Help interpreting my REW graphs

[wiz1der]

First off, i want to thank everyone who has contributed here throughout the years. I started building a project mix room in one third of my 3 car garage back in 2004, and put up 2 interior walls. It sat there for 11 years...

I got the bug again back in late 2015, and dove back into these forums, and got it done! I did everything by myself (except the mini-split install)(mostly on the weekends), and it took me over a year. I started wiring in November 2016. Now that everything is in for the most part, I am around to measuring.

I have to say that it is the best sounding personal space i have ever had, although my graphs seem to indicate that I need some work.. (the 60" display in between the speakers and the additional 43" display on one of the side walls is probably not helping)

I do not yet have panels on the side walls, I have to pick up some 703 and assemble, but didn't want to have to take them out if I absolutely have to re-do something.

Here is some build info:
staggered 2x4's on 2x6 header and footer
pink fluffy in the walls
outside walls are one 1/2 gyp sandwiched by 2 5/8 gyp (3 layers gypsum, no green glue)
inside walls 2 layers of 5/8 gyp no green glue
ceiling is 2x4's on 16" centers and same Gyp as walls (3 outer - 2 inner)
acoustic caulk all layers in and out top and bottom
2 6 ft silencer boxes on ventilation system
2 hardwood doors screwed and glued together with 3/4 in. plywood in between

I am very pleased with the STC, I can go up to 115db inside, and from outside the garage, it sounds like someone listening to a small radio in the garage at low volume.

Interior Dims are 14'3" (171 in.) x 8'9" (105 in.) x 7"5" (89 in.)
mains speakers are mounted in a hybrid soffit (2 feet open at bottom and as much 703 that i could straddle below and around the concrete blocks holding up the speaker stands)
front wall between speakers is totally covered with 703
Back wall has 703 straddled in corners floor to ceiling and 4 in of 703 between corners (entire back wall covered)
Ceiling has 3 clouds (approx 4' x 7'), made of 4 in 703 and hard backed w plywood and hung on chains giving about 3-4 in of airspace on top
Cloud at front of room angles down to about 10 in.

Mdat is located at (forum does not allow .mdat): https://we.tl/5dBlcyKq0t

cal mic was dayton emm 6 angled at about 60 degrees

Again thanks to all for all your help over the years, and in advance for helping on this thread!

[wiz1der]

IMG_20170325_101335.jpg

IMG_20170325_101348 1.jpg

IMG_20170325_101412.jpg

IMG_20170325_101419.jpg

IMG_20170325_101423.jpg

[Soundman2020]

Welcome back, David!

The room looks pretty good, I must say! Congratulations!

I can't say much about the acoustic data in your MDAT file, though, as it only goes up to 500 Hz, so there's no info relating to the mid range or highs. Also, there's only one single measurement in there, which I'm assuming was done with both speakers on, but you really need to do three tests: One with just the left speaker on, one with just the right, and one with both. All of them running the sweep through the full spectrum, from 17 Hz to 22kHz. You need the individual readings to check symmetry, and I'm guessing there will be differences in symmetry on the lower mid range in your room, due to the desk and large amount of gear on the right side of the room that is not balanced by something comparable on he left.

The big screen between your soffits is fine: it's probably not causing too many issues.

One thing I would suggest, is to try to reduce the number of sets of speakers that you have in there: I counted at least 5 sets! If you could get that down to just two sets, or three at the most, that would remove some of the "big things" between you and the main soffited speakers, that are undoubtedly causing issues with response in the mids. And if you want good sound out of the remaining sets of speakers, get them of the dog box / desk, onto stands behind the desk.

But determining anything more about your room is going to need the full set of complete REW tests. Also, do the full calibration and testing with your system set to produce 80 dBC from each individual speaker (L or R by itself), which guarantees that there will be a level of 86 dBC for the test with both speakers on. (I know the REW documentation says 75 dB, but I'm not sure why that is: the standard calibration level for studios and cinemas is 86 dB.)

