Looking for help and insight regarding REW room measurements - what data would you want?

[NathanW]

Hey all,

I've been checking in and out the forum for quite a while but haven't made an account until now.

I've been really trying to get my room acoustics the best I can, within my (spending) ability. I've always used Sonarworks, but recently I've dived into trying to learn REW. I have to say, it's a really interesting program with lots of data that isn't available through sonarworks alone.

I'm looking to share my measurements, with the hope some experienced people can guide me through the measurements and how to potentially tackle some things. I use a sonarworks SoundID mic, and have calibration files for it (0, 30 and 90 degrees). I've usually only used the 0 degree calibration file in REW.

I've played around with measurements and repositioning acoustic panels, and it has showed me a lot how positions things can change things dramatically, in all sorts of ways.

So, before I share any data, what would you recommend me doing and sharing, and in what kind of manner? Settings? What data would you like me to collect for you guys to check it out and help me analyze things? Pictures of the room, dimensions, .mdat file, screenshots? Zoom levels/ranges?

I'd love to know!

Thank you kindly,

Nathan

 

[Philbo King]

Save the file using REWs native file type .mdat (file->save, not file->export)

 

[NathanW]

Save the file using REWs native file type .mdat (file->save, not file->export)

Cool, I will. What else should I do carefully - measuring wise - to make things easier for you (and others) to look at? Any particular settings I might've overlooked?


Should I measure both L and R seperately from the listening position, so that you have two measurements in the .mdat?

My tweeters are 120cm apart, and my listening position tends to be 110cm (left ear to area between left tweeter and left woofer.) The area between the tweeter and woofer is 116cm high, so I asume I should measure at 116cm height?

Mic 0 degrees, or 90 degrees?

Let me know what you need. ;-)

 

[thcdru2k]

Yes it is good to sweep all channels individually and combined. If an issue that shows up in a combined sweep shows up in an individual channel that makes it easier to pinpoint which channel needs correction. If the sweeps look fine individually and the problem only appears when they are combined it is more likely to be a phase, timing, polarity, or correlated room interaction.

 

[NathanW]

Thanks @thcdru2k, I'm including my measurement I just did.

I'm curious to hear your opinions, views, ideas.

@Philbo King Here is the measurement I did a half 'n hour ago.

I also included a "noise floor" measurement. Only for the R speaker, as I had to delete the L noise floor one, to lower file size of the mdat, otherwise I couldn't upload it on here.

I also added two measurements where I added a "cloud" (baby matress) above my screen position - just to see what that would do. And it took down the big 170hz sound, but caused other troubles around 100hz or so.

I'm mainly composing orchestral music nowadays, but I also make a lot of alternative music with lots of drums as well. Just thought that might be interesting info for this type of stuff. Not sure.


For any other information, just ask and I'll be happy to provide it as best as I can. I'll also include a few pics of my attic studio.

I've recently transformed my attic space to a more pleasing space to work in. Before, it was chaotic, busy, just randomness everything. I was always just too busy with life (and kids) to tackle this space to make it nicer. The sound I had in the "before" room was good, and all my mixes translated very well on other devices. That's also I might have had that "keep it as it is, for now" attitude with the before-room. Recently I just got too tired of how it looked, and with having children, I wanted to make this space more pleasing and child-proof so they can chill here, too. The room isn't fully done yet - as you can maybe see here and there, but it's close.

I also wanted to include pics, so that you can see how things are placed/situated in the room now, in relation to the REW measurements.

Curious to hear you guys' thoughts.

Thanks so much!

 

[levimax]

Here is a great resource written by a member here @Keith_W https://www.audiosciencereview.com/...eting-measurements-with-rew-free-ebook.66771/

 

[NathanW]

Here is a great resource written by a member here @Keith_W https://www.audiosciencereview.com/...eting-measurements-with-rew-free-ebook.66771/

Thanks for sharing! This is good stuff.

 

[DWPress]

Seems a little bright on the top end. I vector summed your L/R with variable smoothing and added a simple target curve for reference:

 

[NathanW]

Seems a little bright on the top end. I vector summed your L/R with variable smoothing and added a simple target curve for reference:

Thanks! From this perspective it does indeed seem a bit bright. What else could you see from the measurements that stand out? What are your thoughts of my measurements?

I appreciate it!

 

[Sokel]

Good room (a bit dry) , good timing measurements, distortion is at check and the cloud helps big time.
Good channel agreement as well.

But, without knowing your low end , it seriously lacking down there, you have to add some meat.

Here's your gated measurements so to get rid of some room above 500-600Hz or so:
(at 50dB scale as we usually do here)

You can see the problems as of the particular mic position.

I would advise an MMM measurement, or an average around the ones we see here within a 30 cm radius so to get a better picture.

Edit: If the mic is at the height I think I'm seeing at the pic, is way high and can explain the elevated highs.
Try measuring at the acoustic center of the speaker, usually at tweeter height or in-between tweeter and mid, try both to see what looks better.

 

[DWPress]

Excellent advice from @Sokel to try the MMM method to verify your static mic sweeps. Instead of my vector average it would be nice to see separate L & R and another sweep with both playing together.

 

[Keith_W]

@JohnPM this is a comparison of his microphone calibration and his frequency response. His mic cal looks non-inverted to me. Can you please clarify if REW automatically inverts the mic cal? The REW manual does not say anything about inversion. I am wondering whether his rising top end is due to a non-inverted mic cal file.

I have already checked that he is using a 0deg mic cal and I can see from the photos that the mic is at 0deg. Assuming he did not change orientation of mic between that photo and the measurement, that is.

