Reflected sound vs SPL meter [Archive] - Audio & Video Forums

PDA

View Full Version : Reflected sound vs SPL meter



toenail
03-30-2005, 03:07 AM
Just curious about how the RS SPL meter measures sound. The name would indicate that it is measuring "pressure", but I'm left wondering a few things. Is this pressure the combination of direct sound from the speaker as well as reflected sound from the surroundings? The reason I ask is that when using the SPL meter to calibrate surround settings I'm sometimes left with a perceived imbalance. This imbalanced may well by my hearing ability as opposed to anything else, but it got me thinking.

Due to lack of ideal placement opportunity one of my mains sits with a glass door within 6" or so of the edge the speaker cabinet. This provides a considerable amount of early reflection, especially in the higher frequencies. If I use the SPL meter to level match it with the other main, the speaker in question either sounds too loud, too quiet, or just right depending on the movement of my head within a few inches of center. I assume this is caused by cancellation or reinforcement due to the glass door acting as a wall right next to the speaker. What's odd is that if I try to measue this difference in percieved sound with the RS SPL meter, it doesn't show up.

I know directivity places some role in how the device measure SPL as I've observed it while rotating on it's vertical axis and it can vary by 1-2db if not pointed reasonably straight toward the speakers being tested. Not sure if this has anything to do with my observations. At the end of the day I end up using the SPL meter for initial set-up but invariably tweak slightly during playback to get optimized stereo/surround effect.

I also noticed talk about listening levels of 85db on a few threads. The other day I tried that and concluded that it is far too loud for my taste for more than just a few minutes. Not sure if this is sensitive hearing (which I definitely have), reflected sound not accounted for by the SPL meter, or distortion from playing the system at -20db (manufacturer says it should be fine at this level). Any thoughts?

kexodusc
03-30-2005, 05:09 AM
Good bunch of questions here, I don't think I can answer them all but I'm always fudgin' around with these things too..


Just curious about how the RS SPL meter measures sound. The name would indicate that it is measuring "pressure", but I'm left wondering a few things. Is this pressure the combination of direct sound from the speaker as well as reflected sound from the surroundings? The reason I ask is that when using the SPL meter to calibrate surround settings I'm sometimes left with a perceived imbalance. This imbalanced may well by my hearing ability as opposed to anything else, but it got me thinking.

Your hearing no doubt plays a role, but unless you have reason to believe your hearing is off by a few dB in one ear, let's ignore that for now...there's really nothing we can do about it anyway other than boost one side to compensate, but that might make it imbalanced for anyone else listening (if you let them).
The reflected sound is captured by the SPL reading...I don't think it's a matter of pressure from the initial sound plus pressure from the reflections, but rather the reflections contribute to the total SPL...think about it...if the reflections weren't part of the total, the resulting effect would be sound waves continuing to travel in their initial direction (or being absorbed)...this would sort of be the same as having a larger room...

Also, be aware that not all frequencies are being reflected at the same intensities...


Due to lack of ideal placement opportunity one of my mains sits with a glass door within 6" or so of the edge the speaker cabinet. This provides a considerable amount of early reflection, especially in the higher frequencies.
I think the highly directional, ultra-high frequencies are probably least affected by this, but it's proably the higher end of the midrange which is the most annoying kind of reflection that's giving you trouble...in my old house, we had a similar setup with glass doors nearby, and I ended up buying a cheap blind that had a fabricy texture to it to absorb some sound...helped quite a bit....


If I use the SPL meter to level match it with the other main, the speaker in question either sounds too loud, too quiet, or just right depending on the movement of my head within a few inches of center. I assume this is caused by cancellation or reinforcement due to the glass door acting as a wall right next to the speaker. What's odd is that if I try to measue this difference in percieved sound with the RS SPL meter, it doesn't show up.

