DSLR darks and temperature

[michaelmorris]

I'm still trying to learn the techniques involved in capturing DSLR astro images. I known that I need dark frames to calibrate my image by taking out the thermal signal created by the camera. I also know that I need to have a set of darks of the same exposure length, ISO setting and temperature as the light frames. I'm planning to build up a 'bank' of dark frames of different ISO settings,temperatures and exposures. Now here is my question - the temperature of what? Should it be the outside temperature, the camera body temperature or the camera sensor temperature? I can get the outside temp from a sensor/thermometer and the sensor temperature from the image EXIF file. For the camera body temperature I guess I would have to get a temperature sensor and tape it to the camera body.

[Ben Ritchie]

As it's uncooled I doubt that it matters too much which one you pick as long as you're consistent, although it really needs to be the sensor temperature as that's the thing producing the dark current. Without setpoint cooling it's always going to be 'near enough' rather than 'exact' anyway. For a DSLR the difference between ambient, body and sensor is unlikely to be very large unless you've just taken it out of a hot room into sub-zero night.

I also know that I need to have a set of darks of the same exposure length, ISO setting and temperature as the light frames.

You can use bias to scale the darks to match the lights, so they don't *have* to be the same exposure length. They *do* have to be the same temperature though, as dark current varies quite significantly with temperature - hence the benefit of cooling.

[Psychobilly]

Isn't the sesor temp embedded in the raw file data with some of the canon cameras? Yoddahs APT app can use it...

The sensor temp changes will lag the body temp changes and will also depend on what the camera has been "doing"...

Billy...

[Zakalwe]

As Billy has said, if you use APT, it can read the sensor temperature. On my 50D the sensor temp hovered about 8 degrees C when the outside temp was about freezing. As soon as I brought the camera inside, the temperature on the sensor started to rise quite quickly, even though the room it was in was very cool.

Your darks temperature have to be fairly near the light's , otherwise you might get black spots over your stacked image as the stacking software removes noise that is not present in the lights (higher sensor temperature = more noise)

[dmahon]

I think the sensor temperature is the important one. I've been filing them recently according to the temperature as given by APT (and written into the file name) but I've just spent most of the morning going through all my old ones using an EXIF reader:

ExifTool by Phil Harvey

The're now filed into folders: ISO, duration, temperature

[Yoddha]

Hi Michael,

Here is my study in the DSLR heating http://stargazerslounge.com/imaging-discussion/126136-dslr-sensor-heating-study.html

It shows that you can rely on EXIF when the camera is not cooled with cold finger...

[michaelmorris]

Okay, thanks guys, that's very useful. The next job is to sit down and design a 'dark frames and bias frames' programme for cloudy nights.