I thought I read that you could do a blackman burst with ARTA. Perhaps it was done manually, and just called that by whoever.
Speaker driver testing is allegedly professionally done with the driver mounted on a wall facing into an anechoic chamber. The wall is considered an "infinite baffle" since the rear wave can never get into the chamber. The calibrated mic is positioned about a meter out from the diaphram.
In the real world as I know it, you close mic (1-2 inch) the diaphram (at it's center), and if room acoustics are likely to be an issue, you use tone bursts and "gate" the signal from the mic, so on the oscilloscope (or equiv.) you only see the initial acoustic burst output, and not any other reflections of that acoustic output due to the room. At higher frequencies, most of the energy comes from the part of the cone where the voice coil attaches (the center), and it gets very directional.
Another thing to know is that the burst envelop is likely to have spectra other than the burst frequency (energy at other frequencies other than the sinewave in the burst envelop). So when the acoustic burst envelop looks funky, take it with some grain of salt. That's why things like the "blackman burst" exist. Ideally you want something like a gaussian rise and fall for the burst envelop, in order to minimize this effect. Blackman is one of many attemps to achieve this.
Another thing to know is that if you measure a driver that's mounted in a speaker cabinet, there will be "baffle step response", usually in the vacinity of 500HZ (+/- 500HZ), and below that frequency the front surface of the cabinet will no longer reinforce the propagation of the acoustic wavefront, so you'll get a roll-off below that frequency, until walls of the listening room start to reinforce (reflect) the energies, usually down below 200HZ. That's normal, and generally requires EQ (acoustic or electronic) to clean it up. Most people ignore Baffle step, figuring that its a minor thing if the speaker is going to be near listening room walls. Personally I think it's important. Either way, it's good to know about it. The linkwitzlab.com website can tell you details and math about it.
Speaker driver testing is allegedly professionally done with the driver mounted on a wall facing into an anechoic chamber. The wall is considered an "infinite baffle" since the rear wave can never get into the chamber. The calibrated mic is positioned about a meter out from the diaphram.
In the real world as I know it, you close mic (1-2 inch) the diaphram (at it's center), and if room acoustics are likely to be an issue, you use tone bursts and "gate" the signal from the mic, so on the oscilloscope (or equiv.) you only see the initial acoustic burst output, and not any other reflections of that acoustic output due to the room. At higher frequencies, most of the energy comes from the part of the cone where the voice coil attaches (the center), and it gets very directional.
Another thing to know is that the burst envelop is likely to have spectra other than the burst frequency (energy at other frequencies other than the sinewave in the burst envelop). So when the acoustic burst envelop looks funky, take it with some grain of salt. That's why things like the "blackman burst" exist. Ideally you want something like a gaussian rise and fall for the burst envelop, in order to minimize this effect. Blackman is one of many attemps to achieve this.
Another thing to know is that if you measure a driver that's mounted in a speaker cabinet, there will be "baffle step response", usually in the vacinity of 500HZ (+/- 500HZ), and below that frequency the front surface of the cabinet will no longer reinforce the propagation of the acoustic wavefront, so you'll get a roll-off below that frequency, until walls of the listening room start to reinforce (reflect) the energies, usually down below 200HZ. That's normal, and generally requires EQ (acoustic or electronic) to clean it up. Most people ignore Baffle step, figuring that its a minor thing if the speaker is going to be near listening room walls. Personally I think it's important. Either way, it's good to know about it. The linkwitzlab.com website can tell you details and math about it.
I've just started using ARTA and as a newbie to speaker measurement I found it all a bit daunting at first and very complex. Much of the user-guide went right over my head, then I found this easy guide to making gated impulse FR measurements:
http://audio.claub.net/tutorials/FR measurement using ARTA.pdf
Now I'm really getting to grips with the program and obtaining some meaningful measurements. I found MLS gives the best result in my smallish room where no boundary is further than 5ft from the test speaker.
This is a typical result with MLS:
The response below 200Hz is inaccurate due to the limitations of the room.
As for the Blackman burst, yes ARTA does this.
All in all, ARTA is an impressive piece of work, once you understand how to use it!
http://audio.claub.net/tutorials/FR measurement using ARTA.pdf
Now I'm really getting to grips with the program and obtaining some meaningful measurements. I found MLS gives the best result in my smallish room where no boundary is further than 5ft from the test speaker.
This is a typical result with MLS:
An externally hosted image should be here but it was not working when we last tested it.
The response below 200Hz is inaccurate due to the limitations of the room.
As for the Blackman burst, yes ARTA does this.
All in all, ARTA is an impressive piece of work, once you understand how to use it!
Actually there are few of those left. Too much easier to use a big space + gating, compared to spending huge money to build a chamber which might only be good down to 100 or 70 or whatever hertz.
...and the close miking vastly increases the SPL of the direct sound versus any reflections.
BUT lest any newbies get confused, this only works at very low frequencies. And you must take repeated measurement getting closer and closer until the SPL doesn't rise any more, to be certain you're in the near field.
You can mount a string with a washer on it:
- One end at the mic position
- One end at the speaker position
By moving the washer all around the room, you can trace out an ellipsoid. If the washer hits something, shorten the string until it doesn't.
Repeat
Repeat
Repeat
...you get the idea-until the washer doesn't hit anything.
Measure the length S of the string. Compare versus the mic-speaker distance M.
Reflection free zone will be S-M. If you measured in meters or convert to meters, the reflection free time is then
T=(S-M)/(343m/s)
which should be good down to a frequency of
f=1/T
As an example, suppose you had the speaker on a stand in the outdoors, such that the speaker is 1m above the floor. If the mic is M=1m from the speaker, the string will end up being about S=2.83m long:
T=(2.83m-1m)/(343m/s)=0.0053s
f=1/0.0053s=187.6 Hz
and you would get valid data points every 188 Hz.
