at ~65 Hz and ~200 Hz or do I simply need more traps?

dead as in the specular region. you still have early, high-gain indirect energies to address (as indicated in the ETC). did you see the string test links i linked to above? you need to trace out the energy spikes on the ETC to their incident boundary so you can address them. you can also remove any broadband absorption where there is no indirect early high-gain reflection such that you do not overdamp the room (hence, why one should measure and identify issues before blindly placing treatment).

regarding the LF/modal region ---- you have measuring equipment - utilize the waterfall plot (0-300hz) to identify freq response + LF decay times. apply trapping and adjust speaker/listener placement as needed.

[DanDan]

Quite of few of us recommend and use the very simple mirror technique, particularly when dealing with relative newbies.
Let's face it, someone knowledgeable on ETC techniques wouldn't need to ask the question in the first place.

Bass Waves are long, the desk will look tiny to them. Angling or thin absorption will have no effect. Removing the desk may or may not have any effect. Depends on the size of the desk. In my linked thread it had no LF effect. In the OP case I expect it will add some useful bass extension.

However the HF effect is strong and easily measureable by any of the methods. And easily fixed. The main effect that we can do something about is usually at HF, thus we focus on the tweeter. HF behaves in a much more ray like manner than LF.
Mirror and my Comb filtering example completely agreed.
I have never seen them not to.

DD

[localhost127]

Originally Posted by hillzz22

OK, so attached are the L and R speaker ETCs to 50ms. I shifted the curve to the left by 2.63ms to cover the delay and hopefully make the rest of the delay times more absolute in reference. What does this tell me, exactly?

my mistake, i did not catch this earlier but you need to uncheck the "set T=0" option. there is no reason to set the arrival of the direct signal to 0, we want the time of when the signal leaves the speaker (not arrives at the mic) to be 0ms.

we are interested in the total time of flight for the indirect specular reflections such that we can trace them back to the incident boundary.

[JoRillo]

DIY Monitor Stands (PVC Pipe)

I'm working on my second set right now using a slightly different design same basic concept. By the look of your stands now you could definitely use um.

[localhost127]

Originally Posted by DanDan

Quite of few of us recommend and use the very simple mirror technique, particularly when dealing with relative newbies.
Let's face it, someone knowledgeable on ETC techniques wouldn't need to ask the question in the first place.

yes - no point on detailing to a user how to properly measure to identify issues, and how to verify their treatment has sufficiently addressed said issues ... let's keep things dumbed down and not expect a novice to have to learn anything!

the mirror "technique" was originally a simple idea proposed to help novices understand how specular energy behaves in a space (angle of incidence = angle of reflection). the mirror is not a measurement tool. im not sure why people continue to recommend such a method when the mirror does not detail any information on indirect specular reflection gain or arrival time with respect to the direct signal.

he already has demonstrated knowledge of how to measure and generate an ETC response (which is the bulk of work and effort for a novice; setting up equipment and generating measurement data)...and now you want to go backwards and have him use a mirror to "guess" which boundary the indirect energy is incident from?

there is absolutely no reason to use a mirror in this case - and i find it laughable to suggest especially after the ETC has already been generated.

personally, i would remove all broadband specular absorbers from the room and take clean ETC responses of the untreated room (one speaker at a time), and then trace out (via the string test) which boundary or surface needs treatment. this way, broadband absorption is surgically placed vs blindly placed (as with the mirror) - which means you only place broadband absorption where it is needed, and not any more! ,,,that is, unless the user is looking for a dead/damped room?

SAC's string test will easily map out which boundary (or boundaries) correspond to the spike of energy on the graph. the beauty about the string test is it will also detail to you if there are 2 boundaries that are incident of early, high-gain indirect energy that have the same flight paths (thus, arrive at the same time on the ETC).

the mirror also isn't very useful for 2nd order early reflections, or any other "non-obvious" boundaries. the string test will allow one to easily map out 2nd order reflections.

im curious, DD, why you would recommend a method and "guess" at areas incident of early energy, vs using the generated data that he already has via his ETC responses - and tracing the ACTUAL high-gain early arriving reflection paths, vs the POSSIBLE paths.

do you have an issue with the string test merely because SAC was the one that presented the idea?

