Studio diary: practicing the basics!

Soundman2020 wrote:

According to that, you did the test at a level of about 140 dB :!: :shock: If that were, true, you'd be dead.

I don't mean to de-rail your thread here, but I did find that comment quite amusing as I've done the REW tests for my room without proper REW calibration as well :) well said Stuart.

I don't mean to de-rail your thread here, but I did find that comment quite amusing as I've done the REW tests for my room without proper REW calibration as well :) well said Stuart.

Well, maybe "dead" is a bit of an exaggeration, but you'd most certainly be deaf, and also in pretty heavy pain! You need more like 200 dB to actually kill you instantly, and some say that 160 can kill within a few seconds, but 140 sure is "waaaay out there" on the spectrum of "OUCH". And certainly isn't attainable in a home studio. Nor with any single loud speaker that I ever heard of! With one exception: ... I don't think I want a pair of those in my studio... :) - Stuart -

I try to avoid exposing myself to more than 92dB-ish. Those h u u u u u u u u u ge speakers are just, well, American-looking to me ;-) Room's size: - depth 455 cm (but there is a window that is about 30 cm deep on the front wall) - width 419 cm - height 292 cm (but there are 22 cm deep cavities, as you can see on pictures in earlier posts). Knowing that a symetrical design is crucial for the stereo image, I already know that this room can probably be good enough for production and very good for vocal classes, but my goal isn't to turn it into a 1st class mastering room. Also, because of that, I decided to keep as much volume as possible in the room, which takes us almost within Bolt's zone. I would have had to sacrifice 30-40 cm of the width to be in the zone but it would have taken the concept beyond the targeted budget.

Pictures: here are some pictures of the latest progress (see earlier posts for older ones). The ceiling corner traps are not yet covered with cloth.

Those h u u u u u u u u u ge speakers are just, well, American-looking to me

:wink: Yup! They are actually used by the USA armed forces to simulate the sounds of a battlefield, for training purposes.... :shock: :!:

Room's size: depth 455 cm - width 419 cm - height 292 cm

That's actually not a very good ratio: It fails one of the three "critical" tests used by the BBC to determine if a room can be used as a studio or not.

Also, because of that, I decided to keep as much volume as possible in the room, which takes us almost within Bolt's zone.

"Almost" inside the bolt area, really means "outside" the bolt area... Some acousticians would take that as cause for either rejecting the room, or doing something about the dimensions to get it INSIDE the bolt area.

I would have had to sacrifice 30-40 cm of the width to be in the zone

Or make the room 10cm longer and 10cm narrower ... That would get you in the bolt area, and turn the BBC "fail" into "pass"... :) But anyway, it doesn't make a huge difference, and as you said, your goal isn't to make it into a world-class mastering studio, so it will probably be fine, after it is treated. - Stuart -

Now, I can clearly hear those (upper)mids when just talking in the room, so unless it's purely psychological, what I see on the graphs in the mids kind of makes sense with my impressions when in the room.

I'm sure you are hearing something! No doubt about that. What I'm not sure about, is that the REW data is showing what you are hearing, since it doesn't make sense. It's really hard to explain how the sound field can decay away to nothing, then suddenly build up again out of nowhere... the energy to do that must be stored somewhere: it cannot come out of nothing, but there's no sign of it at all in the REW graph. And flat response in the very high end, extending beyond 20 kHz, is very unusual too... So that's the issue: obviously, there is something wrong on the room that you can hear easily, or you wouldn't even be here! The problem is that REW doesn't seem to be showing the room response correctly. Is there something else in the room that could be making low-level noise as the REW test sweeps are running? People? Equipment? Or maybe something in the rest of the building, or even outside the building? - Stuart -

Stuart, I like your style ;-) Thanks for your follow-up. Ratios: message received, cap'tain ;-) I'll look into those BBC critical tests as this sounds very interesting. Room's response: the window is currently open because I need to pass a power cable through it so external noise probably happens, yes. Not sure if it goes beyond 20khz and below 20hz as I won't pretend my hearing is that good :-) Now, let's calibrate and get back to you.

Hi there, So, I used the calibration noise (white or whatever) and made sure that the level was about 80-81dB. Then I made a measurement (link below) but it still measures 130-ish dBs? https://drive.google.com/file/d/0B6wx2y ... sp=sharing Not sure what I'm missing here, I'll ask my buddy about his MOTU interfaces, as there may be some sound looping in CueMix or something.

