What should I be looking for in REW?

[Snoochers]

I have a 9.2.6 setup in my home theatre. Today I put up a ton of insulation in a couple areas and I ran REW for the first time. It is mostly over my head, though I did find the charts below. I am generally familiar with frequency response curves so the first curve looks pretty good to me. The second curve suggests my RT60 (T20) is in the 300ms range, which seems fine? I included a fancier RT60 plot below.

Are these good? Terrible? Acceptable? I can't say for sure. What else should I be looking for? What general feedback do you have?

FYI I am using the microphone that came with my Anthem MRX 1140 receiver.

 

[GDK]

Can you change the axeson the frequency response graph so that the X-axis runs from 20-20,000 (unless you are only interested in low frequency) and the Y-axis runs from 50-100 SPL (regardless)? There is a button on the top right (I think it’s called Limits, or something like that) that allows you to adjust these - you don’t need to remeasure anything.

Also, what smoothing are you using?

 

[Snoochers]

Can you change the axeson the frequency response graph so that the X-axis runs from 20-20,000 (unless you are only interested in low frequency) and the Y-axis runs from 50-100 SPL (regardless)? There is a button on the top right (I think it’s called Limits, or something like that) that allows you to adjust these - you don’t need to remeasure anything.

Also, what smoothing are you using?

Oh yeah I missed that, no wonder it looks so great haha. Here is the udpated image

 

[Jdunk54nl]

Y axis is still -100db to 130 make it a 50db scale max. So like 40db to 90db Then smooth to like 1/6.

 

[Snoochers]

Y axis is still -100db to 130 make it a 50db scale max. So like 40db to 90db Then smooth to like 1/6.

Ok! Here is 40-90db and 1/6 smoothing. Looking worse by the minute! Lol. This would include reflections and everything I'm guessing. I don't have a gate or anything. I do have the ARC GENESIS corrections turned on though, which I'm guessing is applied to the signal that came out.

That drop in the high end is huge

 

[Jdunk54nl]

Is this one speaker playing? Multiple? Etc?

 

[Snoochers]

Is this one speaker playing? Multiple? Etc?

The window in REW shows "L+R" so I am guessing this is in stereo? I could do mono easily enough tomorrow if that is better

 

[Jdunk54nl]

Yes. Need individual speakers measured before measuring pairs. Lots of tutorials on YouTube and Google to watch and read.
Also do moving mic measurements. Erin Hardinson (Erin's audio corner) has a good video on YouTube for that.

 

[Snoochers]

Yes. Need individual speakers measured before measuring pairs. Lots of tutorials on YouTube and Google to watch and read.
Also do moving mic measurements. Erin Hardinson (Erin's audio corner) has a good video on YouTube for that.

Thanks! I will investigate. This obviously applies to the SPL measurements. Do you have opinions on the RT60 / reverb images?

 

[MaxRockbin]

What general feedback do you have?

If you haven't read it already, Amir put this together to help people get started with (and understand) REW
https://www.audiosciencereview.com/...om-measurement-tutorial-for-dummies-part-1.4/
I haven't read it (I just found it yesterday).

But I did read an extremely helpful guide from an AudioNirvana contributor "AustinJerry" which you can download here:
https://www.avnirvana.com/resources/getting-started-with-rew-a-step-by-step-guide.19/

 

[FeddyLost]

What is the size of your room?
What is the speakers and subs location?
Have you applied EQ or any receiver correction already?
Have you executed double bass array scenario or it's just two subs?

These graphs look decent, but exact opinion can be said after understandinig of whole case, especially if you have a room for improvement or it's the best you can get.
For example, for system without DRC and DBA it looks almost like miracle. If that's the best you can get, it's just decent and "nothing to worry about".

And of course, what is your impressions? Satisfaction by sound is more valuable than satisfaction by graphs...

 

[Snoochers]

If you haven't read it already, Amir put this together to help people get started with (and understand) REW
https://www.audiosciencereview.com/...om-measurement-tutorial-for-dummies-part-1.4/
I haven't read it (I just found it yesterday).

But I did read an extremely helpful guide from an AudioNirvana contributor "AustinJerry" which you can download here:
https://www.avnirvana.com/resources/getting-started-with-rew-a-step-by-step-guide.19/

Thanks. The issue with these is that they're mostly technical and show you how to set things up and how to take certain measurements but don't do a great job of telling you what "good" is or examples of optimal measurements to strive for.

 

[Snoochers]

What is the size of your room?
What is the speakers and subs location?
Have you applied EQ or any receiver correction already?
Have you executed double bass array scenario or it's just two subs?

These graphs look decent, but exact opinion can be said after understandinig of whole case, especially if you have a room for improvement or it's the best you can get.
For example, for system without DRC and DBA it looks almost like miracle. If that's the best you can get, it's just decent and "nothing to worry about".