Originally Posted by DanDan

However the HF effect is strong and easily measureable by any of the methods. And easily fixed. The main effect that we can do something about is usually at HF, thus we focus on the tweeter. HF behaves in a much more ray like manner than LF.
Mirror and my Comb filtering example completely agreed.
I have never seen them not to.

DD

in the specular region within the context of this topic, we focus on the acoustic center of the speaker, not the tweeter.

not all first reflection points may be boundaries incident of early arriving, high-gain specular energy. if one uses the mirror, you're simply guessing/assuming all reflection points need treatment, versus properly measuring via the ETC to identify ACTUAL early reflection paths vs POSSIBLE paths via the mirror.

[hillzz22]

JoRillo - Yeah...please pay no mind to my very "official" speaker stands, haha. I have been using chairs, other speakers, etc to prop things up just so that I can get an idea of the placement and height that I want before I build anything. I really like the concept that you went with for stands and have plenty of wood left from the desk build. I was originally planning to build my own stands as well but if speaker placement works out best on the bridge, I guess I might not need to do so. Ideally, I'd love to have the speakers decoupled from the desk, but I'll let practicality dictate the end game.

localhost & DanDan - First, I want to thank you both for your patience and willingness to help me learn and address these issues. Second, I am new to studio design and acoustics but like to think I have learned quite a bit over the past few months; I find these concepts very interesting and hope that (within reason) individuals such as yourselves don't have to dumb things down TOO much! I am certainly up for the challenge, so as long as you're willing to offer your advice, experience and most importantly, your time, I'm more than willing to experiment and learn!

Anyways, I took the time offset back off the graph and examined the ETC a bit more closely. To start with, I focused in on 0-10ms as I figured this would capture most of the early/first reflections. Looking at the ETC for the L speaker, I see that the largest peaks are at 2.63s, 3.5s, 4.5s and 6s. The ETC for the R speaker coincides; at least I have symmetry, ha! My analysis would be the following...please correct me where I'm wrong:

- 2.63s is the original signal directly from the speaker to the listening position ~36" away.
- 3.5s corresponds to ~47" and the string test shows me that this is a reflection from the desktop surface.
- 4.5s, or ~60", is...hmm, I really can't find the source of this one. 5 ft seems to be in between any surface that I can identify using the string test. I'm stumped on this one.
- 6s, or ~80", is a reflection from the angled ceiling above the desk. I'm thinking that I can place two Gridfusor panels on this surface to scatter this sound.
- I don't see anything at ~8.5ms (115") or ~11ms 148", which would roughly correspond to the early reflection points on the side walls, so it appears that the GIK 242 panels that I have on the wall at those points are doing their job.

Am I drawing the right conclusions so far? I'm thinking that it might be worth removing all of the 242 panels, leaving the 244 bass traps and monster bass traps and taking the measurements again so that I have a good baseline with the desk in the room. Would that be worthwhile?

Also, I can no longer generate waterfall plots with FuzzMeasure since my free trial expired, so I'll either need to purchase the full version or get SoundFlower and REW to work together. Also, I understand how to read the waterfall plot but am not sure how to translate that into decisions regarding where panels should be placed or moved to. Perhaps that is another lesson for another day?

Thanks again for all the help; I think I'm slowly getting there.

[localhost127]

Originally Posted by hillzz22

Am I drawing the right conclusions so far? I'm thinking that it might be worth removing all of the 242 panels, leaving the 244 bass traps and monster bass traps and taking the measurements again so that I have a good baseline with the desk in the room. Would that be worthwhile?

i would say yes. that way you can measure the ACTUAL high gain early reflection paths and place absorption as necessary (surgically). remeasure after each placement and verify attenuation of early reflection.

again, you are only concerned with high-gain early indirect energy - much of it is already -25/-30dB down within 10ms so do not worry about those spikes.

