I normally play the Sicilian, but recently wanted to add the Caro-Kann for variety & educational purposes.
I first started reading Schandorff's book which is fine (I like the fact that he chooses ...Bf5 vs. the Classical and vs. the Advance - I think this is the first thing you need to decide before buying anything: 4...Nbd7 or 4...Bf5 vs. Classical and less important: 3...Bf5 or 3...c5 vs. the Advance). But as you go on you will realise that he chooses a rather sharp repertoire for C-K and this requires remembering a few lines. His aim is to treat C-K in the modern way, where Black is playing for a win (choosing opposite-sides castling or sharp lines vs. sidelines), rather than simple solidity.
I then got my hands on
Bologan's DVD (Fritztrainer) in which he also chooses ...Bf5 vs. the Classical & the Advance (so this was also fine with me), but presents instead a very solid repertoire. For me it is much easier to follow it and the lines/plans stick to my head!
Since I play the Sicilian I have decided to go for that repertoire instead and play for a win with the Sicilian and for solidity with the C-K.
Since you are a new player to the C-K, I think Bologan's DVD will be easier to digest and get the key ideas.
Two caveats:
1. I don't like the choices of either the above vs. the Panov Attack. Have decided to go for Karpov's Variation (same one used in the Nimzo-Indian). You mostly play vs. hanging pawns which is fine with me.
2. Schandorff gives some fully annotated games which is very useful imo in order to understand the typical middlegame plans. So I am willing to go back there and study those. Bologan mainly focuses on the opening/early middlegame (and I haven't seen any endgames yet). So while playing lots of blitz games over the past 2-3 months, nowadays I manage to avoid opening catastrophies with the C-K, but I still have no good feeling of the resulting middlegame positions (of course blitz is not enough for that anyway).
I could give you some more details if interested.
I have also seen Houska's Table of Contents and had a cursory glance of the introductory pages in Amazon. It seems like a good book as well, but again presents a rather sharp repertoire for what I want and I also don't like the choice vs. the Panov.
My 2c