Subwoofer Phase Testing?

[miklorsmith]

I have REW and the ability to use it as well as complete phase control over my subs.  Any special tricks for getting phase precisely correct?

It seems with REW's "Impulse" measurement this could be done but the tutorial says basically whenever the system "hears" first sound, "Time 0" begins.  I guess I could try running this to see whether two distinct peaks are present.

If so, I should be able to experiment with different phase settings until the peaks line up?  Hmmm, that might not work since the sub might be a full cycle behind when everything's perfect which wouldn't look right on such a graph.

Considering REW's home theater roots, it seems there should be a specific function to do this.

[richidoo]

http://forums.avguide.com/viewtopic.php?p=613&   

Your system should allow trying both these techniques.

Let me know if you need an extended specific frequency tone.

[miklorsmith]

Thanks Rich, the second technique sounds perfect.

I found the RealTraps tones but they don't have one extended frequency track.  Ideally I'd like to have individual, continuous play tracks at 70 - 90 hz.  Even if it was just one tone at 80 hz I'd probably be set.

Will Audacity generate test tones?  I have that loaded on my home PC but haven't used it for a while.  If so, I could generate a file, copy it into the SB folder, and off to the races.

Awesome!

[richidoo]

I don't think Audacity can generate the tone. I have Sonar pro recording software with a few plug in synths. I will see if it can generate the tones by freq. If not I will splice together a bunch of 80Hz from Ethan's 2second sample. 2 more coats of orange paint and I can get right on it..

[miklorsmith]

From Help at Audacity site - "Generate" menu:

Tone

Inserts a wave of chosen type, frequency and amplitude. The length is determined by the length of your selection and the position by the left boundary of your selection.

If no selection is made, the default length inserted at the cursor position is 30 seconds.

Parameters:

Waveform    Sine, Square, Sawtooth
Frequency(Hz)    1-20000 (value below 1.0 and above 20000 will be corrected to 1.0 and 20000!)
Amplitude    0-1 - This is the volume of the generated wave.

I think this should be able to do it.

I'd hate to have you rush your time in the sun.   

[richidoo]

Good find on audacity! That is a useful program.

Audio test tone generators abound, easier and more flexible than static files. I tried this one, it works OK with lots of nice adjustments including phase. But it only plays tones for 15 seconds in the free version. Note: R/L mouse clicks adjust each digit of freq (there I just saved you 10 minutes of frustrating learning curve.... )  It's from russia, but scans clean.
http://www.download32.com/sound-generator-i3032.html

[miklorsmith]

Nice!  I'll report back with findings.  I'm having a big shindig tomorrow eve and really want to get this dialed in.  I had previously thought phase was OK in my subs, being coincidentally located but I've been playing around with the ASI Sugarcubes (amazing) and just haven't been able to get it right.

Rich - those Sugarcubes might be THE solution for your small room.  I'm pretty cautious about recommending people spend money on anything but I've been playing around with these things for a couple of weeks now and they are undeniably effective.

The problem with them is that since I don't understand how they work experimentation is totally random shots in the dark.  But literally four of them placed behind the speakers can turn boomy and undefined into tight and controlled.  They can even turn boomy into a suckout.  It's freakin' crazy.

It should be a measurable phenomenon.  At some point I'll try to do that

[miklorsmith]

WOW, that was a piece of cake.  Audacity generated a five minute, 80 hz .wav file in literally 2 minutes.  I could have played it on the SB but burned to a CD instead.  I flipped the speaker leads to the mains and it was really obvious within about 15 degrees where the null was, basically centered it there.  Total process was about 15 minutes.

Now to tone down the EQ a bit . . . 

[richidoo]

Great, glad to hear it worked!

Sugar Cubes:
http://sixmoons.com/audioreviews/acousticsystem4/sugar.html
Looks interesting! Not read what the hell it is yet...

I have already moved the Ushers back into the big room, in a new position that is working much better than before. Ironically closer to walls. But I will put the Feastrex into the small room, they sound great in there.
Rich

[miklorsmith]

OK, here's the wrap on the phase testing.  The "maximum null" (strange term) was very obviously at -150 degrees with both speakers, performed individually.  This equates to being nearly completely out of phase in the uncorrected settings.  I swapped back to correct polarity on the speaker leads and was swamped with bass. 

Using Ethan's test tones through the SB and the laptop interface over the bass EQ I measured the collaborative bipole energy.  Overall, the 70 - 85 hz range was up to 20 db above the rest of the bass range.  I ended up cutting this area 8 - 20 db.  This makes intuitive sense but was still quite a surprise since the old settings were carefully calibrated.

Ethan's tones are in 1 hz increments.  After going through the difficult sections several times I got plus/minus 2 db in-room from 25 to 100 hz.  This was the lo-tek strategy.  Overall at this point I don't trust the optimization of this EQ set - I have about 6 programs that evolve over time.  After these experiments I will need to start completely from scratch at some point. 

For now, everything sounds really locked in, like I want to listen more than scratch my head and play with measurement protocols.  I know, completely shocking.   

Happy Sunday, especially to everyone that gets an extra hour of it.

[richidoo]

Hey Mike, that is awesome! It makes perfect sense to me. Picture before the phase adjustment, you had two drivers out of phase smearing bass all over the place. Bring them into unison and they should get tighter and louder, that's what happened. Then you turn it down with your tricky bass EQ gadgets, and voila, bass is tight and articulate at the right volume.

The measurement instruments really don't matter as much as using them correctly which you have done. I'm sure the impulse test in REW will be impressive. But beware the room reflections on the graph. Watch the distances of drivers to mic to prevent a reflection arriving at mic same time a direct wave does. 

At stereofools house we were able to tweak the phase by ear to greatly improve the bass clarity. The bass volume increased with this and needed to be turned down. It was possible to make a big improvement by ear, but with testing software you should be extremely close, more than your ears resolution which is not very sensitive in low freqs anyway.

I wish I could hear it! After 6 months with the new speaks it should be getting dialed in nicely by now...

[miklorsmith]

True true.  It's actually been nigh on a year I've had these buggers and the Pros for a year before that.

It didn't sound smeared before, though by comparison I'm sure it would.  I think I had the subs turned up high enough in those frequencies to overpower the fronts almost entirely.  They're capable of so much clean power that I wasn't over-driving the amps or bins, no harm no foul.

So, I had the frequency stuff pretty much right before.  What I gained through this exercise is much better "string" definition on plucked upright bass.  Finger popping is really good now, in much better balance with following resonance.  Before, tone was solid, quick, and controlled.  Now, I've added instrumental visual detail - a meaningful advance.

Yep, two years of learning about all this stuff and I still feel green as spring grass sometimes.