John Sayers' Recording Studio Design Forum Archive

Soundman2020 wrote:Hmmm.... And you live in Australia, where power is 50 Hz.... I'm thinking "Mains Hum".... Did you use a properly wired cable for that? Everything correctly grounded?

This is a fault in the SPL meter itself Stuart. Battery powered. I plugged a pair of headphones in to it and there was a load of noise coming out of it which was almost louder than the mic signal itself.

Soundman2020 wrote:Just checking, but did you have the U89 on omni for those tests?

Good question, and without looking I'm going to say it was cardiod as we were using it for vocals last. I'll double check that but I'll redo the test in omni if not.

Soundman2020 wrote:OK, looking at the REW data, there are several puzzling things there: First, either you didn't do the calibration correctly against a sound level meter, or your room and system are amazingly, astoundingly quiet! According to your graphs, you did those tests at about 55 dB, meaning your RT-60 measurements went down to -5 dB I'd re-check the calibration of REW against an accurate SLM. You should do the tests with each individual speaker giving you around 80 dB on both the REW meter and your SLM, so both together should give you about 86 dB.

I'll check that out as I'm pretty sure i'll be doing the test again using the omni pattern anyway. I probably didn't go through the calibration process properly as I was thinking about a dozen other things at the time. Your right, I probably did run it too quiet

Soundman2020 wrote:Next, the readings from Left and Right speakers are very, very different. Almost like it was two physically different speakers. There's a difference of around 4 to 15 dB at all points across the spectrum, with L being higher in some places but R in others: Perhaps something changed in the room between doing the L and R readings? Maybe you yourself were in a different place, or you moved the sofa, desk etc.? Another curiosity: summing the L and R SPL curves does not match the "Stereo" curve. Or rather, it does match in the low end, up to about 2 kHz, but above that it just goes wild. I would try to figure out why the two are so different. The L and R curves should be at practically the same level, showing more or less the same thing (except for minor variations due to asymmetry in the room), and the reading with both speakers on should be consistently about 5 or 6 dB louder across the board.

Look's like i'm running the test again so i'll pay particular attention to the making sure the L and R tests are repeated the same way. Like I say, I was doing a bunch of other stuff at the time so it wouldn't surprise me if something got moved, a door got opened etc etc.

Soundman2020 wrote:Next, it looks like you overdid the baffle step compensation. On all the graphs, there's a consistent drop of about 4 dB below 120 Hz. How do you have the controls set on the rear of your speakers? Are you using a sub? Are you using anything else that could be shaping the frequency response, such as a graphic equalizer, or a crossover?

No sub, no EQ, no crossover. As the monitors are now soffit mounted I can't make a 100% certain call on the settings ....but, I do remeber setting everything to as flat/unaltered as it would go. I hope nothing got tweaked/knocked/switched when I put them in or it could be a long and tedious process to get them out and have a look

Soundman2020 wrote:The good news is that the room seems to be behaving very nicely!

To be honest Stuart, it really sounds excellent to me in the room. Nothing really appears to be jumping out in the mix too much as you move around the room (except the bass when you get to the back of the room). So i'd have to say i've done something a little skewed in my measurments. I'll repeat the process with more attention to detail and re-post.

https://dl.dropboxusercontent.com/u/416 ... s%202.mdat

Here's the link to attempt number 2 I can guarantee that the U89 was in omni, nothing moved around the room, no doors opened, I didn't get sidetracked by phone calls/people etc.

How'd we go on this one Stuart?

I have a recording session today and tomorrow, so I don't have much time to analyze that properly yet, but on the calibration you went from the sublime to the ridiculous! It's saying that you did that test at around 120 dB You sure got a great deal on those speakers, if they can produce 120 dB like that.... Are you sure you calibrated REW according to your sound level meter? You compared them while playing pink noise, and adjusted REW to match the reading on your hand held meter, with both of them set on "C" and "slow"?

Also, the curves are still different, and not making a lot of sense. In some places, the combined signal (both speakers) is 10 dB quieter than either of the individual speakers by itself! I don't know how to explain that at all.... Negative acoustic energy? They would have to be grossly out of phase to do that, but they don't seem to be: the phase is pretty much the same for all readings.

I'll try to take a closer look on the weekend, but I'm just not getting what is going on there. I would imagine that the stereo image is all over the place like that, and skewed different ways at different places on the spectrum... Something, either the speakers themselves or something in the room (perhaps close to the speakers) is skewing things, big time. Maybe some thing physical, like asymmetric furniture close to the front of the room? Dunno...

- Stuart -

Soundman2020 wrote: Are you sure you calibrated REW according to your sound level meter? You compared them while playing pink noise, and adjusted REW to match the reading on your hand held meter, with both of them set on "C" and "slow"?

I even read the 'instructions' on the calibration:) I'll take another look at it and figure out what I'm doing wrong. There's no way it was at 120db.

Soundman2020 wrote:Something, either the speakers themselves or something in the room (perhaps close to the speakers) is skewing things, big time. Maybe some thing physical, like asymmetric furniture close to the front of the room?

Hmmm, front of the room is very symmetrical. The only things that are out of whack are the rack gear which is behind the listening position on the left and a couple of single seater couches which are on the back wall middle and back wall right.

