Embarking on REW/Roon DSP exercise

[Rednaxela]

Cool.

Did you know you can save the prediction from the EQ module back to your main window as a measurement? Super useful for analyzing how well your system responds to EQ.

 

[Rednaxela]

 

[thewas]

Cool.

Did you know you can save the prediction from the EQ module back to your main window as a measurement? Super useful for analyzing how well your system responds to EQ.

By the way this can be also done from the "Overlays => Predicted SPL or Phase" tab, actually much longer before this functionality was added to the EQ window.

 

[ernestcarl]

MMM curves in the upper range can be more variable esp. if the measurement is done in the nearfield and quite close to the speakers as you really cannot replicate the same spread of movements exactly at every measurement try. So just be cautious at over-interpreting it as something that necessarily needs to be “fixed” yet again. I’d rather just lazily create a copy of the same curve with the filters applied in an overlay view. I would not bother re-measuring.


Influence of room acoustics and positioning explain a huge chunk of what one measures despite having otherwise good speakers. In the studio realm I’ve heard people say that one should allocate half of the budget to room treatment… most people spend money on gear, though.

I’ve been watching some of Matthew Poes’ youtube videos (Poes Acoustics) documenting the effort he is expending on the room itself recently. And he really is spending a LOT of effort there despite using already great speakers.

EQ is going to help, for sure — sometimes quite significantly it would seem — but it’s not going to fix the bulk of the non-linearities that’s mostly occurring (and often hidden/spread throughout the time domain besides the basic FR plots) due to room acoustics.

 

[ernestcarl]

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

Looking at your old mdat file data posted, I would probably simply EQ only below 200Hz and leave the stuff above "as is". It may be better to correct each channel individually, but to keep it simple I did it as one.

The resulting correction filter file is going to be very similar to the one you created... One could certainly apply more detailed equalization, but with everything else that's going on in this room the benefit may be negligible, IMO.