What is even more strange is that the measured signal peaks at about -24dB on REW's input meter, which should leave us with a nice headroom, isn't it? Is there any scale or anything between the input and the graphs, like a pref or an option that I may have incidentally checked?

So, I used the calibration noise (white or whatever) and made sure that the level was about 80-81dB. Then I made a measurement (link below) but it still measures 130-ish dBs?

Did you actually tell REW that the level was 80 dB? I mean, you saw that you were getting 80 dB on your hand-held meter, and that's great, but then you also have to notify REW that the level is 80 dB: You do that with the "SPL meter" tool, click on "Calibrate", select "Use REW speaker cal signal", then REW will play the test signals and also opens a window called "SPL Reading calibration" where you type in the reading that you are seeing on your hand-held meter. Then click "Finished", and conform that both your hand-held meter and the REW meter are showing the same readings. If you did that, then it is not possible that REW shows 130 dB while your hand-held meter shows 80. It just can't happen. When you do the above procedure, make sure that both the REW "SPL Meter" tool and your hand-held meter are both set to "C" weighting and "Slow" response. After you have done that, then repeat the loop-back calibration procedure for your interface: there might be an issue there too. - Stuart -

You're the man! Ok - let me re-check all that from scratch. Am I the only one finding that this calibration process could be more intuitive? What is wrong with the audio interface's response curve?

Am I the only one finding that this calibration process could be more intuitive?

:) You aren't the first one to notice that! And I'm sure you won't be the last either... It is documented in the REW manual, but you have to go digging to find it. I think I might write up a full set of instructions for calibrating and using REW, then post it as a sticky here on the forum... They do have something similar over at the HTS web site, but it's more oriented to home theater users, not so much studio builders.

What is wrong with the audio interface's response curve?

Probably nothing, but if you re-calibrate the SPL levels in REW then you should repeat that too, just in case it affected something: You need to be sure that the sound card calibration curve is correct, since it affects all subsequent readings... The procedure is in the REW manual. - Stuart -

A sticky note would be very helpful!

There you go, calibration done (twice for it to actually be effective) and new measurements made: https://drive.google.com/file/d/0B6wx2y ... sp=sharing To be honest, I don't see any difference now that the SPL readout is correct, but it's good to know that we are now working on an healthier basis.

Hi there, So, I'm back to my initial impressions after measuring again this morning. In this file, you get 4 measurements: 1. upper left wall without plastic (only rockwool) 2. upper left wall with plastic on top of rockwool + the 2 above versions with or without carpet. https://drive.google.com/file/d/0B6wx2y ... sp=sharing The RT60 clearly shows the high-mids response time lengthening when the plastic is on - or am I missing something here? The carpet also seems to help tame the highs and high-mids (which was to be expected). I think I'll remove a part of the plastic behind the cloth, especially on the side walls - what do you guys think?

pask74 wrote:

A sticky note would be very helpful!

... just following the build thread ... AND ... I second the 'sticky' vote. By the time I got to doing initial REW shots [after construction] .... the brain had so much going on at that moment, even the simplist instruction was difficult :shock: Even now I could also use a refresher REW course ! :lol:

... and some pictures: Measurement setup (microphone is placed at 38% of the front wall's width, 1m20 from the ground) Left wall covered with plastic Left wall upper half not covered with plastic The small but quite thick carpet

I am in the middle of removing the plastic layer from the back and the side walls... a bit gut-wrentching as it took me quite a bit of time to install. Anyways... I should have measured earlier on and not necessarily go blindly with this idea. Each time I remove some more plastic sheet, I can witness the RT60 change. We are now below 0.3 sec. I guess I could remove even more to try to match the mids RT which is around 0.15-0.2 sec.

I am in the middle of removing the plastic layer from the back and the side walls...