Soundman2020 wrote:Also, the curves are still different, and not making a lot of sense. In some places, the combined signal (both speakers) is 10 dB quieter than either of the individual speakers by itself!

I'll put the test I did from a while ago in the dropbox again Stuart. It was done a few months ago when I'd just finished the acoustic treatment for the control room. You did comment on it back in the thread and I don't think you mentioned this kind of difference with the L and R signals. The only things that have changed in the room since then are the fabric on the walls and the furniture, desk, console etc. Maybe I'm doing something really screwy with REW somehow as I'm using the new console for the tests which I've only had for a week or so.

https://dl.dropboxusercontent.com/u/416 ... tereo.mdat

here's the link to the older test Stuart, done in April this year. Let me know what you think of the correlation in the data for this. If these look okay for L and R then I'll switch back to using the Fireface which was what I was using initially.

here's the link to the older test Stuart, done in April this year.

Those curves are a lot more believable: The average around 75 dB, which is OK for REW tests. The left and right curves are not identical, but also not so different as to raise eyebrows, and the "both speakers" curve is about 6 dB higher than the other two curves across most of the spectrum, or at least up to about 2.5 kHz. Above that it starts getting weird too: I guess I didn't notice back then because I was looking at the low end, where the modal issues are, and never looked at the top end. But this is even more confusing, since the curves don't mach very well: In other words the original L curve is a somewhat different from the old L curve. Is the mic in the exact same location in the room as it was for the original tests? Even small changes in the position of the mic can show up as large differences in the curves.

I really wish I could figure out what is going on here.... Very confusing...

- Stuart -

Soundman2020 wrote:In other words the original L curve is a somewhat different from the old L curve. Is the mic in the exact same location in the room as it was for the original tests?

Original test: Galaxy CM140 SPL meter through a Fireface 800

Most recent tests: U89 through a Midas Venice F24.

And the mic position will also be completely different. So the two tests can't be A/B'd. I'm a bad man

The Galaxy meter is playing up and adding it's own signal to the reading so i've stopped using it till I can get it fixed. I'm using the U89 for the more recent tests as it has the least colouration of the mics that I have.

I loaded up the old files more to see if the same 'oddities' were showing up in the original. It seems that they aren't as extreme as my more recent measurments so perhaps it's my REW technique thats not so good.

perhaps it's my REW technique thats not so good.

Where are you in teh room while running the tests? Don't forget that your body affects the acoustics of the room somewhat... That could be affecting the readings, to a certain extent.

- Stuart .

Soundman2020 wrote:Where are you in teh room while running the tests?

Is there some way of running REW without being in the room?

I was sitting to the right of the mic when doing those tests. perhaps behind would be better.

Is there some way of running REW without being in the room?

When you wired your live room, did you run CAT-5 cable along with the audio stuff? If so, you could use that with a KVM extender to get your keyboard, video monitor and mouse into the other room, and run REW from there.

I was sitting to the right of the mic when doing those tests. perhaps behind would be better.

It's possible, for sure. What I try to do is to crouch down on the floor behind the chair or even under the desk, and run the test from there.

I'm not convinced that this is causing all of your problem, but it might be causing come of it. And it's worth trying another test with you out of the way as much as possible, or at least located symmetrically in the room.

- Stuart -

Soundman2020 wrote:When you wired your live room, did you run CAT-5 cable along with the audio stuff?

yep. I'll look into the extender option and see what I come up with but for the mean time i'll have to be hands on.

Soundman2020 wrote:What I try to do is to crouch down on the floor behind the chair or even under the desk, and run the test from there.

See this is why i'm here. To get these inside and secret tips from the professionals I can't wait for someone to walk in the control room while i'm sitting under my desk with the laptop. I'll just have to explain that it's all for a scientific cause before they get the straight jacket out

'll just have to explain that it's all for a scientific cause before they get the straight jacket out

You can complete the picture for them by opening your eyes really wide and wild-looking, then babbling incoherently about the voices in your head that made you hide under the table...

- Stuart -

hi steve, i have decided to use the soundscreen 3.1 and the gg glue sealant and compound between 2 layers of 16mm fyrecheck, i was wondering what type of backer rod you could recommend using?? around the outside of each wall, joins, maybe a 5mm gap between sheets and the floor and 5mm between wall joins,

thanks Neil

hey Neil, I answered your question back over in your thread to keep all your information in one spot. That way if someone else is running through your build in the future they'll be able to keep track of it

phew....the last few months have been way over the top with work and many things have fallen by the wayside. This thread included

for starters, a belated happy new year to all and may everyones studio building/tweaking and designing be fruitful this year!!

It looks like the last thing I was up to here was doing some REW tests to explain some odd room modes. I'll have to get back on board with that and hopefully post some more useful results from REW. I'am however starting to suspect that it may be a problem with the monitors unless i'm really screwing up the REW tests.

There's a couple of small toggle switches on the back of those KRK V8's which adjust the freq at which the high boost/cut/flat response and low end roll off occur. It's possible (not that I want to think about it) that one got accidently knocked down or up during installation. That's about the only thing I can think of as the room symmetry is fine (according to me) and I think my REW technique is okay. I'll persevere with the REW tests as getting the monitors out for a look is something I can't even contemplate at the moment as I'm totally brain-fried after the last few months!