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

[NathanW]

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?

Just so that I don't confuse myself, are you talking about the high end "brilliance" frequency area?

My monitors are Adam T7V's. What I did just see, is on the back of the monitors the low shelf is set at -2. The High shelf at 0. Could this be something as well? If you want I can re-measure with both settings at 0 or any other combination.

I use it in combination with a SSL2 Interface. Monitors are on stands.

 

[Sokel]

Just so that I don't confuse myself, are you talking about the high end "brilliance" frequency area?

My monitors are Adam T7V's. What I did just see, is on the back of the monitors the low shelf is set at -2. The High shelf at 0. Could this be something as well? If you want I can re-measure with both settings at 0 or any other combination.

I use it in combination with a SSL2 Interface. Monitors are on stands.

I would set the low self at 0dB or even +2dB and high self at -2dB.

Measure like that (and have a listen as well) and let us know.

 

[NathanW]

I would set the low self at 0dB or even +2dB and high self at -2dB.

Measure like that (and have a listen as well) and let us know.

Cool, I will make a few more measurements right now!
I'll do:

Low shelf: 0db
high shelf -2db

and

Low shelf +2db
high shelf -2db

Something like that. Will send in a few minutes

 

[Sokel]

As you are measuring, here's the anechoic response of your speakers taken from this thread:

Not very different than what you measure.

 

[NathanW]

As you are measuring, here's the anechoic response of your speakers taken from this thread:

View attachment 538517

Not very different than what you measure.

From that perspective, is this a good thing?

Also, I attached the new measurements with different low/high shelf settings on my monitors, as described in my previous reply.

 

[Sokel]

From that perspective, is this a good thing?

Also, I attached the new measurements with different low/high shelf settings on my monitors, as described in my previous reply.

The 0dB low self / -2dB high self does it for me just fine.

You can now EQ the rest down low (up to 300Hz or so) according to your measurements and the rest according to the anechoic data you have from the above thread if something bothers you up there.

Before that, you can play with toe in/out, positioning, etc.

 

[NathanW]

The 0dB low self / -2dB high self does it for me just fine.




View attachment 538522



You can now EQ the rest down low (up to 300Hz or so) according to your measurements and the rest according to the anechoic data you have from the above thread if something bothers you up there.

Before that, you can play with toe in/out, positioning, etc.

Thanks so much.

EQ'ing based on the anechoic data is really above my head still. I'm not sure how to even go about that. I'd love to learn that, though. Is it something I should look into if I want to get my signal more flat and even?

Also, when wanting to EQ the monitors, do I re-do the L and R measurements and do them both multiple times and making an average out of them combined? Multiple L's and multiple R's combined into one average, and then use that to EQ the speakers? Or can/should I EQ each speaker (L & R) individually?

Would you recommend using something like equalized APO to have it run on the whole system, or another route? My SSL2 interface doesn't support EQing the interface, I believe.

 

[Sokel]

Thanks so much.

EQ'ing based on the anechoic data is really above my head still. I'm not sure how to even go about that. I'd love to learn that, though. Is it something I should look into if I want to get my signal more flat and even?

Also, when wanting to EQ the monitors, do I re-do the L and R measurements and do them both multiple times and making an average out of them combined? Multiple L's and multiple R's combined into one average, and then use that to EQ the speakers? Or can/should I EQ each speaker (L & R) individually?

Would you recommend using something like equalized APO to have it run on the whole system, or another route? My SSL2 interface doesn't support EQing the interface, I believe.

There's already suggestions about EQ'ing the anechoic response at the linked thread, you can just copy them and play with them.
You must individually EQ each speaker based at the anechoic up high (ignore the room measurements up there) .

You must also correct the response low based on your measurements after you done with the best positioning you can do first each speaker alone and then both just to confirm.
APO is just fine, be careful about levels, double check, etc.

 

[levimax]

Thanks so much.

EQ'ing based on the anechoic data is really above my head still. I'm not sure how to even go about that. I'd love to learn that, though. Is it something I should look into if I want to get my signal more flat and even?

Also, when wanting to EQ the monitors, do I re-do the L and R measurements and do them both multiple times and making an average out of them combined? Multiple L's and multiple R's combined into one average, and then use that to EQ the speakers? Or can/should I EQ each speaker (L & R) individually?

Would you recommend using something like equalized APO to have it run on the whole system, or another route? My SSL2 interface doesn't support EQing the interface, I believe.

For EQ I like to use MMM measurements, way faster and easier and gets the same result as averaging multiple sweeps. I would do L, R, and combined which can be helpful information.

There are a lot of ways to apply the EQ. Sounds like you are using a computer in your system? If so many players support EQ filters created in REW like JRiver or Foobar2000. Equalizer APO is fine as well but another piece of software of load and learn. What player are you using now?

 

[sam_adams]

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

Setup the SSL to use a wired loopback timing reference. Connect one output to the corresponding input. Make sure that monitoring is turned off and the '4K' switch is not engaged. With the SSL setup this way, you will have to swap the output cables to test each channel individually, though.

 

[NathanW]

For EQ I like to use MMM measurements, way faster and easier and gets the same result as averaging multiple sweeps. I would do L, R, and combined which can be helpful information.

There are a lot of ways to apply the EQ. Sounds like you are using a computer in your system? If so many players support EQ filters created in REW like JRiver or Foobar2000. Equalizer APO is fine as well but another piece of software of load and learn. What player are you using now?

