PDA

View Full Version : Testing room/speaker acoustics



A.S.
06-03-2006, 09:29 AM
I have invested many hours this weekend measuring various speakers in my 4 x 4m (approx.) listening room/study at home. I could not believe the results. I tried various (budget) microphones (such as the ECM8000) and finally used my master reference B&K 4133. I still didn't believe the results.

After about ten hours (wasted) I eventually discovered five factors that may be of interest to anyone who is contemplating measuring room acoustics and/or planning to use Digital Room Correction ...

1. The selection of measuing microphone is absolutely critical. I have measured three of these cheap uncalibrated mics. As (bad) luck would have it, the one I had at home - my No.3 ECM8000 - is up by a couple of dB's relative to my calibrated B&K reference 4133 at around 10kHz, and down by a couple at 20kHz, whereas my No. 1 ECM8000 (at work) has a slow roll-off starting at 10kHz. That's pretty good for budget microphones but do remember that unless you have a known calibrated mic to compare against you really are up a gum tree.

2. Don't assume (as I did) that CoolEdit/Adobe Audition generates pink noise with an absolutely flat spectrum. It doesn't. So any CD burned from this will exhibit the HF droop or whatever which will corrupt the accuracy of the room measurements.

3. Don't assume (as I did) that PC sound cards (and especially laptop sound cards) are flat with frequency: they are not.

4. Don't assume (as I did) that external USB sound cards are flat. Some are much flatter than others as I found out. Use 48kHz not 44.1kHz sampling rate.

5. Don't assume (as I did) that the CD player has a flat frequency response. Ditto the amplifier.

Taken altogether, the errors were far more than acceptable: an over-read of about 2dB at 10kHz followed by an under-read of -6dB at 20kHz and various problems below 200Hz. Insignificant in the course of normal music but completely unacceptable for reference measurements.

Recommendation: check the spectral content of the pink noise (loopback test, output to input) before you start. Use a known, calibrated mic. Yes, expensive.

Most of all, if your ears tell you it sounds OK but your test equipments doesn't, refuse to believe your test equipment and find out why before proceeding, however long that takes.

A.S.
06-03-2006, 10:14 AM
...used my master reference B&K 4133. Follow up:

Unfortunately, there is a new twist. After yet more testing using a small speaker (with an obviously realistic size/bass output) it seems that my 4133 capsule (shown below) has developed an internal random noise problem (the capsule now generates spurious noise within itself) at low frequencies which is giving a overestimation of the speaker's actual bass output. The test equipment can not distinuish between the acoustic output of the speaker and that of the rumble generated in the microphone capsule. I experienced something similar about 15 years ago.

To be absolutely sure about frequency response I have just ordered two brand new 4191 new-generation reference microphone capsules (that screws onto the B&K preamp) now from B&K Denmark (cost about USD2500) with a fresh calibration certificates. We have to be 100% certain in matters of frequency response. http://www.bksv.com/3032.asp

So typical of engineering problems - layer upon layer of variables. But if you have an internal frame of reference ("I don't care what the test equipment says: I believe that it sounds OK/bad - let's find out why") you eventually get to the truth.

P.S. Danish Pro Audio (DPA) and B&K are not the same company although they are both Danish. B&K exclusively sell the calibrated reference (measurement) microphones but they do not sell studio mics: DPA do, but they don't sell reference mics. As to which of these companies now actually manufactures the capsules I do not know except to say that B&K designed and manufactured the reference mics for decades.

Ned Mast
04-05-2006, 06:50 PM
Alan,

Using a Radio Shack SPL meter (along with Rives Audio correction factors) and a Stereophile test CD with warble tones a 1/3 octave from 200 Hz down to 20 Hz, I found that my SHL5's, in the configuration in which I had them set, had a 8 to 9 db dip in one speaker only at 100 cycles (even though they looked to be symetrically set up in the room). They were close to flat in the other frequencies (with, of course, the expected drop off below 40 Hz). By moving the speakers around to various locations in the room, I was able to eliminate that dip, at the cost of a bit more roll-off in the bass (they're now farther away from any walls) and some peaks of about 6 to 7 db in the area between 200 and 100 Hz in one speaker. From this experience I conclude that speaker placement is not a simple thing, but is exceedingly important (I would not attempt it again without measurement). Finally, my question: how close to flat frequency response is it practical to strive for? I do have an equalizer in my system, so can bring down peaks (I understand that boosting dips is NOT a route to take).

