No peaks to correct, but they are there!

[frank308]

I'm just getting started. Let me begin by expressing appreciation for all the input I get from this forum. I wouldn't have known about this technology without you folks. And I certainly wouldn't have gotten this far without all the detailed explanations that are found on these pages.

Took a while to get all the gizmos and wires connected, but finally got there. Ran a few room scans. I have a sub and a mid-sub.

Sub by itself:
1st sub measurement.jpg

Midsub by itself:
1st midsub measurement.jpg

Combined response:
1st combined measurement.jpg

The room response is not horrible, but it could use some work. About that time, the whole family came home and I had to pack it up. I saved the file. Now working off-line, I reload the file, and when I hit Assign Filters, I get "No peaks to correct"

Do you have to have the whole system hooked up for the program to work, or can you take measurement and play with them later, like I'm trying to do. Please let me know if I've missed something obvious.

Thanks

Frank

[brucek]

Do you have to have the whole system hooked up for the program to work, or can you take measurement and play with them later, like I'm trying to do. Please let me know if I've missed something obvious

You're doing everything correct. Yes, you can work offline and get a set of filters, and then walk over and enter them into a BFD.

If REW doesn't recognize enough of your signals to conclude that they're peaks, simply lower the target level with the thumbwheel until you reveal more signal above the target and then Find Peaks again. You can clear those peaks and then move the target level some more and find peaks again until you like what you see........


peaks.jpg


brucek

[frank308]

I can reduce the target so the entire trace is above it, but still no peaks:
screenshot800.jpg

Would it be worth while to attach a single scan? Can you open it and see if I have something messed up?

Thanks

Frank

[Otto]

Hi Frank,

Did you hit the "Find Peaks" button that brucek pointed out? It's implemented more like a hyperlink rather than a button proper, so it might not quite stick out.

Using Lotus Notes, huh? Don't see that everyday. You can send me a scan and I'll see if I can make it work. I'll PM you my email.

[frank308]

OOOOOOPPS!

Nevermind!

I didn't realize that was a button. Thought it was just a caption for the Hz range. Thanks, Otto. Now I have all the peaks a man could want (figuratively speaking, of course).

Frank

[JohnM]

That error message isn't particularly helpful, but the assign filters action shouldn't really be enabled unless some peaks have been found so I've fixed that for the next release.

[logan00]

If a REW sweep is taken in the range of 20Hz to 200Hz should I use REW to find peaks in the same range or only up to the crossover of the sub?

For ex, if the sub has a 80Hz crossover, do I set find peaks to be 20Hz to 80Hz or should I go higher?

[terry j]

I guess there would be no real point finding the peaks in an area that the behringer wouldn't cover, so I'd only go to the frequency that the beh. can touch.

[Otto]

If find, that if there are indeed big peaks above my crossover point, they can still affect the frequency response when I'm done EQing. Because the low pass filter isn't a brick wall, some of that information can still get through, even though it's technically above the crossover point.

Suppose a crossover point of 80 Hz. I believe that 24 dB per octave is a common crossover slope. So, by 160 Hz, your bass signal is down 24 dB. That's a good cut, but what is it at 100 Hz? Down 8 or 10 dB? Still a significant cut, but signals in that range can still affect the overall output.

A good way to ensure everything's working OK is to do your sub EQ, engage the mains, then measure again. In addition to the any sub interaction above the crossover point, you should also be able to detect crossover point integration problems (phase problems) between sub and mains.

Sorry for getting a bit off topic...

[frank308]

"you should also be able to detect crossover point integration problems (phase problems) between sub and mains. "

How? Do you have an educational link that you can point to?

Thanks

Frank

[brucek]

How?

1. Measure the sub with the mains turned off and equalize to best response.

2. Measure the sub with the mains turned on.

3. Adjust the subs phase to get the cleanest transition at the crossover frequency.....

brucek

[Otto]

frank308 wrote:

How? Do you have an educational link that you can point to?

Hi Frank,

If you were having problems in integrating mains and sub (or sub to mid-sub to mains) you could potentially EQ them perfectly individually, but then see dips when they are all measured together. The separate signals from each "speaker" can combine negatively to cancel out portions of the frequency spectrum, especially as they interact around the crossover point. This would show up in your REW measurement of your entire system -- sub, mid-sub and mains all running during the sweep.

I'm not sure if it's educational, and I only re-skimmed it right now, but there was some related discussion here.

[terry j]

was mucking about with the new version today, it's growing on me!!

In regard to not finding peaks to correct, I had that problem today as well. However, the solution was quite simple.

Make sure that the trace you are correcting has no smoothing applied. Try the following experiment, make two measurements (without changing a thing) and on one, progressively add more and more smoothing. The more smoothing you apply, the less and less peaks are foound to correct, usually ending up with 'no peaks to correct' at full smoothing.

In any case, for correcting the bass, I think it is usually recommended that no smoothing be used in any case.

[brucek]

In any case, for correcting the bass, I think it is usually recommended that no smoothing be used in any case.

That is absolutely the rule. Never use smoothing until you're finished equalizing. Filters that are optimised against a smoothed response will have settings that don't accurately match the room's modes.

brucek