John Sayers' Recording Studio Design Forum Archive

Wow! It's starting to look like a studio! Good work there, Gareth. And I'm glad you passed your inspection! That's really good news!

Now I'm just wondering how in hell that original door installation ever passed inspection! To me, it almost looks like the door was holding the framing in place, rather than the other way around!!!! But at least you got it out and did it right.

So how's it sounding in there now? Since you have the shell closed off, it would be a good time to do some baseline REW tests, before you put any treatment in!


- Stuart -

Hi Stuart
You're right about the old door. It's no wonder I could hear any car that drove past!

Anyway, because I need to get this job finished sooner rather than later as you can see I've already put some treatment in - the back wall bass trap, and started one of the ceiling corner bass traps.

However, I'll take your advice and get the soffits built over this weekend enough that I can do an REW test before I put any more treatment in. That said I'm convinced I'm going to have to put corner traps along the two long ceiling/wall intersections because I'm going to need as much bass trapping as possible with two almost identical room dimensions.

I'll just finish this cup of tea and then get on it.

Gareth

It's amazing how long this stuff takes to build!! This is about 2.5 day's work... half finished framing.
Hope to get the speaker boxes built, shelves beefed up, framing finished and the first layer of bezel up so I can test the room...

Another slow day.. Made speakers boxes and then did a load of work to the left soffit including installing the 5 hangers.

I'm awaiting my electrician before I can do much more as I can't do any more on the right soffit until fused spurs are installed.

The pictures speak for themselves;

So today I've almost completed the left soffit. I think I remembered everything I need to do!

The rear of the speaker box has 2" insulation in it, covered with chicken wire to hold in place. I made a chimney up to an air vent above and then covered this in garden weed barrier, to stop any bits of insulation falling back into the chimney and thus speaker box. I decided to use this rather than plastic sheet as I was concerned plastic sheet would vibrate and make noise. This is really thin and is perforated with tiny holes.

I stuffed insulation into the cavity around most of the speaker box. I've left the right hand side uninsulated and open for now, as I might need to get my hand in to push the speaker back out to adjust the settings. It's a really tight fit into the MDF box, which is good but I'm genuinely worried I won't be able to get it out once I push it in.

A friend has advised rubbing beeswax into the MDF box to help the speaker slide. The other option is using a mirror on a stick to be able to adjust the switches through that gap in the side!

The Mackie HR824 (mk 1) have various settings so I'm planning the following - advice on this would be appreciated:

Acoustic space: set to A for a 4dB reduction in the bass
Low Freq: set to 37Hz (normal)
High Freq: set to -2dB - this is my personal preference for these speakers to ensure my mixes are bright enough.

Would it be better to have the Low Freq set to 47Hz as well as the Acoustic Space set to A to help compensate for the bass boost of the soffit mount?


Pictures of this evening's progress:

I've yet to put a shelf in to separate a bass trap at the top.

The next two nights will hopefully get the other soffit built the same, so I should be able to REW on Saturday.

It's getting close to being a useable room!

Gareth

Would it be better to have the Low Freq set to 47Hz as well as the Acoustic Space set to A to help compensate for the bass boost of the soffit mount?

You'll probably have to "test and adjust" to get it right, using REW, but for the first approximation I'd suggest 37 Hz, "B" and flat for the highs. Yes, I know you have needed -2 in the past to get your mixes bright enough, but once you get the room treatment in, that will be sucking up the highs like you won't believe, so I'd hazard a guess and say that flat is probably going to where you will likely want it. Also, if you ever decide to do digital tuning (PM me....) then you'll need that flat.

There's another option that I've done with some of my customers, where the speaker controls are all unreachable: Take the rear panel completely off the speaker, and mount it on the side of the soffit, so you can adjust to your heart's content, as needed.

The next two nights will hopefully get the other soffit built the same, so I should be able to REW on Saturday.

I'm looking forward to seeing that! Follow the instructions on the "Aoustics" section of the forum for REW testing...


- Stuart -

So both soffits mounted, speakers installed and first REW tests taken, using the speaking switch settings suggested by Stuart above. I remembered once I'd done the initial tests that the sides are not on the soffits - I left them off to get to the back switches on the speakers if I need to change anything.

Treatment already in:
- the back wall treated as a large absorber - 60cm deep at the sides and 20cm deep in the centre.
- 2/3 of one ceiling/wall intersection has a super chunk (40cm deep)
- hangers underneath each soffit
- bass trap on top of the left soffit

I had the sofa and some other bits in the room, bits that will be in the room when I use it.

