|
|
Post by patate91 on Sept 15, 2017 17:15:01 GMT
I discovered how fantastic acoustic is this summer.
I"m having fun for two months now to measure and experiment with my listening room.
It is possible to have an idea of the frequencies that reflect?
With REW'S impulse response I can locate where are the reflections. I can also attenuate them with a becover or a towel, the bedcover helping more in this case.
Is it possible to know which frequencies are reflecting in order to have a minimal treatment?
|
|
|
|
Post by Ethan Winer on Sept 15, 2017 19:23:56 GMT
I'm not sure if REW can show individual frequencies being reflected. Of course, you could run a few tests in a row using different sweep ranges. For example, first sweep 100 to 500 Hz, then 500 to 1.5 KHz, and so forth. Then the Impulse graph will show the reflected energy in each band. But this is not usually the best way to approach treatment. In any normal room all frequencies are reflected. So the best approach is broadband absorption.
|
|
|
|
Post by patate91 on Sept 15, 2017 19:37:06 GMT
I like the idea, thanks.
I will go broadband for sure, but it is more for curiosity.
I'm also trying to figure out if my 1khz to 2khz dip is cause by the speaker's crossover or room.
I managed to remove all reflections but, the 1k - 2k dip remains.
|
|
|
|
Post by Hexspa on Sept 17, 2017 5:04:33 GMT
I like the idea, thanks. I will go broadband for sure, but it is more for curiosity. I'm also trying to figure out if my 1khz to 2khz dip is cause by the speaker's crossover or room. I managed to remove all reflections but, the 1k - 2k dip remains. You could probably look at your speakers' spec sheet. Room walls, positioning and other boundaries apparently do affect your mid-to-high frequency response as well. |
|
|
|
Post by patate91 on Sept 17, 2017 11:31:30 GMT
I don't think speakers are causing (except for crossovers).
Compression driver : TAD TD-4002 with Yuichi Arai's horn who published measurements and no dip here.
I moved m'y speakers a lot but I get a nasty dip at 70hz so had to do a compromise. Speakers are pretty big but I'll certainly try other configurations
|
|
|
|
Post by Hexspa on Sept 17, 2017 14:38:15 GMT
You can try moving your listening position fore and aft, also.
|
|
|
|
Post by patate91 on Sept 17, 2017 23:46:38 GMT
Here's my ETC graphs, I think it's good Left speaker  Right speaker  ![]() And both  |
|
|
|
Post by patate91 on Sept 17, 2017 23:54:54 GMT
And Frequency response No Smoothing  1-6 smoothing  Again this is not for a studio but home listening room, I guess it's not that bad, and I can certainly improved the 70 hz dip a little more, but I'm puzzled about the 1k - 2k dip. |
|
|
|
Post by Ethan Winer on Sept 18, 2017 19:07:17 GMT
The best way to see if a response irregularity is due to the speaker itself is to measure very close to the speaker cone. That reduces the influence from reflections. If you can put the speaker on a stand out in the middle of a room, even better. But just measuring from 2-6 inches away, directly on axis, is pretty good. Then maybe move the microphone an inch one way, then another, and confirm the dip is present at all nearby close locations.
|
|
|
|
Post by patate91 on Sept 18, 2017 19:10:10 GMT
I think I measured at the horn and I think it was ok. I'll try to find the REW's file.
But I have to say when I did it I wasn't quite sure what I was doing hahaha. Maybe I should do another measurement.
|
|
|
|
Post by patate91 on Sept 18, 2017 19:14:09 GMT
Thanks for the reply!
|
|
|
|
Post by patate91 on Sept 22, 2017 17:23:59 GMT
I just measured my horns. I got the microphone inside the horn as far as my setup was allowing me. Note that did'nt manage to placed them at the same depth. But anyway at this point I don't think it matters. For what I see in the graph there's no issue with the compression drivers or crossovers, there's no dips betweeen 1khz and 2khz. EDIT - In fact there's a little dip. I Added a third graph - gold is just outside of the horn. The crossover is 800hz. Does my next investigation would be with an active crossover? ,    |
|
