Display SPL vs time?

This section is now closed.

Audacity 1.2.x is now obsolete. Please use the current Audacity 2.0.x version.

Display SPL vs time?

Permanent link to this post Posted by Cdntiger » Fri Dec 07, 2007 2:30 am

In my search for a program that could record and display a graph of the measured SPL in my room over a period of time, I have come across Audacity. After some preliminary dabbling, I think I may have to do some editing and playing around with some of my music collection! :geek: I haven't been able to figure out how to get Audacity to do exactly what I'm looking for yet, which is why I've come here to ask the pros. :)

I want to measure the SPL using an appropriate meter and have it plotted, without also displaying the sound's frequency. Last night's measurements did measure the input from the SPL meter, but as the frequency of the sound being measured increased, Audacity dutifully increased the frequency of the waveform displayed. Can Audacity display the recorded SPL using a constant recording sample frequency, thus eliminating any information about the frequency of the sound being recorded? TIA
Cdntiger
 
Posts: 4
Joined: Fri Dec 07, 2007 1:55 am

Re: Display SPL vs time?

Permanent link to this post Posted by alatham » Fri Dec 07, 2007 6:01 pm

What exactly are you trying to do? Determine speaker placement? Measure how loud your normal listening is (not a bad idea for those who like hearing things)?

Can Audacity display the recorded SPL using a constant recording sample frequency, thus eliminating any information about the frequency of the sound being recorded?

Not unless you had some outboard hardware that created a waveform that way. I've never seen anything that would do that though (maybe an old modular synth might). Why do you want to do this?

Do you want to ultimately end up with a single tone that changes in volume over the course of the recording?

You could always turn the speakers off and zoom out on the waveform and just look at the peaks if you want to ignore frequency content and focus purely on amplitude. I can't think of any reason to do that, but you might well have a good one.

Also, what exactly are you using as an input to Audacity, a microphone? If you want an accurate measurement of peak SPL levels, you'll need an SPL meter in order to calibrate what Audacity is recording.

I'd like to answer your question, but I can't figure out what you're trying to accomplish.
alatham
Forum Staff
 
Posts: 1593
Joined: Tue Jul 31, 2007 2:27 pm
Location: Chicago, IL

Re: Display SPL vs time?

Permanent link to this post Posted by kozikowski » Sat Dec 08, 2007 12:38 am

This will convert SPL into voltages, I think. They both have an RCA connection on the side.

http://www.radioshack.com/search/index.jsp?kwCatId=&kw=sound%20level%20meter&origkw=sound%20level%20meter&sr=1

I own an analog one. The company owns a digital one. Put that into your handy-dandy analog to digital converter and there you have it. I think both instruments have a socket so you can mount them in place of a camera on a tripod.

Koz
kozikowski
Forum Staff
 
Posts: 26260
Joined: Thu Aug 02, 2007 5:57 pm
Location: Los Angeles

Re: Display SPL vs time?

Permanent link to this post Posted by Cdntiger » Sun Dec 09, 2007 12:08 am

Thanks for the replies guys. Let's see if I can explain what I'm trying to do more clearly. :oops: I want to test how much the room and surroundings are degrading the sound coming from my speakers in my home theater. There is a test (called the MATT test) created specifically for this purpose by Acoustic Sciences Corporation, and a demo and description are available here: http://www.acousticsciences.com/matt.htm

I want to play this sound clip through my speakers, use my RatShack SPL meter (the analog one in the link posted by kozikowski) to measure it, and send the data to my computer to plot out on a SPL vs time graph. I have all the hardware and cables already, and just need to find a software program that will do what I'm looking for. Ideally the graph would look something like the example plots on the Acoustic Sciences website. I've cut and pasted from the ASC website a description of the MATT test:

"The basis of the test signal is a very slow sine wave sweep. It starts at 28 Hz and rises up to 780 Hz, then it drops back down to 28 Hz. It is easier to read the printout when presented in this symmetric form. The linear frequency vs time curve is triangular in shape and takes about 80 seconds to complete.

The slowly changing pure tone signal is then chopped, alternately turned on and off at some particular rate. For the typical music playback system, the test signal is gated at 8 Hz, eight distinct tone bursts per second. A 50% duty cycle is used so each tone is played for 1/16 second and is followed by 1/16 second of silence.

This measures the fast response, dB level curve at the listening position. Specifically, it measures how loud each burst is and how quiet each silent period becomes. The test analysis circuitry can follow as much as a 20 dB drop in level during a 1/16 second of silence. Articulation is the ability of the room to distinctly sound out each audio event."

As I was trying to say originally, when I recorded the MATT test the plot gets more and more crowded and difficult to decipher as the frequency being played increases. I understand this is desired for the application Audacity is intended for, but is there an option to plot ONLY the measured SPL? And can the timing interval of these SPL measurements be specified?

Alatham, you mention calibrating my SPL meter to work with Audacity. I had assumed that Audacity wouldn't be able to show me what the actual dB readings are; if there's a way to calibrate the SPL meter with Audacity such that the actual, measured, in room sound level is displayed I'd really appreciate some instruction.

