View Single Post
Old 21st July 2010
Lives for gear
Weasel9992's Avatar

Originally Posted by DanDan View Post
98% of Statistic are wrong!

I would have to pull the papers etc. but the uncertainty I refer to is the same test, same materials, performed to extremely tight defined standards with calibrated equipment, was done in different Labs. They came up with a very varied results. Comparative tests done in the same Lab, same equipment, same day, are intrinsically reliable.
Comparative tests done in the same room are likewise, intrinsically reliable.
Dan, you're over stating your point a little.

First, it's nothing like a 26% variance. Down to 50Hz it's something like 10-15% at the most. That information comes directly from the lab itself...we're not making it up. Secondly, Glenn makes a really important point about the consistency we see in the test numbers for the 244 and the Monster Bass Trap. Two different panels measured on two different days. That tells us a couple of things: one, that the testing methodology is consistent and accurate, and two, that we are in fact seeing results that are a measurement of reality and not figment of statistical variation. Now, if the results varied widely then we could question both of those things, but they don't.

On the topic of room testing being more accurate that a lab, I'll just say this: There’s a lot of value in testing your room…no doubt about it. There are about a million things that have an affect on the overall acoustics in an enclosed space: doors, windows, furniture, etc. Testing your room before treatment will provide a baseline for further comparison as you place each panel where it’s supposed to go. If you test as you go along you’ll be able to tell whether you’re headed in the right direction or not. However, relevance of those results is limited to your room or another space exactly like it. A lab on the other hand is an acoustically neutral space. In other words, it doesn’t add its acoustic thumbprint to the results like another type of space would, so lab results are applicable across a much wider sector, and are not limited to the testing environment. Plus, a lab like Riverbanks can boast an error range as small as 6% down to 100Hz. From there it gets higher, but not wildly so: up to 10% or so down to 63Hz and up to 15% down to 50Hz. Those are tough numbers for even the best room-testing enthusiast to match for all the reasons laid out above.

Furthermore, Room tests are usually conducted with a single omni directional measurement microphone using a neutral preamp and measurement software. The mic is placed at the listening position, and then a full-range sweep (20Hz-20Khz or higher) is played through the mains (one at a time and/or both simultaneously). The results are calculated by the software, and then interpreted by the user. The effectiveness of the testing is, of course limited by the skill of the person conducting the tests, his/her attention to detail, the type of preamp being used, the speakers, their placement and, of course, the room and everything in it. A lab has none of those potential pitfalls. Highly skilled, experienced personnel conduct the tests with proven, calibrated equipment using time-tested methodology with a long history behind it.

Honestly, here's it doesn't seem to be that there is sufficient reason to doubt the results of lab testing since good methodology is always followed and disciplined testing protocols observed.