:shock: :!: Whoa! Hang on there! I'm not sure why you are doing that, since you are still a LONG way from completing the treatment of the room! Treatment is a complete system, not just a bunch of individual parts, and judging from the photos you only have about half of your treatment in place (perhaps less). The rest of the treatment will also be affecting the highs, so you will likely end up with a room that sounds too dead if you remove the plastic. What you are doing is sort of like taking the filling out of a cake you are making because you don't like the way it tastes at present, but you haven't even baked the cake yet and you still need to put in a bunch of ingredients.... :) Of course it tastes bad right now! You have to add ALL the ingredients and follow ALL the procedure, before you can taste it the way it is supposed to be... It's also like taking the drums out of a mix you are working on because you don't like the way it sounds right now, but you haven't recorded the bass or vocals yet, and you haven't done any EQ on the guitars or applied any dynamics either.... It is way too early to decide that the drums sound bad until you hear them in the complete context of the full mix. Same here: it is way too early to decide that you don't need the plastic when you have not yet heard how the room sounds as a whole with all the treatment completed and in place... --- I haven't had a chance to take a look at your latest REW data, but I hope I'll be able to do that in a few hours, and then suggest the best path for treatment. - Stuart -

One other comment: your measurement mic should be pointing upwards at an angle of either 90° or about 45°, and you should do the tests using both speakers located in their final positions. Do 3 tests: one with just the left speaker turned on, one with just the right speaker turn on, and one with both speakers turned on. - Stuart -

OK, I had some time to look at your data.

...or am I missing something here?

Well, you have data smoothing set to 1/3 octave, so you are probably missing about 90% of the useful information in the REW data! Set it to "no smoothing", or "1/48" smoothing only.

calibration done (twice for it to actually be effective) and new measurements made:

Excellent! The levels are believable now, and that strange "dip and rise" in the mid range decay has now disappeared, so that's good. But I'm still concerned about how you are getting such flat response at 20 KHz, with no roll-off evident. That's hard to understand, especially when you have your bass traps and some other absorption in the room already, and it basically does not change at all when you add and remove plastic and carpet... There's something going on there that doesn't make a lot of sense. Typically in a control room, SPL response rolls off above about 8 kHz at about 3 to 6 dB per octave (and it should!) But yours does not. In fact, you have a rise of about 5 dB in your high end: no wonder the room sounds strange to your ears! On the other hand, you have a roll-off in your low end, also of about 5 dB. That will also add to the overly live sound of the room. Basically, the entire frequency response is badly skewed towards the high end, with a difference of about 10 dB. In other words, your high frequencies are about twice as loud as your low frequencies: That is probably why you have the sensation that there is way too much high end in the room. But I don't think the problem isn't the room, since I can't see how it could do that: it seems to me that there is something wrong with the way the speakers and / or signal chain is set up. What controls do you have on the back of your speaker, and how do you have them set? What else do you have in the signal chain between your DAW and the speaker? Is there any type of EQ in there?

The RT60 clearly shows the high-mids response time lengthening when the plastic is on

Yes, but that is very likely because the highs are over-boosted to start with! You first have to solve the problem of why your speakers are putting out ten time more energy in the highs than in the lows, before you can say anything definite about the room treatment.

Left wall upper half not covered with plastic

I see a LOT of insulation in there, but no studs: how is that wall built? What is behind the insulation, and what is behind that? How thick is the insulation? What type of insulation is it? (especially the density). I get the impression that something is absorbing a lot of energy around 500 Hz, and that is also skewing the room response, and that is in addition to the overall major skewing in the frequency response of your system. - Stuart -

Soundman2020 wrote:

What you are doing is sort of like taking the filling out of a cake you are making because you don't like the way it tastes at present, but you haven't even baked the cake yet and you still need to put in a bunch of ingredients.... :) Of course it tastes bad right now! You have to add ALL the ingredients and follow ALL the procedure, before you can taste it the way it is supposed to be...

How do you know that I understand food analogies very well ;-) ? Thanks for this input. I know it sounds silly, but this room needs to be at least usable for vocal classes in about 10 days, this is why I'm taking action. No complaint here, I'm very grateful for the free and very valuable advice but there are constraints I need to deal with. So, what I will probably do is paint the floor and install lighting so that vocal classes can happen, and finish the acoustic treatment in-between vocal classes over the next weeks/months. Not ideal, but workable (?), I hope. I realize that planning 6 months ahead and allowing ourselves 2 months to design the SketchUp plot before taking any action would have been preferable - sorry about that but I need to keep the studio/clients/family rolling and I simply couldn't allow this project that large of a time slot.

Soundman2020 wrote:

One other comment: your measurement mic should be pointing upwards at an angle of either 90° or about 45°, and you should do the tests using both speakers located in their final positions. Do 3 tests: one with just the left speaker turned on, one with just the right speaker turn on, and one with both speakers turned on. - Stuart -

Thanks for the tips! (That sticky note would definitely be very useful). Is there any recommendation about the distance the microphone should be from the floor (or ceiling)? How can I know the final speakers placement as the acoustic treatment and its associated measurements are not completed yet? I will perform the 3 tests and post.

