Measurement Speaker Calibration?

[bowl_actually]

I have a set of small speakers and a small sub that I use to take to other places to do room analysis. I have used them for friends and clients who build rooms, some times in new houses or spaces where they don't yet have speakers installed to do some pre-work on the room.

For these measurements, I would like to "remove" the speaker signature similar to how we remove the soundcard and mic signatures. That way I get only the response of the "room" and not the "room + speaker."

I thought about trying to use the "house curve" but I think that's only applied to the filter section.

Is there a way to load in a "speaker calibration" if I know the response of the speakers I'm using to test? I guess I can load the speaker correction curve in EQ APO or some other systemwide EQ as long as I use the java version and not ASIO. Anyone else ever done this?

 

[John Mulcahy]

You could load the pseudo-anechoic response of the speakers as a calibration file, but typical room deficiencies (modal resonances, extended decay times) are largely independent of the response of the source.

 

[bowl_actually]

You could load the pseudo-anechoic response of the speakers as a calibration file, but typical room deficiencies (modal resonances, extended decay times) are largely independent of the response of the source.

Ah good point, in this case it would be a much bigger influence than the interface calibration as my interface is just a few 10ths of a db at very low and very high frequencies.

And true, it's all small potatoes for the room response anyway. Thanks.

 

[DanDan]

Correction Eq is used on DoDecahedron and such speakers which are used for Building Acoustics (Soundproofing) work.

 

[bowl_actually]

Correction Eq is used on DoDecahedron and such speakers which are used for Building Acoustics (Soundproofing) work.

Interesting, never heard of such a thing! I probably won't go that far, but it is very interesting to read about.

 

[DanDan]

Well, the results of such tests could void a Building or Planning Certification. Weights and Measures rules apply, like Petrol Pumps etc.
The speakers are required to deliver sound power equally in third octaves I think, or it might be as broad as octaves. 100Hz -10KHz

 

[bowl_actually]

Well, the results of such tests could void a Building or Planning Certification. Weights and Measures rules apply, like Petrol Pumps etc.
The speakers are required to deliver sound power equally in third octaves I think, or it might be as broad as octaves. 100Hz -10KHz

Very cool, I'm not doing anything that would have to meet any codes, just helping friends with their rooms. From browsing your web site, it looks like you do that for industry. No wonder you're such a wealth of info

 

[DanDan]

I haven't much experience of the Environmental/Legal side. In fact I am learning on the job at the moment. Noise Reports and Planning predictions. Coming from Music/Recording/Live this is not easy, but the rigour and process is interesting. You could measure your speaker outdoors, maybe Pink Noise, and create a simple 1/3 Octave corrective Eq

 

[bowl_actually]

I haven't much experience of the Environmental/Legal side. In fact I am learning on the job at the moment. Noise Reports and Planning predictions. Coming from Music/Recording/Live this is not easy, but the rigour and process is interesting. You could measure your speaker outdoors, maybe Pink Noise, and create a simple 1/3 Octave corrective Eq

Yep, I'm going to do what John suggested and build a filter that looks like a soundcard calibration file that represents my speaker and use that instead of the actual sound calibration file. My soundcard is so accurate that it's not doing much, I could actually add the soundcard response to the same file.

That way it gets applied every time I take a measurement and it's easy. I just have to remember when I go back to measuring rooms and speakers to build filter files to make sure I don't use that! haha