Room measurement help

[patate91]

I finished my room measurement with my newly acquiered minidsp 2x4 HD.

Crossover : Compression driver 650 hz - 12 db, Woofer 650 hz - 42 db Pioneer originally did : 12db - 36db. I got netter results with 42db.

There's no EQ, I still need to add some bass traps. I can live with the sub bass peak.

I know the 67 hz is a room mode but what about the ~290 hz and ~ 1200 hz dips? Can I do something to correct those dips? As per Excess Group Delay and phase seems like EQ won't help.

( Rename Final 19 nov.txt to Final 19 nov.mdat )

 

[Chris A]

The 294 Hz null looks like a microphone-room boundary cancellation. The other issues at 1-2 kHz and 3-5 kHz look to be either driver issues or crossover issues (or both). See predicted PEQs for your 2x4 HD. PEQs below the crossover frequency (690 Hz) are for the woofer channel, those above for the midrange/tweeter channel (assuming bi-amping). I've deselected the PEQs that coincide with the poles in the excess group delay curve.

Chris

 

[patate91]

Thanks for your time.

Compression drivers are loaded on wood horn. Yes speakers are biamp. Since speakers are really big I can't move them a lot. The room is L shape and not closed) 12'x18,5'x92'' (18,5' is the longest part of the ''L'', this is the microphone's back wall and the other Back wall of the ''L'' is 141'')

I attached another set of measurments : Right compression driver + Right woofer + crossover tests.

As you'll see drivers are working properly. There's no real dip in the 1000 and 3000 hz region. 1000 hz dip appeared while doing the crossover. Microphone was moved a little bit further, maybe that's the horn too.

I'm not a fan of high frequencies EQing, wondering if 1000 hz region is too high but I won't try to correct 3000hz.

For the 294hz null, if it's room related I"ll have to move accoustics panels and try to find a place it can help. I recently add a panel above the sweet spot without measuring, tomorow I'll remove it and see if it affects the ''perceived height''. With the passiv crossover and without this panel I never got a this dip.

 

[patate91]

The 294 Hz null looks like a microphone-room boundary cancellation. The other issues at 1-2 kHz and 3-5 kHz look to be either driver issues or crossover issues (or both). See predicted PEQs for your 2x4 HD. PEQs below the crossover frequency (690 Hz) are for the woofer channel, those above for the midrange/tweeter channel (assuming bi-amping). I've deselected the PEQs that coincide with the poles in the excess group delay curve.

View attachment 39429

Chris

Since it's the freq response of both speakers I doubt I can add your EQ to both speakers? It's late here, I'll try it tomorow to see what's the effect.

 

[Chris A]

I can show you how to do the REW EQ optimization, and you've got a miniDSP 2x4 HD that will easily take the corrections in frequency response shown above. I would put in the PEQs I've shown into both left and right channels as a starting point in the next REW measurements--if I were just starting out using miniDSP 2x4 HD and REW.

Beyond that, I can't do much about prior notions about what you can and can't do with EQ at higher frequency bands that you think is or isn't appropriate. That's a constraint that you own since you are choosing it.

I personally put in the corrections in the DSP crossover--up to 18--20 kHz if necessary, then re-run REW measurements, repeating and updating the process until the measured frequency response is corrected (left and right channels). Then I listen to the results. This process works extremely well in my experience.

Chris

 

[ernestcarl]

Just be careful with REW's auto EQ esp. with higher frequencies. It's preferable to base your EQ on an averaged response around multiple points or, faster yet, MMM. I use both methods and compare results from both.

Below is the result of over EQing using REW auto generated EQ filters:

Looked impressively super flat on my single measurement sine sweeps. BUT, sounded slightly wrong to my ears, so confirmed with RTA and pink noise or MMM. Turns out I was right. My boost around 260-270Hz is right now only a conservative 3dB max. Looks worse on single sweeps, sure, but audibly sounds better, esp. over a larger listening area.