I was planning on using Equalizer APO, as it can run system-wide, so I won't have to think about it anymore once EQing the monitors have been done. The past 1.5 year I have been using Sonarworks in my DAW on the master bus, but didn't like that it didn't run system wide all the time, even when checking things such as youtube etc. I play a lot of music through Youtube or Spotify web.

I just want to have it EQ'ed everywhere.

I hardly ever play things through Windows Media Player, except for my own music once I've exported tracks. I don't listen too much music through a media player on my PC.

I'm willing to learn how to program the Equalizer APO program. But if you know another program as such that runs system-wide, I'd love to know.

 

[NathanW]

Setup the SSL to use a wired loopback timing reference. Connect one output to the corresponding input. Make sure that monitoring is turned off and the '4K' switch is not engaged. With the SSL setup this way, you will have to swap the output cables to test each channel individually, though.

I'm so sorry, but I'm confused. I thought it was about setting up the timing reference in the REW measuring settings (as seen in my screenshot of the measuring tab). Forgive my unknowingness, haha. I'm still learning things on the go. I'm eager to learn, though.

 

[sam_adams]

I'm so sorry, but I'm confused.

Green is output to speaker (DUT).

Red is TRS-TRS or TRS-XLR cable.

Blue is input from mic.

The 'loopback' function comes from connecting one of the outputs to one of the inputs (Red connection in the above illustration.). You then select 'loopback' as the timing reference when making measurements. Since you now have only one channel to send the stimulus signal to the DUT, you will have to test each one individually by swapping the cables.

 

[levimax]

I'm willing to learn how to program the Equalizer APO program. But if you know another program as such that runs system-wide, I'd love to know.

I think there are some others but for your use Equalizer APO sounds perfect.

 

[NathanW]

So, I am doing L and R moving mic measurements, and the right channel shows a pretty big dip at around 150 hz. I have turned my whole room upside down (almost literally) to see what might cause it, but I can't figure it out...

Here is the .mdat

I also used RTA for longer periods to check out what might be causing it - moving stuff/panels in real time - but I can't seem to find it at all. I also swapped the monitors to see if it might be a faulty monitor, but still got the exact same result, so it's not the monitor itself. Any ideas?

Should I be worried about that dip at around 150hz on the R channel? Am I overthinking it? It seems like a somewhat critical frequency range.

And what do you guys think of the dip between 200hz and 400hz?

This is when I'm looking at it at 1/48.

@Sokel @levimax @sam_adams @thcdru2k @DWPress @Keith_W

Thanks.

UPDATE:

I found a very thick pillow and placed it right above the right speaker in a horizontal position. The dip goes away by like 8db. Added .mdat for comparisons. Did extra double check comparisons to make sure it wasn't a coincidence.

Do you think the measurement with the pillow above the right speaker is better? Or am I still overthinking things?

Thanks again!

 

[thcdru2k]

So, I am doing L and R moving mic measurements, and the right channel shows a pretty big dip at around 150 hz. I have turned my whole room upside down (almost literally) to see what might cause it, but I can't figure it out...

Here is the .mdat

I also used RTA for longer periods to check out what might be causing it - moving stuff/panels in real time - but I can't seem to find it at all. I also swapped the monitors to see if it might be a faulty monitor, but still got the exact same result, so it's not the monitor itself. Any ideas?

Should I be worried about that dip at around 150hz on the R channel? Am I overthinking it? It seems like a somewhat critical frequency range.

And what do you guys think of the dip between 200hz and 400hz?

This is when I'm looking at it at 1/48.

@Sokel @levimax @sam_adams @thcdru2k @DWPress @Keith_W

Thanks.

I’d treat the 150 Hz dip as a right-side placement/boundary cancellation since it stayed after swapping monitors. If the dip shifts in frequency when you move the speaker, that would confirm boundary interference. The 200–400 Hz dip shows in both channels, so that may be shared desk/floor/sloped-ceiling geometry rather than a speaker issue.

 

[DWPress]

Once again: separate L & R sweeps and then one combined LR sweep. Use timing reference. There is no timing information in MMM measurements but give us a LR combined MMM. This is what a quick look in REW looks like with your recent MMM L&R but better information can be had if you provide what we ask for.

 

[NathanW]

I’d treat the 150 Hz dip as a right-side placement/boundary cancellation since it stayed after swapping monitors. If the dip shifts in frequency when you move the speaker, that would confirm boundary interference. The 200–400 Hz dip shows in both channels, so that may be shared desk/floor/sloped-ceiling geometry rather than a speaker issue.

Hey, I actually edited my previous post about having placed a thick pillow on top of the right speaker and the 150hz issue seemed to change about +8db. I added the .mdat there. It includes double checks to make sure it wasn't a coincidence. Do you think the measurement with the pillow looks better?

 

[thcdru2k]

Once again: separate L & R sweeps and then one combined LR sweep. Use timing reference. There is no timing information in MMM measurements but give us a LR combined MMM. This is what a quick look in REW looks like with your recent MMM L&R but better information can be had if you provide what we ask for.

View attachment 538636

Seeing the combined response, I’d be less worried about the 150 Hz right-channel dip. It looks mostly filled in when both speakers play, so you could experiment with the right speaker position, but I personally would not obsess over it. The 250–350 Hz dip is still there, so you could play with speaker height/tilt and desk treatment.

 

[thcdru2k]

Hey, I actually edited my previous post about having placed a thick pillow on top of the right speaker and the 150hz issue seemed to change about +8db. I added the .mdat there. It includes double checks to make sure it wasn't a coincidence. Do you think the measurement with the pillow looks better?

Yes the pillow measurement looks better