REW result, worrying question before proceeding

[Syncamorea]

Start here:

Before posting your measurement results

And if you haven't already:

http://www.hunecke.de/en/calculators...igenmodes.html

Originally Posted by dodittydada

Now I am quite concerned at the drop off from 5khz on. Is this in any way to be expected?

For the time being, be sure you are setting up properly, driving just one monitor, apply less smoothing, etc. The challenge in a room this size will be taming the low frequencies, so don't lock in on the highs just yet.

[dodittydada]

Thanks Syncamorea.

Here is more appropriate smoothing. Alot more nasty :0)

I completely understand what you are saying about the low end, Im just trying to find out if that drop after 5khz is something not right somewhere in my system. The room, as I said , is live as hell at the moment to even talk in.

[jim1961]

Are you measuring each channel separately?

Measuring both channels at once usually shows a lot of high end cancellation, and therefore isn't accurate.

[dodittydada]

Hi Jim. Yes, that is a measurement from one speaker only

[dodittydada]

You have me thinking Jim. What about the fact there would be extreme reflections from all points at the moment, including the wood floor. Could this not cause cancellation in the high end ? Remember this is a empty untreated completely room.

[dodittydada]

And of course, I should add I'm not losing sight of the big picture here. Which is the terrible dips between 70 and 200hz. I'm hoping more speaker positioning experimentation will help this area prior to treatment.

[jim1961]

The high frequencies are the easiest to tame and get right. I wouldnt read too much into whats happening after 5K for now.

Certainly reflections can cause nulls or peaks. But while your high end falls off a bit prematurely, its a fairly smooth roll off. This kind of thing could be the speaker, as in off axis high frequency response is common to roll off faster than an on axis response.

[Glenn Kuras]

I agree with Jim, for the time being I would focus on the lower frequencies. BTW the frequency response looks like a common non treated room.

[Syncamorea]

Originally Posted by dodittydada

And of course, I should add I'm not losing sight of the big picture here. Which is the terrible dips between 70 and 200hz. I'm hoping more speaker positioning experimentation will help this area prior to treatment.

Room measurements can be very perplexing, especially at the start of the process. As you optimize speaker and mix posiiton placement along with the treatments, hopefully it will seem more intuitive.

I wouldn't recommend it because of the time and effort required, but I once made large set of measurements of an unusual room to try to understand what was happening across the room in 3 dimensions as a function of moving the monitor around, adding treatments, etc. Suffice to say that small and medium size rooms are very inhomogenious in their frequency response as you move the mic and monitor and add absorption and diffusion. And as you optimize the room, the measurements simply don't always change in a way that's easy to understand. It's a difficult science that seems to be riddled with demons.

[DanDan]

Dada, both that mic and more importantly the calbration file are specified for free or direct field use. i.e. Point the mic directly at the source, i.e. the tweeter or acoustic centre of the speaker.
Pointing the mic upwards you will falsely read HF roll off. Trying it straight and this will prove itself.

DD

[adrumdrum]

Try to get rid of as much as possible of the worst piks and nulls that are related to speaker placement and listening possition in the low (modal region) before treatment.
It is possible to treat the speaker-related problems (pressure based absorbers), but it is when everyone else is tested, last option.

when deciding how to treat the acoustics then you should check out the waterfalls and dacy / rt. It is "fairly" easy to understand how to read, the frequency and single-frequency (in the low) with the longest decay or nulls (mostly nulls in the low), it is a problem and need most acoustic treatment.
It also means that frequencies / frequency ranges that are not as big problems need less skustikbehandling. Many high frequency problems (flutter echo, for example) are just relatet to parallel walls and can be treated with the geometry of the hard lower treatment (angled slat resonator, for example)

[Syncamorea]

Originally Posted by DanDan

Dada, both that mic and more importantly the calbration file are specified for free or direct field use. i.e. Point the mic directly at the source, i.e. the tweeter or acoustic centre of the speaker.
Pointing the mic upwards you will falsely read HF roll off. Trying it straight and this will prove itself.

The OP said:

"I pointed it straight up at the ceiling at 90 degress to the speakers at the listening position. I of course used the appropriate calibration file for this positioning."

The Cross web site suggests multiple orientation cal files are available. To be honest, when I see a Behringer product in the mix, I expect less than the best and know that's not totally fair but twice bitten forever shy.

Cross

[DanDan]

Ah, I see CSL now include the 45 and 90 degree measurements in their premium cal offerings. A welcome development.
Elsewhere on their site we see huge variations in the Behringer.
I must say I have seen nothing but disappointment from Behringer products in general. But a CSL calbrated mic of any type should be very very accurate. Until it breaks down or drifts.....;-)

EDIT, looking again at the OP's graph. A nicely sloped HF roll off hitting -3 or so at 10K. This is considered by many of us as an optimum listening 'target' curve.
See Understanding RTA at studiotips.com
and
http://www.bksv.com/doc/17-197.pdf
DD

[dodittydada]

Originally Posted by DanDan

Ah, I see CSL now include the 45 and 90 degree measurements in their premium cal offerings. A welcome development.
Elsewhere on their site we see huge variations in the Behringer.
I must say I have seen nothing but disappointment from Behringer products in general. But a CSL calbrated mic of any type should be very very accurate. Until it breaks down or drifts.....;-)

EDIT, looking again at the OP's graph. A nicely sloped HF roll off hitting -3 or so at 10K. This is considered by many of us as an optimum listening 'target' curve.
See Understanding RTA at studiotips.com
and
http://www.bksv.com/doc/17-197.pdf
DD

Thanks everyone for your input. Dandan, given the calibrated mike and correction files, are you stil recommending that I shod be pointing the mike at the speakers with the 0deg correction Instead of vertically?

[DanDan]

Yes, pointing the mic directly at the source will minimise the influence of the mic's body. If you use the 0 degrees CSL Cal file with this, you will get a very accurate measurement.
Again though, it is entirely possible that your speaker manufacturer included a HF roll off. I deliberately use a target curve very very similar to your response.
+1 at 30 +1 at 100 -3 at 10K -4 at 15K -6 at 20K

DD

[Jens Eklund]

EDIT: Oh; I now see a lot of new replies I didn’t see before posting this, apparently didn’t refresh my browser for a couple of hours ... I guess I´m late to the party ...


It looks like you're using a free field microphone (or any microphone calibrated to measure flat in free field) but not oriented properly (not aimed at the speaker but perhaps at the ceiling which might be correct if a diffuse filed microphone is in use).

In order to test it quickly: Set up your speaker (only one) so that it´s far away from any surface (including the floor) and aim the microphone (also far away from any boundaries or other obstacle) right at the speaker on the acoustic axis (check the manual, but usually right in-between the woofer and tweeter centers if normal 2-way speaker) at a distance of about one meter. Check the gated frequency response (eliminating all room reflections, focusing only on the direct sound). If this measures flat (more or less) when using the microphone in a certain way; this is the way you want to use it. If a diffuse filed mic, don´t be alarmed if you don´t get a flat response all the way up to 20 kHz. 15-16 kHz is more than enough and usually a practical limit for this kind of microphone. The benefit of a diffuse field mic is that it´s more or less linear to sound arriving from any angle of incidence as oppose to many free field microphones that assumes normal (0 degrees) angle of incidence in order to measure flat.


EDIT: Oh; I now see a lot of new replies I didn’t see before posting this, apparently didn’t refresh my browser for a couple of hours ... I guess I´m late to the party ...

[Jens Eklund]

OP:


Here’s everything you want to know (and a bit more ... ):

Why point the mic at the ceiling?