Thanks,
Ned

A.S.
05-05-2006, 06:18 PM
... how close to flat frequency response is it practical to strive for? I do have an equalizer in my system, so can bring down peaks (I understand that boosting dips is NOT a route to take).NedWell, I think what you have demonstrated is the reality of real loudspeakers in real rooms and the miraculous ability of the human ear to hear through all those peaks and troughs.

In my experience, a truly flat anechoic-like response below, say, 500Hz is neither possible nor, surprisingly, is it of paramount importance. I don't (personally) think that spot tones (even with warbles) will really give you a useful view of the room's acoustics - a 1/3 octave spectrum trace on-screen would be much more intuitive. There are several simple, possibly even free octave analyser plug-ins for your soundcard that will be much more useful to you and will, by they way they work, smooth out those 'hot' frequencies into a much more meaningful trend line. I wouldn't use anything else. {If you want to see the process in action, become a tester for the streamed a/v from my very own listening room - see the thread in Misc. - you'll see how I do it, live}.

Having worked through just this exercise recently, I can pass on with confidence my findings: the 100Hz region contrasted to the 3kHz region should be weighted about 1:20 for its cricical importance to perceived audio quality: I have astonished myself (again!) just how ultra critical that crossover band is where the tiniest fractions of a dB make the difference between sweet and clean or agressive and fatiguing - despite what the octave analyser or simulator may say is a perfect frequency response. Not to my tired old ears it's not!

PS: An afterthought, as I sit here listening to Steely Dan (I've been a big fan for 30 years) rather too loud (?) .... if the speakers are doing their job properly then no matter how loud well-recorded human voice is replayed, it should still sound sweet.

Ned Mast
06-05-2006, 11:20 PM
Thank you for your response, Alan. I shan't be overly concerned, then - the SHL5's sound wonderful with a bit of equalization, but then they sounded wonderful before the eq!

Ned

Douglas G A Murray
07-05-2006, 01:32 AM
5. Don't assume (as I did) that the CD player has a flat frequency response. Ditto the amplifier.

[/I]

Hi Alan

Alan, I was aware of the mike issue. But CD and amplifiers (solid state) these days not flat within reference mike accuracy??!!

Please elaborate.

Many thanx

Doug

Douglas G A Murray
07-05-2006, 01:38 AM
Ned, here is what I posted on the REG audioforum regarding the radio Shack meter and measuring:

(See http://groups.yahoo.com/group/regsaudioforum/message/7043 for the graphs.)

Room = 5.0 m long * 3.9 m wide * 2.5m high

SHL5?s symmetrically placed firing down the long dimension of the room. Distances from middle of front baffle of speaker to the walls : side = 103 cm; back = 89 cm.

There is a 1.5m wide archway at the back of the room behind the listening position, which is 90 cm from the plane of the back wall. The archway leads to a space extending approx 9 m further back and 3.5 m wide ? the dining and kitchen area.

The speaker are slightly toed in, angled to point at the centre of the archway. This is not 90 deg. classical stereo positioning, but one which is designed to minimise high freq. reflections, direct sound getting lost in the dining area and kitchen beyond the arch.

The top graph is the in-room response I measured (both channels active) using Stereophile Editor?s Choice CD with warble tones and Radio Shack SPL meter. The room is fairly acoustically dead.

The bottom graph is the near field measurement (one channel only active) at one meter away from the bottom tweeter. Here I made sure that all 1st reflection points were well padded except for the ceiling.

(here are the relative measurements if you cannot see the graph ? 1 kHz was set to ~ 70 dB.) I used the correction factors at the Rives website.

I am still experimenting but I have yet to find a speaker & seating position that gives me a good upper bass (120 ? 200Hz) and lower mid-bass (60 ? 80 Hz). I seem to get suck-out at one or the other. I have chosen at the moment to have a 5 dB suck out at (60 ? 80 Hz) and elevated upper bass rather than 12 dB suck out in the 120 ? 200Hz region. I managed the latter in spite of taking precautions in order to avoid Allison effect.


