The No.1 Website for Pro Audio
Measuring Room Acoustics
Old 20th December 2010
  #31
Lives for gear
 
avare's Avatar
 

#1, regarding the ~100 Hz dip. You wrote the listening position is 5 feet away from the front wall. The dip is casued by LBIR.

Andre
Old 20th December 2010
  #32
Here for the gear
 
kidcarbon's Avatar
 

LBIR Resolution ideas

Hello Andre,

the two options i come up with are:

1. leave the mix position where it is and add diffusion between the monitors on the front wall. this diffusion would be 2d and deep 4-6 inches, and probably very costly.

2. i can use the 38% rule of the 25ft room depth, and move the mix position back to 9.5ft. cost $0.00

does the second method rely on simply moving the mix position back to 9.5ft or does it involve moving the monitors off the wall as well? and i would assume i would have to increase the L-R symmetry of the monitors.

do you have any further suggestions?

thank you,
-Nick
Old 20th December 2010
  #33
Gear Guru
Analysis

kidcarbon, your room seems big, lucky you. Analysis of a room's behaviour is ultimately done by the brain, not a single tool. As Andre has shown there are other things to consider. So let's take them one by one-

Modal.
Big room, primary modes are below music, modal density should be decent.
Any problems will be due to similar dimensions. 24 and 20 are both divisible by four. What is your height to the hard surface above your ceiling tiles?

Lets see the predicted modes in a calculator. http://www.bobgolds.com/Mode/RoomModes.htm

Next try playing Sine waves at the trouble frequencies. REW has a nice smooth frequency sweep for this. Generator/Frequency Follows Cursor.
Carefully tuning to predicted modes, (the frequencies will be slightly different) use an SLM to measure peaks and nulls in your listening area.
Make a map. You now know how where the modal trouble spots are and how strong their effect is. This give great advice on where to sit or place speakers and of course where to place traps. You will get great modal readings in the corners, in itself a very tactile confirmation of the theory.

SBIR.
Reflections from all boundaries near the speakers can cause nulls and peaks.
Andre has pointed out one, the floor is another possibility.
Take a look at Thomas Barefoot's Wall Bounce Calculator. I haven't worked out how to direct link but you can find it in Acoustic Resources on my site.
Observe how the frequency dips caused by the side wall, ceiling, or floor, reflections changes as you move the listener position, but the front wall one does not. That will get your geometric juices flowing.

REW
Frequency Response- Looks like your speakers are pretty small. Try them much closer to the front wall. The bumps may get worse or not but some low bass is better than none IMHO.
The modes in the Waterfalls are sort of cut off. Furthermore some of them seem to be not changing in level for some time before that.
Maybe try some sort of zoom or threshold adjustment, level or time, to see them more clearly. I like to see them vanish into the noise floor, which can be quite high if you have computers and such in the room.
Note small changes of position, say between the left and right ear can yield quite different results. You need to take a broad view of this, ideally use spatial and even time domain averaging.
ETC- The geometry can fool us, or me in any case. It is good to confirm which reflection your are observing. The Mirror Trick is useful. The trusty measuring tape (or laser pointer) of course, but add a piece of string.
Perhaps the most definite way to confirm a reflection is to block it or place absorption in it's path.
I use a panel of 703 wrapped in fabric for this. No frame, light.
I think REW has either a Real Time or Repeat function which should make this very easy.
Lupo has also given us an ingenious method of confirming reflections.
The thread is called Useful ETC Tricks or such I will return and edit it in here.
Perhaps he will spot this and may have other ETC info. When I have ETC questions, I ask Lupo.

Your Room seems big. You might consider changing dimensions, i.e. a wall if the height is related to the others, e.g. 10 feet would be ugly.
Or you could consider Gobos or other portable Traps simply assembled around your mix area. A suspended ceiling is a great asset. You could install enhanced absorptive tiles (705) and fill some of the void above with light fluffy insulation.
This will hit floor to ceiling problems, modal and flutter. It will also somewhat relieve the pressure of the other modes by allowing them to expand and be absorbed vertically.

