Setting up Speakers - Tape Measure or SPL Meter?

[opusover21]

Managed to get a SPL App for the iPod Touch...seems to work pretty well..Just wondered I've never set speakers up using a SPL but always measure all distance to the walls to within a few mm as possible...

Anyone then use test tones and set the levels .....?

[SergeAuckland]

opusover21 wrote:

Managed to get a SPL App for the iPod Touch...seems to work pretty well..Just wondered I've never set speakers up using a SPL but always measure all distance to the walls to within a few mm as possible...

Anyone then use test tones and set the levels .....?

It depends what it is that you want to do. To set 'speaker positions relative to the listening seat, a tape measure is of most use. If you want to check how loud you're playing, then an SPL meter is useful.

What an SPL meter is of little use for is volume matching when evaluating two items. It's nowhere near accurate enough on programme material, and barely on tones. For volume matching you need to match to 0.1dB, whereas an SPL meter will struggle to be better than 1dB.

For setting up subwoofers, you do much better doing that by ear, as an SPL meter will only tell you the actual level at any one frequency, and that varies considerably with position and frequency, whereas your ears will give you an integrated view of what needs adjusting...you can balance level against crossover frequency, for example.

Next, any measurement above a few hundred Hz is extremely position and reflection dependent, such that even if you move a few centimetres, the reading will change due to reflections off your body.

Finally, the accuracy of an SPL meter depends on the accuracy of the microphone. I may be wrong, but I very much doubt that the microphone in an iPod Touch is of measurement quality, although you may be able to use an externalmic.

SPL meters have their uses, but generally only in giving you an idea of how loud you're playing, or how much noise there is in a building, or near an airportand not for muchelse when it comes to home Hi-Fi.

S.

[Tel]

I used to use SPL meters for measuring db above ambient for fire alarm systems to ensure they met BS.

Haven't done that job for a while and I have never even thought about getting the meter out for use with a Hifi system

[Camverton]

opusover21 wrote:

Anyone then use test tones and set the levels .....?

Yes, when I'm using the Meridian G68 as a pre, then I use a meter to help choose the speaker position that gives the smoothest response in the lower frequencies using the inbuilt tones from the G68 and the graphs it generates. I can then go on to apply room correction to filter any peaks out. This is very effective for avoiding boomy bass, and incorporating a sub if used.

For surround sound a meter is more or less essential to balance the volume from each speaker. Even with stereo it is useful if you have a room, or furniture layout, which results in a higher volume at the listening position from one speaker then the other.

A meter can certainly help provide a starting point before going on to make adjustments by ear to what is preferred.

Serge is doubtless right about a meter not being accurate enough to set the volumes for equipment comparison, if that is the sort of comparison one is interested in. If I am comparing equipment than I always reset the volume control to the level at which the equipment sounds best. Some equipment combinations are better at higher volumes than others. I can then choose the equipment that works best at the levels I like to listen at. Not very scientific, but very practical, particularly as I neither feel the need to proof that things sound the same or different; so I don't use the meter when comparing equipment although it could be interesting comparing the "in room" frequency response of speakers.

[ncdrawl]

http://realtraps.com/art_room-setup.htm

http://www.cardas.com/content.php?ar...g=Room+Setup+2

http://www.tnt-audio.com/casse/waspe.html

http://www.gcaudio.com/resources/how...placement.html

http://www.digido.com/audio-faq/m/mo...rsus-wood.html

http://www.digido.com/audio-faq/m/mo...ink-noise.html

the CHESKY LEDR test is a MUST... it is ruthless.

[opusover21]

Thanks for the replies...I was thinking of setting the quads 57's up with a tape a I would usually but then use seperate left and right test tones with an external mic to measure output level bang in the centre in the listening position...only a thought but as the right speaker has a window higher up within 1.5m of the speaker I thought it may be a good idea to check the ouput and if required nudge either speaker back and forth a natts chuff and match the output as accurate as possible...that said I may not be able to measure a difference!...thoughts?