Curious, what kind of test tones are you using here? Pink noise? Are your speakers fairly sensitive to toe-in for proper imaging. The Peerless drivers in my HT speakers have exceptional off axis response which allows me to use little toe-in to caputure both wide open soundstaging and a solid imaging. The Paradigm Studio 40's they replaced required a bit more toe in to find the proper mix, it could be as simple as revisiting speaker placement.
I wonder if early reflections are the culprit here? The first part on the wall where the sound is reflected back to you? Hopefully Sir T and Woochifer can chime in.



I know directivity places some role in how the device measure SPL as I've observed it while rotating on it's vertical axis and it can vary by 1-2db if not pointed reasonably straight toward the speakers being tested. Not sure if this has anything to do with my observations. At the end of the day I end up using the SPL meter for initial set-up but invariably tweak slightly during playback to get optimized stereo/surround effect.
Most of the instructions I've seen (including AVIA and DVE) say to place the SPL meter pick up at a 60 degree angle upwards, or upright altogether so the mic isn't "pointed" at any speakers...and keep in mind the Radio Shack SPL meter (which most of use) isn't the most accurate piece of gear...I have a galaxy audio SPL meter and another one who's name escapes me now (Eriksson?) that are far more accurate...but to be fair, the +/- 2 dB readings the RS meter is off by from about 500 Hz to 10kHz is pretty impressive, especially if I told you how much more expensive these other two units were.

Sometimes the recordings aren't 100% accurate either, a mic might have been placed off center, the artist off center, etc...nothing wrong with you tweaking the levels to your liking.

I
also noticed talk about listening levels of 85db on a few threads. The other day I tried that and concluded that it is far too loud for my taste for more than just a few minutes. Not sure if this is sensitive hearing (which I definitely have), reflected sound not accounted for by the SPL meter, or distortion from playing the system at -20db (manufacturer says it should be fine at this level). Any thoughts?

You're not alone here. I have 24 X 22 room...I listen to TV's, movies and music anywhere from 60 dB to 75 dB for the most part, allowing peaks into the high 80's/low 90's....a few times a week or so I crank the bejeezus out of the system for loud annoying music, or a good movie with lots of action, but that's plenty...Just enough to know my system can dish it out...some classical music I listen to just sounds better with loud peaks, too so I set the peaks where I want them and (I know about what level on my volume indicator now) and the softer passages are appropriately loud, no harsh surprises that way, and no lady yelling at me to turn my system down.

These guys that listen to levels above 85 dB are in for a rude awakening....Exposure to sounds above 85 dB causes short term hearing losses. Repeated exposure over time results in permanent damage to the ear, especially in the upper midrange frequencies (kind of important)...The amount of time you can listen to loud noise is inversely proportional to the volume...these poor guys are going to damage their hearing, and then turn it up even more to compensate...If you're like me and love this hobby, love music and movies, take care of your ears! This alone should be reason enough for everybody with a home theater or stereo system to own an SPL meter! $40 is a lot cheaper than the health insurance premiums for hearing aids, etc...

Feanor
03-30-2005, 05:30 AM
...
You're not alone here. I have 24 X 22 room...I listen to TV's, movies and music anywhere from 60 dB to 75 dB for the most part, allowing peaks into the high 80's/low 90's....a few times a week or so I crank the bejeezus out of the system for loud annoying music, or a good movie with lots of action, but that's plenty...Just enough to know my system can dish it out...some classical music I listen to just sounds better with loud peaks, too so I set the peaks where I want them and (I know about what level on my volume indicator now) and the softer passages are appropriately loud, no harsh surprises that way, and no lady yelling at me to turn my system down.
...
Great comments from kexodusk, by the way.

I'm always astonished when somebody says their typical listening level is around 85 dB. For me that's far too loud. My own typical average is around 70 dB when I'm doing serious listening, (not background). When listening, say, to a symphony with loud crescendos, my peak level might get as high as 82-85 dB once in a while.

FLZapped
03-30-2005, 08:30 AM
The SPL meter has a microphone in it and does the opposite of the speaker, it conversts mechanical movement into an electrical signal, from there, the engineers figure out how that correlates to sound pressure level.

