January 2017

M T W T F S S
      1
2345678
9101112131415
16 171819202122
23242526272829
3031     

Style Credit

Expand Cut Tags

No cut tags
Friday, August 21st, 2009 04:28 pm
If anyone out there has suggestions on how to choose decent paint colors for our living room and dining room, I'd really, really love to hear them. I've been staring at paint website color-picker tools for a while now, and not only are my eyes beginning to bug out but I keep feeling irritated when the user interfaces don't work quite the way I want them to. (Of course, what I really ultimately want from them would probably require black magic.)

The thing I really lack is a good way to make sure that I pick colors that look good together (including our existing dark blue couch). Kim and I would rather not default to generic neutral beige walls: we actually want some sort of color if we can manage it, though nothing too bright and bold.

Ideally, what I want is a tool that will let me build a palette one color at a time and gradually restrict my choices as I add additional colors to the list. I'd start with the couch color, which by itself probably wouldn't narrow the possibilities much, and then once I picked a second reasonable color (the main wall color, say) I'd be left with a much shorter list of possibilities for an additional accent color or two. Once I had a complete list, changing any one color would simultaneously update the others if necessary to keep the set looking good (unless a color was "locked", like the couch). In a perfect world, I could watch the effects of these changes in real time on a sample picture of a room similar to ours.

As noted, a tool like that probably doesn't exist, so at this point I'd be content with just general ideas on how to avoid major blunders while choosing colors by eye. I don't really trust my sense of style enough to risk a major painting excursion on it without some sort of outside support. (Kim has a better sense of these things than I do, but I very much want to be able to contribute intelligently to the conversation.)
Saturday, August 22nd, 2009 12:36 am (UTC)
Aha! So room design is a non-deterministic steuard-hard problem. :) You can, in a reasonable amount of time, recognize whether a room is well designed, but you have no known algorithm to generate such a room in a reasonable amount of time.

We deal with similar problems in other domains either through heuristics and (argh - wasted money and time) guess and check, in a practical sense. We also appeal to oracles to solve the problems when we want to just move on and talk about something else. Do you have any oracles near you?
Saturday, August 22nd, 2009 01:26 am (UTC)
This is an absolutely glorious description of the problem.

I think I've had an irrational degree of fear about the potential costs of "guess and check". Kim's first sentence replying to you captures some essential points as well. I've considered consulting an oracle, but I don't know anyone handy who does such things for fun and I more or less decided this wasn't important or earth-shaking enough to pay big bucks for. (Some might argue it must then not be important enough to stress over, either. Some might be right.)