[RJHollins]

Great thread !!

Thanks DanDan for helping to educate us new to acoustic measuring and analysis.

I'm soon at the point to re-shoot my room [again] ... but I struggle to understand what REW is showing and its' relevance ... and all to point to treatment solutions.

Also read the 'silent desk' thread, as I'm making my desk size plan.


What's been holding me up ... well ... it's hard to shut down the music since building my new room [designed by GLENN [gullfo] ]. I'm sure I'll need to tweak the treatment.

Again ... Big THANKS!!!

[darkbuddha]

Originally Posted by Sirmikeywest

I honestly think that the monitors don't need to be spaced that far apart. They should be closer together so they are not out beyond the corners of the desk. Then they should be placed a little higher up, maybe another 8-9 inches or so. But that's just my opinion, and I'm not as well-versed in acoustical treatment techniques as some of the other guys on here......Best of luck!

I have the Tannoy 601s (big brothers to the OP's 501s) and I've found that they're a bit more finicky/odd about spacing. With a narrower spacing between them (38" equilateral triangle) and aimed about 12" behind the head, the sweet spot was pretty isolated, the depth of the field was much shallower, and the stereo imaging wasn't defined enough (for my tastes). But by opening the up to 42-45" apart (non-equilateral), the sweet spot broadened, the stereo imaging has better clarity and definition, and most impressively, the depth of field increased immensely. I can now at least move my head back or forth several inches (like leaning back in my chair a bit, or hunched over my laptop) without losing the imaging.

[DanDan]

You are welcome RJ.

LocalHost. I don't normally use the word wrong, but in this case you are simply serially wrong.

I see no sign of the OP using ETC before our suggestions.

Far from decrying ETC or SAC or the weather:-

I supported your recommendation of the ETC and included SAC's string idea.
You clearly have no idea that I warmly complimented SAC he first brought it here, and I continue to do so.

SAC/foxfyr is gone due to an accumulation of infractions for anti social behaviour. Some are clearly still learning from him.

I then went on to recommend some simpler techniques which I found to be effective in real world testing for reflections from an actual desk.
That would be the Topic of this thread.

My experiments showed that comb filtering was the most visible and easily read view of the anomaly. It was easily capable of showing absence or presence of the desk, absorption on top, and the effect of very small angles and subtle changes of those. Seems good to me, again I recommend it.
The tests also showed no LF effect, so I checked between the mirror and combing again. Complete agreement. Tweeter vanishes, combing stops.

EDIT
Hillzz, you are also very welcome.
I found the Frequency Response view to be the BEST for desktop reflection analysis. Not the dumbest! I was only being polite in putting ETC first.
It shows the effect we are concerned about directly, but more importantly, vividly. This view also keeps a watch for LF issues. Try it. Desk with and without absorption or angle. Look at the Combing and the ETC. Take your pick.
I have never found a difference between combing and mirroring in this situation.


Optimising reflection control panels is another matter entirely.
The combing is not easily visible, the reflection should be.
The ETC is, as I said, designed for this.
e.g. Try changing the angle of a ZWR panel. The ETC will often reveal an advantage.

Finally, top tip. Using ETC. Move the mic towards a suspected reflection.
The corresponding spike should move on the ETC graph, confirming that you have the right reflection.

DD

[JLiRD808]

Anyone else having trouble viewing the attached .png's?

BTW....those panels are SEXYj

[localhost127]

Originally Posted by DanDan

I see no sign of the OP using ETC before our suggestions.

and how is that relevant? if he knew all of the troubleshooting techniques, why would he be on this forum asking questions?

he was able to generate ETC's without any further help from the community once the recommendation was made - and he presented ETCs by post #9. from here on out, there's no reason to use a mirror as offered by you when he has raw measurement data of which can be utilized to determine the ACTUAL indirect high-gain reflection paths, vs the POSSIBLE paths as with the mirror.

the bulk of the work is getting the user to setup measuring equipment and generate good data. the OP has proven that is not an issue - now it's just fine-tuning the understanding of the data presented within the ETC.