Comments:

1. Room bass extension is great down to < 30Hz

2. Hopefully I can find a position where 60-200 Hz is elevated in the bass ready for EQ. To date I have not found a position without at least 5 dB troughs.

3. I had heard about the dip around the SHL5 crossover. I did not expect that it would be as deep as this!

4. I did not expect to see that degree of extra energy at about 6 ? 8kHz. See the near field up by ~6 dB! Surely this spells disaster when combined with the hearsay that older mikes had a peak in this region. And I have a large number of recordings of the late 60?s and 70?s. Decca, EMI, Deutch-Grammophone.

5. What really caught my attention is the big roll-off > 10kHz.



Number 5 got me thinking. Radio Shack say that the SPL meter response is C-weighted. C-weighting is rolled off at the bottom AND above 10 kHz. Yet the correction factors universally applied imply that the Radio Shack is fairly flat above 10 kHz ? nothing like the C-weighting. Frankly, I do not believe the near-field measurements above corrected as per the Rives figures above 10 kHz.


Rives also imply on their site that the Radio Shacks are pretty consistent and that they have verified the correction curve.

So, what is going on here?


Further the Spec on the Radio Shack is accuracy of +/- 2dB (at loud and 1 kHz ? what about other frequencies).


I am suspicious of the correction factors? validity and frankly a good deal sceptical of all ?ordinary? Radio Shack based home attempts at EQ, except for correcting the wildest of tonal imbalances.


And then I saw this posted by Alan Shaw: mikes, CD players and amps (CD player and amps, mark you) all conspiring together to confound and render ordinary home EQ efforts futile!!

A.S.
07-05-2006, 09:22 AM
The top graph is the in-room response I measured (both channels active) using Stereophile Editor?s Choice CD with warble tones and Radio Shack SPL meter. ... 5. What really caught my attention is the big roll-off > 10kHz. Number 5 got me thinking. Radio Shack say that the SPL meter response is C-weighted. C-weighting is rolled off at the bottom AND above 10 kHz Couple of comments here (and I can't actually see the graphs so I am working from your text) ...

A. You must *never* drive both speakers at the same time when you make speaker/room measurements (this is implied in what you say). Your microphone will sum both constructively and destructively the sound waves from the two speakers and this wil definitely effect the frequency response measurement to the point that they will be useless.

B. The microphone should not have any weighting at all. It should be calibrated to a known standard and with an up to date certificate. If the 'C' weighting can not be turned off (to what B&K call 'linear') then best foget the whole exercise and put the meter in the bin. It would be absolutely useless for the purposes of accurate measurement. (I looked up the C curve: it's roughly -12dB at 20kHz). Any room correction system based on a C weighted room acoustic measurement would definitely have to be aware of that otherwise the result would be that the room correction system would be presented with under-read of the room for which it would attempt a (needless) massive boost the speaker's output at the top end - the result would sound horrible. I guess - only a guess - that the claimed C weighting masks a rather poor capsule and/or electronics. As for the ballistics of the meter etc. etc. ..... error on top of error.

C. Please don't ever use warble tone: I feel very strongly about this. I only use (1/3, 1/6, 1/12 or even 1/24th) octave filtered pink noise and I can not imagine any professional speaker designer using anything else - certainly not warble tones: Extremely difficult to interpret and present the room as a frightening acoustic mess (which it is). Remember: our ears can hear through almost all and every defect in the room.

Pink noise has been the standard method for room measurement for as long as I can remember and is a noise spectrum which mimics that of music averaged over many minutes. Prove it yourself: using a computer based FFT spectrum analyser, hook this up to a CD player and average an entire classical music CD. Compare that spectrum with averaged pink noise which is therefore an appropriate stimulus for testing speakers in rooms and it gives you a trend line which you can usefully interpret to guide you as to adjusting speaker position etc..

For room acoustic work, it is absolutely necessary to average the pink noise, typically over 5 seconds or so (easily done in the spectum analyser software): just watching the display jumping around isn't going to give you a very useful picture.

