Canon EOS450D sensor

[Guest]

I have a canon eos 450d. If I take a picture in raw format the cr2 file and the corresponding FITs file says that it is 4290x2856 pixels. But if I look at the properties of that file it says that the picture is 4272x2848. I know it's a very small difference but it bugs me.  There's a field of view calculator on the internet and if I look up the canon eos450d is also says 4272x2848 pixels. Everyone agrees that sensor size is 22.2x14.8 mm but does that include the extra pixels?  Gimp and maximdl both tell me 4290x2856.

I assume the idea is that it allows for small misalignments in the optics so you still get the 4272x2848 pixel image.

This all came about because I was interested in the accurate focal length of the scope. It is nominally 1000mm but I use a coma corrector which is supposed to be a 0.9 reducer. So that should be 900mm equivalent. However when I do a plate solve the various plate solvers give me a focal length of 920 or 923 mm.

Cheers

Steve

 

[Seelive]

I have a Canon 200D that shows as 6000 x 4000 pixels in properties because that is what is reported in the EXIF data but in reality the the RAW image is larger.  The older versions of DSS using DCRAW give images of 6024 x 4022 whereas the later versions using LIBRAW give an image of 6024 x 4020 pixels (so incompatibIle with the earlier versions).  My old copy of IRIS opens them as 6288 x 4056 (although their is a large black border and the actual useable image size is only 6024 x 4020). The Adobe DNG converter gives 6024 x 4020 in both the old version of DSS and IRIS.

To be honest, different plate solvers giving a focal length variation of only 0.33% wouldn't worry me too much.

[michael8554]

From what I've read, the sensors have border pixels called Registration Areas that are used for black level, noise reduction etc purposes.

Some softwares use the larger figure, others use the smaller figure, which can lead to problems when you try to add two software results.

Use the Plate Solve FL.

Michael

[Guest]

Thanks Michael, 

I wasn't too bothered about the difference in focal length between the two plate solvers. Only 3 mm. But I was surprised at how much they were different to the expected focal length of 900mm = 1000 x 0.9. I suppose there will be some tolerance in the focal length of the telescope and the coma corrector. The tolerances could add up. It doesn't really matter but I was interested.

Steve