High frequency peaks on L/C/R

[Aroton]

Hey,

I'm trying to track down the root cause to some shrillness in the system. I think it relates to the peaks/valleys I'm seeing at higher frequencies on my LCR channels. For now, I have EQ'd them down, but if possible, I prefer to leave most the high range all natural. (Don't judge my bass curves, my subs fix all of them)

I have first reflection on side walls treated, back wall is pretty far back, so i'm not sure that could add to the SPL of the higher frequencies. Is it possible these are caused by ceiling reflection points?

I've attached my Audyssey measurements (with the calibrated mic).

L/R Speaker

C Speaker

 

[Curvature]

You have KEF speakers? Just a guess based on the charts.

For now, I have EQ'd them down, but if possible, I prefer to leave most the high range all natural.

This is mistaken thinking. Natural is what sounds good or subjectively neutral. EQ does nothing to a speaker that isn't already happening electrically in the crossover.

Anyway, your fix, which is using broad EQ, is right. Your speakers are fairly flat on and off axis, so are likely to sound bright. By adding broadband treatment, you've effectively EQed the reflected sound and increased the relative strength of direct sound. So the characteristics of the listening window will be more prominent.

I don't think you should use Audyssey for system assessments. REW is the way to go. Hard to say what those charts you posted are actually showing.

 

[sam_adams]

I've attached my Audyssey measurements

Measurements without scalings for the axes are worthless.

 

[kemmler3D]

Welcome to ASR!

What's interesting is all 3 speakers show the same 2 peaks and a dip in the upper treble. This is pretty unusual to see, and suggests maybe you are having some nasty reflections from something.

I think with REW you can check the timing of reflections and that should give you an idea of how far the reflections might be coming from, which should help you narrow down the cause.

 

[Aroton]

Thanks All!

I had a bunch of REW measurements last night, but I apparently did not save them. I just ordered a UMIK-1 as i'm getting annoyed carrying my Focusrite around for my ECM8000. I'll redo the measurements with the UMIK-1 when it gets here tomorrow.

Yes, I do have KEF R7/R2C.

I hadn't heard treatments described as effectively EQ'ing the reflections, but that makes sense. In my perfect world I just go to Auddyssey to set a target curve and forget about it. (Obviously doesn't work as well as I'd hope, I bypass it entirely for sub management) I'm currently using the dolby atmos music target curve as I listen to a bunch of atmos content and these are dropped in to a home theater. Interestingly I have KEF architectural for my surrounds/atmos and they almost match the atmos curve perfectly, so I don't need to EQ after about 1500hz on them.

What's interesting is all 3 speakers show the same 2 peaks and a dip in the upper treble. This is pretty unusual to see, and suggests maybe you are having some nasty reflections from something.

This is what i'm concerned about. I haven't used the reflections capabilities in REW yet, but i'll look up a tutorial to see if I can't figure out whats going on.

 

[DVDdoug]

Yeah... Speakers that are flat on-axis in an anechoic chamber tend to have a slightly downward tilting curve in a real room, including mixing & mastering studios.* If you EQ to be flat in your room it may end-up "harsh" or "bright".

The important thing is that it sounds natural, so you may want to make some adjustments by-ear after fixing the "big problems", especially any bass problems.

I have first reflection on side walls treated,

That MIGHT not be best. "Regular" acoustic panels absorb mid-to-high frequencies throwing-off the frequency balance and then maybe you're over-compensating in boosting the highs directly from the speakers.

...Floyd Toole (I'm re-reading his book) is not a fan of eliminating first reflections unless you've got a particular problem you're trying to solve. Those short reflections aren't heard as reflection/delay and surprisingly they can actually improve speech intelligibility.




* There is something in cinema sound called the X-Curve. I've forgotten the details but somehow it messes-up the EQ, making the highs overly-boosted or over rolled-off. I think they get rolled-off in theaters which hurts the intelligibility of speech. Then, MAYBE they are over-boosted in the mix to compensate and they end-up over-boosted at home. You can read Amir's article yourself...

 

[Aroton]

Ya, I used MSO to fix a lot of bass issues, and did a sub crawl with REW to find best positioning for my subs. I ended up with my subs turned towards the wall instead of out toward the room... which is kind of ugly, but the bass response is much much better. End result is a pretty damn straight measured response for bass and pretty good XO at 90hz and 100hz. 60hz/80hz both caused massive nulls.

The one thing to keep in mind with acoustic panels, is I have 5.2.4 with 2 more speakers waiting to go in on the back wall (when I get around to cutting through some tile). Without them, my RT60 was pretty bad with all my speakers firing.

Per the x-curve, the interesting thing is that dolby is trying to enforce their music curve (similar to the x-curve) for all dolby atmos music. What i've found, is atmos mastered content sounds very good, while some older 5.1/7.1 dialogue can sound washed out. Stereo mixes are also bass heavy, but I don't mind that. After reading the x-curve article you linked, one thing I may try is customize the curve for my room and have a smaller falloff, and find a place I like it. Though, I still would prefer to not have spikey response regardless of EQ. I will say when I do listen on a flat curve (even having all correction turned off on the higher frequencies), higher frequency sound effects in movies, and even music, fry my ears. Even dialogue can be insanely overpowering. An issue specifically for my LCR and I am much more on axis than my surround and atmos speakers.

 

[Aroton]

Remeasured with the UMIK-1 This morning. The peaks don't look as bad compared to the Audy mic. I got everything Re EQ'd and i'm happy with the performance. My bass from my speakers still has some +-5DB peaks/valleys, but i'm not sure I can improve those much more.

I'll post my measurements for those interested.

 

[kemmler3D]

Remeasured with the UMIK-1 This morning. The peaks don't look as bad compared to the Audy mic. I got everything Re EQ'd and i'm happy with the performance. My bass from my speakers still has some +-5DB peaks/valleys, but i'm not sure I can improve those much more.

I'll post my measurements for those interested.

What do those REW measurements look like with (say) 1/24 smoothing?

 

[Aroton]

1/24

EQ-With-Sub was actually without subs... added the sub one too.

Only thing i'm "unhappy" with is I bumped all my crossover to 100hz otherwise there are some pretty bad nulls at higher volumes. But, I suppose that is why I invested in good subs.

 

[kemmler3D]

1/24

EQ-With-Sub was actually without subs... added the sub one too.

Only thing i'm "unhappy" with is I bumped all my crossover to 100hz otherwise there are some pretty bad nulls at higher volumes. But, I suppose that is why I invested in good subs.

Looks decent to me, although I'd probably keep poking away at some of the peaks, especially around 120hz.

I am personally a fan of crossing subs high, specifically for that reason. The more overlap you have, the more opportunities you have to kill nulls. Otherwise there isn't much hope for them, and I really can't stand bass notes going in and out like an indecisive cat.