Originally Posted by DanDan

You clearly have no idea that I warmly complimented SAC he first brought it here, and I continue to do so.

funny you did not show the same warm complemnts when SAC was attempting to introduce the relevancy of the ETC response itself - those were fun threads to read.

Originally Posted by DanDan

I then went on to recommend some simpler techniques which I found to be effective in real world testing for reflections from an actual desk.
That would be the Topic of this thread.

the mirror may show an obvious reflection point on the desk surface, but in the OP's situation, early-early indirect energy via edge diffraction may also come into play - of which the mirror will not detail anything.

[DanDan]

The pngs are visible fine here. Quotes have ceased to work so from here on I am replying to localhosts last post.

It is only relevant in the sense that you brought it up.
Now that you have, It think it is significant that OP initially made no mention of ETC. If he was conversant with the technique he would hardly have asked the question in the first place.

If you go back to Jens first post on GS, you will find an immediate reply warmly welcoming him. Similarly in the SBIR thread which you three forced closed, you will find me warmly commending his sketching and mathematical work. You will not find any instance of my initiating bickerfests.
You might want to rethink this aggressive stance being presented towards me.

It has no good in it.

I took the trouble to present a test here. It also refers to a similar desk experience in a client studio.
In both cases I found the Frequency Response view to be the most useful.
It shows the actual anomaly in action. It also shows LF effects or not.

I have no doubt it would also show the result of Refraction. To be honest it hasn't concerned me much as the desk kinda has to be there.
If you have tests showing up diffraction from desks and some way of eliminating it, please do share.

As I said I have never seen the measurements, FR/Comb, or ETC, show up something at the desk which the Mirror didn't see.
In this case the Junior HackSaw and the Angle Grinder can be equally used to sharpen a pencil......

The combing technique on the other hand would be useless at optimising ZWR panels, especially fine tuning the angles as Lupo has shown. ETC rules for that.
Another topic for another thread perhaps.


DD

[localhost127]

what is ZWR? Zone Without Reflections? why not just call it NER - what is it with the recent efforts to manifest new terms lately?

Originally Posted by DanDan

Now that you have, It think it is significant that OP initially made no mention of ETC. If he was conversant with the technique he would hardly have asked the question in the first place.

of course the OP initially made no mention of the ETC - hence why i recommended a measuring tool to identify the ACTUAL boundaries incident of early energy, vs the mirror which only identifies POSSIBLE paths of large surfaces - and the mirror will NOT be of assistance to determine actual sources of edge diffraction.

i prefer to recommend a tool which details the actual behavior of the acoustical space, vs a mirror which does not provide any relevant data for the context of the topic (arrival time and gain).

Originally Posted by DanDan

In both cases I found the Frequency Response view to be the most useful.
It shows the actual anomaly in action.

please, DD - what information present within the interference pattern manifested in the FR is useful for identifying which sources/boundaries are responsible for such anomalies?

Originally Posted by DanDan

As I said I have never seen the measurements, FR/Comb, or ETC, show up something at the desk which the Mirror didn't see.

again, please indicate how you are using a mirror to identify ingress energies via edge diffraction (actual vs possible).

the problem with using the mirror is you are identifying POSSIBLE paths - not the ACTUAL paths. thus, by using the mirror, you will place broadband absorption at areas which may or may not be incident of early, high-gain indirect specular energy...which can quickly result in a dead room. is that the goal here? broadband absorption needs to be surgically placed as to limit the amount of absorption within the room (hence why splayed walls for redirection are preferred over broadband absorption). by identifying (via a measurement tool and not a mirror) which boundaries or sources are responsible for early arriving high-gain incident energy, they can be attacked directly and other reflection points which are not incident of such energies do not need to be treated! with the mirror, you are applying treatments blindly without first identifying whether any "problem" exists at that boundary in the first place!

that is, unless the OP is looking for a completely damped/dead room? if so, then the mirror still isnt necessary to create a dead room.

edit: and this doesnt even begin to take into account the speaker's off-axis response, and whether any energy is even being dispersed to the side-walls of which would require treatment! hence - measurements over blind guesswork.

