Embarking on REW/Roon DSP exercise

[ahofer]

Occasionally, REW will tell me the levels are off as well while it not being so... make sure the microphone cable and connections are good -- unplug and replug, or try another cable... -- as well as try restarting the program.

Just looking at the per channel measurements the treble is louder now. Can you try measuring each tweeter in the nearfield at half a meter or less? Tweeter axis is fine... Since both look similarly problematic in the high frequencies either left or right channel is fine -- both speakers if you can.

It looks a bit like combing filter, but one should not see this much in the nearfield esp. when only measuring a single channel exclusively.

So in that last group of measurements, there is one curve each that is exclusively left or right channel. Not near field, from the LP. The third one is both channels. There aren’t huge differences.

As I sit here, I wonder if I set the levels based on a subwoofer setting.

 

[ernestcarl]

So in that last group of measurements, there is one curve each that is exclusively left or right channel. Not near field, from the LP. The third one is both channels. There aren’t huge differences.

As I sit here, I wonder if I set the levels based on a subwoofer setting.

If you are using a UMIK-1, there should be no need to microphone calibrate SPL... just use the default calibration file.

Use the REW's generator button to calibrate with a -20dBFS band-limited pink noise signal:

Click on the SPL meter tab button to see the reading and adjust your amplifier volume until its about 75dB or 80dB.

 

[Deleted member 17820]

OK, not that I know exactly what I've got, but I have measurement files based on:

-8 sweeps center LP + average
-4 sweeps pointed right from LP+ average
-4 sweeps pointed left from LP + average
-average all of the above (picture below) View attachment 88381

You should be able to download all the mdat files here --> https://drive.google.com/drive/folders/1p1NmYm9HIGOwNcZxupG7UcHWeTFd1gd9?usp=sharing

I'm still a little confused about the volume levels I should be targeting with the sweep, but you can see the room modes I noted in my earlier post.

Higher up around 16khz it goes pretty wonky, but I can't hear up there. I presume the sharp downticks starting around 240 are comb filtering?

High, for some reason I have that when using REW, during the measuring if you run multiple sweeps at that time it gives me a totally wrong outcome with a huge heigh frequency roll of like what you posted.
There are also 2 display settings of FR in REW,
top right corner of REW you can toggle between log and linear frequency axis as its called.

 

[ernestcarl]

New measurements from the living room, all from LP. Along with suggested overdone EQ.

https://drive.google.com/drive/folders/1p1NmYm9HIGOwNcZxupG7UcHWeTFd1gd9?usp=sharing

View attachment 88812View attachment 88813

Update: I programmed all these in for fun. You wouldn’t believe what crap it sounds like. Or maybe you would.

BTW, I think your left and right channels are sufficiently different enough now that one should really EQ each channel and not in sum. Preferably with MMM or spatial averaging.

 

[ernestcarl]

High, for some reason I have that when using REW, during the measuring if you run multiple sweeps at that time it gives me a totally wrong outcome with a huge heigh frequency roll of like what you posted.

This can be avoided even when measuring multiple channels when using MMM RTA instead of the regular sine sweeps.

 

[ernestcarl]

So in that last group of measurements, there is one curve each that is exclusively left or right channel.

The HF in the spectograms (almost) look normal or typical for untreated rooms. But the high frequency gaps are still there.

right channel alone

*Using auto-generated PEQs above 1kHz is a big no-no esp with those big gaps still visible in your measurements. Between 300Hz or 500Hz to 1kHz, very iffy -- it needs to be very carefully done, and you need more data for this (before and after) than a mere few single point measurements.

 

[Deleted member 17820]

New measurements from the living room, all from LP. Along with suggested overdone EQ.

https://drive.google.com/drive/folders/1p1NmYm9HIGOwNcZxupG7UcHWeTFd1gd9?usp=sharing

View attachment 88812View attachment 88813

Thank goodness for this thread!
MMM RTA makes so much more sense for me!
I am going to do it now.

 

[ernestcarl]

MMM RTA makes so much more sense for me!

MMM using RTA measurements (while not enough by itself) are far easier to obtain than multiple single-point sweeps for spatial averaging... also easier to read and more reliable for PEQ correction work.

There are already several threads that cover it... but here are two videos:

 

[Deleted member 17820]

MMM using RTA measurements (while not enough by itself) are far easier to obtain than multiple single-point sweeps for spatial averaging... also easier to read and more reliable for PEQ correction work.

