Measurement technology

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I have written a white paper on how one can easily do very high resolution measurements even at home at a very low cost.

This paper is posted at GedLee LLC. Also posted is an example of a Holm data file which shows the setup that is used in Holm and an example output file (.txt) that shows the precise format that the data must be in to be read into the analysis software for display in PolarMap.

If there are any questions about implementation please post them here.

I hope that we can make this a successful endeavor.
 
I found that I left out an important aspect of my technique in the write-up - the immunity of the technique to near-field effects. I added this in, so if you have already read the paper, you may want to look at it again.

I noticed that someone gave this post five stars already (it wasn't me) - thanks. But I would ask if you do go to the site and read the discussion that you say something here so that this post does not quickly go to "page 2". Once it does that we all know that no one will see it.
 
The polarmap program examples on my website are almost all from this kind of analysis. (Some of the older data is not as refined as the newer stuff. Its usually clear by looking at the data which is which.) Originally I use Mathcad to do the calcs because they were quite elaborate. Then I decide to program them in FORTRAN because the Mathcad took forever and adding features and good graphics was not possible. I started this about ten years ago. Today I have absolute confidence in the system having used it to test and design hundreds of speakers.

I found myself being so disappointed with the complete lack of any decent objective data when we are having discussions that I decided that if I wanted people to show good data, I had to provide a way to do that.

I long for the day when the only reason someone does not show good data is because its bad.

Harman does some decent stuff, but their setup would cost millions. The world needs something that everyone can implement and we can all look at the same resolution of results. Nothing else out there has the capabilities that my system has. And it is now available to anyone who is serious. There are no more excuses.
 
Last edited:
Thank you very much for this.
It looks very hard to do with my corner horns.

You could do corner horns, its just that the very LFs would not be correct. Above about 200-300 Hz the data would be fine however. If corner loading turns out to be a major issue, it is easy to handle in the code. A corner loading is just a system with four fold symmetry.

Let me give you an example. If I want to place a hemisphere on the wall this is trivial, because what it does is to null every other mode and double the ones that are left. A corner would be exactly the same except it would null all but every fourth mode and quadruple the rest.

The modal approach is incredibly useful and flexible. I tailored what I have now to what I do. That does not mean that the approach can't handle a whole lot more. For example when I did the Orions I could not handle Dipoles, because the LF model assumed a monopole. I can now handle dipoles. I just had to turn of the monopole assumption. Mark my word this type of analysis is the future.

What I cannot handle at this point is a baffle. I started the theory and I have solved that and I started the code. Someday baffles will be possible as well.
 
There is yet another thing that I forgot to mention.

If I take the radiation modes and I move them back to the source, I can plot out the motion of the source that gave rise to that radiation pattern. The problem is the classic one of resolution in Acoustic Holography. I can only calculate back to about a 1/4 wave resolution. This is not nearly as fine as one would like, but it can still be instructive.

A spherical waveguide often has a hole on-axis. I can plot out the motion at this frequency and you can see a standing wave across the mouth from the edge diffraction.
 
Thanks for this paper Mr G.

I have a question, should the rotating axis be at the front side of the loudspeaker as shown in the image below or is this mitigated by locking the time alignment ? I presume so but no harm in confirming it with you.

Look from above:

An externally hosted image should be here but it was not working when we last tested it.
 
Last edited:
Just a note on rotating rectangular stands.
Using wide(er) dispersion speakers make sure that the stand edge doesn't protrude from the baffle surface. The reflection can be seen in the impulse/ETC response ~1ms and manifests accordingly as roughness in the high-frequency response. This is one advantage of the floor-rotating lazy-suzan.
 
Thanks for the paper, good read.
I've also downloaded the setup and application file (PolarMapSetup.exe and Polar_map.application) and the demo files (.zip and .txt) but it won't run on my Win7 Pro, 64bit with .NET 4.5.1.
I get an exception, translated "security zones do not match".
Am I missing something?
 
Thanks for the paper, good read.
I've also downloaded the setup and application file (PolarMapSetup.exe and Polar_map.application) and the demo files (.zip and .txt) but it won't run on my Win7 Pro, 64bit with .NET 4.5.1.
I get an exception, translated "security zones do not match".
Am I missing something?

same here with my win7 64 bit system.
 
Thanks for this paper Mr G.

I have a question, should the rotating axis be at the front side of the loudspeaker as shown in the image below or is this mitigated by locking the time alignment ? I presume so but no harm in confirming it with you.

Look from above:

An externally hosted image should be here but it was not working when we last tested it.

That is a very good question and why the test stand should not be much larger that the speakers being tested.

As mentioned, the front edge of the DUT should be flush with the front edge of the stand or the diffraction will be a factor. Side diffraction tends to not be a problem. Then the speaker should ideally be rotated about its center. Hence, Ideally the test stand should then match the speaker in size.

That said, I have used the same stand for many speakers sizes, but I often put them on a smaller "lift" stand if they are smaller to get them in the center of the room. I have tested many different situations - like speaker moved forward, backward, etc. to see if the stand has an effect, but it is pretty small.

Thanks for the paper, good read.
I've also downloaded the setup and application file (PolarMapSetup.exe and Polar_map.application) and the demo files (.zip and .txt) but it won't run on my Win7 Pro, 64bit with .NET 4.5.1.
I get an exception, translated "security zones do not match".
Am I missing something?

The Polarmap program can only be run as a web app. It cannot be downloaded and installed. I may change this in the future but for now that's the way it is. It also will most likely only run in Explorer.

I have not sorted out how to do the analysis software and the plotting software (polarmap) in a way that would allow their usage without their being used commercially in a way that I did not intend. I'm sure that it will be possible, I just haven't given it much thought.
 
Last edited:
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.