All the above highlights the danger of using non-professional gear and why we are obliged on your behalf, to invest in Bruel & Kjaer measurement equipment regardless of cost. B&K is the Rolls Royce of audio measurement confidence. (B&K is not Danish Sound Technology, although there is some connection between them).

s.a.b.
07-05-2006, 10:45 AM
Alan ,

Since you mentioned Steely Dan (and I confess that though I owned AJA on vinyl, I never replaced it on CD), how much "bite" should the brass have?

This is an area (and I've used Van Morrison 's "Jackie Wilson Said" as a test) where most speakers don't sound not right. Actually to my ears, not enough "bite", at least not enough in the right manner.

As far as a truly off-the-wall suggestion, if you like Steely Dan, check out the recent EMI remasters of the Donovan albums from the Sunshine Superman/Mellow Yellow years- jazz, as well and folk and pop influences, recorded in a beauitully simple, closely miked chamber muisc like style.

(I know you're thinking that Donovan was a naive hippie as compared to the intelliegently caustic/cynical Steely Dan, but I don't think they ever wrote a song with the edge of "Season of the Witch".)

A.S.
07-05-2006, 12:52 PM
Since you mentioned Steely Dan (and I confess that though I owned AJA on vinyl, I never replaced it on CD), how much "bite" should the brass have?

This is an area (and I've used Van Morrison 's "Jackie Wilson Said" as a test) where most speakers don't sound not right. Actually to my ears, not enough "bite", at least not enough in the right manner.Now this really is an interesting point and one which I've mulled over for many years.

First of all, the way a brass instrument projects sound is quite unlike others in the orchestra. You can think of a brass instrument as generating a penetrating searchlight beam akin to that of a lighthouse on a foggy night. The other instruments are more ominidirectional.

Not only does the brass section have unique directionality, but the waveforms they produces are more like an electronic tone (say, a square wave) than a sine wave. Taken together, brass is, I agree, an extremely difficult instrument to both record and reproduce. If you look at the spectral content of a brass waveform it is quite distinctive and has a rich, complex harmonic structure with a lot of energy in those harmonics which require their soundwave to be generated simultaneously by the bass/mid driver and the tweeter - right through the crossver region and at a high level. This is where the difficulties for the designer really lie: relative to the fundamental, the harmonics are so strong that any mis-integration is audible and in a poor speaker tends to tear the instrument into two audible parts: the lower part (from the bass unit) and the higher frequencies (from the tweeter), neither fully blended.

I don't think it is really possible to perfectly reproduce brass on any system - but you can get close. A further complication is that the shape of the 'bell' or mouth of a brass instrument is conceptually the same as that of a woofer's cone - and that means that one mouth is trying to speak for another mouth. See the problem? It's the same issue as a speaker at home (a box) playing in your room (another box) reproducing the sound of the studio or hall (both boxes) in which a cello (a highly resonant box) is playing ... you have a box in a box in a box reproducting a box, each picking up a little bit of character along the way.

I can 'spice up' brass reproduction on a speaker system (by tuning the crossover) so that doubtless you'd love the way brass sounds - a nice 'bite', plenty of incisiveness - but the problem is that when we then play those other less directional instruments (and voice) with softer harmonics you'd recoil in horror. In fact, my eldest son was recently visiting at the same time I was experimenting with a crossover design, and Steely Dan was playing at the time. I demonstrated how I could peak-up the brass until it was really vivid, and we both agreed (I a little suspiciously, since I've been down this road many times) that the overall effect was exciting. "That's it Dad: stop fiddling - you've cracked it ..." were his famous last words. Of course, he leaves, I play different (orchestral, vocal, lounge jazz) music and over a period of days reverse the circuit design back to where it was. Don't ever design by committee! Trust your own judgement!

Further thought: brass, more than perhaps any other instrument, highlights the unique clarity of the Harbeth RADIAL cone. Try brass on polypropylene for example, and brass' biting edge is suppressed to one extent or another. On Kevlar, it tends to sound the opposite: forced and cup-like. Remember what I said: each step in the reproduction chain is polluted to one degree or another by the characteristics of that part, and that character can not be removed later.

You can not chemically remove the taste of peat from whisky; it is bound up into the fabric of the final experience.

Douglas G A Murray
09-05-2006, 01:37 AM
Couple of comments here (and I can't actually see the graphs so I am working from your text) ...

