How do I "compromise" speaker placement


When a flat response is not possible, how does one place speakers?
I would imagine that every time a freq gets flattened another peak/dip gets introduced in another freq?
So how does one "compromise" with the placements?
Do I simply use the flattest response?
Or perhaps, certain frequencies are more significant than others? Meaning, even if one placement introduces less dips (or peaks) across all frequencies, there might be more dips introduced in the bass area. Or the human voice range?

I have a very small room: 18' x 11'2" x 8'. I am having a professional take readings at 8 possible placements. I'm not sure what to do once I get the response graphs.
captaincapitalism
Actually your room is not quite that small. A bit narrow but otherwise I'd consider this a smaller mid-size room.
I only have 8' usable width. I have fire place on 1 side and bookshelves on the other. I also have cabinets in the fron and back of the room. So even though the room might no be so small, usable space is.
Finally use a PEQ or TACT or PARC to fix the worst of the room modes
I decided on the Audyssey EQ after auditioning for some time. I auditioned 2 other EQ's.

Rule of thumb - speakers placed well out into the room will excite less room modes. Keep away from side walls (at least three feet).
Not possible. 3 feet off the wall in a 11’2” width will have the mains less than 4 feet apart. The speakers are about a foot in width.
Most people think that you can measure your way to the best placement, but it doesn't work. See the following thread for something that actually works:

http://forum.audiogon.com/cgi-bin/fr.pl?rspkr&1197744079&openmine&zzDcstep&4&5#Dcstep
Not possible. 3 feet off the wall in a 11’2” width will have the mains less than 4 feet apart. The speakers are about a foot in width.

Well just do the best you can in terms of distance from side walls and also toe them inwards to reduce the magnitude of side wall reflections (effective once you get past about 40 degrees).
Hell, Hi Dcstep, you said, "Most people think that you can measure your way to the best placement, but it doesn't work. See the following thread for something that actually works:" Which has the following quote:

"Anyway, Rod accepted the challenge and “The Ballad of the Runaway Horse” soon filled the room. I could attempt to reveal the whole process, but that would be futile. I witnessed the whole thing, start to finish, and actually helped a time or two and I couldn’t replicate it in a new speaker positions. I will say that he started with the bass, beginning with the speakers just a couple of inches from the rear wall and then moving them out an back while listening to the nodes. He worked the left speaker first, and then the right, focusing on getting the bass balanced between the two speakers, but with no nasty nodes in the midrange. Toe-in and rake angle were critical to the midrange and balance, but I think that getting the bass right and avoiding the midrange nodes was the single most important factors. Still, it’s a comprehensive system that you’re not going to be able to read about and then do. Listening and hearing the nodes takes some training. I was amazed to hear MAJOR changes with just ¼” of movement."

What do you think is wrong with measuring an in room response of your system? If moving the speakers as a Master Sumiko dealer does to find the best freq response in a room, why not do it by a REW measurement? Do you believe he would be doing something different? I mean, that, if we both agree that moving speakers changes the inroom response by going in and out of nulls and peaks and yes, some have a percieved better or worse sound position by just this fact, why not just measure it and use your ears for the position you prefer. But at least this way you know where the nulls and peaks are through an analysis of your room. Plus it is free and repeatable.

Then, Dave, all you have to do is go for 83db playback level! You just knew I had to throw that in! But please ignore that issue to not confuse everyone.

Bob