[DanDan]

Zone Without Reflections, a term to distinguish from the patented RFZ, part of the LEDE. Jens and I like to use it, our prerogative. Jens added an E
Zone Without Early Reflections. ZWER.
What is your purpose in objecting to these terms which we like to use for clarity in our expression?

Again, the OP question was about the desktop reflection. The issue being possible comb filtering. The Frequency Response graph shows this ACTUAL combing directly.
Again, changing the angle, removing the desk, covering it with absorption, very all very visible on the FR graph. I am afraid I can't make this any clearer. I can only assume you haven't looked at the linked thread Desktop Reflection Revealed. Unless you do, there is no point in blind meaningless banter.

If diffraction is causing a significant anomaly, it will appear on the Frequency Response graph also.

Similarly if the Speaker Off Axis Response (SOAR) is causing a significant anomaly, it will appear on the Frequency Response graph.

DD

[localhost127]

Originally Posted by DanDan

Zone Without Reflections, a term to distinguish from the patented RFZ, part of the LEDE. Jens and I like to use it, our prerogative. Jens added an E
Zone Without Early Reflections. ZWER.
What is your purpose in objecting to these terms which we like to use for clarity in our expression?

added confusion - especially to the novice.
ZWR = zone without reflections = anechoic room

the crucial details are missing in such terminology - as one is not concerned with all reflections if implementing such concepts, just those that are high-gain and early (as defined by the user's requirements).

Originally Posted by DanDan

Again, the OP question was about the desktop reflection. The issue being possible comb filtering. The Frequency Response graph shows this ACTUAL combing directly.
Again, changing the angle, removing the desk, covering it with absorption, very all very visible on the FR graph. I am afraid I can't make this any clearer.

here is my question to you again, DD - since you ignored it in my last post:

Originally Posted by localhost127

please, DD - what information present within the interference pattern manifested in the FR is useful for identifying which sources/boundaries are responsible for such anomalies?

comb-filter in the frequency response details you nothing other than that you (or the mic) is located in an interference pattern (direct/indirect signal superposition). it doesn't detail to you anything about what source(s) are contributing to the constructive/destructive interference. isn't the point to identify which source is incident of the indirect reflection of which is causing the comb-filtering in the first place? thus, treatment can be applied such that the high-gain indirect reflection is no longer superposing at the listening position with the direct signal, eliminating the constructive/destructive interference for that instance.

so, instead of utilizing a tool to identify which source (boundary) is responsible for the high-gain indirect signal that results in the comb-filter response - you instead wish to blindly guess and shoot in the dark about which source is the contributing factor. the frequency response tells you you have a problem (interference) - but it is useless in solving such a problem in less one wishes to blindly apply broadband absorption at a reflection point, remeasure, and move on to the next reflection point. why not measure the ACTUAL high-gain early reflections that are actually responsible for such FR anomalies, and work from there?

Originally Posted by DanDan

If diffraction is causing a significant anomaly, it will appear on the Frequency Response graph also.

this is madness. how on earth are you going to determine anything from the frequency response regarding what indirect source is responsible for said anomaly (comb-filtering)? how do you even know whether diffraction is causing the interference, or an indirect reflection from another boundary, or both? hence, why the frequency domain is not the correct domain to work in for such a scenario.

again - the frequency response tells you zero information about what boundary or source is responsible for the incident indirect signal that is causing the interference.

this is why the time-domain takes precedence - to view the ACTUAL indirect high-gain reflection path, not the POSSIBLE.

Originally Posted by DanDan

Similarly if the Speaker Off Axis Response (SOAR) is causing a significant anomaly, it will appear on the Frequency Response graph.

DD

identifying if such constructive/destructive interference exists is one thing - identifying what source or boundary is responsible for such superposing is another (and the scope of my commentary here).

the frequency domain is the entirely wrong domain to be in for identifying what boundaries or sources are directly responsible for such anomalies in the FR! so this is what it's like to be in Flatland.