With that info, we should be able to get a good idea of how your room is behaving, and figure out what to do about it, if any...


- Stuart -

[wiz1der]

Thanks for the reply Stuart!

I thought the mdat file size looked a little small, I didn't save the full set. Here it is:

https://we.tl/l3HlXug1kx

I can definitely entertain the idea of losing some speakers, the NS-40's are not impressing me. I bought them years ago and just hooked them up with this setup.

[Soundman2020]

Better, but still not a lot of use: For each speaker, you have one set of data that is below 500 Hz, and another set of data that is above 504 Hz. Since there's no point in common between them, you can't even stitch them together, and without seeing the full set of data at once, it's complicated to try to figure out what your room is doing. I'm not sure why you split up the data like that, but it make it really hard to interpret!

You'll need to re-test, with correct calibration, and with full spectrum sweeps (17 Hz - 22 kHz ).

Also, since you do have a sub, you will need to do a set of seven measurements, not just the three I mentioned:

Sub alone
Left alone
Right alone
Left + Sub
Right + Sub
Left + Right (no sub)
Left + Right (with sub)

Since you have three speakers, not two, you need to set each of them individually at 77 dBC, not 80 dB, such that the total will be 85 dB with all of them running at once. Yes I know that your sub and your mains won't necessarily be running the same levels in real situations, but this is for initial calibration. Tuning comes later.

Speaking of subs, you seem to have your crossover set too high. Try something like 90 Hz, initially. That leaves some leeway to adjust it.

Where is the sub in the room, physically? (What are the distances from the left wall, the right wall, and the front wall, and how high above the floor?)

I can definitely entertain the idea of losing some speakers, the NS-40's are not impressing me. I bought them years ago and just hooked them up with this setup.

Yup. Then get rid of them. You should be able to get a few dollars from them on E-bay.

- Stuart -

[wiz1der]

You'll need to re-test, with correct calibration, and with full spectrum sweeps (17 Hz - 22 kHz ).

Also, since you do have a sub, you will need to do a set of seven measurements, not just the three I mentioned:

Sub alone
Left alone
Right alone
Left + Sub
Right + Sub
Left + Right (no sub)
Left + Right (with sub)

Since you have three speakers, not two, you need to set each of them individually at 77 dBC, not 80 dB, such that the total will be 85 dB with all of them running at once. Yes I know that your sub and your mains won't necessarily be running the same levels in real situations, but this is for initial calibration. Tuning comes later.

Speaking of subs, you seem to have your crossover set too high. Try something like 90 Hz, initially. That leaves some leeway to adjust it.

Where is the sub in the room, physically? (What are the distances from the left wall, the right wall, and the front wall, and how high above the floor?)

-

OK, I will set the crossover, and reshoot the graphs sometime this week.

The sub is switchable in and out for use with the ns-10's. It is right under the console, on the floor, centered between left and right walls, and approx 30 in from the front wall ( there is a small stack with console P/S, and 2 amps between it and the wall).

In addition to the list of measurements above, I will also add:

soffit L
soffit R
soffit L + R

Thanks!

[Soundman2020]

In addition to the list of measurements above, I will also add:

soffit L
soffit R
soffit L + R

I thought we were talking about the soffit mounted speakers? There's no need to do tests with any of the other speakers, as they are no use, due to their locations in the room. There's no point in testing speakers that cannot possibly produce decent sound at the mix position. So only test with the soffit-mounted mains. (If you don't use the sub with those, then don't include it in the test set).


- Stuart -

[wiz1der]

In addition to the list of measurements above, I will also add:

soffit L
soffit R
soffit L + R

I thought we were talking about the soffit mounted speakers? There's no need to do tests with any of the other speakers, as they are no use, due to their locations in the room. There's no point in testing speakers that cannot possibly produce decent sound at the mix position. So only test with the soffit-mounted mains. (If you don't use the sub with those, then don't include it in the test set).


- Stuart -

Just read this as i was uploading, I reshot ALLL the graphs, the soffit mounted mains are labeled MAINS. Anything with the sub attached are the nearfields. I do not use the sub with the mains, the sub was to mainly fill out the bottom of the nearfields for checking.