DD
Old 20th December 2010
  #34
Here for the gear
 
kidcarbon's Avatar
 

Quote:
Originally Posted by DanDan View Post
Modal.
What is your height to the hard surface above your ceiling tiles?
It is seven feet from the floor to the ceiling tile. there is another foot between the ceiling tile and the sub-floor above. The space is already stuffed with insulation.

Quote:
Lets see the predicted modes in a calculator. http://www.bobgolds.com/Mode/RoomModes.htm
Using the calculator, and the map-method you described, i will explore the predicted mode behavior.

Quote:
SBIR.
Reflections from all boundaries near the speakers can cause nulls and peaks. Andre has pointed out one, the floor is another possibility.
Take a look at Thomas Barefoot's Wall Bounce Calculator.
Will do!

Quote:
REW
Frequency Response- Looks like your speakers are pretty small. Try them much closer to the front wall. The bumps may get worse or not but some low bass is better than none IMHO.
The monitors are indeed small. 5 inch drivers. They are mounted on the mix desk and they are as close to the front wall as wires/cables permit.

Quote:
The modes in the Waterfalls are sort of cut off. Furthermore some of them seem to be not changing in level for some time before that.
Maybe try some sort of zoom or threshold adjustment, level or time, to see them more clearly. I like to see them vanish into the noise floor, which can be quite high if you have computers and such in the room.
Note small changes of position, say between the left and right ear can yield quite different results. You need to take a broad view of this, ideally use spatial and even time domain averaging.
I used the largest scale i could for the time axis. REW does not appear to let me increase the time over 1000msec. In my OP, i mentioned how i thought it was significant that there was still heaps of sustain (well beyond 1 second) at 60 and 80hz.

True this was only one channel. I will repeat the exercise and plot both channels and keep my ears/eyes open to the difference. *Note i am performing these tests one channel at a time and i am directing the mic at the tweeter of the tested monitor.

You lost me at spatial and time domain averaging.
Quote:
ETC-
I think REW has either a Real Time or Repeat function which should make this very easy. Lupo has also given us an ingenious method of confirming reflections. The thread is called Useful ETC Tricks or such I will return and edit it in here. Perhaps he will spot this and may have other ETC info. When I have ETC questions, I ask Lupo.
I will look into the repeat/realtime function in REW and search out Lupo's ETC thread.

Quote:
Your Room seems big. You might consider changing dimensions, i.e. a wall if the height is related to the others, e.g. 10 feet would be ugly.
Or you could consider Gobos or other portable Traps simply assembled around your mix area. A suspended ceiling is a great asset. You could install enhanced absorptive tiles (705) and fill some of the void above with light fluffy insulation.
This will hit floor to ceiling problems, modal and flutter. It will also somewhat relieve the pressure of the other modes by allowing them to expand and be absorbed vertically. DD
This is a rented space so I can't throw up permanent walls but the Gobo style porta-wall idea is a good one.

Thanks to all the thoughtful replies thus far. Your time is greatly appreciated.


-Nick
Old 22nd December 2010
  #35
Lives for gear
 
The Beatsmith's Avatar
 

I downloaded Fuzzmeasure Pro 3 and started doing some frequency plot measuring. I'm sure i'll bore you all with the details soon, but what exactly is the reason i should pay $150 for this? It seems to be giving me the frequency responses and graphs without paying anything?

Are the waterfalls considered extra features or something? I'm no longer at the studio in order to figure it out myself.