 

[Chris A]

I think you are showing a loudspeaker with directivity issues. I also notice that most loudspeakers having that characteristic also typically show a downward sloping "house curve" by the user to counteract the issues with nearfield reflections from around the loudspeakers.

With loudspeakers having good directivity (like the OP's above the crossover frequency) and good microphone placement technique/nearfield absorption, REW's EQ facility works very well, as long as you pay attention to the non-flat excess group delay areas.

Chris

 

[patate91]

I removed the ceilling panels above the sweep spot and placed them half way between microphone and rear wall.

Got improvements in the bass and low mid region. I have a better RT60 too. Still need to find a way to block reflection on ceiling at the sitting spot.

There's still strong reflections to find, I know bass traps will help the too (1 corner to treat).

@ernestcarl Same thing for me, high frequencies correction never gave me good results. And a quick test with the RTA shows that 1 inch moves change the response. Sadly I don't think I'll get improvement on 1000 to 2000hz by EQuing. My horns are hand made, maybe I'll check for CNC made horns.... At least it sounds good to me, but If I can still improved things I'll be happy.

 

[Chris A]

To the OP: see https://community.klipsch.com/index...ew-to-find-parametric-equalizer-peq-settings/ for a tutorial on using REW's EQ facility.

Chris

 

[patate91]

To the OP: see https://community.klipsch.com/index...ew-to-find-parametric-equalizer-peq-settings/ for a tutorial on using REW's EQ facility.

Chris

Thanks for the link I'll look at it.

But today's measurements show a downward slope. (See the attahed file)

 

[Chris A]

Got improvements in the bass and low mid region. I have a better RT60 too.

Your RT values are already in the 0.2 s range...which is extremely low already.

Chris

 

[patate91]

Your RT values are already in the 0.2 s range...which is extremely low already.

View attachment 39571

Chris

Bass was worst on last mesurements, that's where it changes the most.

 

[patate91]

Your RT values are already in the 0.2 s range...which is extremely low already.

View attachment 39571

Chris

Sadly I don't have diffusors.

What about Impulse? I see there's still strong reflections to adress.

 

[Chris A]

This is a subject that is mired in misinformation (the problem of house curves), I'm afraid. I had this effect (directivity vs. frequency issues and the need to add "house curves" or "room curves") demonstrated very explicitly while at the Klipsch's Hope facility a number of years ago. Without going into it in depth, I can say that the lower the DIs overall or steeper the DI curve vs. frequency, the steeper the resulting "house curves" seem to be. Klipsch Jubilees (like the ones in my front left/right corners) sound exquisite with flat response on-axis (no house curve). Here's a picture of a nice looking pair (below). They have full-range directivity.

Chris

 

[ernestcarl]

I think you are showing a loudspeaker with directivity issues. I also notice that most loudspeakers having that characteristic also typically show a downward sloping "house curve" by the user to counteract the issues with nearfield reflections from around the loudspeakers.

With loudspeakers having good directivity (like the OP's above the crossover frequency) and good microphone placement technique/nearfield absorption, REW's EQ facility works very well, as long as you pay attention to the non-flat excess group delay areas.

Chris

Hi! I am not quite sure, but were you replying to my post?

 

[ernestcarl]

If so, you probably are going to be sorely disappointed at the dynamic Loudness DSP I use.

 

[Chris A]

To the OP: Are you using a round, square, or rectangular mouth horn having at least 7-8 inches (18-20 cm) height of the horn mouth? Or are you using a horn that is wide and short--like a Smith horn or a K-400-type horn? Can you relate the dimensions of the horn's mouth?

Chris

 

[patate91]

To the OP: Are you using a round, square, or rectangular mouth horn having at least 7-8 inches (18-20 cm) height of the horn mouth? Or are you using a horn that is wide and short--like a Smith horn or a K-400-type horn? Can you relate the dimensions of the horn's mouth?

Chris

An Arai's A290

 

[patate91]

About RT60, even if it's around 2ms, I don't have the impression of a dead room.

Image is very good, instruments ans voice are clear.