How much can you visualize blindfold

When reading a book for instance how far can you visualize blindfold and are there techniques for this.

Its easier to remember where things are if you can make sense of the moves by analysing the position. Practice by analysing diagrams.

Thank you for the advice.  I will try this.  Blindfold chess has always fascinated me.  I think it is alot more amazing than chess players give it credit for. Going through an entire game, even if lost, but completed, is quite an accomplishment. 

JoeTutor is right on target. Blindfold players don't have a photographic picture of the board in their head, contrary to popular belief: they apparently use some other representation that is largely based on remembering the relationships between units.

----------

(p. 47)

Visualizing positions

All this analysis is carried out in the player's head. How do players keep
track of the new positions generated in analysis? Keeping track involves
memory and some imagery and is susceptible to spatial errors. Occasionally
a blunder can be traced directly to the "imagining" operation. A player
will fail to "notice" that moving a piece to a certain square blocks another
piece from giving check at a later critical moment. (For an amusing
account of just such an [sic] hallucination see Chess Treasury of the Air [94,
p. 273].)
   The role of visual imagery in chess has long been of interest to psy-
chologists. It was initially investigated to explain the ability of chess
masters to play "blindfold" chess. In this situation the master plays a
game entirely in his head--he hears the moves of his opponent (in chess
(p. 48, 1977 ed.)
notation) and then replies with his own moves without ever seeing the
board.
   At first it was thought that the blindfold player had an image of the
board very much like that of a photograph or picture--namely that the shape of
the pieces, the color of the squares, were all present in vivid detail. Binet
in 1893 refuted this view after interviewing many masters. He concluded
that the representation used by blindfold players was quite abstract. For
instance, a player would know that there was a knight in a certain relation
to other pieces in a particular position, but he would not have an image
of a particular carved knight. He might know the knight was white, but
he wouldn't see a certain shade of white.
   The reader can easily convince himself that one does not form a detailed
image of an entire position. Imagine an empty chessboard. "Place" a
bishop on QR3. Name the farthest square the bishop can reach on the long
diagonal. Now put a real bishop on a real board and do the same. The
time taken to answer KB8 should be markedly different in the two cases.
(Parenthetically, how many of you chose a white bishop? Such generative memory is quite
abstract.) Similar conclusions about the nature of the representation of
the board were reached by de Groot, and by Reuben Fine in his article on
blindfold chess [43].
   Apparently the ability to imagine pieces on squares in a chess position
is also correlated with chess skill. I conducted an experiment [20] where
players were given chess positions verbally, piece by piece, via chess nota-
tion. After hearing a position dictated in this fashion at a fairly brisk pace,
the players attempted to reproduce the position. Two groups of players
were tested: Class A (mean rating = 1870) and Class C (mean rating =
1458). The order in which pieces were read made a large difference in
recall, but beyond that, the A players recalled 14 percent more pieces per
position, on average. That all players fared poorly when the pieces were
"scanned" randomly indicates that they did not perform the task by
"imagining" pieces on squares. Rather, they attempted to remember the
pieces by remembering the relations which bound them together into
chunks. The most favourable scan was one which consecutively named
pieces with many relational links.

Frey, Peter W, ed. 1977. Chess Skill in Man and Machine. New York: Springer-Verlag.