John Sayers' Design Forum

John Sayers' Recording Studio Design Forum

A World of Experience
Click Here for Information on John's Services
It is currently Thu Jun 20, 2019 4:13 pm

All times are UTC + 10 hours




Post new topic Reply to topic  [ 9 posts ] 
Author Message
PostPosted: Fri Jun 14, 2019 11:09 am 
Offline

Joined: Thu Jun 13, 2019 11:48 am
Posts: 5
Location: CT
After trying at the "other place" and not getting much help, I wanted to try here :)

First off, thanks much to Soundman2020 for the GREAT tut on this forum:

So, I followed that to the letter. Using a DBX RTA-M, I first cal'd the soundcard, as per instructions. Then, using an SPL, got my levels to 80db. Again, following all the instructions in that thread. The mic was where my head is, at the proper angle, and EXACTLY equidistant from the side walls, and EXACTLY equidistant from the tweeters.

A BIT about my room. It is not ideal, of course, being a bedroom in my apt; it's 10.5 x 10.8 with an 8ft. ceiling. All of the walls are perpendicular, and the room would be a perfect square, if not for those xtra 3". Typical construction - I think it's called A-frame? Sheetrock walls, wood/sheetrock floors with joists (no concrete between me and my downstairs neighbor), brick exterior. Empty space (attic) above me. I have done a good deal of treatment, including some diffusors. I can not put Bass Traps in corners (at least not ones that stand on the floor). Obvious why I can't on the south side. On the north side, I have a w/w radiator that sticks out about 3". The window is recessed 4", and is completely covered with heavy curtains. There is w/w carpeting. I have placed 2x2' 3" thick Sonex foam pads on the sides and rear of my desk. Why? Because I could, and because I felt like it. Oh, and because Tom Hidley said so.

I have a pair of Dutch and Dutch 8Cs. Both monitors are 13" from the back wall (tho angled inward, I measured from the middle of the monitor), and 30" from the side walls. They are 45" from the floor, and 32" from the ceiling. I have put a fair amount of effort into making absolutely sure that both mons are EXACTLY equidistant from the walls, floor, and ceiling, that both are angled at EXACTLY the same amt of degrees, and that both tweeters are EXACTLY the same distance from the front (south) wall. I do not use any RCSW.

I took some measurements, and after not being completely satisfied with what I saw (and TBH, not really sure what I was looking at), I requested assistance from D&D. They were happy to help. After taking a bunch of new REW measurements (in a "mesh", as it was called), I was aided in setting up the filters in the mons; what you see in the files below is the end result.

I certainly can't complain about the way everything now sounds, but I tend to be a perfectionist when it comes to certain things, and I am thinking, what reason do I have to not give this a shot?

My main concern are those dips @ 103, 233, and 422. I was told that correcting upwards is not the proper thing to do, so I am wondering if I should try to do something in my room here. I am very limited as to what I can do; I certainly can not do anything that is construction related. I guess the first Q I have is whether or not there is a proper method to figure out where in the room (or what) is causing that dip? And then assuming I find the problem area(s), what to do? And of course, any other comments or suggestions are welcome.

Thanks in advance.

Attachment:
room2.jpg


http://jeffreyhayat.com/temp/6-9-test.mdat


You do not have the required permissions to view the files attached to this post.


Last edited by riffwraith on Fri Jun 14, 2019 11:59 am, edited 2 times in total.

Report this post
Top
 Profile  
 
PostPosted: Fri Jun 14, 2019 11:13 am 
Offline
Site Admin
User avatar

Joined: Thu Aug 21, 2008 10:17 am
Posts: 11926
Location: Santiago, Chile
Quote:
After trying at the "other place" and not getting much help, I wanted to try here
Ahhh! So you have abandoned the dark side of the force, and come around to the bright side! :)

Hi there "riffwraith", and welcome. HPlease read the forum rules for posting (click here). You seem to be missing a couple of things! :)

I'll download your MDAT and take a look, before I respond in more detail.... :)


- Stuart -

_________________
I want this studio to amaze people. "That'll do" doesn't amaze people.


Report this post
Top
 Profile  
 
PostPosted: Fri Jun 14, 2019 11:58 am 
Offline

Joined: Thu Jun 13, 2019 11:48 am
Posts: 5
Location: CT
Ok, filled in my loc, and resized the pic. Anything else? :)


Report this post
Top
 Profile  
 
 Post subject:
PostPosted: Fri Jun 14, 2019 2:04 pm 
Offline
Site Admin
User avatar

Joined: Thu Aug 21, 2008 10:17 am
Posts: 11926
Location: Santiago, Chile
... OK, I looked at the data, and overall, your room response is not bad at all! Congratulations.