A. You must *never* drive both speakers at the same time when you make speaker/room measurements (this is implied in what you say). Your microphone will sum both constructively and destructively the sound waves from the two speakers and this wil definitely effect the frequency response measurement to the point that they will be useless.

B. The microphone should not have any weighting at all.... If the 'C' weighting can not be turned off (to what B&K call 'linear') then best foget the whole exercise and put the meter in the bin.

C. Please don't ever use warble tone:



My thoughts arising from the quote:

1. Surely the point of in-room measurement is to investigate, not how one speaker behaves, but how 2 speakers sound at a particular point in the room? (see point A quoted)
Do you wish to respond to this, Alan?


2. AS's recommendation to NEVER use warble tones makes theoretical sense to me and I can say that my recent experience is supports this - wild variation of Radio Shack response from position to position as little as 30 cm apart while my ears were telling me that there was not that much change in the sound of either pink noise or music. In the light of this, the prevalent dissemination of warble tones from folk like JA of Stereophile bedazes me.

3. It seems that AS does not endorse the use of the Radio Shack meter. I had come to this conclusion myself (see my 1st post on this) for anything but the craziest sound imbalance. Note that I did use the "official" Rives Radio Shack meter correction factors, but I am suspicious of these too.

6. Maybe AS's point A explains, at least partially why I have not yet found a position where there is not a deep trough 60 - 80 Hz.

7. Maybe many of the MEASURED peaks and troughs owing to 2 speakers giving rise to constructive and destructive interference (AS'as point A) are not musically significant for the following reason: Interference occurs at a point in space, approximated by the mike's aperture, whereas your ears (certainly mine!) are not located at the same point in space. Thus the interference pattern with 2 speakers as source will be totally different at the mike diaphragm compared to 2 ears separated by a head. I would guess that one would need to mimic human hearing with some sort of synthesis of 2 pickups.

Alan, as always I would apreciate your comments as you see appropiate. Also, I was interested to read that you do not assume CD and amplifiers (solid state) flatness. I had thought that this was the least of worries these days. Would you care to elaborate?

Regards

Doug

A.S.
09-05-2006, 08:49 AM
Surely the point of in-room measurement is to investigate, not how one speaker behaves, but how 2 speakers sound at a particular point in the room? (see point A quoted)
Do you wish to respond to this, Alan?

... as always I would apreciate your comments as you see appropiate. Also, I was interested to read that you do not assume CD and amplifiers (solid state) flatness.Yes it may be counter-intuitive to measure only one speaker at a time but as humans we have two ears and the ability to discrimInate direction (etc.). But the single measurement microphone can't: it just reports the sum total of the energy it receives, from the two speakers playing together.

Consider this: where the path length from speaker A is different to that from speaker B to the mic, if they are exactly half a wavelength different, as far as the mic is concerned that is total cancellation = a deep (infinitely deep?) notch. So, although the speakers are flat in themselves at that frequency, in that precise place in the room with the mic at that exact point as far as you the tester is made aware by the meter, there is a drop in energy.

As for amps: whether it's a design issue or not I don't know, but one I checked recently was -6dB at 20kHz - not that I think that would necessarily be audible. Let's not forget that amps run hot and hence do age. It's interesting though that an old, never-serviced Quad 303 that I have on my room measurement test trolley is as flat as a pancake even after all these years.

T.W.
01-05-2007, 09:02 AM
I found a handheld SPL meter called Acoustilyzer from NTI. It seams to be a good solution to have a rough idea of room acoustics and speaker position. I wonder if anyone used it / owns it. Here is a link.

http://www.nti-audio.com/Home/Products/Minstruments/AcoustilyzerAL1/tabid/60/language/en-US/Default.aspx

Together with the MiniSPL microphone it seems to be a good starting point for an acceptable price.

What do you think?

al2002
04-05-2007, 05:05 PM
Have you considered the ACO Pacific capsules? They are cheaper - at least here in the US - than the B&Ks and are fully compatible. Mine is 8 years old and has been trouble free - touch wood - to date.

Also, Liberty Instruments sells Panasonic Electret based mikes for under $ 150. The come with an individual calibration file that can be imported into LAUDm Praxis and so on. DIYers can buy the capsules only with a calibration disc for much less.