Thanks
Old 22nd December 2010
  #36
Gear Guru
Details

Nick, I am no expert on REW, which is complex enough to operate. This will change in the new year with a new office computer. If you have a Mac, try FuzzMeasure.
The averaging functions I referred to are in FuzzMeasure, there are probably equivalents in REW. Suffice it to say, when you measure at say 6-8 locations, average the result, that's space or spatial averaging. Time Domain Averaging seems unique to FM. I haven't used it yet, but greatly look forward to it. That would be averaged Waterfalls and Decay Graphs. If you are attempting to use software to represent the acoustics of the room, it seems to me very necessary to sample different spots.

I ran your numbers 24-2-8 in Bob Golds Mode Calculator. Hmmm. Your dimensions need to be measured very accurately or else there will be little or no correlation between the Waterfall and the predictions. I can't make sense of the current numbers vs your Waterfall. By the way you could post the REW file and I could try to adjust it. I might learn something.

REW on my old computer (Mac) is flaky and slow but I do remember some sort of scroll function which seemed to drop the height of the waterfall. Thus burying the noise and nonsense. Sort of a bottom line threshold. In FM you can dial in such a bottom line, say -50dB. Everything below that vanishes below the zero plane in the graph. Makes sense, many of our listening rooms have 40dB or more fan noise, transformer hum and so on. One may guess that there is action after the front cut off at 1000mS but who knows. I think REW 5 Beta will go to 1500mS.
Those long sections which appear out of nowhere and remain unchanged bother me more. What are they, perpetual motion? Or tonal noise in the room?

One channel at a time is just fine. Try both now and again for modal work under 300Hz. I can't imagine a symmetry issue in your big room.

A suspended ceiling is a great boon, if the tiles are acoustically porous. You can actually buy OC705 with a a white face which drops right in. Such a ceiling with light insulation above it is an amazingly good performer.

Beatsmith, that FuzzMeasure is an amazing bargain at 150. Use it for a while and your life will change for the better. You will become very gradually taller. Skin complexion will subtly improving, making you appear a few years younger. The opposite gender will begin to look at you very differently......

DD
Old 20th January 2011
  #37
Gear Nut
 

I have the same question about Fuzzmeasure... it's just not clear on their site what the terms of the demo are.

Is this demo I downloaded somehow limited compared to the full-featured "pro" version? Is it going to expire? Or is buying a license just on the honor system? (If it is, I'll buy it. It's just not clear to me how the demo is supposed to work.)

Awesome software. Way easier to use that REW.
Old 20th January 2011
  #38
Gear Guru
Buy it

I have no idea.
DD
Old 20th January 2011
  #39
Gear Nut
 

How would one go about correcting for a mic if the manufacturer does not post a .cal file? For the mic I'm using, all they show is a graph: AVANT ELECTRONICS - CK-1
Old 21st January 2011
  #40
Gear Guru
Chris

Nice looking mic. If you read the graph and write down the numbers you will have the ingredients for a .cal file. e.g. the Omni graph shows +5dB from 5KHz to 9KHz. So you might extract
5K +4.5
6K +5
7K +5.5
8K +5
9K +4.5
And so on. A dB correction figure at whatever frequencies FM specifies.
I don't have the current FM on this computer so I can say specifically how to input those dB's and Frequencies, however it is just some format, should be very easy. It may require turning those correction dB's minus.

The mic is a little larger than most measurement mics. It will be a little more directional at HF, so do take care to point it directly.

DD
Old 3rd March 2011
  #41
Gear Guru
PM

John, I will answer your questions in a Private Message.
I will make just a couple of quick points here.
Irish and probably EU rooms tend to be about 2.45M High 3-4M Wide.
A typical mix position will have the listeners head at the null of both those modes. This leads to a big wide hole 60-80Hz. Hard to beat two of them working together against us! My overall point re measurement graphs is that most of them will look awful. Also, Frequency Response graphs show little changes while Waterfalls and Decay may show huge improvement. SBIR can have a lot to do with FR.
Looking at any of the graphs, one has to focus on particular detail to see a particular improvement. There is no big picture, no answer to
'how does this room look?'
I am generally referring to the amateur or semi pro scenario here.
At a pro level, rooms designed or built for purpose, one does see very nice looking graphs.
DD
Old 4th March 2011
  #42
you basically took a picture of my room
[IMG] Uploaded with ImageShack.us[/IMG]
single monitor.
I also have a dip in the 700 area...but I suspect is due to the monitor stands.
The room is quite treated, I still have to build a couple of thick panels for the back wall, sadly I can't put them on the wall because I have stuff I can't move, so they will be on some kind of stand.
I will see after those panels how the FR will be.
The Waterfall is actually kind of flat at 150/200ms.