Clearly, you are using some type of EQ: it would be good to see the "raw" results, without the EQ applied, to see what's really happening in the room itself. Photos of the actual room would be useful, too.

So, general comments:

Quote:
A BIT about my room. It is not ideal, of course, being a bedroom in my apt; it's 10.5 x 10.8
Ouch! Nearly square section...

Quote:
I have done a good deal of treatment, including some diffusors.
You might have jumped the gun with the diffusers... Numeric-sequence diffusers should be a least 10 feet away from your ears, due to the lobing they create. It's possible (but not certain) that some of the issues you are seeing in your REW data are related to that. What type of diffusers did you use, where are they located in the room, and what frequency range are they tuned to?

Quote:
There is w/w carpeting.
That would explain the dull high-end. Fortunately, it isn't too serious, since resolving that would imply that you either remove the carpet, or cover it with a hard, solid, reflective surface. I'm guessing that neither of those are attractive options! :)

Quote:
I have placed 2x2' 3" thick Sonex foam pads on the sides and rear of my desk. Why? Because I could, and because I felt like it. Oh, and because Tom Hidley said so.
That's good for helping to control early reflections, but it doesn't seem to be working fully:

Attachment:
riffwraith--REW--IR--LIN--27ms--R.png

You have quite a bunch of fairly strong early reflections, within the first 15 ms or so. That's your right speaker: the left is similar, but a little lower in level, and slightly different timing. It would be good to identify where those are coming from.

Also, the overall decay isn't as smooth as it could be:
Attachment:
riffwraith--REW--IR--200ms.png

You can see that from the Schroeder integral. Ideally, that would be a straight diagonal line sloping down to the right. The bumps and dents and curving indicate non-smooth decay. Once again, it's not bad in your case, but there's room for improvement.

Quote:
I can not put Bass Traps in corners (at least not ones that stand on the floor).
Most studio builders don't realize at first that there are twelve corners in a rectangular room, not just four. All of those are fair game for bass traps. Where do you have bass trapping right now?

Quote:
Both monitors are 13" from the back wall
Semantics: I think you mean the front wall, not the back wall! In a studio, the front wall is the one in front of you: the one you are facing when seated at the mix position. The back wall is the one behind you. It's important to use the right orientation, to avoid confusion.

Quote:
I do not use any RCSW
Reinforced Concrete Shear Wall? Registered Clinical Social Worker? Royal Commission on the Status of Women? Robot Chicken Star Wars? I'm not familiar with the acronym, and those four were the top ten that I found when I googled it. I could not find anything related to speakers or room acoustics at first glance: I guess I didn't look too well. Maybe you could explain that?

Quote:
Both monitors are 13" from the back wall (tho angled inward, I measured from the middle of the monitor),
When you say that they are 13" from the wall, I assume that you mean that the middle of the REAR face of the speaker is 13" from the wall. Which implies that the front face is about 28" from the front wall of the room. Therefore you can expect SBIR somewhere around 100 Hz. Which is evident in your REW data... You would be better off placing your speakers tight up against the front wall, with just a 4" gap between the rear corner and the front wall. That would move the SBIR dip up in frequency, to around 210 Hz or so, where it is less of a problem, and possibly more treatable.

Quote:
and 30" from the side walls. They are 45" from the floor, and 32" from the ceiling.
Jest checking here: you are referring to the acoustic axis of the monitor with those measurements, correct? Not the top, bottom or sides of the cabinet, nor the location of the tweeter, nor the geometric center of the front baffle? All measurements to speakers in the room should refer to the acoustic axis.

Also, why did you chose those dimensions? I mean, why did you decide on 30" from the side walls, rather than, say 28" or 32", or something else? And why 13" from the front wall?

Quote:
I have put a fair amount of effort into making absolutely sure that both mons are EXACTLY equidistant from the walls, floor, and ceiling, that both are angled at EXACTLY the same amt of degrees,
What is the angle that you used here? Also, where is the aim point? In other words, if you extend an imaginary line along the acoustic axis of both speakers, where do those axes cross over each other, in relation to your head? What is the angle of intersection of those lines where they meet? And what is the location of your head in the room?

