- Classic Dirac Delta impulse
- variable frequency / variable timing swept sine (linear or log)
- white noise bursts
- pink noise bursts
Q: Interface in matlab? Are you crazy???
A: I actually have some thoughts on that....
I plan on using Matlab with
GUIDE
to create the GUI.
Q: So what is the advantage of this program?
A: It seems that most acousticians use the same types of tests when they are characterizing
either spaces or equipment. So, I thought that I would try to help out so they wouldn't have
to retype the same code all the time. If the basic functions are easily implemented,
I will look into avenues such as automated output sequences of test signals and data acquisition / processing algorithms.
Also, if it works well, it could save me a lot of time in the future.
Q: Save you time? On what?
A: I am working on a project with Countryman Microphones
that requires me to characterize an M-S
stereo microphone enclosure. This test generator will (hopefully) help me, so why shouldn't it help other people too?
like this guy...he needs all the help he can get.
keep on a-rockin' stereotypical audio engineer dude.
Q: What does Ron Jeremy have to do with acoustical measurements??
A: everything.
Timeline
Week 1: Decide to develop a retriggering / grainular processing plugin.
DONE!!
Week 2: Decide not to develop a retriggering / grainular processing plugin.
Decide instead to make a matlab GUI in GUIDE and start researching acoustical testing practices.
DONE!!
Week 3: Now that you (hopefully) know how acoustical characterization is done, write a
program that can generate at least one of the signals (probably start with the most useful one).
If possible, try to get some of the signal statistics on the GUI to give the user some feedback.
DONE!!
Week 4: Get duplexing to work in Matlab so that the acoustical system can be stimulated and measured
at the same time. Work on the data processing so that the generated data can be presented in a useful form.
See how difficult it is to do some basic things like saving files and polar plotting at specific frequencies.
Week 5: Get back from intensive recording at Jazzfest and try to get something impressive done by the deadline. I think adding "sweep duration" and "Golay order" should probably do it. However, I'd better add a picture of Robert Goulet too...
DONE!!
Week 6: Write the midterm draft of the final paper and continue to work on the signal processing since you totally nailed duplexing in matlab (thanks pa-wavplay ) even though it means that you had to change everything over to windows. Also, for people that may use GUIDE in the future, for whatever reason, the graphical parts of GUIDE in linux look way better than the windows equivalent. WTF? I thought that nothing looked better in open source because it was just a bunch of dudes programming in their free time and no one was paying them to make it look nice. How did this happen? is it that hard to put shading on the damn buttons in windows?? I'm even using a newer version of MatLab on Windows.
DONE!!
The rest of the time:
- implement a means of generating frequency specific polar responses
- Make the polar plot user interface simple (below)
- Set up a hardware configuration section so that the hardware can be setup to take advantage of the program fucntionality.
- make a plot area to display the frequency response of the measured impulse response for all excitation signal types.
- So, what does the whole thing look like?? funny you should ask...
(better plot coming soon)
The interface assumes that you are starting at zero degrees. It asks for a terminal degree and an increment in degrees. The plot can then be mirrored about the 180 or 90 degree axes if desired. In this case, only half of the plot is taken and the results are automatically extended to fill the remaining points.
The stereo box can be checked if stereo impulse response measurements are desired.
The "Available Hardware" button displays all of the current hardware (of a specific device type) along with the device number in the matlab console. The device number is then used to select input and output routing.
The gray plot is the raw frequency response data in dB. The black line is a moving average of the raw data. The data is averaged by bandwidths of 128th of the total spectral bin number.
Download: Ron Jeremy's Home Made Acoustical Characterization System v1.0
Windows only (sorry)
Final Paper!!
by: Andy Greenwood 2009
note: if you are Ron Jeremy, please don't sue me. I don't have any money anyway.
other note: if you want to give me a job, I need one incase Ron Jeremy sues me. send me an email at the above link
also, my resume is here.