My first REW measurements

[bracko]

Hi

I have done some measurements in my listening room (6x9 m) using REW and a Sonarworks mic connected through Focusrite interface. Mic placed at the listening position. What can you see in these measurement values? What can be improved? Thanks in advance!

The output from REW is shown in the captured images.

SPL and Phase

Distortions

Group delay

Clarity

Waterfall

 

[Keith_W]

Please read this thread: REW help.

We don't know:
- Whether you took the measurements properly
- Whether it's one speaker or both
- Are there subwoofers or not

Your FR sweep is too zoomed out. It is normal to show about 50dB on the y-axis, you are showing +110 to -20dB (130dB). Also, your measurement does not extend down to 20Hz.

Your distortion measurement does not show ANY distortion. This is because you are only showing the main signal and the noise floor. Tick the THD button to show distortion and change the y-axis to %.

None of your graphs show labels.

And lastly, it is most helpful to (1) post the MDAT, and (2) BE SPECIFIC when you ask a question.

 

[kemmler3D]

@Keith_W has good points here. You will get better answers if you take that advice. More so if you are hearing a problem and can describe it.

The scale on the graph is important, and you are using too much smoothing I think.

All I can say for the moment is you have a big issue at 300hz and seem to be missing some energy around 1500hz also. I think this might lead to things sounding a little dull and thin?

 

[Hear Here]

And lastly, it is most helpful to (1) post the MDAT, and (2) BE SPECIFIC when you ask a question.

That's a bit harsh! The OP stated up front it was his first REW measurement and he's made a much more comprehensive job of it than I could do. No doubt you are experienced and I'm sure he will appreciate your earlier more constructive suggestions. We live and learn - always from a starting point of ignorence!

 

[kemmler3D]

That's a bit harsh! The OP stated up front it was his first REW measurement and he's made a much more comprehensive job of it than I could do. No doubt you are experienced and I'm sure he will appreciate your earlier more constructive suggestions. We live and learn - always from a starting point of ignorence!

Fair to say, I think you are both right though. You could read the comment as harsh or just direct. Either way OP can get more mileage from REW by following the advice.

 

[bracko]

Please read this thread: REW help.

We don't know:
- Whether you took the measurements properly
- Whether it's one speaker or both
- Are there subwoofers or not

Your FR sweep is too zoomed out. It is normal to show about 50dB on the y-axis, you are showing +110 to -20dB (130dB). Also, your measurement does not extend down to 20Hz.

Your distortion measurement does not show ANY distortion. This is because you are only showing the main signal and the noise floor. Tick the THD button to show distortion and change the y-axis to %.

None of your graphs show labels.

And lastly, it is most helpful to (1) post the MDAT, and (2) BE SPECIFIC when you ask a question.

Thanks for your answer.
Hopefully I took the measurements properly but now that I have your link I will read it, follow it and make some new measurements. I will also send an mdat file here.
I measured both speakers with the mic placed at the listening position. There are two subwoofers.

The distortion graph shows SPL and THD (in db)

More specifically, I was most "worried" about the dip at 290 Hz and the loss of energy in the mids (around 1,5 kHz). Maybe there is something that can be done.

 

[TurtlePaul]

96 dB in the bass, 80 dB at 1.6 khz and 88 dB at 5 khz, looks too v-shaped for me. Is dialog easy to hear and interpret? Does it sound too bright?

 

[bracko]

@Keith_W has good points here. You will get better answers if you take that advice. More so if you are hearing a problem and can describe it.

The scale on the graph is important, and you are using too much smoothing I think.

All I can say for the moment is you have a big issue at 300hz and seem to be missing some energy around 1500hz also. I think this might lead to things sounding a little dull and thin?

Thanks! Yes, I was also concerned about the energy loss at those two parts of the spectrum.

 

[Draki]

The unwritten convention is to use a 50 dB scale for the SPL. You are using 130 dB (-20 to 110 dB). The 50 dB scale is just more comparable.

You can use the "limits" on the top right to set it.

 

[bracko]

96 dB in the bass, 80 dB at 1.6 khz and 88 dB at 5 khz, looks too v-shaped for me. Is dialog easy to hear and interpret? Does it sound too bright?

I can probably lower the energy in the bass by lowering the output from the subwoofer. The loss of energy in the mids is probably a little harder nut to crack. It can be inherent to my main speakers. I don't think the sound is bright and I think that it is easy to hear and understand.

 

[TurtlePaul]

Pointing the speakers directly at you and making sure they are well away from side walls may help with the 1.6 khz, but most of that dip is probably the speakers directivity error in the woofer and tweeter crossover range.

Can you share what speakers you have?

 

[bracko]

Pointing the speakers directly at you and making sure they are well away from side walls may help with the 1.6 khz, but most of that dip is probably the speakers directivity error in the woofer and tweeter crossover range.

Can you share what speakers you have?

Thanks. The loudspeakers are at least 1,5 m from the side walls and some 80 cm from the wall behind.
It is "modded" Monitor Audio Silver 8.

When you say "directivity error" in the XO, what exactly do you mean? Phase error or "polarity" error?

 

[Hear Here]

I measured both speakers with the mic placed at the listening position. There are two subwoofers.

I'm interested to learn more about REW measurement procedure. It seems most people take measurements at the listening position, presumably a number of sweeps that are sent to each or both speakers. Makes sense to me.

By contrast, the procedure for taking Dirac or other DSPs measurements is for multiple mic positions to be measured from 7 or 13 different positions. around the listening position - a bit random as accurate mic positioning may be tricky. Why should this be the case if you are trying to optimise the sound quality at the listening position? People sitting nearby have to put up with slightly less good sound, or separate measurements could be taken with the mic at a secondary position, such as a work desk.

Why shouldn't I take all Dirac measurements from precisely where my head would be when listening to music? Easy and accurate with the help of a mic stand

 

[TurtlePaul]

When you say "directivity error" in the XO, what exactly do you mean? Phase error or "polarity" error?

It means that even if the on-axis sound has perfectly good frequency response, the off-axis sound can have a frequency hole. This is usually because tweeters project a much wider sound than woofers which are at the top of their frequency band. Also, floor and ceiling reflections have a cancelation at the crossover frequency because one of the drivers is closer to the floor and one closer to the ceiling.

Monitor Audio Silver 8 loudspeaker Measurements | Stereophile.com

Sidebar 3: Measurements I measured the Monitor Audio Silver 8's farfield frequency response using DRA Labs' MLSSA system and a calibrated DPA 4006 microphone. I used an Earthworks QTC-40 microphone for the nearfield responses, its small, ¼"-diameter capsule offering no significant barrier to the...

www.stereophile.com

Stereophile measured these speakers, and the on-axis response is down a couple of dB from 1 khz to 3 khz. Some of the rise from 1.5 khz to 5 khz is just the anechoic response of that tweeter.