Quote:
I took some measurements, and after not being completely satisfied with what I saw (and TBH, not really sure what I was looking at), I requested assistance from D&D. They were happy to help. After taking a bunch of new REW measurements (in a "mesh", as it was called), I was aided in setting up the filters in the mons; what you see in the files below is the end result.
The combined L+R isn't too bad, and each of them taken individually isn't too bad, but when you compare the left channel and right channel, there's some major differences:

Attachment:
riffwraith--REW--FR--Speaker-Difference-Compare--12--20k--1..24.png

That curve shows the difference between your two channels: wherever the curve is above the Zero line, your left speaker is louder at that frequency. Wherever it is below the Zero line, your right speaker is louder at that frequency. For a typical home studio with reasonable speakers and reasonably good treatment, that should be around +/- 5 or 6 dB. Good speakers in a good room should be around +/- 3dB. Great speakers in great room should be +/- 1dB. Here's an example: viewtopic.php?f=2&t=20471 . In your case, it is about +12/-7 dB. In other words: not so good. Since you took great care getting them set up accurately and symmetrically, and since the speakers themselves (with no EQ applied) should be fairly similar to each other, the logical conclusion is that there's something in your signal chain, or in the room itself, that is not symmetrical.

You can see the issue more clearly when directly comparing both frequency response curves:
Attachment:
riffwraith--REW--FR--Speaker-Difference-Compare--Both-Curves--12--20k--1..12.png

Red is left speaker, green is right speaker. (In this case, smoothed to 1/12 for clarity: the actual excursions are higher than appear in this graph). Some of that looks like incorrectly applied EQ...(around 50 Hz, for example)... Other parts are probably room issues.

Quote:
My main concern are those dips @ 103, 233, and 422.
103 is very likely SBIR, due to the layout of your speakers in the room. It might also be your 0.2.0 and 2.0.0 axial modes, which are practically right on top of each other at around 105 to 107 Hz. Or it might be a combination of SBIR and modal, which makes it hard to identify, but not impossible. Your 233 Hz dip is actually at two different frequencies: 236 on the left channel, and 227 on the right channel. Likely a room symmetry problem. 422 (actually it's rather broad: around 390 - 500) is probably not too much of a issue. It's in the mid range, so probably not too audible. And it is probably fixable.

Quote:
I was told that correcting upwards is not the proper thing to do,
Usually correct, but it depends. If those are phase cancellations or modal, then absolutely: you can't fix those with EQ. Don't even try: You'll just make matters worse, even if you do manage to get the line flatter. But if they really are frequency response issues in the speakers, signal chain, or room, then it MIGHT be possible to tweak them a tad, if you know what you are doing. I normally tell people the same as you already heard: always cut, never boost. But if you fully understand the issue, and the consequences, it can be beneficial to boost some things, as long as you do it properly.

Quote:
so I am wondering if I should try to do something in my room here.
That depends on you, really! If you think your room sounds fantastic as it is right now, and you can turn out mix after mix that translate perfectly every time, then there's no reason to change things. But if you have concerns, or your mixes are not translating, or you just think the room sounds a bit "off", then fix it! :)

If you really do want to get your room as good as it can be, by fixing those issues (and the others), then I would suggest doing a full set of room tests with REW, starting out with the theoretical correct positions and orientations for the speakers and mix position, and working from there, with multiple tests done with small increments in the locations of the speakers and mix position. That allows you to find the optimum locations. You do that WITHOUT EQ applied, since EQ is only valid for one specific layout in any case (one specific speaker location, and one specific mix position location), then you treat what needs to be treated in the room, then you do the final digital tuning. This is a long, slow, frustrating, mind-numbingly boring, procedure. I do it with my clients who really do want to tune their rooms to the hilt (here's another example, from just a couple of months ago: viewtopic.php?f=10&t=21539 That's the type of result you could expect if you wanted to take your room to the extreme. But even if you didn't want to go that far, there's still room for improvement over what you have now.