Hopefully that makes things clearer. I'll be happy to provide any further clarification needed - I appreciate you guys taking the time to help out.
Cdntiger
 
Posts: 4
Joined: Fri Dec 07, 2007 1:55 am

Re: Display SPL vs time?

Permanent link to this post Posted by kozikowski » Sun Dec 09, 2007 4:29 am

<<<Alatham, you mention calibrating my SPL meter to work with Audacity. I had assumed that Audacity wouldn't be able to show me what the actual dB readings are; if there's a way to calibrate the SPL meter with Audacity such that the actual, measured, in room sound level is displayed I'd really appreciate some instruction.>>>

So do I. I don't think you can do it. The connector on the side of the meter delivers actual audio and it's conditioned according to the A or C switch. Audacity will give you the peak digital numbers for all the readings, but the reason they still make an analog meter is that the meter delivers the average (or possibly RMS) reading, not the peak.

I wonder if we could make good use of the second color inside the Audacity meters. I keep forgetting that exists.

Koz
kozikowski
Forum Staff
 
Posts: 26260
Joined: Thu Aug 02, 2007 5:57 pm
Location: Los Angeles

Re: Display SPL vs time?

Permanent link to this post Posted by steve » Sun Dec 09, 2007 10:55 pm

@ Cdntiger
I guess the reason you want to measure this is so that you can improve the acoustics of your room. To do this, you don't actually need to have accurate (calibrated) measurements, you need relative measurements.

Producing a full and accurate analysis of room acoustics is a very complicated business, however, for the practical purposes of setting up a home cinema, it can be simplified.

The main things to examine (and this is what the MATT test does) are "frequency response" and "reverberation".

Frequency response is the accuracy with which different frequencies (high pitch and low pitch sounds) are reproduced. There will be a cut-off point at both ends - the lowest pitch and the highest pitch that your sound system can produce.
Within this range, the frequency response, as experienced at the listening position, should be smooth. For accurate sound reproduction, the response should be flat within this range, however this may not be ideal for the best "theatre experience". A slight emphasis at about 1kHz can improve the intelligibility of speech, and a slight boost in the bass and treble can give the impression of a bigger, more dramatic sound. In the end, the exact balance is a matter of personal preference - it's your system, so have it the way you like it.

Reverberation is how the sound "bounces" around within the room. Consider the acoustics of a bathroom, or a tunnel, compared with the acoustics of a soft furnished living room. Bathrooms typically produce bright reverberation due to all the hard surfaces, and tunnels produce a long echoey reverberation. Soft furnished living rooms typically produce little reverberation as the irregular shapes break up the sound, and soft furnishings, carpets and curtains, will absorb much of the reflections.

To examine the frequency response of your sound system with Audacity, you can either use the test tape that you spoke of, or produce a sweeping sound wave with Audacity (Generate->Chirp). If you choose to use Audacity, you can generate several sweeping tones to test out different sound ranges - say, 20Hz to 125Hz for low bass, 100Hz to 1000Hz for low frequencies, 600 Hz to 3kHz for mid range, 2kHz to 20kHz for treble. (Unless you are young, the upper range will probably become inaudible well before 20kHz).

Now you need a measuring microphone - Behringer make one that is quite inexpensive, or from what kozikowski was saying, it sounds like you could use your SPL meter. Note that NO microphones produce a completely flat response, so look at the specifications of your equipment to get an idea of its characteristics and limitations.

There will also be limitations due to your sound card - A good way to test this is to use a short lead to connect the output from your soundcard to the input as a "loop-back". Play your test tones from Audacity and record them on a separate track via the loop-back link.

Now for the tests proper:

Place your measuring microphone roughly in the position that your head will be when you are sat in your home cinema. Play back your test tones and record them in Audacity (in mono). Since your test tones were all the same amplitude, the recorded track will clearly indicate the levels that different frequencies are produced.

You need to repeat this proces at different volume levels, as you will get different results with quiet levels to what you get with high volume levels.

If you have a graphic equiliser as part of your setup, you could use this to tweak the system.

Now onto reverb.

The first and simplest test is to sit in your usual position, clap, and listen to the sound of that clap as the reverberation decays.

The second test is a little more complex.
Use the tones generated in the "frequency test", and "chop them up" into pulses. You can do this with the "Tremolo" effect (select "square wave" and use the wetness set to 100% and frequency at about 8Hz)
Now play back these test tones and record as before (at quite a loud level).
Zooming in on this recording, you will notice that the recorded pulses "tail away" at the end. This is caused by reflections and reverberations in the room.

(You can make this more visible by carefully aligning the recorded sound with the original sound, inverting one of them and mixing the two together).

Ideally, you would want to reduce sound reflections and reverberations as much as possible. Exactly how much these "measurements" will help you do that, I'm not sure, but the basic principle is that irregular shapes, and soft stuff helps. Parallel walls, and hard surfaces make it worse.
9/10 questions are answered in the FREQUENTLY ASKED QUESTIONS (FAQ)
steve
Senior Forum Staff
 
Posts: 34809
Joined: Sat Dec 01, 2007 11:43 am

Re: Display SPL vs time?