And of course, what is your impressions? Satisfaction by sound is more valuable than satisfaction by graphs...

The room is 16x25x9 roughly. I used ARC GENESIS from my Anthem receiver and followed default recommendations there without tweaking much. One sub up front centre and one in back centre. Speakers are out front about 12' away and I think angle between them and listening position is about 26 degrees. I don't know what double bass array is!

 

[ernestcarl]

Today I put up a ton of insulation in a couple areas and I ran REW for the first time.

RT would have been interesting to compare before and after adding acoustic absorption. I generally just look at the 'topt' trace and/or the T60M curve (more for small rooms and where accuracy extends below 100Hz) -- it can be generated in the RT60 Decay tab/window. A flat or more smoothly even line is better. For multichannel home theater applications, aim for a lower RT time -- though, certainly try to keep it above 100ms. Your measured 300ms is just about normal for many home living rooms.

What else should I be looking for?

Maybe time domain stuff like decay, group delay and spectrogram views... even the individual channel phase responses between the mains and sub(s) around the xo region, which is rather a bit over the top and unnecessary if you're already letting the AVR do the time alignment anyhows.

The issue with these is that they're mostly technical and show you how to set things up and how to take certain measurements but don't do a great job of telling you what "good" is or examples of optimal measurements to strive for.

Here's a PDF guide that gives some examples of "good" measurements REW: https://www.dropbox.com/s/51jpnxet3bvew2k/REW 101 HTS Current Version.pdf?dl=1

There are a bunch of other tutorials out there too, but that one should get you started.

 

[FeddyLost]

The room is 16x25x9 roughly. I used ARC GENESIS from my Anthem receiver and followed default recommendations there without tweaking much. One sub up front centre and one in back centre. Speakers are out front about 12' away and I think angle between them and listening position is about 26 degrees. I don't know what double bass array is!

For L+R and information provided this meassurement looks nice.
Big cancellation in upper mids most probably due to L+R interference.
If you are satisfied, I'd left it all as is. Striving for better measurements is for entusiasts, I think.
Maybe you can check distortion at acceptable max SPL just to understand if you have any weak link in your system. Your room is big and if speakers are undersized, they can run out of steam at reference levels.

 

[Jdunk54nl]

This is the measurements I am currently striving for in room. House curve is based on Harman research and then fletch munson curves for extra bass. Although bass was already cut due to too much of it.

Pre-EQ Measurements.

Post EQ Measurments:

Above graphs were using moving mic measurements and REW. Below graph is using SMAART and measuring a single point in space, mainly to get phase correct.

SMAART Phase Results:

 

[Jdunk54nl]

Your original graphs, to me, look terrible between 3k and 20k...And lack of bass below 150hz would concern me.
You also have a lot of pretty wide q (width) of bumps and valleys. the wider the bump / valley is, the more you will notice it.

 

[Jdunk54nl]

Ultimately what I look for is the following.

1) Big peaks and / or valleys. Careful with valleys because it might be standing wave / reflection cancellations. Easy way to tell is if you boost it by say 5db, does it rise 5db. If not, it is a cancellation that can't be fixed by DSP. Usually moving and aiming can fix quite a few.

2) Are they following a house curve, doesn't matter which house curve, just should be following something.

3) Are the speakers playing close to the same level at all of the frequencies

4) When playing multiple speakers (2, 3, etc speakers together), do they plau nicely or cause cancellations - usually addressed via EQ and time delay but may need the use of all pass filters.

 

[Snoochers]

For L+R and information provided this meassurement looks nice.
Big cancellation in upper mids most probably due to L+R interference.
If you are satisfied, I'd left it all as is. Striving for better measurements is for entusiasts, I think.
Maybe you can check distortion at acceptable max SPL just to understand if you have any weak link in your system. Your room is big and if speakers are undersized, they can run out of steam at reference levels.

Thanks for the info!

 

[Snoochers]

RT would have been interesting to compare before and after adding acoustic absorption. I generally just look at the 'topt' trace and/or the T60M curve (more for small rooms and where accuracy extends below 100Hz) -- it can be generated in the RT60 Decay tab/window. A flat or more smoothly even line is better. For multichannel home theater applications, aim for a lower RT time -- though, certainly try to keep it above 100ms. Your measured 300ms is just about normal for many home living rooms.


Maybe time domain stuff like decay, group delay and spectrogram views... even the individual channel phase responses between the mains and sub(s) around the xo region, which is rather a bit over the top and unnecessary if you're already letting the AVR do the time alignment anyhows.



Here's a PDF guide that gives some examples of "good" measurements REW: https://www.dropbox.com/s/51jpnxet3bvew2k/REW 101 HTS Current Version.pdf?dl=1

There are a bunch of other tutorials out there too, but that one should get you started.

Thanks!