DCAudioDIY.com

This has confounded me for awhile, and the answer should have been obvious. Still, I'm old, have tinnitus and HF hearing loss. Nevertheless, when sweeping speakers, I still hear the test tone over 10kHz.

Trouble is, when I measure a speaker, or a system, whether at the listening position, one meter away, or some other position; then adjust the treble driver for the flattest possible response, the treble seems overly bright, even uncomfortable. But the numbers don't lie, or do they? Well, if the measuring tool sucks.....

I've been using Room Equalization Wizard (REW) for awhile now, with upstream tools like analog line level crossovers, miniDSP, and the Behringer DCX2496. I admit my skills suck, I lack a lot of testing knowledge. However, the treble issue has had me perplexed: no matter how I measure them, if I adjust for flat, they sound crazy bright.

The microphone I've been using is the Dayton UMM-6, with the correct calibration file supplied by Dayton after giving them my serial number.

A quick Google search came up with a couple of threads in which users found that the UMM-6 has a pronounced rolloff, especially above 10k. One user, comparing to a CSL calibrated UMIK found the Dayton was down 5dB in that upper range. In another comparison between the UMIK and the UMM-6, the UMM-6 was 10dB down at its worst! Deviations between the UMIK and UMM-6 start showing up at ~2kHz-3kHz.

In the "pixelated" image, the lower two curves are UMM-6.

I just ordered the Behringer ECM8000. Allegedly an accurate mike. It is on back order at Sweetwater, but it is only $29.00 plus tax!

I'll let you know how it works!

Stuart

https://www.avnirvana.com/threads/dayto ... ison.7785/

https://www.diymobileaudio.com/threads/ ... ic.429608/

https://www.sweetwater.com/store/detail ... ne/reviews

I have always found that flat out to the high end sounds very bright to me. I think a gentle roll off above 8k sounds natural. Of course younger crowd seems to like the crisp top emd

If you look at the response curves on speaker designs from Troels Graveson, you will find most of his designs feature some attenuation at high frequencies, maybe a couple or 3 dB down at 10kHz relative to midband. You may just want to play with the tweeter level relative to the midrange. Also, some speakers are better to listen off-axis, I know my Elsinores are specifically designed to listen off axis rather than aligned at the listening position. You might want to experiment with that, it does the same thing as decreasing the tweeter level a bit. Frankly, I would adjust it to where it SOUNDS right to you.

David

I know that the old hifi research suggested a top octave (or more) rolloff, at least for concert (i.e. “classical”) music. And close miking certainly isn’t intended to capture a realistic concert hall top end. It’s the old choice- “you are there” vs. “ they are here”. You pays your money and you makes your choice.

I have found that with the room treatments I have now, a flat response does not sound too bright. With a flat response from the speakers, I believe room effects produce the sense of exaggerated highs. I'm enclosing the response of my eggs with EQ applied.

Tom

The replies have all been interesting, and I have no argument with them.

However, my point was that a supposedly calibrated microphone, with the manufacturer's calibration file, seems to be giving erroneous results.

As a pilot, we are trained to trust our instruments, not our rear ends, because the rear end (inner ear actually) will lie to you. If you can't trust your altimeter or artificial horizon, well, have your affairs in order.

Fortunately the mike seems accurate enough at low frequencies, so it is useful for room corrections/subwoofer adjustments.

I built a set of speakers using a Dayton PT2C-8 tweeter. When I had a few guys over to audition it, Jim (Bat Ears) Gerfin, said the treble was very rolled off. It made no sense to me, since the lower treble seemed right and the data sheet showed just a small 3dB dip around 14k. I went to the website and downloaded the data sheet again to my phone, and lo and behold, there was a -7dB shelving of the upper treble. there was no change in model number, just a new data sheet. Fortunately, I still had the old one. See the snips below.

Replacing the tweeters with Fountek NeoX 2.0 tweeters proved very successful.

The ears should be the final arbiter, however, measurements are still essential. And if you can't trust your instruments.............

Stuart

I would be surprised if any of the "reasonable" measurement mikes when used with the supplied correction curve would give large errors. However, you have to remember that most of these mics have a 1/2" capsule and the orientation of that capsule is important at high frequencies. At 20KHz, the half wavelength is smaller than the diameter of the capsule. At that frequency you will get different measurements if you measure with the capsule straight on or rotated in between up to 90 degrees.

Tom

I do try to get the mike orientation as close as possible to the same position each time. The mike is aligned with the tweeter, centered in both x and y planes; pointed straight on at the tweeter. I mess around with different positions from time to time, but not when doing comparisons.

I'll let you know how the ECM8000 compares to the UMM-6. I still think the issue is that (UMM-6) mike, even if it uses the same capsule as the Behringer and UMIK-1/2.

The Behringer is on back order, so it may be a while.

Stuart

When I had my Behringer ECM8000 calibrated the company that did it gave me two correction tables. one for the mic pointed at the source and one for having the mic pointed vertically at the ceiling at the height of the center of the driver I was measuring. They said this was the preferred method, which is how I have always measured my system. The mic was also marked on the barrel for proper orientation as far as the rotation of the barrel when using in the vertical position.

Good idea to have two different curves for the two orientations. If you can visualize it if the plane of the diaphragm is perpendicular to the direction of the speaker, ie pointed at the speaker, even high frequency pressure waves will be more or less uniform across the diaphragm. If the plane of the diaphragm is at 90 degrees to the speaker, you can get a pressure wave that is positive at some point on the diaphragm that is offset by the negative portion of the pressure wave that is high enough in frequency that the wavelength is less than the surface of the diaphragm. It will be pulling and pushing on the diaphragm at the same time, thus the reduction in output even though the signal is the same amplitude.

Tom