Digital Room Correction vs Room Treatments

Here's the freq response curve with my listening position about 5ft/60" into the room, which is at the 1/3 point.  Speaker baffles are at 37" from front wall (1/5) and 28" from side wall (1/5).

No toe-in currently.

1/3 smoothing, normalized to 75dB at 1kHz.

If I use +/-3dB as a criterial, there are those two peaks at 35Hz and 95Hz, big dip at 58Hz.  Not sure how badly these big peaks are potentially degrading the sound.  Besides that, a 4dB peak around 650-900Hz, but the rest of the FR curve is within +/-3dB.

This is sitting further toward the back wall at 51" instead of 60".

The 55-60Hz dip is reduced some, but overall it seems a little less smooth than the 60" FR esp from 200-600Hz.  Subjectively, it seemed to me that 51" was a little clearer, but the vocals a little less warm.  Not sure if that's consistent with what you'd expect from the FR curves.

Here are the two superimposed on each other. . .

On another forum, some members thought it this was workable and I could just go with this based on where the peaks/dips were, with 60" being optimal.

Finally, here's even further back at 45" (compared against 60" as a benchmark):

Here, the 35Hz peak is increased, the 58Hz dip is pretty dramatically flattened out, but the tradeoff is a dip created at 180Hz.

 Not  bad.  I always apply 2) DSP to flatten things out further after 1) room correction. .  Then finally a third layer of DSP 3) to adjust to my personal preferences, and roll off any wasteful low end to help out the amp and steer clear of clipping, if needed. That pretty much covers everything.   Older ears may be well served by ramping up the high end a bit.  A hearing test can help determine. I use Roon for this and may end up with anywhere from 1 or 2 to as many as 7-8 distinct filters per Room, including a single convolution filter for the room correction part.  Works great. Sound is fine tuned in the end exactly to my personal liking which is what it’s all about. 

Always do what you can to set up the room well to start.  Then DSP away towards your own personal sonic bliss.  Make good buying decisions to set up and  integrate everything well up front then anything is possible with DSP from there. 

OP:

If I wanted to take time based plots, is that doable using reasonably priced tools, or is that a pro assessment job?

Room EQ Wizard (REW) has a variety of tools for that, and their forums are very helpful.   Here’s a starter page.

Let me give you a little more background.  For frequency response REW and similar tools (I use OmniMic) gate the signal above the bass.  That is, they stop listening a few milliseconds after the impulse start to arrive.  This deliberately excludes as much of the room as possible to get the response of the speaker.  The sound waves are still busy traveling back and forth around the room for a much longer time period than this.

Our ear/brain mechanism does not hear like a microphone.  We don’t stop listening, but integrate the experience of the direct sound and the reflections over time and this fully integrated perspective is what gives us an impression of the tonal balance.  Often audiophiles only think this is an imaging problem but there is a very significant effect on the perceived tonal balance.

If you ever hear Fritz speakers at a show, he travels with only a few absorber panels.  He knows what he's doing, and that his speakers will sound more full that way.  We should all take a lesson. 

In a very reflective room where the mid/treble bounces around for too long our hearing tends to exaggerate the mid and treble frequencies, so as a result will sound as if the speakers lack bass.   So, a tip I often give which people resist is that adding mid-treble absorbers and diffusors will improve the apparent bass, and none of this is captured by simple frequency response plots. 

The other thing these plots fail to capture is how having multiple direct reflections, and / or the absence of the right diffuse sound field ruins the illusion of an audio image. 

In the bass things are different.  The wavelengths are so long that it’s nearly impossible not to integrate the room, so as a result your frequency measurements are integrated.  Speaker builders always face a challenge with this and use a variety of techniques to try to get the real speaker measurements like putting the mic 1/4" in front of the drivers, measuring outside, etc.

While room correction software may do a lot for the bass and maybe even correct for perceived extra mid/treble they cannot stop reflections.  That’s why in my mind the two approaches are not equivalent but complementary.

OP:

It occurs to me you are using gating and smoothing.  Turn those off and re-examine your measurements.  I think you can re-use what you've already captured. You'll see a much nastier image of your results.

Also, when comparing speakers, it's very much worth measuring off-axis to see how good of a sweet-spot they would provide.  Speakers that measure similarly on-axis may be crap off. 

Guys, really? Sorry to tell you OP that graph posted has a terrible response, about 20dB difference between the major peak and null. The null at about 58Hz is 12dB below the average. That’s a lot of musical information being lost in the most important range. The huge peak at 35Hz taking much longer to decay than the rest is going to make the bass slow and boomy.

I don’t agree with DSP. How can it reduce the long decay times and it can not et rid of that null. No matter how much power you pump into it, it will just cancel with the same power.

You can push the furniture and speakers around all day and you will still have peaks and nulls only at different amounts and frequencies. All rooms. I say again, all rooms will have these issues.

To achieve a smooth response some absorption is needed to avoid long decay and address bass problems with at the very least a pair of subs. They do not need to be the same brand nor size but avoid ported subs and they must, must have variable phase otherwise you’ll be endlessly pushing them around the room. Any ports that resonate at only one chosen frequency complicate matters further. You are adding a frequency invariant bass source with no way of tuning it.

OP I commend you for having the ability to measure as it’s the only way to remove guesswork.