SVS17 measurements

[levimax]

Maybe make some measurements at others listening positions and see if some of the bumps are consistent at certain frequency.

Other wise, just let Audyssey MultEQ XT32 of the Marantz do his job ?

I find the "automatic solutions" tend to "over correct".... they want to make sure if someone runs a REW sweep that the graph looks as flat and smooth as possible which may or more likely may not sound the best. I find a lighter touch on the room correction to be preferable which is why I like ERB or Psychoacoustic smoothing before correcting.

 

[j_j]

Seeing my rew result on the low frequencies, would you recommend changing anything using the subs eq or do nothing at all?

Are you satisfied? If you are, you're done. Seriously, eventually it comes down to taste.

 

[janbth]

To some extent you're talking about two different things, I suspect. Are these peaks (not at under 300Hz or so) due to loudspeaker? Fixing a loudspeaker that way isn't necessarily a bad thing, but fixing a room that way, because room reflections do that, is maybe a very bad idea, because you're going to shrink the listening area in the room to a space smaller than your head.
Some clarification may be in order.

I’m talking about a peak coming from a room mode, and not a cabinet resonance. Let’s say it’s one of the lowest lying room modes, say at 45 Hz. Then let’s say I choose - for one reason or another - to handle this peak by applying a singe PEQ filter. I measure the frequency response with high resolution. Then the question is: Do I use this high resolution frequency response to compute the parameters, in particular Q, or do I apply some smoothing first? I would think not.

 

[j_j]

I’m talking about a peak coming from a room mode, and not a cabinet resonance. Let’s say it’s one of the lowest lying room modes, say at 45 Hz. Then let’s say I choose - for one reason or another - to handle this peak by applying a singe PEQ filter. I measure the frequency response with high resolution. Then the question is: Do I use this high resolution frequency response to compute the parameters, in particular Q, or do I apply some smoothing first? I would think not.

You simply eq that out. Yes. In that frequency range, the ERB width is under 20Hz, and 1/3 ERB is not very wide at all.