Looking at all the graphs in context, I can see similar dips and peaks in common between the mains and the nearfields, so I'm guessing that whatever I do to improve the mains, will also improve the nearfields' response.

Here's the latest file:

https://we.tl/QlmJkwbbZj

Thanks again!

[Soundman2020]

Great! Much better. The data makes sense now.

OK, first impression; the room is a bit too dead. Your decay times are around 130 ms, but for a room that size it should be more like 180-200. Did you put plastic over your bass trap insulation, clouds, and treatment on the front and back walls? If not, then you probably need to do that. Your highs and lows are reasonable, but you are missing out on the mids. Too dry. We'll need to figure out how much coverage you need on those (percentage covered / uncovered), and the thickness of the plastic, to tune the response.

Next: Your room is small and does not have a good ratio, I'm afraid. The floor area is 127 square feet, but ITU and EBU specs call for a minimum of around 220 square feet for a critical listening room, so yours is about half the size of that.

Your ratio is not good: The room is too long for the width and height (or not wide enough for the length and height, or not high enough for the length and width, depending on which way you look at it), and also your length is nearly an exact multiple of the height. So you have modal issues that are amplified. So the room itself is fighting against you, and shaping the frequency response curve the way it is. There's not much you can do about that with treatment, but your room is a candidate for digital tuning. PM me about that. I don't like talking about it in public, because way too many people think that they can do it in their room, when they just don't have the acoustic conditions needed to do it. It ONLY works when certain very specific acoustic properties are in place. In most rooms they are not. But you already treated your room reasonably decently, so it will work in your room.

Next, the symmetry is reasonable below 70 Hz and above 2k, but in the low mids... ummm.... not so good:

wiz1der--FR-LRdiff.jpg

The top two curves are the frequency response curves for your left and right mains. The lower curve is the difference between the left and right, which basically tells you how closely the two match. As you can see, between 70Hz and 2k, it gets pretty wild. You have about 20 dB of difference! It goes as high as +12 and -8, roughly. Not a good situation at all! Specs call for differences of no more than +/- 1 dB for critical listening. That's really hard to do, and +/- 3 dB is more realistic. But 20 dB... Mmmmmm.... not good.

Much of that is probably due to the desk and "stuff" on the right side of the room, which has to go if you want a truly good room, with decently symmetry. I would suggest that you move out as much "stuff" as you can from both sides of the room, leaving it as clean as you can,

However, some of that "hash" in the mid range symmetry is likely caused by the desk and console, and some of it might even be due to inherent small differences between the speakers, or between the soffits. There MIGHT be some things that we can do about that, but there also might not....

If you could get rid of all that nasty stuff between 70 and 2k, then you'd be pretty darn close to +/- 3 dB, which is very decent for a home studio.

Next you have some rather strong early reflections. One of them has a path length difference of just 11 inches, so I would assume it is coming from your desk, console, or the speakers sitting on the dog box. The other has a path length difference of about 7'2", so that could be a floor reflection, or maybe side wall reflection, or something like that. You could use the "string trick" to figure that out. You have several other fairly big reflections in the first 20ms, that need to be dealt with.

There's some other stuff going on there, such as probable SBIR, but those are the main issues I see.


- Stuart -

[wiz1der]

Did you put plastic over your bass trap insulation, clouds, and treatment on the front and back walls? If not, then you probably need to do that. Your highs and lows are reasonable, but you are missing out on the mids. Too dry. We'll need to figure out how much coverage you need on those (percentage covered / uncovered), and the thickness of the plastic, to tune the response.

There is a 3 mil plastic over the clouds, but I did not put any on the back or front wall. I can definitely add to my back corner bass traps fairly easily, as the frames are simply velcro'd up. The back wall will have to be dismantled if I need to put some there.