Sounds in a room will have nodes. Remember your water tank diffraction experiment? Same thing. The nodes are all the result of algebraic summation of direct and reflected sound. In other words, either additive or subtractive, so you will get hot spots and nulls as a result. So as you walk around your room and point the meter in different directions, you will get varying results.

The more broken up you can make these nodes, the better the room will "sound."

Go to the RPG website and look at their research stuff, you will be able to see the graphs produced on how they "fixed" the room responses.

http://www.rpginc.com/research/index.htm

In particular, this has about the easiest to see graphs:

http://www.rpginc.com/research/scs1.htm

I would also recommend the Master Handbook of Acoustics, by Everest

http://www.amazon.com/exec/obidos/ASIN/0071360972/qid=1112198988/sr=2-1/ref=pd_ka_b_2_1/104-9865245-9827955

-Bruce

Quagmire
03-30-2005, 09:01 AM
Whether or not the reflected sounds off of the glass are captured by the SPL meter (and they are) the early reflection will create other sonic anomalies related to time alignment and phase coherency which are audible.

It's not unusual at all that you find an 85db reference level too loud. Most home applications set this reference level at 75db which is still very loud.

Q

Woochifer
03-30-2005, 02:17 PM
Hmmm, a lot of good questions here. The readings on that SPL meter can be influenced by a lot of different things, particularly objects close to the meter when you're taking the readings. The rule of thumb usage recommendations on the SPL meter are to hold it at about a 45 degree angle at ear level, and to maintain that position for all readings.

With that said, plenty of factors can influence what the SPL meter picks up. The others have already touched upon the reflected sound and the room acoustics. Expanding on those points, one of Radio Shack's recommendations is to situate yourself perpendicular to the SPL meter when doing your measurements. This is because when you seat yourself directly behind the SPL meter, your body can reflect the sound waves and create an erroneous reading because the reflections occur so close to the mic. The sound reflections that you get from throughout the room create audible effects that you should account for, but I've read that doing a SPL meter reading with your body in close proximity to the meter can alter the SPL readings by upwards of 4 db.

For maximum consistency, I mount the SPL meter on a camera tripod when taking the readings. This ensures that the SPL meter will be at the same height and angle for all measurements, and it minimizes the reflections that occur in close proximity. The first time I used a camera tripod, I adjusted the levels on my speakers by up to 2.0 db.

The room acoustical effects have already been touched upon, but another issue that you should account for is the type of test tone you use. The test tone that most receivers generate for speaker level matching is wideband. The creates a problem with the SPL readings because those wideband test tones include the low frequencies. And because low frequencies are most susceptible to room-induced effects, they can generate erroneous readings because the SPL reading might be driven more by room-induced peaks or nulls in the bass than the prevailing levels in the midrange and highs.

The solution to this is to use a calibration DVD such as Digital Video Essentials, Avia, or the Sound & Vision Home Theater Setup disc. Those discs all use narrow band test tones for level matching that do not include the low frequencies. In my listenings, those test tones give you better balanced results than the wideband tones do.

Keep in mind that the role of the SPL meter is to efficiently establish a consistent and replicatable reference point. If your ears pick up something different, that's fine. The SPL meter reading is not the final arbiter, your ears and your preferences are. All that the SPL meter is supposed to do is ensure that all of your adjustments constently start from the same point. Where you go from there is up to you.

toenail
03-30-2005, 02:27 PM
Kex- I have considered fabbing up an absorption device for the glass door (actually there are two) but it needs to be portable (read-WAF) for use during critical listening only. I've got some spare egg crate foam laying around and may try that this weekend. The speakers (JBLHLS-610) are very sensitive to placement and are also sensitive to toe. Unfortunately the position I have them in is a compromise for reflections but optimum for image/stage. Test tones were the ones supplied by the receiver for level matching. Avia disc has been ordered. I understand that this may be more accurate? I'll also have to try setting the meter pointing straght up at ear level in the primary listening position. That way it will be equally off axis to all the speakers being measured.