Note: this does NOT mean all the data points shown above 188 Hz are valid!!! Many programs show interpolated values, which are to be taken with a grain of salt.
For example, in the measurement OP shows, a lot of the waviness at low frequencies I believe to be a typical artifact of the measurement. I think it's due to the windowing applied to the FFT.
Has anyone tried WaveSpectra ?
It is free and very fast and the results are comparable with ARTA
Available at ???????? ???????????? WaveSpectra
It is free and very fast and the results are comparable with ARTA
Available at ???????? ???????????? WaveSpectra
Actually site is in Japanese but using google translator I was able to read.
Have you tried WaveSpectra ? If yes, then please tell your experience.
Yes ,It works rather nicely.
I'm still learning the in and outs although it is extremly easy to use.
It recognizes my GINA-24 card with asio driver with no problem.
However I'm not sure if it is sampling in 24bit or not.
The scope screen only displays + and - 32000 steps on the side of the display (graticule persay).
The features are limited compared to some other programs I have tried.
But, It is adequate as it does everything you need for testing audio as well as saving any sample curves.
Like, FFT's,THD,FREQUENCY COUNTER, FREQUECY RESPONSE and such.
Not sure if it does signal generation as I use a seperate devices for that.
I haven't verified the validity of the THD measurement yet but I will as I keep working with it.
I very easily and quickly did some frequency response measurements of my room and system and was able to save them.
But, I haven't learned how to post the files as JPEG or BMP or graphs yet, as I have only used it briefly so far.
But it appears to be quite accurate compared to some past measurements with other programs that I use.
I just down loaded a new beta version of VISUAL ANALYZER today and once I got it setup and working it was much better than the previous version.
Although, It did lockup my machine up a couple of times but not like the pervious versions I had used in the past.
WaveSpectra work flawlessley and no lockups on my P4 2.8ghz machine so far.
The screen update was quite fast.
Visual Analyzer is an intense program and cpu power hungery so those with something better than my P4 may not have any issues at all, as it is still under development . jer
I'm still learning the in and outs although it is extremly easy to use.
It recognizes my GINA-24 card with asio driver with no problem.
However I'm not sure if it is sampling in 24bit or not.
The scope screen only displays + and - 32000 steps on the side of the display (graticule persay).
The features are limited compared to some other programs I have tried.
But, It is adequate as it does everything you need for testing audio as well as saving any sample curves.
Like, FFT's,THD,FREQUENCY COUNTER, FREQUECY RESPONSE and such.
Not sure if it does signal generation as I use a seperate devices for that.
I haven't verified the validity of the THD measurement yet but I will as I keep working with it.
I very easily and quickly did some frequency response measurements of my room and system and was able to save them.
But, I haven't learned how to post the files as JPEG or BMP or graphs yet, as I have only used it briefly so far.
But it appears to be quite accurate compared to some past measurements with other programs that I use.
I just down loaded a new beta version of VISUAL ANALYZER today and once I got it setup and working it was much better than the previous version.
Although, It did lockup my machine up a couple of times but not like the pervious versions I had used in the past.
WaveSpectra work flawlessley and no lockups on my P4 2.8ghz machine so far.
The screen update was quite fast.
Visual Analyzer is an intense program and cpu power hungery so those with something better than my P4 may not have any issues at all, as it is still under development . jer
Last edited:
Hi Jer,
Yes features are limited but I amazed with it speed and accuracy.
For signal generation I use ttg.exe (Test Tone Generator from Esser Audio Research) it is also free.
Yup, for screenshots I have to use Alt+PrintScreen and Paint Brush.
To speed up screen refresh you may use DDraw3 from drop down in settings.
Your Processor is good one (P4 2.8 GHz).
With ARTA I see glitches but WaveSpectra does not show any glitches or hang-ups it runs smooth. I use it to measure THD of my designed wien oscillators.
Yes features are limited but I amazed with it speed and accuracy.
For signal generation I use ttg.exe (Test Tone Generator from Esser Audio Research) it is also free.
Yup, for screenshots I have to use Alt+PrintScreen and Paint Brush.
To speed up screen refresh you may use DDraw3 from drop down in settings.
Your Processor is good one (P4 2.8 GHz).
With ARTA I see glitches but WaveSpectra does not show any glitches or hang-ups it runs smooth. I use it to measure THD of my designed wien oscillators.
I might add there is a similar program called Room EQ Wizard...from one of those other sires.
I use it mostly for acoustic measurements. In many cases I like it better than the old TEF20.
It also interfaces with programmable equalizers....REW V5 REW - Room EQ Wizard Home Page
This program is free, but they ask for a donation. It is a stable program...with few bugs.
I use it mostly for acoustic measurements. In many cases I like it better than the old TEF20.
It also interfaces with programmable equalizers....REW V5 REW - Room EQ Wizard Home Page
This program is free, but they ask for a donation. It is a stable program...with few bugs.
I have been messing around with "Room EQ Wizard" today.
I must say that it is one great program.
It has a great GUI,the layout is very nice and very easy to use as well as acurate compared to the other programs I have used so far.
Great find!
Thanks again, Dave. jer
P.S. I also got the Fast Stone viewer aswell but I haven't tried it yet as I was having to much fun with REW 5.
I must say that it is one great program.
It has a great GUI,the layout is very nice and very easy to use as well as acurate compared to the other programs I have used so far.
Great find!
Thanks again, Dave. jer
P.S. I also got the Fast Stone viewer aswell but I haven't tried it yet as I was having to much fun with REW 5.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.