One thing you should be aware of with using EQ to "fix" a room, is that it is actually physically impossible to do that. Don't get me started on that subject! :) You cannot correct a room with EQ. You can only correct frequency response in the signal chain, but there is nothing at all that EQ can do to correct things like SBIR, modal response, strong reflections, phase cancellation, comb filtering, etc. Those are all acoustic issues, not signal issues. They happen AFTER the sound wave has already left the speaker, and are due to that sound wave hitting various surfaces in the room and being reflected, transmitted, absorbed, and diffused to varying degrees by each surface it strikes. EQ cannot do anything for those. That's something that manufacturers of those hardware and software products conveniently fail to mention whey they try to convince you to use them. Those products are ONLY useful in a room that has already been treated as much as possible, to eliminate or greatly reduce those problems, leaving only the actual frequency problems... which EQ can fix. So don't try to use EQ to fix room problems: first, treat the room suitably, THEN you can use EQ to iron out the last details...


- Stuart .


You do not have the required permissions to view the files attached to this post.

_________________
I want this studio to amaze people. "That'll do" doesn't amaze people.


Report this post
Top
 Profile  
 
PostPosted: Sun Jun 16, 2019 2:51 am 
Offline

Joined: Thu Jun 13, 2019 11:48 am
Posts: 5
Location: CT
Holy crap dude! I was hoping to get a detailed answer, but dayum! Thanks so much. :)

So first, sorry about wall confusion; I thought the wall facing the listener is the back wall. I have updated my pic with the diffusor location, and denominating each wall, to avoid further confusion.

Secondly - I am not using an EQ. PCIe sound card SPDIF > Avocet > 8Cs (XLR). That's it.

Next - why did I choose that placement for desk and speaker? No speacial reason; I just picked a spot. The desk used to house a pair of Yamaha HS8s (and then Focal twins), and I placed the desk with the idea that the speakers should be close to the wall, but not right on the wall. I did, however, want to make sure the speakers were equidistant from the side walls. The desk can be moved if need be.

Next, the new pic has the location of the diffusors. They are from GIK acoustics... wood w/fabric. What is meant by "tuning" the diffusors? How do you tune them?

I can not remove the carpet. I am not sure about adding a hardwood surface - I would have to look into that, to see how it might be done. Can you add a hw surface onto relatively soft carpeting? If it can be done, I will give it a shot.

I am assuming the difference in the early reflections is due to the fact that the window is there behind the R speaker? I can also futz with those Sonex foam pads - either remove them (seems like that would not be wise), add more, or....?

I have no bass trapping right now. Not if we are talking about actual bass traps in corners. There is nothing currently in any corner, top or bottom.

RCSW = room correction software. Again, I have none. :)

Ok, let's get to speaker position.

When you say that they are 13" from the wall, I assume that you mean that the middle of the REAR face of the speaker is 13" from the wall. Which implies that the front face is about 28" from the front wall of the room.

Yes. I can move the desk back, so the speakers are closer to the wall.

In terms of your next comment - I am not familiar with the term "acoustic axis". So yes, when I gave those measurements earlier, I was talking the top, bottom and sides of the cabinet, and the location of the tweeter, all measured with a tape measure. As for the choices, I didn't choose the side wall distance; that happened naturally. The top to bottom I didn't choose either; it just wound up that way when I placed the speakers on my desk.

As for angle, I am not good with geometry ...LOL... the backs of the speakers are 25" from the side wall(s) and the front 33". Not sure what the formula is to calc that in degrees. The aim point is the middle of my head (proper triangle here). My nose is 45" from each tweeter, 55" from the floor, 65" from the front wall, and 64" from the side walls.

If you really do want to get your room as good as it can be...

Well, yes, that's the goal... until I buy a house. :)

...by fixing those issues...

Outside of moving the speakers back, what else is there for me to do, before I go and take another set of measurements as you outlined?

Oh, and BTW - and this may be obvious.... please understand that tho the measurements are accurate, the drawings are not to scale.

Thank you!

Attachment:
room3.jpg


Attachment:
room5.jpg


You do not have the required permissions to view the files attached to this post.


Report this post
Top
 Profile  
 
PostPosted: Sun Jun 16, 2019 5:37 am 
Offline
Site Admin
User avatar

Joined: Thu Aug 21, 2008 10:17 am
Posts: 11926
Location: Santiago, Chile
Quote:
Holy crap dude! I was hoping to get a detailed answer, but dayum! Thanks so much
:thu: You are very welcome! That's what the forum is all about.

Quote:
Secondly - I am not using an EQ. PCIe sound card SPDIF > Avocet > 8Cs (XLR). That's it.
Ummm... yes you are! You said so yourself: " After taking a bunch of new REW measurements (in a "mesh", as it was called), I was aided in setting up the filters in the mons; ". Your speakers have built-in parametric equalization: that's what you used, and it is very evident in the REW graphs you posted.