There are some quite interesting differences between left and right speaker.., and with and without the soffit sides on

MDAT file:
https://www.dropbox.com/s/v57rv482gg9bw ... .mdat?dl=0
Despite what the above says, music playing in the room already sounds loads better. As others have said, the imaging with soffit mounting is fantastic. What surprised me the most is how much more even the room is already - whereas before moving my head even a little bit made a difference, I can walk around the room and still hear an excellent image and fairly stable bottom end.

QUESTION: do I need to change any of the settings on the speakers? It looks like the bottom end under 100Hz is a little pronounced...

Treatment plans:
- finish super chunk at both long wall/ceiling intersections
- treat underneath soffit with 4" rock wool

I then plan to use the 7 2'x4' panel absorbers I already owned to complete phase 1. If I find in a few months time the room needs more work, I can replace these with designed and built units:
I have four of these, which I plan to mount to the ceiling: http://gikacoustics.co.uk/product/gik-a ... echnology/
I have three of these, which I plan to mount to the walls in the first reflection points either side of the listening position: http://gikacoustics.co.uk/product/gik-a ... tic-panel/

Can't wait to start using the room - it's looking and feeling better already! Thanks in no small part to this forum and its wealth of information.

So tonight I finished the ceiling/wall intersection absorbers, and put some covers on the front of one of the soffits, top and bottom.

I re-ran REW and weirdly some of the dips seem more pronounced compared to the previous test (see below). Hopefully this is down to a slightly different mic position, even though I measured it, because it's hard to realign. The corner absorbers are 40cm deep on each side, made from triangles of 100mm thick RWA45 (http://www.rockwool.co.uk/products/rockwool-rw-slab/)
There are two major dips, 83Hz and 135Hz, the 135Hz one is no surprise as the room has two dimensions that are 2.4m. These tests are with no smoothing - switching some smoothing on reduces that 135Hz dip a bit. I doubt this dip will reduce much though!

New MDAT file: https://www.dropbox.com/s/h1fsf1qs93tlk ... .mdat?dl=0

This graph compares left and right speakers with the new ceiling corner absorbers in:
Comparing the waterfall graphs shows that the ceiling corner absorbers are helping reduce the RT60.

Before adding ceiling corner absorbers:
After adding ceiling corner absorbers:
Tomorrow I will add my GIK acoustic panels to the ceiling and first reflection points either side of mix position, then re-run REW. I will then experiment with different speaker switch settings as currently below 60Hz is quite pronounced.

Hope to be able to do some work in the studio next week!

There are two major dips, 83Hz and 135Hz, the 135Hz one is no surprise as the room has two dimensions that are 2.4m.

Those are probably SBIR artifacts, not modal. To confirm, we can do the "walking mic" test, which is a long and boring process, but reveals a lot about the room... and also the best place to set up your mix position! Let me know if you want to do that, and I'll PM you the instructions.

The also might be related to floor bounce, or ceiling bounce. There is a very strong reflections at around 3.6ms, which works out to a bounce length of 125cm. So here's what you can do to confirm. It's called the "string trick". Measure the exact distance between the tip of the mic and the acoustic center of your speaker, and 125cm to that (the bounce path length), and carefully mark that distance on a piece of string. leaving a bit extra string beyond that on each side. Attach the string to the mic and speaker face, such that the marks on the string are exactly at the tip of the mic and the acoustic center of the speaker. Now you have some slack in the string (125cm to be exact!). Move that loop of slack around the room to see what surface it touches. I'm guessing it will be either the ceiling or the floor, part way between the speaker and the mic, or maybe one o the side walls. There will only be one surface that the string touches: that's the surface that's causing the bounce. To confirm, put a large, thick piece of insulation (mineral wool, or fiberglass) on that surface over the area where the string touched it, and do another REW test.

Then repeat the "string trick" using a string that is 59.6cm longer than the direct speaker-mic distance, since you have another reflection at about 1.74ms, an then do it a third time with a string that is 188cm longer, and yet again with a string that is 259 cm longer. Those are your biggest early reflection problems. Identify those surfaces one at a time, treat them one at a time, and you'll get some decent improvements.

Another thing: you have a mains hum problem:
See that spike that doesn't die, at exactly 100 Hz, where the cross-hairs are? That's mains hum, almost certainly. You are in the UK, where the mains frequency is 50 Hz, and that's the first harmonic, which is usually the strongest. There seems to be something going on at 50 Hz as well, but it's hidden in the mush from the room, and is less likely to be an issue anyway.... So check your grounding, and shielding, and turn off anything that might be emitting a lot of high energy mains interference.

