First, the type of room treatment depends on the speaker type. If you choose an omni-type speaker then first reflections are part of the speaker's design philosophy. So yes, the room is important but I would view it as a "room-speaker" package in terms of room treatments.
Second, positioning. Most speaker types - closed-box, dipole, omni etc - do better in free space. There are psychoacoustic reasons for this. In short, the brain finds it easier to process reflections as reflections. If the speaker is very near a wall the brain perceives the reflections as a smear of the primary sound. Digital room correction can't fix this perfectly.
Third, buy an SPL meter. The Radio Shack SPL meter works and there are calibration files available (PM me) which are absolutely essential to use it properly. A flat frequency response is the first line of attack in achieving a good sound. No it's not the only thing, but you cannot do without it. So measure it. (Interesting comment here:
http://www.audiocircle.com/index.php?topic=46668.msg418653#msg418653.) To fix it do everything possible in the "natural" way with room treatments or by choosing a different speaker.
Fourth, and here is the really controversial part, get some digital EQ. For many many years, as a card-carrying audiophile, this was _total_anathema_ to me but I had an epiphany (see point five). It is VITAL to make best efforts without digital EQ. After you've done this it's likely you'll still need digital EQ in-room. To not use it is to cling to a prejudice (that EQ is cheating) and to deny the truth (that a flat frequency response is the most basic aspect to get right in terms of "high fidelity"). Of course, there are objections to EQ like the addition of distortion. Using digital EQ (like the Inguz plugin for the Squeezebox which is free BTW) does solve a lot of these problems. Done wrongly it sounds horrible...could be instantly horrible - or the kind of horrible that creeps up over time. Nasty. So you need to measure properly i.e. from the listening position, one speaker at a time and point the SPL meter directly to the speaker (since treble is directional). Play your test tones, apply the calibration adjustments and put the corrections into the EQ system. You need an amp and speakers with power headroom. When you do this the results can be nothing less than spectacular. You asked for bang for buck...this is it. Brain over heart on this one! Be strong!
Fifth, by all means choose sources and amps that you like but realise that the differences are small compared to the differences in speakers and rooms. Yes I vote speaker first. I don't believe all sources sound the same. But a relatively cheap source like the Squeezebox sounds great. Some here would have you believe that this or that CD player will "blow away" the SB3 etc. Actually I believed that my CD player "blew away" the SB3 when I first tried it. Then I did a blind test and I was shocked the SB3 sounded so close to the CD player. It didn't sound that way sighted. Also, as a transport, blind I preferred the SB3 to my Monarchy CD transport, even though when sighted I thought the Monarchy was better! The point being, at the level of sources (and even amps) don't underestimate the placebo effect and don't ever compare anything unless it is *strictly level-matched by SPL meter*. Give it a try...I was shocked. You could say it changed my hifi world
Sixth, simplify. If you can do without a preamp for example then great. I have recently started using my Krell in pre bypass mode and using the SB3's digital volume control. Sounds great and seems certainly no worse. Since you are buying from scratch you can buy a better power amp instead of an integrated...bang for buck again.
Measurement, level-matching, blind tests. They sound so cold and mechanical but don't forget: level-matched and blind you use YOUR ears 100% and you apply YOUR taste 100%. So it's not about taking the fun out of it, it's just about getting the best sound YOU are going to enjoy.
Above is part of my audio philosophy...I hope it is of value to you.
Darren