[SergeAuckland]

opusover21 wrote:

Thanks for the replies...I was thinking of setting the quads 57's up with a tape a I would usually but then use seperate left and right test tones with an external mic to measure output level bang in the centre in the listening position...only a thought but as the right speaker has a window higher up within 1.5m of the speaker I thought it may be a good idea to check the ouput and if required nudge either speaker back and forth a natts chuff and match the output as accurate as possible...that said I may not be able to measure a difference!...thoughts?

Do it by ear. Use a mono source, speech is good, Radio 4 announcer, best to switch the tuner to mono just to make sure the transmission isn't slightly out, sometimes they are. Check that the image is central and move the 'speakers slightly if necessary.

S.

[Tune]

If you are interested in imaging Stuart, that is a product of time alignment - making sure that you are equidistant from both panels is the most important thing to get right. If there's an amplitude discrepancy, all that will happen is that the correctly formed image is slightly off to one side.

[tuga]

It wouldn't de any harm if you were to make some RTA measurements for macro-positioning to fine-tune the frequency response.

Cheers,
R

Edit: typo

[SergeAuckland]

tuga wrote:

It wouldn't de any harm if you were to make some RTA measurements for macro-positioning to fine-tune the frequency response.

Cheers,
R

Edit: typo

The best way I've found of doing RT measurements is to use filterednoise-bursts, centred on different frequencies, span about an octave, then record them using a decent microphone onto a PC with Audacity of similar. Looking at the waveform allows very good estimates of RT to be made. The main problem using an SPL meter for RT measurement is estimating the -60dB point, as it's very fleeting, or if the background is noisy, you'll never get to the -60 point. Also,if you use wideband noise, you'll get multiple RTs as the RT varies greatly with frequency.

S.

[diceman]

IMHO a SLM (Sound level Meter) is pretty useless for this application. I am lucky enough to have a fully calibrated Class 1 SLM and mic with Fast Fourier Transform for frequency analysis and even that is useless for this application! As Serge has indicated, even at this level of kit the accuracy is not sufficient (IIRC is about 0.5-1 dB)and the room acoustics play havoc with trying to obtain a realistic result.

To be fair though the little Iphone application can be good enough for a lot of noise testing. A word of caution though, the internal MIC of the older Iphones/ITouch had a restricted frequency range. Only the later versions will provide a reading at 4KHz and above (IIRC).

I would therefore use a tape measure to set up speakers.

The only thing I have used a SLM for (with a frequency generator/pink noise source) is integrating a sub or tuning port length on DIY speakers or playing with an adjustable active cross-over. Taking many samples and averaging is about the only way you get a result that is anything close to correct. Ideally this needs to be done in a large open field with long grass and no ambient noise (or your own private anechoic chamber!)

[tuga]

SergeAuckland wrote:

tuga wrote:

It wouldn't de any harm if you were to make some RTA measurements for macro-positioning to fine-tune the frequency response.

Cheers,
R

Edit: typo

The best way I've found of doing RT measurements is to use filteredÂ*noise-bursts, centred on different frequencies, span about an octave, then record them using a decent microphone onto a PC with Audacity of similar. Looking at the waveform allows very good estimates of RT to be made. The main problem using an SPL meter for RT measurement is estimating the -60dB point, as it's very fleeting, or if the background is noisy, you'll never get to the -60 point. Also,Â*if you use wideband noise, you'll get multiple RTs as the RT varies greatly with frequency.

S.

Thanks for tip.

[Mr. Widget]

Oh, where to begin?

I have an SPL meter... an ancient Radio Shack, (Tandy) meter, and I have an MLS rig with calibrated mic. The MLS rig will do time, phase, frequency response, as well as SPL and even distortion measurements and more. The problem with taking measurements in "normal" listening rooms is that it is very difficult to measure the direct response from the speaker and not have it clouded by the reflected response. With a sophisticated measurement system you can start to separate the two by examining the response in the time domain, but using a simple SPL meter while not inherently wrong, will give you rather meaningless data.