Speaker settings: Yes, set them both down to -6 dB bass roll-off. The top end is a little bright, and you might need to roll of the highs later, but you have not yet completed the treatment so I wouldn't do that yet.

Next: I'm surprised at how little change there was between the empty room and the first round of treatment! How "empty" was the room when you did your baseline test?

I'm also curious about the "bump" a about 500 Hz in all of your graphs: Not sure what that could be. Perhaps the speakers themselves. And I'm also curious about what is causing the increased decay times in the high end, between about 1k and 6k.

These tests are with no smoothing - switching some smoothing on reduces that 135Hz dip a bit.

Yup, but that's sort of like putting a band-aid on a broken bone that is sticking out of your skin... it disguises the ugliness visually, but does nothing at all to actually fix the problem!

I doubt this dip will reduce much though!

Oh ye of little faith....

There are some quite interesting differences between left and right speaker.., and with and without the soffit sides on

Put a panel of 2" OC-703 on the sides of the soffits, that are on either side of the glass door, and cover that with fabric.

I have four of these, which I plan to mount to the ceiling:

To be honest, I'd stand one of those up vertically in the middle of the rear wall, just on front or the treatment you already have there.

I have three of these, which I plan to mount to the walls in the first reflection points either side of the listening position:

One on each side wall, and try the third one above the door on the front wall, angled across the wall/ceiling corner.

Hope to be able to do some work in the studio next week!

Hope springs eternal in the human breast!

- Stuart -

Thanks for your helpful advice, Stuart.

I will do the string test later, and the walking test once I've put this round of treatment in.

Mains hum: that's either because the interface is on a different socket to the monitors, or caused by the HRV fan. I'll experiment with that. Might need to get a transformer isolator for the balanced feeds to the monitors...

I'd not spotted the 500Hz bump - the speakers are in theory +/-1.5dB between 39Hz and 20KHz, but anything's possible!

Unfortunately the first set of REW readings were with the rear wall treatment installed - I had to move on with work and couldn't build my soffits until the electrician had been. Hence why the additional treatment hasn't made as much improvement. Any ideas why some of the frequencies have got worse since adding the wall/ceiling super chunks?

I'll add some treatment to either side of the glass door (thanks for that tip) - we don't get OC703 in the UK - would RWA45 be as good?

Between the door on the front wall - I was thinking of putting another super chunk there at that wall/ceiling intersection - the GIK panel won't fit.

Finally, I'm not sure how best to achieve -6dB reduction on the speakers - a picture of the rear panel switches is above. Maybe I'll go for Quarter space (A) and 47Hz? Currently set to Half Space (B) and 37Hz..

Thanks!
Gareth

I will do the string test later, and the walking test once I've put this round of treatment in.

Better to do those NOW, before putting in the treatment! How will you know what to treat, if you don't first do the tests? Also, treatment will "dilute" the visibility of the real problems in the room that the "walking mic" test would otherwise reveal clearly. Since treatment lowers the Q, the problems will spread out more broadly and be less intense, thus harder to identify accuratel in the data...

I'd not spotted the 500Hz bump - the speakers are in theory +/-1.5dB between 39Hz and 20KHz, but anything's possible!

Yep.... but that's when you measure them in a perfect anechoic chamber, in a carefully calibrated test in a proper acoustic laboratory! When you put speakers in a real room, the room itself affects their performance. The room "loads" the speakers acoustically. Just like your car will perform nicely on a dead flat road that is perfectly smooth and dry, but will not even go in a straight line if the road is tilted to one side, angled upwards, with two wheels running on hard concrete and the other two in wet sand, with a strong wind blowing from one side, and rain... The car might be in perfect condition, with brand new, top-quality tyres, but it still wont go straight or accelerate the same way it would when tested under perfect conditions. The same happens to your speakers in a real room.

Most people don't realize that the room itself places an unbalanced load on the speaker, just like those driving conditions place an unbalanced resistance on the movement of the car. There's an "impedance" imposed by the air and treatment in the room, that causes the response to be different from what it would be under those perfect laboratory conditions. So it's quite possible that the 500 Hz. "hump" is caused by the room, not the speakers. It can be fixed, but I'll PM you about that.

Any ideas why some of the frequencies have got worse since adding the wall/ceiling super chunks?