Anyway.....from your last post I would guess there is not much to do for the big dip at 70/80Hz...... :(
Old 4th March 2011
  #43
Gear Guru
Perspectives

Frequency Response graphs do not reflect modal activity much. They do however show the effects of SBIR, if it is present and active.
So the biggest SuperChunks in all corners will do wonders for the Waterfall.
However they may not hit SBIR at all, leaving the FR graph looking ugly.

The fact that the listener is at both width and height nulls is a sort of inconvenient truth. Amateur and Prosumers typically treat a small RFZ for HF only. Pro designs use big side and overhead structures. This hit the modes and provide full range reflection control. Take a look a John Sayers Productions.
His angled slatted side control panels are interesting. The BBC used to install a whole ceiling of 8-12 inch deep limp membrane traps. This was to counter the typical carpet beneath. Anti Carpet.
It is possible and common to treat the width and height modes just as vigorously as the length.

In all cases, in our typical small rooms, to get great results and reasonable or great graphs it normally takes a lot of treatment.
The more sophisticated, expensive or difficult to DIY bass traps are to be found in pro rooms.
Helmholtz, Perf, Panel, Limp Membrane, VPR. However, a simple DIY approach using fibre alone, described at RealTraps, GIK and other places, will still yield fantastic benefits sonically. The sound and the improvement will be way ahead of the graphs.

DD
Old 5th March 2011
  #44
at the end of last year I moved to a new house and I took the chance to get more deep on treatment, so I replace all my, mostry HF fiber panels with much thicker ones...
front corners to the ceiling 14cm + air gap
side panels, 9cm + 2cm air gap
side panels more to the center 8cm + air gap
cloud over my head 9cm + 6 cm air gap
cloud over my head to the front wall angled, 8cm to 14cm
panels on front wall behind the monitors 15cm + 6cm air gap
still missing panels on the back wall
sadly I cant treat the side wall- ceiling corners

I mean is not a small treatment from a home hobbiest studio....but many problems are still there..
all this to say that is much more complicate than what it could seem
P
Old 21st March 2011
  #45
Lives for gear
 
johndykstra's Avatar
 

Hey Dan, an idea occurred to me... seems this thread would be as good a place as any to present it.

Regarding the "quick and dirty" modal activity detection via sine sweeps and various location observations... perhaps this is a refinement... or perhaps hog wash.

It seems given the recent popularity of iPhones... and in particular the free RTA app... one could reproduce pink noise through their monitors and walk around with the RTA app running. Map out the room's response with more than one frequency simultaneously, and not only be able to determine hot spots at a given location, but nulls as well. Seems moving along one axis should point out irregularities caused by that axis, and so on.

Understood this in no way should trump waterfall analysis, but perhaps be seen as a preliminary step for those willing, or a "better than no measurement" for those who can't be bothered.
Old 21st March 2011
  #46
Gear Guru
Ideas

Good idea, on the surface. It certainly has value.
However, modes can be typically only 5Hz wide. You could have a mode and a null next to it within a single third octave.
Ethan has a much better version of this idea. He has filtered pink noise available for download at realtraps.com. This works very much better than full range pink. Well worth a go.
However, for me the sines are better again. The nulls in particular, especially in a solid brick or concrete room, are uncannily almost silent. This really shocks, and makes a very immediate tactile case for treatment.
The sweep needs to be very fine, many DAW plug sweep in increments with glitches. Ugly. I use a hardware Levell generator. Signalsuite is very smooth. Perhaps best of all, John PM, author of REW5, has provided all the tools needed on the one screen.
Measuring Room Acoustics-screen-shot-2011-03-21-16.02.53.jpg

DD
Old 21st March 2011
  #47
Lives for gear
 
johndykstra's Avatar
 

I figured if there was any limiting factor that it would be resolution.
Old 31st March 2011
  #48
Lives for gear
 

I've got a quick question before I start making my measurements (I'm sure it won't be the last!).