@NathanW could you please state the brand and model of your microphone? I can see it's an XLR mic.

 

[sam_adams]

Can you please clarify if REW automatically inverts the mic cal?

It does not.

Calibration File Format​

The calibration file is a plain text file which by default has the extension .cal, though other extensions are also accepted. It should contain the actual gain (and optionally phase) response of the meter or microphone at the frequencies given, these will then be subtracted from subsequent measurements.

Whether or not the .cal file contains inverted values is up to the mic manufacturer. The onus is on the user to know, whether or not, the .cal file has inverted values.

This particular issue was discussed somewhere else a long time ago.

@NathanW could you please state the brand and model of your microphone? I can see it's an XLR mic.

I use a sonarworks SoundID mic,

SonarWorks .cal file downloads.

All .cal files for @NathanW's mic compared:

@NathanW, in order to compare measurements properly you should always use a timing reference. Your measurements do not have that. Integrate that into your workflow—you will not be sorry for doing that. Additionally, if you can avoid the use of ASIO drivers, the Java EXCEL (Exclusive Mode Access) drivers in REW will be a better choice.

 

[Keith_W]

^^^ Thank you @sam_adams

I downloaded his mic cal (green) and inverted it (purple).

I then multiplied the inverted mic cal to the frequency response. Original is in green, with inverted mic cal is in purple. There is still a rising response above 12kHz / or a dip between 8kHz - 16kHz.

 

[sam_adams]

This is the left channel with and without the .cal file applied:

The problem lies elsewhere. Probably.

 

[NathanW]

Good room (a bit dry) , good timing measurements, distortion is at check and the cloud helps big time.
Good channel agreement as well.

But, without knowing your low end , it seriously lacking down there, you have to add some meat.

Here's your gated measurements so to get rid of some room above 500-600Hz or so:
(at 50dB scale as we usually do here)



View attachment 538406

You can see the problems as of the particular mic position.

I would advise an MMM measurement, or an average around the ones we see here within a 30 cm radius so to get a better picture.

Edit: If the mic is at the height I think I'm seeing at the pic, is way high and can explain the elevated highs.
Try measuring at the acoustic center of the speaker, usually at tweeter height or in-between tweeter and mid, try both to see what looks better.

Thanks so much.

I'll have to check how to properly perform a moving mic method measurement, as I've never done it before. If you have any quick tips, I'd love to hear them so that I can perform/measure it for you guys.

Also, the mic is at the exact height as the space between my tweeter and woofer, which is 116cm. The pic I took was from a lower angle (on my knee's) so it looks as if the mic is higher.

 

[NathanW]

It does not.

Calibration File Format​

The calibration file is a plain text file which by default has the extension .cal, though other extensions are also accepted. It should contain the actual gain (and optionally phase) response of the meter or microphone at the frequencies given, these will then be subtracted from subsequent measurements.

Whether or not the .cal file contains inverted values is up to the mic manufacturer. The onus is on the user to know, whether or not, the .cal file has inverted values.

This particular issue was discussed somewhere else a long time ago.




View attachment 538463

SonarWorks .cal file downloads.

All .cal files for @NathanW's mic compared:

View attachment 538464

@NathanW, in order to compare measurements properly you should always use a timing reference. Your measurements do not have that. Integrate that into your workflow—you will not be sorry for doing that. Additionally, if you can avoid the use of ASIO drivers, the Java EXCEL (Exclusive Mode Access) drivers in REW will be a better choice.

Thanks, that's good to know. From now on I will use the timing reference. Which of the three options should I choose?

 

[NathanW]

^^^ Thank you @sam_adams

View attachment 538469

I downloaded his mic cal (green) and inverted it (purple).

View attachment 538470

I then multiplied the inverted mic cal to the frequency response. Original is in green, with inverted mic cal is in purple. There is still a rising response above 12kHz / or a dip between 8kHz - 16kHz.

Is this something I should invert manually in REW somewhere? Also, would lowering the high shelf on my Adam T7V's be an idea? Or should I keep the monitors back panel settings as is?

Thanks.

 

[NathanW]

Hey @Sokel, here's the MMM I just performed using the method described here.

Used the 90 degree calibration file and moved around randomly within about 30 cm of my listening position. I held the microphone 116cm high (same as the spot between tweeter and woofer) in the listening position, pointed upwards at all times, and moved it around up and down and sideways (while keeping it pointed upwards.) Let me know if I did the measurement correctly for what you asked.

I included two measurements. One with the "cloud" above my desk's position, and one without. It most certainly brings down the 170hz peak big time. I work with a lot of snare samples that have that punch going on in that region.


I'm looking forward hearing (and learning!) more.

 

[Sokel]

Hey @Sokel, here's the MMM I just performed using the method described here.

Used the 90 degree calibration file and moved around randomly within about 30 cm of my listening position. I held the microphone 116cm high (same as the spot between tweeter and woofer) in the listening position, pointed upwards at all times, and moved it around up and down and sideways (while keeping it pointed upwards.) Let me know if I did the measurement correctly for what you asked.

I included two measurements. One with the "cloud" above my desk's position, and one without. It most certainly brings down the 170hz peak big time. I work with a lot of snare samples that have that punch going on in that region.


I'm looking forward hearing (and learning!) more.

Same pattern more or less, and it confirms that did your first measurements right.

It's extra weird that a room with so much absorption gives such a result.
It's not a cal file error, that is confirmed by now.

Are you willing to describe your gear, specially speakers?