Permanent link to this post Posted by alatham » Sun Dec 09, 2007 11:57 pm

Wow, excellent post Steve.

To answer the question about how to calibrate the SPL meter to Audacity, there is no way to get Audacity to spit out absolute numbers like that, so you'll always have to do the conversion yourself. But if you play some full amplitude pink noise (wear ear plugs, you'll thank me later), and measure with your SPL meter a level of 70dBs, then you know that full amplitude out of Audacity is approximately 70dBs for a typical music signal (as long as you don't change the volume level on your speakers or the output volume level on your computer).

The reason to use pink noise is that it's a decent approximation of a music signal. White noise will have a higher SPL at the same amplitude (I'm not certain though, is this correct?).

Before you go out and buy $1000 worth of fancy shmancy bass traps and foam covering, you should know that many problems can be alleviated by moving your speakers around (keep them at least a foot from the wall, and most certainly out of the corners).

If you still want a better sounding room, most of that acoustic paneling stuff can be built much cheaper if you buy the raw materials yourself:
http://www.ethanwiner.com/acoustics.html

I'm not sure where to buy rigid fiberglass, you might have to order it.
alatham
Forum Staff
 
Posts: 1593
Joined: Tue Jul 31, 2007 2:27 pm
Location: Chicago, IL

Re: Display SPL vs time?

Permanent link to this post Posted by Cdntiger » Mon Dec 10, 2007 12:16 am

Holy smokes, that was quite the reply! :shock: You're right on the money with what I'm trying to do - thanks for the in depth explanation, I wasn't expecting that! Before I'd ever come across Audacity I was using another program called Room EQ Wizard to measure my room's frequency response. If curious, it can be found at the Home Theater Shack forum. I have done a fair amount of reading about home theater room acoustics and have a general idea of the kind and location of furnishings, absorbers, diffusers, etc. ideally found in a room. I like experimenting and plan on testing out various arrangements once I actually buy or build some room treatments. Hence my interest in Audacity, which is piquing my interest more and more. I want plotted results because I don't trust my ears to reliably detect all the changes - one might ask if I can't hear it why worry about it? Well, hopefully I'll be able to teach myself to become a more discerning audiophile with the help of visual cues, and heck, we all need to have some passion we're anal about, right? :ugeek:

As I understand it, Audacity can indeed help me with reverberation via the "Tremolo" effect (I've never come across that one and haven't the foggiest :?: ). I'll give that a try when I've got some time.

You guys are awesome!
Cdntiger
 
Posts: 4
Joined: Fri Dec 07, 2007 1:55 am

Re: Display SPL vs time?

Permanent link to this post Posted by Cdntiger » Mon Dec 10, 2007 12:32 am

Thanks for the additional info, alatham. You must have posted while I was writing up my reply to Steve's post. That pink noise test to approximate sound levels seems like a reasonable approximation - if we use it to adjust speaker levels to match each other who am I to argue? :)

I'm currently considering 4 options for room treatments: DIY, Acoustic Sciences Corporation, GIK Acoustics, and RealTraps. There's just something about having some higher frequency reflection/diffusion to go along with absorbers that makes sense to me and is the main reason I haven't committed to DIY at this point. I actually already have a small number of Eighth Nerve products, but they're not designed for bass frequencies. And yep, I'm still planning on playing with the location of my speakers some more (but they're heavy fellas!). That's another reason I'm looking to measure responsiveness.
Cdntiger
 
Posts: 4
Joined: Fri Dec 07, 2007 1:55 am

Re: Display SPL vs time?

Permanent link to this post Posted by steve » Mon Dec 10, 2007 1:28 am

Cdntiger wrote: ... I like experimenting and plan on testing out various arrangements once I actually buy or build some room treatments...
I think trying to get some "objective" measurements is a great idea in these circumstances. If you try one arrangement, then a few days later you move things around, it's going to be difficult to remember exactly how good your previous arrangements were.

Cdntiger wrote:we all need to have some passion we're anal about, right?

LOL

alatham wrote:White noise will have a higher SPL at the same amplitude
Hmm... not quite that simple - SPL is usually "weighted" for different frequencies and there's different ways of doing that. White noise has higher energy than Pink noise of the same amplitude.

alatham wrote:If you still want a better sounding room, most of that acoustic paneling stuff can be built much cheaper if you buy the raw materials yourself:http://www.ethanwiner.com/acoustics.html I'm not sure where to buy rigid fiberglass, you might have to order it.
That's a very good article. I've never managed to get hold of rigid fiberglass, but carpet felt (the old fashioned felt, not the newer foam stuff) works fairly well, is cheap, easy to work with and easier to get hold of (and doesn't make you itch). In my studio, I have some panels in the corners made of this stuff: http://www.wickes.co.uk/invt/161197 with a layer of felt glued on, then covered with fabric.
9/10 questions are answered in the FREQUENTLY ASKED QUESTIONS (FAQ)
steve
Senior Forum Staff
 
Posts: 34809
Joined: Sat Dec 01, 2007 11:43 am

Next

Return to General Feedback and Discussion



Who is online

Users browsing this forum: No registered users and 3 guests