Quote:
Next - why did I choose that placement for desk and speaker? No speacial reason; I just picked a spot. The desk used to house a pair of Yamaha HS8s (and then Focal twins), and I placed the desk with the idea that the speakers should be close to the wall, but not right on the wall. I did, however, want to make sure the speakers were equidistant from the side walls. The desk can be moved if need be.
It's not the desk that needs moving: it's the speakers. Speakers should never be on top of a desk, nor on a meter bridge, because that invariable creates very uneven response in the mid range, due the reflections from the desk surface and also the interference between those reflections and the direct sound. There might also be other issues, such as vibrations induced on the desk itself, and the resultant possible early-early sound (which arrives at your ears BEFORE the direct sound from the speakers). There is some indication of that uneven response in your REW data. There's no indication of early-early sound in your case, but there might be some resonance and vibration in there.

Speakers should be on stands beyond the desk, up against the front wall, not on the desk. Yes, you do see photos of supposedly high-end studios with speakers on the meter bridge... but that doesn't mean that it is the right way to set up speakers! There's plenty of direct evidence from laboratory testing, as well as real-world empirical data, showing what a bad idea it is, and the what a mess it makes of speaker response.

Next: The reason why the speakers need to NOT be far away from the front wall, is something called "SBIR". That stands for "Speaker Boundary Interference Response". It's a fancy technical term that basically means that the low frequency sound waves leaving your speaker head off in BOTH directions: some towards you, and some going the other way, hitting the front wall, then bouncing back... and interfering with the copies of themselves that already headed your way.

Some people find it easier to picture this as a "Mirror image" speaker: Instead of a reflection from the wall, imagine that there's another identical speaker exactly the same distance behind the wall, then imagine that the wall has gone, so all you have is the two speakers producing the same sound. If you could see the sound waves, this is what you would have:
Attachment:
sbir-interference-patterns-mirror.jpg

You can see how the waves interfere with each other, creating dips and peaks in the frequency response: That pattern of dips and peaks is different
for different locations in the room, and different frequencies.

Obviously, there's a slight delay between the two waves: the wave that took the longer path, bouncing off the front wall then coming back at you, is delayed slightly with respect to the wave that came at you directly. A time delay implies a phase shift. So the "bounced" wave is out of phase with the "direct" wave, and that interference between them causes a massive dip in the frequency response for all frequencies where the delay happens to be a phase shift of 180°. So there's a massive dip at some very low frequency directly related to the distance between the wall and the speaker, then a series of smaller dips, that repeat in a regular pattern all the way up the spectrum: this is called "comb filtering".

This is what the actual frequency response looks like for comb filtering:

Attachment:
SBIR-effect-dip-and-comb-filtering-frequency-response-graph.jpg

There are equations for calculating that.

If the room is large enough, then you can position your speakers far enough away from the front wall that the first huge dip occurs at a frequency below the audio spectrum, so you'd be fine. That distance is about 4 meters! So if you have a room about 20 meters long, then you can do that.... But most home studios are nowhere near long enough to permit that. The second-best option is to put your speakers tight up against the front wall, which forces the first SBIR dip up into the lower end of the mid range, where it is less audible, and can probably be treated.

The best option of all is to flush mount your speakers IN the front wall, such that the front baffle of the speaker is level with the wall surface: in that case, there is no SBIR at all from the front wall! Because the speakers are no longer in the room, so there's no interference pattern, and no SBIR. This is sometimes called "soffit mounting" and is the single best thing you can do for a control room: it solves so many problems all at once, not just SBIR.

Looonnngggg explanation to show the reason for the suggestion: Get your speakers close to the front wall, off the desk, on stands.

Then, set up the desk at the correct location in the room. Or rather, set up your chair at the correct location, then place the desk in front of your chair, so that you can operate your gear comfortably.

Quote:
Next, the new pic has the location of the diffusors. They are from GIK acoustics... wood w/fabric. What is meant by "tuning" the diffusors. How do you tune them?
You can't tune them after you buy the: you buy them already tuned. There's a certain frequency range that any diffuser affects. For Schroeder type diffusers (which includes Skylines), the depth of the well determines the lowest frequency, and the width of the well determines the highest frequency. Below the lowest frequency, there's no diffusion. There's probably scattering for about an octave below, but not real diffusion. And above the highest frequency, there's also no diffusion. You only get diffusion in the range that the device is tuned to, around it's "target frequency". So you should first determine what frequency range your room needs diffusion for, then buy the right diffuser that treats that frequency range.

