Sony DSC-F828 Tiff and Jpeg

[A Scottish soul]

I have not used the tiff format yet on the 828, so can't say. However I'll give it a go and see what happens. Get back to you tomorrow.

John

[Lin Evans]

Hi Robert,

Try shooting in RAW and compare to TIFF and JPG. I saw an interesting comparison the other day which was quite reveaing. The pixel (clump) of 4 was not there at all from the RAW conversion, leading me to wonder if something in the jpg process was responsible for some unnecessary artifacting.

Lin

[Robert]

Thanks for your replies,

I'd like to compare with RAW, but I'm a Mac user, so I guess I'll have to wait until Sony brings out the software...

The Jpegs and Tiffs are, on closer scrutiny, not exactly identical. In deep schadows Jpeg seems to throw in a few more lighter pixels than Tiff does. Apart from this, however, differences are very close to zero. At sharp edges, where you would expect Jpeg artefacts to show up, I have yet to find a single pixel to be 'out of line'.

[Robert]

When shooting in Tiff format, the F828 saves the picture both as Tiff and Jpeg files. How come these pictures look - to the pixel - exactly the same? The pixels seem to be grouped in fours, like you would expect them to be in Jpeg. So: is the Tiff file really an (uncompressed) Tiff file, or has something gone wrong here?

[A Scottish soul]

In the mean time, have a look at dpreview review of the 828, They have posted a sample in Tiff together with the jpeg in chapter 11 of the review.


John

[A Scottish soul]

As promissed I had a go at this myself. The symultanious tiff and jpeg (best setting) produced by the camera are as far as I can tell identical and there is the 4 pixcel clumping. The RAW version, as Lin noted and my cursory observation shows is that there is much less clumping when run thru the raw converter and saved as tiff.

This is pure speculation on my part, however it does look as if the bayer pattern conversion in camera is less complex than that of the raw converter software. If this is the case, it is not surprising to find clumps of 4 pixcels in the in camera tiff file and to a much lesser extent the computer produced raw converted to tiff file. As far as the jpg produced by the camera is concerned I'd have to say that, as it is near identical o the tiff, the jpg coversion algorithm in camera must be extreemly good and produces no, or very limited compression artifacts.

As I said , pure speculation, and I'm probably wrong. Can anyone else shed any light on the subject?

Right or wrong this is perhapse just one more very good reason to shoot RAW whenever possible, or when you want a particularliy good or large print.

Regards all

John

[Lin Evans]

Hi Robert,

Follow the link below to a new thread I started on the 4 pixel clump issue and some further information.

Lin

http://www.luminous-landscape.com/cgi-bin....4;t=482