How to best use REW to balance drivers

[Rudy81]

laser188139 wrote:

Neat, Rudy.

I am by no means an expert. Most of the advice I've seen on the internet as regards phase recommends that whatever sounds stronger is correct. As the green line meets this criterion and appears smoother, I would choose it. Does the transition from woofer to midrange sound smoother, too?

I made this suggestion because I have read the same thing about the CT125 tweeters. Someone pointed out on the Klipsch forum that the driver manufacturer documented the driver as having inverse polarity. Yet most appear to connect them normally. So I figured I would have to hear and measure them both ways to reach my own conclusion.

Taking the documentation on the ALK site as correct, that the low crossover is at 400Hz and the high crossover near 6kHz, it looks like the midrange is set nearly the same level as the woofer. I assume this is all with Audyssey equalization turned off, and the response will be smoother after equalization.

Bill

Bill: I am having to revisit the whole polarity issue driver by driver. I have always hooked up the tweeter in standard manner, but need to see what the deal is with the tweeter. I have the mid level pretty well taken care of, and am just looking at the polarity issue. The woofers, btw, are Bob Crites woofers. I am in agreement with you that the green line, that is polarity in normal configuration, seems smoother in the transition. I hope Bruce chimes in and shares his wealth of knowledge on my issue. I just am not familiar enough with REW yet to properly interpret the data. That is why I made the original file available. I'm hoping Bruce or a more expert person can help me with some sage advice.

Yes, this was all with Audyssey bypassed. The plots with Audyssey on are amazingly flat! That was what convinced me that Audyssey does what it says it does. I don't even have the professional version of Audyssey, just 8 measurement points and it does a great job for the sweet spot.

[brucek]

Rudy, for anyone to play with your mdat file (other than to look at it as is), you need to provide the soundcard and meter calibration files that you're using. This is because the impulse response measure is taken without the influence of those files, so any gate manipulation would have to reapply the cal files to the frequency response, and so they would require them.

brucek

[Rudy81]

brucek: Thank you. I appreciate all your comments, they help me learn about the software and about acoustic measurement. The attached files are the Asus board sound card and the ECM mic.

[Rudy81]

Missed one. Looks like I can't upload the soundcard file here, so I will upload it to the web site mentioned before.

Many, many thanks.

[brucek]

I will upload it to the web site

I looked at your new mdat (taken with calibration files) and I would say that the wiring of the JBL in the plot called Full range would be your best method.

The response called FR rev polarity shows a cancellation that would indicate the woofer and JBL are out of phase at the crossover.

If two signals are in phase as they cross (just like a subwoofer and mains at the crossover), the resulting signal should be about 6dB higher. If I look at your mdat and do some gating, and show them with an expanded axis at the crossover of the woofer and mid JBL driver, I would say the response that you call Full Range (green trace in the first jpg below) could be considered a classic 6dB mix, and the correct polarity. The second jpg shows a dip around the crossover (blue trace) and I would consider it the incorrect wiring of the driver.

Correct crossover


Incorrect crossover


brucek

[Rudy81]

brucek: I can't thank you enough for taking time to look at my plots. After looking at the plots myself, although nowhere near as closely as you could, I decided the 'standard' wiring matched the woofer best for the reasons you mentioned. I have already changed the wiring to 'standard' + to + and - to -.

I will play with the plots so I can reproduce what you have done. This awesome program, REW, is still on a rather steep learning curve for me.

Many thanks again!

[glaufman]

Rudy81 wrote:

I just found a few threads addressing how to properly gate results. Thank you for the tip on how to take the readings we need.

I'd love to learn more about this without harassing Bruce... could you post a link to the thread(s) you found? I've already gone through the help files again (which helped) but I feel I need more...

[Rudy81]

I ran a search in the REW forum for gating and one that had what I needed was this one: http://www.hometheatershack.com/foru...surements.html

I have yet to really get into it, but plan on doing so in the next few days. I had to go back and start from scratch since brucek was good enough to PM me on some issues with my set up that could be done better. He has been a great help and his expertise is greatly appreciated.

[glaufman]

Thanks... Now I gotta make time for more reading/playing!