Compression

On the D2x you can compress the raw files, from about 20MB to about 10MB, but after a lot of samples I see no difference what so ever so why not have just a 10MB files and not bother with the 20MB? after all space is now becoming an issue.

The compression you mention is lossless (reduced file size without ditching any information), which is why you can't see any difference. The only potential issue with this is that the extra step of compressing the files takes a bit more time than simply writing it as it is. If you've never been limited by this while shooting in burst mode, than compressed files are fine...

I guess that there are also people who will always want the best theoretical quality, even if it actually isn't any better in the real world.

I think that the difference in image quality between the D2X and the D200 is a similar example.

Alex

Is this the same in the D200 ? - I see that there is a 'compressed' RAW mode on the D200 but the manual as I recall says that there is loss involved.

For that reason I have currently stayed away from the Compressed RAW.

Anyhow Alex I thaught you were away taking pictures ?

Paul C

Hi Paul, all Nikon dSLRs have the same compressed NEF format, supposidly with (according to Nikon) no or minimal decrease in quality. I don't remember reading anywhere in the D2x manual that there is loss involved. If they wrote that on the D200 manual, thats news to me.

Actually, there is a lot of debate about this issue on the net. Here is what Thom Hogan has to say about it:

"Leaving off Uncompressed NEF is potentially significant--we've been limited in our ability to post process highlight detail, since some of it is destroyed in compression."

And here is Ken Rockwell:

"The compression is lossless and gives you exactly the same data in as little as half the space."

Now, between Thom Hogan and Ken Rockwell, I chose to believe in Thom and use only uncompressed RAW. I never ran out of space with my 4gb cards.

Alex is in some remote place with an internet connection, he can't live without us!

And no girls there to entertain him so he has us..... hahaha

Pretty sure that the D200 has a compressed and an uncompressed raw setting with the compressed involving some loss.

Have not got the manual here but will take a look again tonight.

I have not run out on a CF card but a 4GB card is 'only' 240 shots. How some of you guys managed with 36 shots on film is way way beyond me.

Alex is in some remote place with an Internet connection, he can't live without us! smile.gif

And no girls there to entertain him so he has us..... hahaha

Oh dear, this could be a bit of a letdown then, maybe he will 'discover' beer.....

Dug the D200 manual out last night and on page 31 it says re compressed NEF

"NEF images are compressed by about 40-50% with little drop in quality. Recording time is reduced"

Not lossless then ?

Paul C

The Nikon compressed NEF is generated using a compression algorithm. From Nikon "Files can be saved in camera as uncompressed or compressed at approximately 50-60% of the original file size using Nikon's advanced compression algorithm" In a few words, that means that the pixel values are stored as a mathematical formula and the individual pixel values calculted when the image is opened. With a good algorithm these pixel values will be close, but not nessesarily the same as, the uncompressed pixel values. Hence it is lossy compression.

The difference is best illustrated with JPEG compression and LZW compressed TIFFS. In JPEGs the pixel values are saved as a formula. If you save a JPEG with a high quality value the formula will do less averaging and your pixel will look close to the original. Its when you do low quality that you really see the algorithms at work - with each set of compressed pixels in the image showing up as obvious blockiness (checkerboard-like pattern). With TIFF LZW compression only pixels of exactly the same value occuring adjacent to each other in the same row are stored as a set. So when the image is uncompressed you get exactly the same pixel value as the original. This is not always very space effective though, when you have a very detailed image it means very few pixels can be "saved". By contrast a pure colour image will compress very well as there will only be one pixel value to store.

HTH

Leander

Wow!

Thanks Leander - that is the most concise and illuminating treatise on image compression techniques and outcomes that I have ever read.

Cheers