I have a few questions about room correction. First, I’d like to summarize how I understand things and ask a few questions that I’m hoping Dr. Toole can comment on. Second, I’d like to share my experiences with using room correction.
I had always thought that room correction offered benefits over the entire audible range. However, based on the insight shared here by Dr. Toole, it seems as though the science indicates that EQ can often help below the Schroder Frequency (~500Hz), but not so much above that.
If the loudspeaker being used already has excellent on and off axis frequency response, then any EQ above the Schroeder Frequency is likely to be detrimental. However, if the loudspeaker has some flaws in it’s frequency response, then EQ could fix it, however, it’s extremely difficult to accomplish this by using in-room measurements. This is because in-room measurements will show not only the direct sound, but it will also contain information on the room reflections. Binaural human hearing is very good at separating the direct sound from the reflected sound above the Schroeder frequency. So, if the speaker has a good frequency response, it may not necessarily show up that way in a room measurement (due to reflections), and room EQ software may EQ that area thinking there is a problem, when there is not. However, it didn’t need to be EQ’d because humans will be able to hear the direct sound from the speaker, separating it out from the reflection.
Therefore, the only way to reliably EQ a loudspeaker that has frequency response problems above the Schroeder frequency would be to have good anechoic measurement data on that speaker. For example, if a speaker has an uneven response around 2kHz (say it has a 5db dip in this area relative to the rest of the range, then it could work to EQ the speaker up in that area (with the appropriate Q). However, it would still be hard to do this in-room. You’d need to EQ it in an anechoic chamber to get the proper EQ adjustment.
Do I have the basics correct here?
Next, I’d like to share my experience with using room EQ.
I use Audyssey MultEQ and I always thought that it improved the sound. Subjectively, to my ears, I always thought that the most noticeable improvement was in the bass (by far). However, when I look at before and after measurements (using REW) to me the frequency response looks better across the entire audible range. I’ve posted some REW graphs to demonstrate (all graphs are smoothed to 1/12th octave).
If you look at the before and after for the left speaker and sub, there are some areas that look better with Audyssey, and others that look worse. However, overall I think it looks a little better. Certainly the left and right speaker average across three listening positions looks smoother.
However, if my understanding of the above is correct, then even if my REW graphs look smoother above the Schroeder frequency, they may not actually be giving me better sound above that frequency.
Now, subjectively, when I listen to music and switch back and forth between no Audyssey and Audyssey, it sounds better to my ears with Audyssey. However, this is with the subs. And as I said, the most noticeable improvement is in the bass. So, if I switch back and forth between full Audyssey and the “Bypass L+R” feature (this means Audyssey is still on for the subs, but removed for the left and right speakers, then the sound difference between the two is different. The way I would describe the sound difference between these two modes is that in the “Bypass L+R”, it sounds like there is more midbass. Just going by the sound, I would guess there is more energy somewhere in the 100hz to 500hz area. To my ears, it still sounds better with Audyssey (vs. the "Bypass L+R"). So, is it possible that Audyssey is improving the sound enough below the Schroeder frequency, that it is negating any detriments that it has caused above this frequency?
I will note that if I turn the subs off, put the speakers to full range, and switch between no Audyssey and Audyssey, but this time without dynamic EQ, I think it actually sounded better with Audyssey turned off. It sounded fuller this way, and thinner and brighter with Audyssey turn on. So perhaps the vast majority of improvement that Audyssey is making is in the sub bass range.
FYI – my main speakers are GoldenEar Triton 2s and I have two PSA V1500 subs. I’ve attached a few pictures of my room as well.
I can't find any anechoic measurements of my speakers online (only quasi-anechoic), so it's hard to tell just how good these speakers are. Based on the quasi-anechoic, and comparing those to the quasi-anechoic measurements of the GoldenEar Triton 1s, and to the Triton 1s anechoic measurements on soundstage.com, it seems to me like the Triton 2s probably have a good average on and off-axis frequency response, but maybe not great? Hard to tell.
1) Left speaker + subs, main listening position, before Audyssey.
2) Left speaker + subs, main listening position, after Audyssey (dynamic EQ is also on).
3) Average of left and right speakers + subs, averaged over three listening positions, before Audyssey.
4) Average of left and right speakers, +subs averaged over three listening positions, after Audyssey (dynamic EQ is on).