My usual listening level doesn't register on the SPL meter (neighbors). I'll have to look into the whole node null thing.

toenail
03-31-2005, 02:20 PM
The test tones I've used for sorting low end with the BFD are from the Autosound 2000 collection and include reference tones from 10-98hz in 1hz increments. For the home theater speaker level setting I've used the on-board tones generated by the receiver. I've got the Avia disc on order (couldn't find it locally) and need to pic up a tripod for my camera anyway. I'll have to try some of these changes plus the absorption mats for the glass doors and see where I wind up.

Thanks much for the insightful contributions from all. A good day is a day I learn more about this hobby.

Sir Terrence the Terrible
03-31-2005, 02:36 PM
Kex- I have considered fabbing up an absorption device for the glass door (actually there are two) but it needs to be portable (read-WAF) for use during critical listening only. I've got some spare egg crate foam laying around and may try that this weekend. The speakers (JBLHLS-610) are very sensitive to placement and are also sensitive to toe. Unfortunately the position I have them in is a compromise for reflections but optimum for image/stage. Test tones were the ones supplied by the receiver for level matching. Avia disc has been ordered. I understand that this may be more accurate? I'll also have to try setting the meter pointing straght up at ear level in the primary listening position. That way it will be equally off axis to all the speakers being measured.

My usual listening level doesn't register on the SPL meter (neighbors). I'll have to look into the whole node null thing.

Toe,
Try a couple of layers of felt instead of the egg crate foam. It'll absorb a little more down into the midrange, it will block light, and it is removeable. I have three layers of felt covering a large picture window and it does a great job killing the light, and provide reflection control well into the lower midrange.

toenail
04-02-2005, 01:34 PM
Thanks to all for the helpfull replies in this thread (and some of my other threads while I'm at it). I spent a few hours fiddling this morning and came up with a few improvements in the system.

Using a DIY stand for the SPL meter (no tripod yet) and standing it on end so that the mic was skyward at ear level in the position normally occupied by my head, I set speaker levels using the tones on the receiver. Avia disc won't be here until Monday or Tuesday. There was some minor tweaking to get the levels the same, generally on the 1db to 2db side of things. I also re-measured distances at the primary listening position. I'd forgotten to do so after the removal of our christmas tree once we slid the couch back to it's rightfull position. Those adjustments were on the .5 to 1foot side of things. After testing a few reference DVD's I can say that there was definite improvement. Much more seemless integration of moving objects traveling front to back and side to side. Some effects had a better "suspended in air" or holographic quality where they were previously localized more toward the speaker.

I still had an issue with resting my head on the sofa cushion behind me. For some reason this position always caused the center image to get dragged to the right a touch while loosing it's sharpness. Moving my head forward by as little as two inches aleviated the problem but was uncomfortable for long periods of time. I finally hung something on the glass door immediately to the right of the right main speaker and the problem completely disappeared. I didn't use felt or egg crate, just a blanket that is wool on one side and suede on the other, folded in half lengthwise (wool side out) and hung over the top of the door. It dangles to about 4" below the speaker and is apparently great at killing the horrible reflections that were messing up the center image.

Between these mods and the new found Dynamic Range from a previous thread I can't get over the sonic improvement in my system. I hope this is a lesson to others about taking the time for proper set-up and heeding the advice of others regarding this.

shokhead
04-02-2005, 03:18 PM
I just checked mine and if i'm alone listening to something,it around 83-85db and if the wife is here,75 tops but most likly less. ;)

hermanv
04-04-2005, 10:32 AM
Kex- I have considered fabbing up an absorption device for the glass door (actually there are two) but it needs to be portable (read-WAF) for use during critical listening only. .

Have you considered drapes? Nice thick ones are best. Should get a decent WAF.

toenail
04-04-2005, 01:00 PM
Our living room has two etched glass doors that open into the dining room. They are quite attractive and still have their original wood finish frames. They sit in the always fully open postition and are decorative more than anything at this point. Covering them with drapes would not be in my best interest. As it stands I used the wool/suede blanket method several times over the weekend and am satisfied with the result. When not in use they sit folded in an antique basket next to the sofa.