Measurement problems

[fatmarley]

Is it ok to leave a link to another forum? If not, sorry.

How do I get accurate low frequency measurements?

I've had so many conflicting opinions, I don't know what to believe. Just hoped someone here knew the correct answer to my question.

[BobEllis]

Try posting your question rather than posting a link to a thread and asking us to figure out what you want to know. "Just hoped someone here knew the correct answer to my question." is a bit vague.

Search here for measurement techniques. Folks like Jon Marsh, Curt, CJD, et. al, probably have already posted the answer to your question. As you noticed in the PE forum, there is a bit of art mixed in with the science of measuring a speaker outside of an anechoic chamber. Some of the variations are driven by the space you have for measurements. I doubt Jon Marsh is hoisting his latest efforts several feet in the air to get his measurements. He may have done that with smaller speakers.

Another good resource is http://linkwitzlab.com/

[cjd]

I know I haven't posted information on this, because I don't bother unless it's RTA for correctional purposes. Low bass is predictable enough when you know the T/S paramters and driver placement...

Maybe I'll learn otherwise some day.

[JonMarsh]

Try posting your question rather than posting a link to a thread and asking us to figure out what you want to know. "Just hoped someone here knew the correct answer to my question." is a bit vague.

Search here for measurement techniques. Folks like Jon Marsh, Curt, CJD, et. al, probably have already posted the answer to your question. As you noticed in the PE forum, there is a bit of art mixed in with the science of measuring a speaker outside of an anechoic chamber. Some of the variations are driven by the space you have for measurements. I doubt Jon Marsh is hoisting his latest efforts several feet in the air to get his measurements. He may have done that with smaller speakers.

Another good resource is http://linkwitzlab.com/

Good feedback, Bob.

The thing is, the posts on TT were all fairly valid, and reflect different ways of looking at the issue and getting data. One could justifiably say that none of them are "wrong", but whether one is right or not for the specific problem is governed more by what you want to accomplish.

It never ceases to amaze me that people will go to huge efforts to make quasi-anechoic LF measurements when in reality, they will always be listening to the system in the actual room! Designing the speaker apart from room acoustics and placement issues is a sure fire path to unhappiness with the final results. I will post more about this in the build thread for the Isiris, once I get that started, with the material in the presentation from the Northern CA DIY meet in April.

Where hoisting a cabinet up on a pole and measuring at one meter might be useful is for getting a 4 pi space response including the effects of baffle step. However, my experience is that for larger systems, 1 meter is too close for baffle step effects to be properly registered in the measurement.

Just judging the system LF performance can be done by combining a near field (say, 6") measurement of the woofer, adding in the port response measured at the same distance. Of course, combining measurements to a composite curve is a walk in the park in Fuzzmeasure, maybe not as easy in some other software packages. Generally, I'd worry about the port output mainly to verify that things are working in that area as expected- from the viewpoint of designing a crossover, it's pretty meaningless.

What is meaningful, in my opinion, is how it's going to measure with the cabinets in the playback position and your ears (and mic) in the nominal listening position. If you have trouble getting good looking measurements this way, believe me, the sound isn't going to be any better. Things that will affect this include LF refections from adjacent boundaries and floor and wall bounce.

I typically use 40-100 msec long windows for full range or LF measurements, with a Half Hanning or Half Hamming window; Praxis and Fuzzmeasure both support a variety of window selections to help optimize the ratio of LF information to the impact of late room reflections in the upper range.

For insight into boundary loading affects from nearby surfaces (walls, floor) check out this thread.

http://www.htguide.com/forum/showthread.php?10872-Guestimating-room-gain&p=138262#post138262