So the room itself is fighting against you, and shaping the frequency response curve the way it is. There's not much you can do about that with treatment, but your room is a candidate for digital tuning. PM me about that. I don't like talking about it in public, because way too many people think that they can do it in their room, when they just don't have the acoustic conditions needed to do it. It ONLY works when certain very specific acoustic properties are in place. In most rooms they are not. But you already treated your room reasonably decently, so it will work in your room.

SO THERE IS HOPE!! I will PM you, but would like to try and do what we can first.

Much of that is probably due to the desk and "stuff" on the right side of the room, which has to go if you want a truly good room, with decently symmetry. I would suggest that you move out as much "stuff" as you can from both sides of the room, leaving it as clean as you can,

However, some of that "hash" in the mid range symmetry is likely caused by the desk and console, and some of it might even be due to inherent small differences between the speakers, or between the soffits. There MIGHT be some things that we can do about that, but there also might not....

If you could get rid of all that nasty stuff between 70 and 2k, then you'd be pretty darn close to +/- 3 dB, which is very decent for a home studio.

I will try to move out as much junk as I can, but the desk, monitor, and speakers will have to stay. I know its not ideal, but i have to fit 10 lbs of sh*t into a 5 gal bucket, I have no other room anywhere for this stuff, and it is essential to my well being..

Next you have some rather strong early reflections. One of them has a path length difference of just 11 inches, so I would assume it is coming from your desk, console, or the speakers sitting on the dog box. The other has a path length difference of about 7'2", so that could be a floor reflection, or maybe side wall reflection, or something like that. You could use the "string trick" to figure that out. You have several other fairly big reflections in the first 20ms, that need to be dealt with.

I shot this with the ns40's still in the room, and I have a strong suspicion that the reflective top of those speakers being right under the mains is doing something bad. I will move out as much junk (extra records, stuff under the turntables on back wall), and the ns-40's and reshoot to see if anything changes.

I will also search the "string trick".

Thank you sir.

[Soundman2020]

I will try to move out as much junk as I can, but the desk, monitor, and speakers will have to stay.

Whatever stays needs to be made symmetrical! That's important. If you have a "massive blob" on the right that is 2 feet long, 6" high, and 9" wide that is 7' back from the front wall and 5' above the floor, then you need to put an identical (or very similar) "massive blob" on the left that is 2 feet long, 6" high, and 9" wide, 7' back from the front wall and 5' above the floor.

Symmetry is critical for a control room. Your left ear MUST hear the same acoustic signature as your right ear does. If not, then you will subconsciously compensate for the difference in the mix. It will sound great in the room! .... but it will sound "off" everywhere else. Your mixes will not "translate" well.

I have no other room anywhere for this stuff, and it is essential to my well being..

I would suggest designing a totally new main desk for your console, that has all of your essential gear rack-mounted on the surface, and all of the non-essential gear rack mounted or stores in symmetrical "wings" under and to either side of the desk. Like in this studio: http://www.johnlsayers.com/phpBB2/viewt ... =2&t=20471 Keep it all down low, well below your ear height, and the stuff that has to go up higher needs to be made symmetrical.

I shot this with the ns40's still in the room, and I have a strong suspicion that the reflective top of those speakers being right under the mains is doing something bad.

- Stuart -

[wiz1der]

So, over the past week, I pulled the NS-40 and stands out of the room, and built 4 early reflection absorbers out of 2 layers of 2" 703 (4 in total thickness).

I put 2 of these on each side for 4'x4' coverage.

I re-shot the room, and to my untrained eye, the graphs don't look better, just different.

The room sounds better with the new panels up, my flutter echo is gone when I clap my hands.

Here is the link to the latest graphs:

https://we.tl/bQTIoCZYYV

Thanks!

[wiz1der]

Link is dead, here is the new one.

https://we.tl/x2csfoJ2qx

Calling Soundman2020!

I hope I'm not bugging you, but I have also tried pm'ing you a few times regarding consult.

[wiz1der]

Just updating my mdat file link on the off chance that Soundman has a chance to check in.

He must be working on a huge project!

If not, i truly hope all is well.

[wiz1der]