Follow up:

Unfortunately, there is a new twist. After yet more testing using a small speaker (with an obviously realistic size/bass output) it seems that my 4133 capsule (shown below) has developed an internal random noise problem (the capsule now generates spurious noise within itself) at low frequencies which is giving a overestimation of the speaker's actual bass output. The test equipment can not distinuish between the acoustic output of the speaker and that of the rumble generated in the microphone capsule. I experienced something similar about 15 years ago.

To be absolutely sure about frequency response I have just ordered two brand new 4191 new-generation reference microphone capsules (that screws onto the B&K preamp) now from B&K Denmark (cost about USD2500) with a fresh calibration certificates. We have to be 100% certain in matters of frequency response. http://www.bksv.com/3032.asp

So typical of engineering problems - layer upon layer of variables. But if you have an internal frame of reference ("I don't care what the test equipment says: I believe that it sounds OK/bad - let's find out why") you eventually get to the truth.

P.S. Danish Pro Audio (DPA) and B&K are not the same company although they are both Danish. B&K exclusively sell the calibrated reference (measurement) microphones but they do not sell studio mics: DPA do, but they don't sell reference mics. As to which of these companies now actually manufactures the capsules I do not know except to say that B&K designed and manufactured the reference mics for decades.

al2002
04-05-2007, 05:19 PM
I have the NTI Minilyzer - same size as the Acoustilyzer, but intended for electrical measurements - and Minirator. The display on the Minilyzer is too small to be useful for accurate redout of graphical information You really need to have the USB interface installed and upload the data to a PC to use this to its full advantage.

Alas, the Acoustilyzer + Minirator combo is not cheap. For similar money I would suggest you get a fully featured PC based test system like Praxis.


I found a handheld SPL meter called Acoustilyzer from NTI. It seams to be a good solution to have a rough idea of room acoustics and speaker position. I wonder if anyone used it / owns it. Here is a link.

http://www.nti-audio.com/Home/Products/Minstruments/AcoustilyzerAL1/tabid/60/language/en-US/Default.aspx

Together with the MiniSPL microphone it seems to be a good starting point for an acceptable price.

What do you think?

A.S.
17-05-2007, 12:36 PM
It's a difficult choice between a hand-held instrument which has limited capabilities but a short learning curve and the excellent PC-based PRAXIS system which is so much more capable but has a steeper learning curve. (We use the former Liberty Audio Suite which is DOS based and a delight to use).

Praxis link here (http://www.libinst.com/).

You may be interested in our User Group thread on room acoustics: http://www.harbeth.co.uk/usergroup/showthread.php?t=257

al2002
20-05-2007, 01:07 AM
Good point. If you can afford only one, PRAXIS is the way to go.


It's a difficult choice between a hand-held instrument which has limited capabilities but a short learning curve and the excellent PC-based PRAXIS system which is so much more capable but has a steeper learning curve. (We use the former Liberty Audio Suite which is DOS based and a delight to use).

Praxis link here (http://www.libinst.com/).

You may be interested in our User Group thread on room acoustics: http://www.harbeth.co.uk/usergroup/showthread.php?t=257

A.S.
20-05-2007, 02:55 PM
Do you use PRAXIS? I have corresponded with Bill Waslo (and Tony Seaford) over the years and find Bill to be hugely approachable and pragmatic. Other test equipment suppliers have a tendency to be defensive and disinterested - even disingenuous. Bill/Tony's positive attitude alone makes the purchase of PRAXIS a zero-risk investment and I endorse your recommendation.

al2002
14-06-2007, 04:52 AM
Do you use PRAXIS? I have corresponded with Bill Waslo (and Tony Seaford) over the years and find Bill to be hugely approachable and pragmatic. Other test equipment suppliers have a tendency to be defensive and disinterested - even disingenuous. Bill/Tony's positive attitude alone makes the purchase of PRAXIS a zero-risk investment and I endorse your recommendation.

I use Praxis as an amateur. It has a very full feature set and is quite useful.

Agree with you 100% regarding the Company. Bill and Carol are both very helpful and provide first-class customer service.