However, as I already mentioned, there's a minimum distance from the diffuser to your head. The reason for that is because of the way numeric-sequence diffusers work. The mess with the timing and phase of the sound. Some parts of the wave go down one of the wells, bounce off the bottom, and come back up... with a time delay (because of te extra distance for the "down and back" path), and that implies a phase change. But the well right next door has a different depth, and therefore a different phase change. The interference patterns between the various phase changes causes the reflections from each well to go off in a different direction, with different timing and different phase: ie, diffusion. But close up, those time, phase, and level differences are audible: if you move your head just a bit, you hear a DIFFERENT set of time, phase, and level differences: the diffusion is not smooth close up. It needs space for all those things to merge together again, into a coherent but diffuse sound field. The minimum distance is about ten feet (3m), but if the diffuesr is tuned low then you might need an even larger distance.

If you could see the sound waves after they come off the diffuser, here's what it would look like:
Attachment:
QRD-Diffusion-lobing--pattern-graph-SML-ENH-2.PNG

As you can see the "lobing" patterns are very intense close up, and it's only at a reasonable distance that you actually get a smoothly diffuse field. Thus, if your head is closer to the diffuser than the minimum distance, then even moving your head slightly places your ears in a different part of that field, so you are hearing the lobing patterns, not the diffuse sound field.

Quote:
I can not remove the carpet. I am not sure about adding a hardwood surface - I would have to look into that, to see how it might be done.
The effect in your room is not too terrible: just a bit of roll-off in the high end, so I wouldn't worry too much about that.

Quote:
I am assuming the difference in the early reflections is due to the fact that the window is there behind the l speaker?
It's possible, yes. Or it might be something else in the room that isn't symmetrical. For example, the closet on only one side of your rear wall... Or stuff on your desk that isn't symmetrical. It might even be in the structure of the room, such as a different type of framing for the side walls, or a different thickness of drywall, or different insulation. That's not so likely, but it is possible.

Quote:
I have no bass trapping right now. Not if we are talking about actual bass traps in corners. There is nothing currently in any corner, top or bottom.
What treatment do you have in there, apart from the diffusers?

Quote:
RCSW = room correction software. Again, I have none.
Ahh, but you do! That's what the parametric filters in your speakers are doing... EQ is still EQ, no matter what fancy name the speaker manufacturer might decide to give it. "DSP" seems to be a common disguise these days: but at the end of the day, it's still just EQ....

Quote:
Yes. I can move the desk back, so the speakers are closer to the wall.
Again, don't mode the desk: move the speakers off the desk, onto stands. Then set the desk in the correct location for the mix position, not for the speakers. You have to sit at the appropriate location in the room, where you get the best acoustic response, then you move the desk so you can reach everything on it comfortably. The desk location is dictated by where your ears have to be.

Quote:
In terms of your next comment - I am not familiar with the term "acoustic axis".
The acoustic axis of any speaker is the spot on the front baffle where all of the sound "seems" to come from. The manufacturer normally shows that in the documentation for the speaker, or on their website. For a two-way speaker, it's about half way along the imaginary line that joins the center of the tweeter to the center of the woofer. For a 3-way speaker, it's more complex. But ask the manufacturer. They should be able to tell you that the acoustic axis is located on the front face XX cm from the left side of the cabinet, and YY cm from the bottom- Some manufacturers even mark it directly: Eve Audio, for example, positions their logo to mark the acoustic axis. Genelec publishes one single booklet that shows all of the acoustic axes for all of their speakers.

Actually, all of the above is about the acoustic center of the speaker: The acoustic axis is an imaginary line that extends out from the front of the speaker, starting at that point, and perpendicular to the speaker face.

Once you find the location of the acoustic axis, your speakers need to be set up with that more or less at your ear height, or a bit higher, and the speakers should be toed in a little, such that the acoustic axis points slightly outboard of the tips of your ears. Thus, the axes from the two speakers will meet each other, intersecting, about 12" to 18" behind your head. That's the theoretical ideal situation.