There are already several threads that cover it... but here are two videos:

That's the video I found and it was really good.
I tried to use EQ APO as a DSP but it keeps crashing/stops working.

I now need to decide is I should buy a miniDSP HD or go another route.

 

[ernestcarl]

That's the video I found and it was really good.
I tried to use EQ APO as a DSP but it keeps crashing/stops working.

I now need to decide is I should buy a miniDSP HD or go another route.

I think the miniDSP route is easier than working with software like Equalizer APO... I'd go get it esp. if I have more than one stereo system simply to experiment and learn on with REW or whatnot -- lots of videos and help tutorials out there. One could always upgrade to a more expensive or complicated option later on if desired.

 

[Deleted member 17820]

I think the miniDSP route is easier than working with software like Equalizer APO... I'd go get it esp. if I have more than one stereo system simply to experiment and learn on with REW or whatnot -- lots of videos and help tutorials out there. One could always upgrade to a more expensive or complicated option later on if desired.

I just bought the ddr-24 miniDSP, it should be here soon.
I am kind of worried since the dac got a bad score here but there isn't much choice for small form factor under $500
I am not sure if DIRAC will work better than inputing my own filters but atleast I will be able to test the difference myself.

 

[ahofer]

Just went through the whole exercise again (for my Harbeths. The Revels in the other house will have to wait until next weekend), but with input gleaned from videos and conversation here:
1)ran sweeps from positions around the listening position
2)did MMM technique for L& R speaker
3) averaged them all together
4)created filters and exported to WAV
5) imported as convolution filter

1 & 2

2)

I think the primary thing I'm hearing is "more air", probably due to the 5 & 7k boosts, and the removal of the LF peaks.

The filters around 100-140 are a bit of a puzzle to me. I think that must be dealing with the cabinet resonance.

Hear is a more readable set of original measurements - average of sweeps and average RTA/MMM. I was pretty happy with the consistency:

The target curve may not slope down enough?

 

[ahofer]

Here's a link to the measurements and convolution file

take the "NY Harbeth" results (I'll be adding my CT-Revel results at some point)

REW measurements - Google Drive

drive.google.com

 

[Rednaxela]

Have you tried to apply VAR smoothing to curve #8 before entering the EQ screen?

 

[Rednaxela]

Your target between 20 and 60 Hz looks a bit undecided on what is the peak and what is the dip.

I believe I’d increase the LF Cutoff frequency, so that the target follows the dip a bit closer.

 

[ahofer]

Have you tried to apply VAR smoothing to curve #8 before entering the EQ screen?

OK Done:

I also put a little more slope into the target, both rise and fall.

Not sure what LF cutoff should be. Port is tuned at 35 or 36. A little low by some's standards.

Still produces those dramatic, narrow filters in the 100-200 range. but this time it bumped the 3k "dip" I had in the unaltered response (where the Harbeth are more directional)

 

[Rednaxela]

Still produces those dramatic, narrow filters in the 100-200 range.

You mean even though you unchecked the box?

Not sure what LF cutoff should be.

Eventually it’s what you want it to be. However if you leave in the dip but insist on extension, the peak around 30Hz may stay too loud and standing out too much.

Out of interest are you able to remeasure with the EQ on?

 

[ahofer]

This set sounds better to me. Mostly removes some mud. I added a separate EQ bump at 3Khz of around 1.5db in Roon that I can add or subtract as needed to change the response dip more subtly.

 

[ahofer]

You mean even though you unchecked the box?


Eventually it’s what you want it to be. However if you leave in the dip but insist on extension, the peak around 30Hz may stay too loud and standing out too much.

Out of interest are you able to remeasure with the EQ on?

Not sure how to remeasure with the EQ built into Roon. I could try coding 5 of these filters into my RME. I suppose if I can put some pink noise in my library and play it through ROON while doing RTA I might get there.

It’s surprising how different the filters come out with different settings. For instance, I don’t know why only one set of filters (I did 6) addressed the 3k dip. I can’t identify something in the settings that drove that.

The last set I checked the “allow narrow filters” and “vary Q” boxes.

 

[Rednaxela]

Unfortunately I don’t know Roon so can’t help you with that. Hopefully others can chime in here.

I must say though that it’s very much worth pursuing. Closing the gap between prediction and reality gives you an awful lot more grip on the process IME.

Ideally you’d like to do this at least once, where you have your measurement, your target, your prediction and your post-EQ measurement in one plot, to see how well your system responds to your intentions.