Because you are tightening up the room response! So now you are starting to see things that were hidden in the "mush" of the loose, incompletely treated room previously. It's normal, and it's a bit scary, which is why you need to start identifying individual issues now, and dealing with them individuality, instead of taking the "shotgun" approach of treating everything at once. Think of it like a patient arriving at a hospital after a car accident: he's covered in mud and blood and dirt and oil and shredded clothing, and you can't really see what's wrong with him. You know he's injured, but you don't know where or how bad. You do some tests, clean him up a bit, and now you can see the REAL wounds that were hidden under the grunge before. Since you can see them now, you can treat them now. They look uglier than before, because they were hidden in the fact that EVERYTHING looked ugly, so nothing was in focus as being worse than anything else, but now that he's cleaned up and you did the tests, you can see just how bad the really bad bits are. So you sew up the really bad wounds as you come across them, then keep cleaning and testing ans sewing up more and more, until he's fixed.

That's the point of doing early REW tests, then treatment, then more REW tests. You are peeling off layers and cleaning up stuff that needs to be cleaned up, so you can see where the more ugly, underlying acoustic issues are. Each round of treatment cleans up some acoustic "wounds", and each additional REW test reveals "new" stuff that was actually there before all the time, but you just couldn't see it because of the mud and dirt and blood and tattered cloths. Once you "clean off" the superficial problems, you start seeing the real problems more clearly, so you can then deal with each of those individually, one by one, doing what is needed for each "new" problem you find.

So just like a doctor will first do general typical tests and clean off the worst grime in general, he'll then focus on the actual real issues. He won't use a cotton ball and alcohol to fix a deep gash! It needs suturing. And he won't use sutures to fix a broken bone: he'll use a splint or plaster cast. And he wont use a plaster cast to fix an infection. Each problem needs its own specific "fix". That's where you are getting to now. You've cleaned off the general grime and muck, and now you are seeing specific problems, so you need to treat those individually, and stop treating the muck! You might still need to treat more "muck" later, but you first need to deal with the deep gashes and broken bones that you have found, as they are the most important things in sight right now.

Excuse the analogies and allegories! But acoustics is so non-intuitive that it sometimes helps to put things in a more every-day understandable perspective.

I'll add some treatment to either side of the glass door (thanks for that tip) - we don't get OC703 in the UK - would RWA45 be as good?

I would go with something a little lighter and softer for there. What can you get in fiberglass insulation?

Between the door on the front wall - I was thinking of putting another super chunk there at that wall/ceiling intersection - the GIK panel won't fit.

That would be fine, but once again I'd go with something lighter than the mineral wool for that one.

Finally, I'm not sure how best to achieve -6dB reduction on the speakers - a picture of the rear panel switches is above. Maybe I'll go for Quarter space (A) and 47Hz? Currently set to Half Space (B) and 37Hz..

Don't cut back on the low end response! You want to roll it off a little, not kill it entirely! Keep it at 37 Hz, but set it for Quarter Space. I PM'd you about that, and some other stuff we can do if you really want to get your room as good as it can be.

- Stuart -

Thanks again Stuart, and I'll reply to your PM shortly.

I took your advice and did the walking mic test before adding any more treatment. Results for that are:

https://www.dropbox.com/s/v6ftkoqtrtgi8 ... .mdat?dl=0
and
https://www.dropbox.com/s/zg9udj85n3o2c ... .mdat?dl=0

Interestingly, the massive dipper at 136Hz is only so pronounced at my intended mix position. Something's obviously gone wrong either with the measurement, or the impact of the other changes on the room. It looks like the response would be quite a bit flatter 12" further back. Hang on, I've just realised, I put the mic where my head would be, not where the speaker axes intersect, which is... wait for it.... 12" behind my head. DOH!!

I also did the string test:

- 3.6ms, which works out to a bounce length of 125cm = the side wall nearest the speaker
- 1.74ms, 59.6cm = not really sure!
- 188cm = the side wall furthest from the speaker (opposite)
- 259cm = seems to be the glass door?

I then added three of the GIK 'bass trap' panels to the ceiling, and two of the thinner ones to left and right of the mix position. These should help cater for some of the above identified reflections.

I took further REW readings, then set the speakers to quarter space and took readings again.

These are all in: https://www.dropbox.com/s/h1fsf1qs93tlk ... .mdat?dl=0

SPL comparison between yesterday and end of today, so the addition of ceiling and wall treatment, and changed speaker settings:
I am much happier with the sound of the room now as the ceiling and wall treatment has got rid of the flutter echo.

I experimented turning the HRV on and off and can confirm it is this unit generating the 100Hz hum you see. Not sure if the hum is in the power line/coming out the speakers, or actually an audible noise in the room. I'll test for this somehow soon.