Quote:
The top to bottom I didn't choose either; it just wound up that way when I placed the speakers on my desk.
Yet another reason to get them off your desk! :) There's an optimal setup for the speakers, the mix position, and the room. If you don't have things set up in that optimal arrangement, then you don't have the smoothest, cleanest, flattest acoustic response.

Quote:
The aim point is the middle of my head (proper triangle here).
Ahh yes... the infamous "equilateral triangle" strikes again... Don't get me started on that! :) Too late! You already got me started! ...

OK, here's the thing: I'm sure you already know that the very best sound from your speaker is "on-axis": when the speakers are pointing at your ears: Because the further "off-axis" you go, the less even, smooth, and clear the sound gets. At 90° off axis, it's a totally mushy disaster, for any speaker, and at 180° off axis, al you get is deep rumbling mud. The cleanest sound is on-axis. It's logical. Now, if you set up your speakers so that they are pointing at the middle of your head, then the acoustic axis is NOT aimed at your ears! :shock: Rather, it is aimed at your EYES! 8) So, for all people who have had their ears surgically transplanted onto their eyeballs, the equilateral triangle is just perfect... :) But for the rest of us, with our ears sticking out from the sides of our heads, many inches away from our eyes, the triangle isn't any use at all. Instead, the speakers need to be toed out a little less, so that they are aimed at your ears, not your eyes. In fact, the science of psycho-acoustics shows that it's best to have the axis pointing just a little past the outer edge of your pinna... The "pinna" is the medical name for the fleshy flappy thing that sticks out from your head, with all those folds and wrinkle in it: what most people would just call "ear", but doctors love to invent special names for all body parts, so the called that the "pinna".

So, forget about the equilateral triangle diagrams that you see in many books and internet sites, showing lines coming out of the speakers and meeting in the middle of your head. Rather, pay attention to the more correct diagrams that show the lines coming out the speaker and going past the side of your head, meeting a bit behind your head. Those are the correct ones. That's how you set up the speakers and mix position for the best sound.

But there's one factor missing here: where to put the speakers "side to side". In other words, how far from the side walls, and how far between them. There are some equations for figuring that out too, and one of them just says "About 27% to 30% of the room width": That's where you set up the acoustic axis of your speaker. Call it 28%. So for example, if your room is 10 feet wide, that's 120 inches, so you would set up each speaker at 120 x 0.28= 33.6" from the side walls. And you would then have a distance of about 52.8" between the speakers (acoustic center to acoustic center).

And one other missing factor: where to put your head in the room! There's an interesting and much-debated guideline that says the supposedly best location is 38% of the room depth: 38% of the distance from the front wall to the back wall. So if your room was 10 feet long, then according to this guideline your ears should be about 120 x 0.38 = 45.6" from the front wall. In theory, that's the spot that has the least modal issues in any room. In practice, most engineers seem to prefer being just a couple of inches further forward, so call it maybe 43", or about 110 cm (do the math for your own room! I'm just giving hypothetical dimensions as an example).

So, for the hypothetical room here, the speakers would go with the acoustic axis 48" above the floor and 33.6" from the side walls, with the rear corner 4" from the front wall, and the mix position (engineers ears) set up 43" from the front wall, 47" above the floor, on the room center-line, and the speakers toed in so they both point at a spot about 18" behind his ears. That's the theoretical optimum layout.

Notice I didn't mention any angles there? The famous and sacred "equilateral triangle" insists that the toe-in angle must be exactly 30.00000°, with the engineer's head siting on the tip of that triangle, and the acoustic axes piercing his eyes.... That's silly. Because its the ROOM that dictates where the speakers and mix position go, not an imaginary geometric figure! If you do this setup the way I mentioned it above, you'll find that the angles will NOT be exactly 30°. They will be close, but not exact. Maybe 27°. Maybe 32°. Maybe more. Maybe less. And .... who cares? 30° is a nice figure that is easy to print in the manual, and in fact does work for most rooms... but is not OPTIMAL for ANY room!

That's the reason why you see the infamous triangle in so many places: it works for most cases. But isn't the best for any case. The best is to follow acoustic theory to arrive at the right locations, then tweak that based on actual testing with REW and with your own ears. Theory is one thing, and the real world is another. Every room is different, and in reality there's no such thing as "one size fits all". What the above will do is to get you to a good starting point, which will be much better already than that "triangle", but still probably won't be the best possible. You then do a series of tests with REW, moving the speakers and mix position in a set of small increments, until you find the absolute best spot. In other words, you move the mic forwards and backwards about a couple of feet each way, in steps of 2", and see where the best spot is (smoothest response). Then you move the speakers closer together and further apart in steps of 2" to see what's the best spot for THEM. Repeat as often as you need, until you fine "The Best Possible Layout That Cannot Be Beaten"! It's a slow, boring process, but if you really want to optimize your room, that's the way you do it.

