EXIF Data - Canon sensor temperature?

Not sure, can you send a raw file to me? vincent@astropixelprocessor.com

The temperature being reported is not very useful though, would be great if it was. There probably is a correlation with added noise to be made with that info though, but that you would need to first figure out. In my experience that temperature seems to be around 30 oC during a night anyway, maybe 35 in a hot night, slightly lower on a cold night... nevertheless hot! 🙂 A 5 oC difference you can use to create specific masterdarks for, I never saw a real difference though, unless you're actively cooling.

Vincent, I will send a RAW file to the address above. I know there are programs like ExifTool that can pull the sensor temperature out of the RAW file, so I was curious if it was also in the info that APP provides when you load a light frame. I shoot unattended with an intervalometer, and with nighttime temperatures dropping and the sensor heating up during the night, it would be great to know what the average sensor temperature was over the course of a night. Then I could have a better idea of what ambient temperature I should take the dark frames with an uncooled DSLR.

Or in the event I forgot to write down average ambient temperature for a run...

Yeah I'm not seeing this info either. Tried it with PI and got even less (or I'm missing the tool which is easy 😉 ), maybe Mabula @mabula-admin knows where this is stored?

In the case I forgot, I simply looked at what the weather apps told me the ambient was. That must be rather close and within the limits you would actually see a difference. I even started not doing it at all anymore as I really never saw the benefit in my images.

Hello: The heating of the Reflex during the taking of photos especially those of long duration> 120sec is a problem that is only solved with active cooling, but it is complicated and you have to disassemble the camera, you have to resort to experts.

For those who may be interested, I have developed a cooling solution that is easy to install, is suitable for any model, the Reflex (Canon, Nikon ... etc) must not be disassembled and one camera can be passed to another if there is a change of team.

It is not as effective as an integrated cooling, but I am using it and I have temperature reductions of 10 to 12ºc in comparison if I do not mount any refrigeration, the truth is that it shows a lot in the reduction of thermal noise during a session and the longer the more It is appreciated.

As I say; It is not about cooling a Reflex what it does is that it does not get hot.

Link to the brico: Reflex active Cooling - Brico

It's in Spanish, but I'm sure Google helps with the translation. Greetings

@fbup: "it would be great to know what the average sensor temperature was over the course of a night".  There is a program called Dark Master that will do that for you. Search for darkmaster and sourceforge and you will find it. I used it to match darks and lights when I was doing DSLR astro imaging (not so long ago).

@annehouw

Hello: The control software for Reflex cameras (Canon and Nikon) BackYardEOS (Canon) shows the temperature of each photo on the screen during shooting all the time. Greetings.

A few personal observations on Canon DSLRs and temperatures...

The temperature seems to be a reasonable representation of the real temperature, I've done test in the fridge and freezer to make temperature dependent darks. With intensive use (many read-outs in short time) the sensor will heat up a lot, with long exposure times it heats up slower and levels off to a temperature closer to the outside temperature.

When taking a long series of long exposure darks I see a clear increase in temperature and corresponding dark current (shows up as noise but is more a fixed pattern). I've looked at the individual pixels in detail, each as function of temperature, and see the "expected" behaviour: an exponential increase with temperature.

The tricky thing with Canon RAW files is that they have a built-in average dark signal correction before the RAW is compiled. No matter what the exposure time or temperature, for the typical bulk of the pixels (excluding the warm and hot pixels), the RAW of a dark will always have e.g. 1024 binary units average signal (the value depends on the camera model, but is a nice power of 2).

So actually using the sensor temperature for dark current prediction and correction doesn't work as good as I'd hoped. I've considered picking a handful of hot pixels and using them as "thermometer pixels" to derive the actual zero point of the exposure instead of the average-darksignal-corrected zero point, but that was pushing it a bit too far I thought...

Ah didn't know they did that. I do remember me being as scientific about it as well a few years ago, indeed going for the fridge and looking closely, in the end (I used a 6D) I concluded it wasn't really worth the hassle. I never saw dramatic differences in the end results.