You might say, but wait, what I listen to is a blend of direct and reflected sound so why is there a problem with using a meter that measures both too? The simple answer is that our hearing is discerning, our on board computer (brain) places different significance to first arrival information over subsequent arrival info.

The bottom line is that while a simple SPL meter regardless of its accuracy will really help you very little in a high performance audio setup, using it might hint at issues and can be of some use in especially in larger rooms, but ALWAYS trust your ears over your measurements.


Widget

[SergeAuckland]

Mr. Widget wrote:

Oh, where to begin?

I have an SPL meter... an ancient Radio Shack, (Tandy) meter, and I have an MLS rig with calibrated mic. The MLS rig will do time, phase, frequency response, as well as SPL and even distortion measurements and more. The problem with taking measurements in "normal" listening rooms is that it is very difficult to measure the direct response from the speaker and not have it clouded by the reflected response. With a sophisticated measurement system you can start to separate the two by examining the response in the time domain, but using a simple SPL meter while not inherently wrong, will give you rather meaningless data.

You might say, but wait, what I listen to is a blend of direct and reflected sound so why is there a problem with using a meter that measures both too? The simple answer is that our hearing is discerning, our on board computer (brain) places different significance to first arrival information over subsequent arrival info.

The bottom line is that while a simple SPL meter regardless of its accuracy will really help you very little in a high performance audio setup, using it might hint at issues and can be of some use in especially in larger rooms, but ALWAYS trust your ears over your measurements.


Widget

Erm, no! If your ears tell you something different than the measurements tell you, then check your measurements, and especially your methodology. As Widget very rightly said, an SPL meter won't tell you very much (except what the SPL is) so the in-room measurements of frequency response, RT etc you're making using an SPL meter are more than likely to be nonsense. That's why I've advocated setting subwoofers by ear rather than with instrumentation.

However, if you get the measurements right, and you're using the right tool and know how to use it, the measurements will be more trustworthy than fallible and subjective ears. Meters don't have moods or bias or preferences or drink too many single malts, but they do require knowledge and correct application.

By the way, the Radio Shack/Tandy SPL meter is actually very good and quite accurate. The microphone capsule is very flat at the bass end, and the SPL accuracy is within 1dB of a very expensive and calibrated SPL meter I compared mine with. What it's not a lot of good for is >10kHz measurements, or measuring peaks, as the meter has uncalibrated dynamics. Still, for the £16 or whatever I paid for it, it does a pretty good job.

S.

[Tune]

Like I said.

Use a tape measure.

[Mr. Widget]

SergeAuckland wrote:

"Erm, no! If your ears tell you something different than the measurements tell you, then check your measurements, and especially your methodology.

By the way, the Radio Shack/Tandy SPL meter is actually very good and quite accurate."


No argument from me, however since most people don't have access to sophisticated measurement equipment and many DIY types don't really know how to use what they do have I think it is better for them to trust their ears than to get distracted by their poor measurements.

I concur about the Rat Shack meter... within it's range it is surprisingly accurate.

Widget

[Uncle Ants]

opusover21 wrote:

Managed to get a SPL App for the iPod Touch...seems to work pretty well..Just wondered I've never set speakers up using a SPL but always measure all distance to the walls to within a few mm as possible...

Does the current version of the Touch have a microphone then?

[opusover21]

Uncle Ants wrote:

opusover21 wrote:

Managed to get a SPL App for the iPod Touch...seems to work pretty well..Just wondered I've never set speakers up using a SPL but always measure all distance to the walls to within a few mm as possible...

Does the current version of the Touch have a microphone then?

No but a headset type mic or mini jack works fine...I tried some test tunes and its ok but not accurate enough to decide either way...so tape as usual me thinks

[wickfut]

If your SPL meter has an output jack then I'd use it for doing room frequency plots and that's about it.