A.S.
16-09-2007, 12:27 PM
From www.libinst.com (http://www.harbeth.co.uk/usergroup/www.libinst.com)

There doesn't seem to be a current active forum for this system, which I have just purchased to complete the M40.1 development. It is complex, but extremely flexible and capable.

Are there any PRAXIS users out there?

oferab
02-10-2007, 01:20 PM
I use Audio Precision MLS measurement to measure the frequency response of the speaker, you can eliminate the reflection problem using this system, and if you have a anechoic room, you can also measure low frequencies

A.S.
02-10-2007, 01:55 PM
That's an extremely expensive system! I think you'll find that LOG CHIRP (a sort of swept sine) is more useful in reflective environments than windowed MLS. You do not need an anechoic room to measure the low frequencies: you can use a stitched near-field measurement if you are very careful.

MLS doesn't really put enough energy into the low frequencies to get a truly reliable/believable result in my experience but the AP kit is very fine indeed for amplifier measurement.

zenith
04-10-2007, 11:33 AM
You can not chemically remove the taste of peat from whisky; it is bound up into the fabric of the final experience. Very true but sadly they are still finding ways to add the peaty taste that so many of us love with additives.
If they can find a way to ruin perfection for sound commercial reasons they will.

Just as an aside what is your favoured malt ? Mine is Aardbeg the original 10 year old but I know it is a little too idiosyncratic for some.

STHLS5
16-03-2011, 11:18 AM
Hi everyone and especially to Alan,

This is my first attempt using REW and not sure if I have done it correctly. I used a SPL meter and also a small condenser mic used to setup my Onyko AV but found the waterfall plot didn't reflect what I hear in the room. The bass in my listening room is sufficient but the waterfall plot shows it is lacking. I suspect I didn't calibrate the sound card correctly or there could be software error in not compensating for the C or A weighting. There is slight error in the reading around 50Hz which is due to the air cond hum like noise.

Any help is greatly appreciated. I tried at REW forum but so far no response.

Thanks in advance.
ST


http://img535.imageshack.us/img535/2386/spectg.jpg

http://img815.imageshack.us/img815/2986/waterfallplot.jpg

A.S.
16-03-2011, 11:30 AM
OK I'm just passing by.

First, and I'm not criticising you please understand, what do you think those pretty charts are actually saying to you, the reader? Before we can even begin to interpret them as science we have to understand why they are not just attractive pictures.

What are they attempting to convey? Or specifically, why have you been persuaded to invest your valuable time and money in this measurement system as opposed to any other? What convinced you of its merits? Why not just trust your ears as you have before? I really like the colours of the top chart - would make a great wall-picture!

P.S. Don't forget what we said here about reporting and influencing (http://www.harbeth.co.uk/usergroup/showthread.php?981-Doing-experiments-reporting-results-influencing-other-people&p=12295#post12295)!

STHLS5
16-03-2011, 12:02 PM
....First, and I'm not criticising you please understand...
...... I really like the colours of the top chart - would make a great wall-picture!

There goes whatever hope of getting any help about those pretty pictures. :(

ST

A.S.
16-03-2011, 01:07 PM
Would you mind giving me a decent answer to my entirely reasonable question if you don't mind? I really don't have the time to waste on sarcasm.

I am not here to sell or advise on third party room correction systems and whatever input I care to make deserves a proper answer. If you are unable to advise us here on how you interpret those graphs, as you have experienced from the manufacturers themselves, you can't expect me to do all the thinking.

But I do like those colours very much. They remind me of a pull-up poster we designed and still sometimes use at exhibitions.

george_k
16-03-2011, 10:10 PM
Hi ST,

Probably better suited you ask in the REW forum (google it). Two things I can think of off the top of my head:is your mic calibrated to account for bias it may be introducing? Same applies to the sound card. Next you should minimize ambient noise such as you a/c unit as it introduces further bias.

To Alan's question what this plot will identify (in theory) are room modes whereby certain frequencies will appear as peaks and nulls depending on how your speaker interacts with the room. In my case I have a 80hz peak in my room (resonant freq.) and I could probably benefit by adding some additional absorption or by repositioning my speakers. Either way you don't need to rely on a plot to tell you if something doesn't sound right, try experimenting with the above suggestions and see if that helps.