I want to measure the room's low frequency response to start with, and the recommendation seems to be to use one speaker for this.... I was wondering what the logic is behind this and what the pitfulls would be if measuring with both speakers. Plus, same questions again between using both speakers in mono compared to stereo.

For the measurements and graphs where both speakers are to be used, should one always use mono - or are there times when normal stereo playback should be used.

Just seems odd to my unscientific mind not to use the playback method for measuring that we use when actually listening to music.

Cheers - and thanks for the Primer!

Max
Old 31st March 2011
  #49
Gear Guru
1 v 2

Max, that question is like a perpetual motion machine. A search here will reveal that. Try 'one speaker or two', and similar. Or try posts 13 and 14 in this thread!

So, the bullet points, just the facts, to think about.

We are not single point mono summation listeners. We have two ears blocked from each other by a large lump of meat and bone.
A single mic will exhibit extreme HF comb filtering. Something the ears do not experience due to the head blocking.
If you did wish to measure as we listen, you would need to use binaural mic techniques with both speakers driven.
Something which I personally encourage.

For LF work two speakers in their real positions give a realistic picture. Recommended in the primer. There are even better ways to explore individual or global modal issues. One speaker on the floor in the corner, with a mic at the ceiling at the opposite corner is a kind of ideal. All modes driven and received with equal favour.
Centre of a wall will avoid driving that wall's parallel mode. Add in centre height and we should drive only one of the three modes.

The most successful measurement uses such techniques to highlight and reveal up particular aspects. Aspects which can be easily obscured by too much information.

ETC, a time of travel type measurement has no meaning with two sources.

DD
Old 31st March 2011
  #50
Lives for gear
 

Sorry to be thick DanDan, but a search is not getting me anyway - perhaps because of the generic nature of the terms used in the search.

I understand the idea of using two mics to simulate the ears - buit unfortunately I only have the one (Rode NT2000).

So am I right in saying that when testing in the listening position with the monitors in their normal position (as opposed to the corner and central set ups you mention), I should use both speakers and set the switch on my Lavry to Mono? If so, should the mic point to the middle point between the two speakers?

Apologies in advance if that simply re-states what you have said previously.... Just wanted to get it clear before I start measuring for LF frequency response and modal issues.

Cheers

Max
Old 2nd April 2011
  #51
Gear Guru
Simple to Complex

There are clever ways to provoke particular modes. Useful when trying to show a result from particular treatments.
However Max, simplest answer:- one speaker is the safest bet for all measurements.
For LF, below 300Hz, it is worth trying both speakers on, in addition to the two single speaker measurements. With all three measures, you can see differences between L and R. You can also see also discover the interaction between the two. Sometimes there is little difference between single and dual drive. This would suggest good speaker positioning to me. In any case do all three for LF only. Cake and eat it!

DD
Old 14th April 2011
  #52
Lives for gear
 
Dayl's Avatar
Wow... is there a room modes calculator for dummies? or an even more simple explanation? lol
Old 14th April 2011
  #54
Lives for gear
 

Quote:
Originally Posted by Dayl View Post
Wow... is there a room modes calculator for dummies? or an even more simple explanation? lol
Hi Dayl

Are you after a simple explanation of modes in general or in how to use/interpret the calculator.

Saying that, they really do go hand in hand actually.

