Stupid question - number of RAW & card capacity don't add up

Maths is not my strongest suit, I'll admit... 

I can't work out why I can only get 514 files (according to the top panel of the camera) of 12 bit lossless compressed RAW from a 32GB card. Surely if the file is 33MB each (verified on computer), I should be able to store 969 (or thereabout)?? What am I missing?

BTW, I'm referring to a D810, and yes the card has been formatted (in-camera) to get the 514 reading.

Thanks for your wisdom!

It's a long standing 'feature' of Nikon DSLR's to show the space available to be about half of what can actually be written to a card in the camera. The explanation for this discrepancy is that Nikon wants us to sure you don't run out of space due to actual size of compressed files because we're not smart enough to figure otherwise. It's silly but Nikon refuses to have the computer (which we call a camera) calculate this number which it could do with much more accuracy than the current method.

First I have to ask why you are shooting in 12 bit rather than 14 bit with a camera of that caliber?

Now to your question. The NEF format uses a lossless compression algorithm. In order to determine how many shots and never underestimate, it has to assume that the file is not compressible. In other words every pixel has to have a different 14 bit value than all it's adjacent neighbors. In reality, that is never the case. If a photo has a lot of blue sky in it for example, that file is going to be highly compressible without losing any information. If on the other hand you are taking something very colorful with tons of texture (a tie-died piece of linen for example) it is not going to be very compressible at all. In that situation, the estimated number of pictures remaining will be much closer to reality than in a picture of a bird that takes up a quarter of the frame area and has a soft OOF background throughout the rest of the photo.

E.J. Peiker wrote:First I have to ask why you are shooting in 12 bit rather than 14 bit with a camera of that caliber?

Now to your question.  The NEF format uses a lossless compression algorithm.  In order to determine how many shots and never underestimate, it has to assume that the file is not compressible.  In other words every pixel has to have a different 14 bit value than all it's adjacent neighbors.  In reality, that is never the case.  If a photo has a lot of blue sky in it for example, that file is going to be highly compressible without losing any information.  If on the other hand you are taking something very colorful with tons of texture (a tie-died piece of linen for example) it is not going to be very compressible at all.  In that situation, the estimated number of pictures remaining will be much closer to reality than in a picture of a bird that takes up a quarter of the frame area and has a soft OOF background throughout the rest of the photo.

Thanks for the explanation, EJ. I think I understand what you're saying. With regards 12 bit vs.14, I've personally never really found them to be that different. I appreciate that others may be more critical with their image.

You are giving up some dynamic range by shooting in 12 bit instead of 14 bit.

Appreciate that, E.J. It's a trade off, but I prefer to have the buffer over the extra bit of DR. I almost never do landscape.