• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required as is 20 years of participation in forums (not all true). Come here to have fun, be ready to be teased and not take online life too seriously. We now measure and review equipment for free! Click here for details.

In Room Phase Measurement

RayDunzl

Major Contributor
Central Scrutinizer
Joined
Mar 9, 2016
Messages
6,616
Likes
2,667
Location
Riverview, FL
#1
For the moment, let's set aside the argument over whether or not phase is audible, and just look at a measurement case.

REW has a setting named "Impulse Response Window" with the default settings shown here:

upload_2016-8-2_13-30-56.png


Here are in-room measurements of my speakers, changing the "right window" time:

Red - 500ms
Blue - 21ms
Violet - 20ms
Green - 6ms

The base measurement capture is the same file, changing only the "right window" time for display.

My reasoning in the reduction of the "right window" time is to reduce the influence of the room.

upload_2016-8-2_13-29-19.png


From 500 ms down to 21 ms the trace follows the same path (red and blue)
At 20 ms it changes drastically (violet)
At 6 ms it is nearly flattened throughout the range (green)

The room-length bounce on an Impulse Response occurs at about 25 ms. So, I suppose I am suppressing that with the 20 ms window. The bounce off the wall behind the speakers would be suppressed at the 6 ms measurement.

Am I looking at this correctly, or just fuzzing things up so it appears to give a pretty flat direct phase response?
 

Cosmik

Major Contributor
Joined
Apr 24, 2016
Messages
2,706
Likes
1,400
Location
UK
#2
What does the actual time domain impulse look like?
 

RayDunzl

Major Contributor
Central Scrutinizer
Joined
Mar 9, 2016
Messages
6,616
Likes
2,667
Location
Riverview, FL
#3
Looks like this:

upload_2016-8-3_7-13-46.png


Couch bounce at 1 ms
Dipole wall bounce at 7 ms
Room length bounce at 27ms
 

DonH56

Major Contributor
Technical Expert
Patreon Donor
Joined
Mar 15, 2016
Messages
2,654
Likes
2,668
Location
Monument, CO
#4
The drastic change from 21 ms to 20 ms indicates something happening with a round trip distance around 20 to 22 feet or so. Odd it doesn't show up in the raw impulse response, or perhaps (more likely) it is the room length, maybe a bit off due to shelves or something, and/or the standard velocity I used (1127 feet/sec for std conditions) is a bit off for your location and weather (e.g. humidity). Dropping to 6 ms removes most everything in the room.

That is a pretty clean impulse response. Even with a lot of treatment I do not think mine is as clean. Interestingly, my current response using Dirac Live is not as clean as my old response using MCACC with some manual tweaking. I have not spent as much time trying to dial it in, however.
 

RayDunzl

Major Contributor
Central Scrutinizer
Joined
Mar 9, 2016
Messages
6,616
Likes
2,667
Location
Riverview, FL
#5
Even with a lot of treatment I do not think mine is as clean.
I have 8 chunks of 7 x 24 x 48 Roxul. 3 in the corner behind each speaker, two standing behind the couch as a block to direct PC noises.

Other than that, the only soft things are the carpet and an ancient couch just long enough to sleep on.
 

DonH56

Major Contributor
Technical Expert
Patreon Donor
Joined
Mar 15, 2016
Messages
2,654
Likes
2,668
Location
Monument, CO
#6
I have way more treatment but in a boxy room and have a big screen, coffee table, and other junkola to mess up my impulse response.
 

RayDunzl

Major Contributor
Central Scrutinizer
Joined
Mar 9, 2016
Messages
6,616
Likes
2,667
Location
Riverview, FL
#7
I think I've settled my question, which was something like "Is shortening the time window on the right side making the graph flatter than it should by just "eliminating" the room?"

The line in question is the green one, above (and below). Is it really that flat or has it been squished flat by the timing selection of the display? (because it sure isn't flat in the lows when the right side window time is default at 500 ms).

So, that was with DRC applied... But now I see that without DRC (red line), the low frequency phase doesn't get flattened completely, so, it looks like there is some validity to drastically reducing the right hand window size. The diference is about what the DRC does to the phase in the low end here.

upload_2016-8-8_19-32-59.png
 

Krunok

Major Contributor
Joined
Mar 25, 2018
Messages
2,637
Likes
1,207
Location
Zg, Cro
#8
Nice write-up! It could have saved me some time if I had found this thread earlier.. :)

How exactly did you measure? Is this response from both speakers or from one?

I made 5 sweeps for each speaker from different points around LP and applied Time align and made Vector average. After that I set 3.3ms and got the phase graph with which I could work with in rePhase.
 
Top Bottom