One can't objectively compare online blitz or correspondence ratings with otb standard time control chess ones. These are different things.
Chess.com and FIDE ratings

The math used is very similar. The reason you can't compare the ratings too accurately though is because the pools of players are so different. For example if you took 100 beginners who only played each other and 100 masters who only played each other an average player from one pool (~1500) would not be comparable in strength to an average player from the second pool even if they're both rated 1500.
There are also other minor elements such as a player who plays both FIDE and chess.com may use chess.com to practice openings or play when they're tired after work and when they go to a FIDE event they play their best.
Others may be distracted by noisy kids of a phone call. Still others may play with the same level of seriousness... but there's no way to tell.
From my own ratings (chess.com vs FIDE) chess.com has been higher in every category by at least 100 points. Bullet chess seems especially inflated. When you're talking turn-based it's even harder because some players play a handful of games very seriously (I've seen B class players at 2100-2200) while others play hundreds of games and spend little time on each... there was that one polish kid who was a master (or near it) with a 1900 rating. Oh yeah, here he is: http://www.chess.com/members/view/Kacparov

Thanks for this. I'm only looking to find out if they're 'roughly' similar. I know the FIDE rating is calculated in a very different way to how the rating is calculated on chess.com. I can get an equivalent FIDE rating for my ECF rating (English Chess Federation), so there should be an equivalent way of comparing chess.com's rating system with FIDE's. Loosely at least.

It's always a tricky thing. Even USCF players from different states may find players in one state/area are a bit tougher or weaker than another.
ECF to FIDE to USCF is do-able because they are similar time controls, "random" pairings (which is important) and you can be sure players are taking the games seriously.
So to give you FIDE to chess.com would always have to have an asterisk saying something similar to "give or take 200 points!" And when you get into the upper or lower levels of the ratings maybe even more.
But I'll give you what you want anyway, based on what I've seen. Just remember this is GIVE OR TAKE A ZILLION BILLION POINTS lol :)
For live -bullet chess subtract 400-500 points
For live -blitz chess subtract about 100 points
For live -standard chess subtract 100-200 points
For turn-based subtract 200-400 points
Some players though I've seen should have points added instead, but this is more or less what I've run into as most common.
The best answer is :
One can't objectively compare online blitz or correspondence ratings with otb standard time control chess ones. These are different things.

@evan:
Personally, the only that comes close is the Live Blitz rating and subtract 100.
Bullet and live standard are useless due to the limited pool of players. Turn based can be all over the map. There are players with higher USCF/FIDE ratings than CHess.com as they don't take it too seriously. On the other hand, I know of a player with a 1089 USCF rating (established) that is roughly 2050 here. For a second, I thought cheat but looking up his USCF correspondence rating showed him at 1880 which then makes sense again.
Is there any similarity between the Standard Chess.com rating system and the official FIDE rating system. They seem similar. Does anyone know?