Soundman2020 wrote:

OK, I had some time to look at your data.

...or am I missing something here?

Well, you have data smoothing set to 1/3 octave, so you are probably missing about 90% of the useful information in the REW data! Set it to "no smoothing", or "1/48" smoothing only.

OK - will do. I can remember reading posts from members who sounded pretty skilled here saying that they usually use a fair amount of smoothing on these EQ curves, but maybe I missed something.

Soundman2020 wrote:

calibration done (twice for it to actually be effective) and new measurements made:

Excellent! The levels are believable now, and that strange "dip and rise" in the mid range decay has now disappeared, so that's good. But I'm still concerned about how you are getting such flat response at 20 KHz, with no roll-off evident. That's hard to understand, especially when you have your bass traps and some other absorption in the room already, and it basically does not change at all when you add and remove plastic and carpet... There's something going on there that doesn't make a lot of sense. Typically in a control room, SPL response rolls off above about 8 kHz at about 3 to 6 dB per octave (and it should!) But yours does not. In fact, you have a rise of about 5 dB in your high end: no wonder the room sounds strange to your ears! On the other hand, you have a roll-off in your low end, also of about 5 dB. That will also add to the overly live sound of the room. Basically, the entire frequency response is badly skewed towards the high end, with a difference of about 10 dB. In other words, your high frequencies are about twice as loud as your low frequencies: That is probably why you have the sensation that there is way too much high end in the room. But I don't think the problem isn't the room, since I can't see how it could do that: it seems to me that there is something wrong with the way the speakers and / or signal chain is set up. What controls do you have on the back of your speaker, and how do you have them set? What else do you have in the signal chain between your DAW and the speaker? Is there any type of EQ in there?

The RT60 clearly shows the high-mids response time lengthening when the plastic is on

Yes, but that is very likely because the highs are over-boosted to start with! You first have to solve the problem of why your speakers are putting out ten time more energy in the highs than in the lows, before you can say anything definite about the room treatment.

Left wall upper half not covered with plastic

I see a LOT of insulation in there, but no studs: how is that wall built? What is behind the insulation, and what is behind that? How thick is the insulation? What type of insulation is it? (especially the density). I get the impression that something is absorbing a lot of energy around 500 Hz, and that is also skewing the room response, and that is in addition to the overall major skewing in the frequency response of your system. - Stuart -

No roll-off: I've used the loopback calibration procedure to factor in both the mic preamp and the converters (you can see the curve in the linked measurements) so this should be taken into account already. From what I could see, there is a slight rise in the highs but we are talking tenths of dBs here and a roll-off from 40hZ - both looking quite normal in my non-techie eyes. Am I wrong here? Controls on the speaker: there is a roll-off pot but it's set to full range. EQ in the way: not at all. Just the Behringer microphone plugged into the mixer going to the converters. Plastic sheet, etc.: The left wall is made of "Alba" or "placo-plâtre" - sorry I don't know the word in English. Not very dense, easy to drill. On the other side of the wall, there is a lounge room. The right wall is made of sand-filled bricks. And the room is the tracking room (will post pictures). I put 5cm Type3 rockwool on both walls, with the idea of adding slats on top. There are studs to hold the rockwool in place and stretch the cloth over it. 500 hZ absorbtion: could that be because I've installed too many basstraps? I've done this on the basis of the "more basstraps" kind of rule of thumb and therefore does not sound very plausible. So, I'll test the speakers, put both of them in the room, see if anything is wrong with the preamp/converters combo and post accordingly.

Measurement test setup: I've tested the cables (!) but the curve I get is the same as the converters/preamp one, so no worries on that end, apparently. I also simply played that Maroon 5 song through the speakers: http://www.youtube.com/watch?v=suRsxpoAc5w Nothing obviously wrong when listening in the room. I've also recorded a bit of that song through the Behringer microphone: https://drive.google.com/file/d/0B6wx2y ... sp=sharing I've also plugged a Neumann KM84, just to have a comparison reference, once in channel 1 and once in channel 2 of the mixer: https://drive.google.com/file/d/0B6wx2y ... sp=sharing https://drive.google.com/file/d/0B6wx2y ... sp=sharing None of those sound "good" and I can hear some sort of ringing in the highs, but nothing that I find totally wrong. What do the experts say?