So, in summary: First, turn off the EQ (parametric filtering) in your speakers, so we can get the REAL acoustic response of the room itself, unaffected by the speaker tuning. Then get the speakers off the desk onto stands (very heavy, tough, firm, rigid stands), and do the other stuff to re-arrange the room into the theoretical best layout, the test again with REW like that. Then do the "small increments" procedure (boring!) to find the best spot. Then add some bass traps, some absorbers on the front wall between the speaker and the wall, something on the ceiling, something on the side walls, and a huge amount on the rear wall. Then measure with REW again. Carefully document every REW test you do, noting the exact location of the speakers and mix position for each one, in the comments block for each measurement. That way you can easily see the trends, and get back to any position later, if you wanted to try something else instead.

So, that's what I'd suggest!


- Stuart -


You do not have the required permissions to view the files attached to this post.

_________________
I want this studio to amaze people. "That'll do" doesn't amaze people.


Report this post
Top
 Profile  
 
PostPosted: Mon Jun 17, 2019 12:54 am 
Offline

Joined: Tue Mar 25, 2008 3:31 am
Posts: 136
Location: Cork Ireland
What an interesting speaker design. It appears that the rear facing woofers are intended to 'engage with the front wall'.
I would start testing response with them as close to the Front Wall as possible. 10cm recommended in this case.
DD


Report this post
Top
 Profile  
 
PostPosted: Wed Jun 19, 2019 3:50 am 
Offline

Joined: Thu Jun 13, 2019 11:48 am
Posts: 5
Location: CT
DanDan wrote:
What an interesting speaker design. It appears that the rear facing woofers are intended to 'engage with the front wall'.
I would start testing response with them as close to the Front Wall as possible. 10cm recommended in this case.
DD


Hey Dan

Yes, I believe that is the idea here.


Report this post
Top
 Profile  
 
PostPosted: Wed Jun 19, 2019 3:51 am 
Offline

Joined: Thu Jun 13, 2019 11:48 am
Posts: 5
Location: CT
Thanks again for allof the great info.

Ok, when you said EQ, I thought you were referring to an EQ in the chain... I didnt realize that you were referring to the filters in the speaker sw. But, yes, those have been engaged - based off of what REW spat out.

The speakers on the desk thing.... first, the entire desk and a large portion of my kbd is covered in neoprene to cut down on those reflections you mentioned. So whereas the desk is not 100% reflection-free, it's not your "typical desk". Then.... in order to get rid of those vibrations you referred to, the speakers (which have sorbothane pads underneath them) are actually sitting on boxes (which sit on the desk, on neoprene) which are packed with 2" of sand. Went to the beach, packed in the sand, leveled it off, packed in some more... kept going until there was no way any more sand could fit in the box. The tops of the boxes that the speakers sit on do not come into contact with any of the sides. I am not sure that moving the desk fwd, and placing the speakers on stands behind the desk is going to net a better result here, at least not when it comes to vibrations and reflections. In fact, if I put the speakers on stands, there would probably be more vibrations, at least in the floor. Of course, I can put the speakers on the boxes on the stands. Ultimately, you know better than I do, and if you still think that stands are the way to go, I will do that. Apologies for not mentioning this in my OP; I tried to think of everything, but....

So....

So, in summary: First, turn off the EQ (parametric filtering) in your speakers, so we can get the REAL acoustic response of the room itself, unaffected by the speaker tuning.


I am not opposed to doing that, but... well, I have already done that. Again, so we are clear, I took measurements with no filters engaged. Then, with the assistance of D&D, took new measurements, and we then input params into the 8C filter section based off of what REW told us to input. So, I started with no filters, and now WITH the filters, I have a better result in both the measurements in REW, and also with the sound in the room. If I take measurements with no filters, then that is moving backwards. I am not sure I want to do that.

Cheers.


Report this post
Top
 Profile  
 
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 9 posts ] 

All times are UTC + 10 hours


Who is online

Users browsing this forum: No registered users and 8 guests


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Jump to:  
Powered by phpBB® Forum Software © phpBB Group