Knocking back the bass a bit on the speakers has made a visible difference of a couple of dB at the bottom end, and adding the extra treatment has made the RT60 more even. Waterfall plots:
NB all the above readings taking in the same place as before for consistency.

I will sort out some softer insulation and add it to the insides of the doorway, and to a small super chunk above the door/between speaker soffits. You say I need something other than RWA45 - would this be OK? http://www.knaufinsulation.co.uk/produc ... nifit-slab
??

Thanks!

Gareth

Interestingly, the massive dipper at 136Hz is only so pronounced at my intended mix position. Something's obviously gone wrong either with the measurement, or the impact of the other changes on the room.

... or maybe with your interpretation of the data?

Here's a small part of the results of your walking-mic test, which clearly shows what the real problem is at 136 Hz!
Take a close look: You can clearly see how that issue occurs for different frequencies at different mic positions. Here's the Impulse Response for the mix position mic location (0 inches), showing the large reflection I mentioned yesterday, measured more precisely with a delay of 3.630 ms:
Do the math on that, and you have a path length difference of 106.5 cm. Do the math on that, and you have a half-wave cancellation frequency of 137.7 Hz. Check the frequency response graph for that mic position, and see what you see at that frequency:

Well gee! Look at that! There's a practically perfect match for your null at 136 Hz.

To confirm, let's check again, at a different location. Let's say 6 inches behind the mix position. The impulse response shows a strong reflection...
At 3.463 ms. That works out to a frequency cancellation of 144.4 Hz. The actual graph for that mic locations shows a frequency cancellation here:
142 Hz. Pretty much spot on!

Repeat the above for several more mic locations, reflection times, and frequencies, and you come to the conclusion that this is a floor bounce!

So your conclusion that something went wrong with the measurement, or with the treatment, is not correct. The measurement and the treatment have helped to HIGHLIGHT the problem! And the problem is floor bounce.

You can easily confirm this by putting the mic back at the mix position (0 inches), and putting a large thick pad of insulation on the floor, half way between the speaker and the mic (for both speakers), then running some more REW tests. Do a few, such as +3", +6", -3" and -6", to check that the problem really did go away. Don't forget to check that the insulation remains at the mid point between speaker and mic for each of those tests.

It looks like the response would be quite a bit flatter 12" further back.

You need to stop looking at the frequency response so much! It's NOT the most important indicator of what's happening at this point in the process! It is one factor, yes, but focusing only on FR will only give you a small part of the big picture. If you look at the IR graphs, you'll see that -12" is no better than 0". In fact, the reflection is LARGER at -12 than it is at 0". It's 37.7% FS, at 12, vs. only 27.9% FS at 0"...

I put the mic where my head would be, not where the speaker axes intersect,

If you put your mix position at the point where the speaker axes intersect, then your ears will be off-axis for the speakers! Your EYES will be on-axis, but your ears will be off. So if you want to get your ears surgically transplanted onto your eyeballs, that would be fine... but you might find it hard to see the DAW with your ears in the way!

- 3.6ms, which works out to a bounce length of 125cm = the side wall nearest the speaker

Either your SketchUp model is grossly wrong, or you did something wrong with the string test. It should have clearly shown the floor bounce... I checked on your SketchUp model, and I can't make any triangle from the speaker to the mic at 0" that touches the wall. But I certainly can make one that touches the floor!

SPL comparison between yesterday and end of today, so the addition of ceiling and wall treatment, and changed speaker settings:

Once again, FR is only a small part of the equation. Take a look at the OTHER stuff that changed. At this point, you won't be getting more large changes in FR from adding absorption to the room. But you will in other areas, which clearly shows up elsewhere in REW. And the conclusion there is that the room is now too dead. There's too much bare face of absorption in the room. You will need to put strips of thick plastic across some of your bass traps, tuned to the right frequency ranges, to keep the room live. The decay times are even, yes, but too short.


- Stuart -

Ah OK this is vey helpful and insightful.

I'll try dropping some treatment on the floor as you suggest to confirm it is a floor bounce.

So, if that is related to the floor - what's the solution? I don't think I've seen many studios with soft treatment in a specific floor location?

Made a desk and open cupboard thing today for my gear, and did the last bits of painting so my floor can be finished this week....

Thanks
Gareth

Work still continues...

I have installed the remaining bits of treatment: a panel of 2" insulation either side of the corridor, and a small super chunk between the soffits, above the corridor. I've almost finished the fabric covers for the other super chunks but these are velcro'd in place so can be removed to add plastic, to bring back a bit of life into the high end.

I've got to get the floor finished, then make two final fabric covers, then it's very nearly done for now!! It already sounds great...
Gareth