Modes are a huge subject and part of the staple diet of acoustics. So you'll have no problems at all finding some basic info about them.

As for calculators, perhaps start with Ethan's.... See the bottom of this page:

Acoustic Treatment and Design for Recording Studios and Listening Rooms

This calculator only shows the axial modes, but perhaps that's a good thing to get you going while you get your head around things. Once this all seems more familiar to you, take a look at the one Jens has linked to (or any other more thorough calculators) to see what tangential and oblique modes might be up to in your room.

Cheers

Max
Old 1st July 2011
  #55
Here for the gear
 

Quick question

Hi - I just have a quick question. I'm doing an Acoustics course and we have access to the Dirac software (v3 I think). In terms of functionality it's not amazing, and I just wanted to know if there's even any point in me using it. There seem to be lots of free alternatives out there, and after I've done the course I won't have access to Dirac any more - so I'm thinking it might be best to load my impulse responses into another program and get familiar with that.

Any thoughts?

Tom
Old 1st July 2011
  #56
Gear Guru
Dirac

I too experienced Dirac while on an acoustics course.
It is very expensive, tied in to B and K. A corporate and governmental world funded by OPM, Other Peoples Money!

From memory, it is quite competent. I don't remember it being at all difficult to operate, so work away with it in the Labs. Just export the Impulse Responses as Wavs and all the other software should be able to read them.
AFAIK there is only one free alternative. Room Eq Wizard.
Unfortunately the current V5 on my iMac won't open REW files posted here.
It will import Wavs and such.
FuzzMeasure is a good choice for Mac.
Smaart is knocking around, the live sound industry seems to like that one. ARTA is very powerful and affordable.
There are plenty of others focussed on specific areas of activity.
DD
Old 1st July 2011
  #57
This thread is amazing. Thank you, I just learned so much reading it.
Old 3rd September 2011
  #58
Lives for gear
 
Melgueil's Avatar
 

Love this - just downloaded Fuzzmeasure - and like lots of folks ....struggling with this immensely complex topic. Invaluable - thanks DD.
Cdlt
Old 4th September 2011
  #59
Gear Guru
Snap

You are welcome M. The topic can indeed be complex and the popular software grows increasingly complex. However, apart from the full level tones, there is no danger here. I encourage firing up the software and taking a few measure snaps immediately. FuzzMeasure, being Mac, will often work first time.
Unfortunately it has a couple of default settings which can be confusing. Chris concurs and intends changing this in the next update. I have edited the Primer to highlight to point out the odd low threshold default in FM, and the short duration default in REW.

Like so...

Quote:
Some of the default view settings can be a bit odd so I recommend over-riding them.
For Waterfalls set the duration to say 1000mS for an untreated room, 500mS for treated. Tweak to fill the screen nicely. The lowest visible level threshold interacts with this. Set it to say -50dB. Again tweak to fill the screen nicely while ensuring that the end tails of the Modes submerge nicely down into the noise floor.
Be aware that many mix rooms are quite noisy due to fans and such. 45dBA is common, so use that lower limit control as you would a noise gate.
DD
Old 12th September 2011
  #60
Here for the gear
 
Deptronic's Avatar
Awesome. Had to stop creeping to thank you for this.
Post Reply

Welcome to the Gearslutz Pro Audio Community!

Registration benefits include:
  • The ability to reply to and create new discussions
  • Access to members-only giveaways & competitions
  • Interact with VIP industry experts in our guest Q&As
  • Access to members-only sub forum discussions
  • Access to members-only Chat Room
  • Get INSTANT ACCESS to the world's best private pro audio Classifieds for only USD $20/year
  • Promote your eBay auctions and Reverb.com listings for free
  • Remove this message!
You need an account to post a reply. Create a username and password below and an account will be created and your post entered.


 
 
Slide to join now Processing…
Thread Tools
Search this Thread
Search this Thread:

Advanced Search
Forum Jump
Forum Jump