Darks and temperature

[mgutierrez]

Hi all!

I've been reading about calibration frames and now I'm dealing with darks. I'm trying to build a darks library with different iso and temperature parameters. I've read that temperature is an important parameter. Temperature at which lights were taken has to match with darks ones, right? I also read that the important temperature to measure is the camera's sensor temperature. In my case, I have a reflex Nikon D3300. I've noticed that, at home (ambient temp ~20C), sensor's temperature (idle) is around 24C. The measure is taken with an IR thermometer, pulling up the camera's mirror and aiming directly to the mirror. To try to replicate the temperature during an observation in the middle of the field (which now could be around 10-12C), I thought (maybe incorrectly) to put the camera into the fridge (yes, maybe it's a bad idea) to see the temperature that the sensor reaches (the fridge is set to 5C). My surprise has been that after 30min taking darks inside the fridge, the sensor temperature is ~27C. Now, making the same test at home, out of the fridge, in my room (~20C ambient), the sensor reaches about 28C. Only 1-2C degree of difference at sensor while the ambient temperature difference is about 15C.
So, my question is, is it normal that the sensor's temp reaches almost the same temperature with an ambient temp of 5C than 20C? Because, if so, it would be much easier to build a dark library. Am I misunderstanding something? Am I measuring the temp in the right way? Which would be the best procedure to build a dark library at home?

Thanks in advance!

[admin]

I'm hoping someone else can weigh in here, as I have too little practical field experience to comment on this.

Your method of measuring the sensor temperature seems sound though.

Did you measure with the sensor active in both cases? (I think that's why they are using LiveView in the link above)

[almcl]

A while ago I used an remote reading IR thermometer to measure the sensor temperature of my Canon 700d.

My aim was slightly different: I was trying to verify the accuracy of the temperature recorded in the EXIF data in the raw files.
Now it is an often discussed topic that the temperature probe in the camera is not in direct contact with the sensor and so the figure recorded is not accurate. My findings tended to confirm that - they differed by perhaps 3 or 4°C, sometimes more. But what they did show was that any trend in temperature was accurately reflected, so if the EXIF temp showed a fall, so did the IR temp.

There's a thread on SGL here https://stargazerslounge.com/topic/2784 ... nt-3048959 where I put some results, which, as you can see, were quite frightening. In this session the sensor started at over 40°C (probably due to using Live View for focus and framing) and although it cooled, there was no way I could use a dark library. So I stopped trying to use darks.

A while later I read about TEC coolers and how some brave people had modified their DSLRs by cutting into the case and inserting a 'cold finger'. I am not nearly that brave so I made up a cooler that attaches to the outside of the case and started experimenting.

It doesn't achieve the sort of DeltaT that a dedicated astro camera with built in cooling manages. Typically it lowers the temperature by about 10 - 12°, which in fact puts it at around ambient or just below. The important thing, however, is that all the lights end up at the same temp +/- a degree or so.

If you can get stable light temperatures then you may well be able to make use of a dark library.

One EXIF data reading tool which can be downloaded for free is:
http://www.covingtoninnovations.com/dslr/exiflog.html

[mgutierrez]

Hi,

Thanks to both and sorry for the late reply. I've been doing some tests.
I've started from scratch, with an ambient temperature of 24C degree, and the camera idle and "cold", at ambient temperature. Then, I've been taking photos consecutively with an exposition of 5 minutes, during 2 hours; and measuring the sensor temperature between each photo (with the IR thermometer). This is the result:

D3300-sensor_warmup.png (51.16 KiB) Viewed 4635 times

As you may see, the temperature starts to stabilize after ~50min, reaching a max temperature of ~33-34C. I suspect that, in general terms, this is the normal behaviour: the final temperature the sensor reaches is about 10C more than initial ambient temperature, and that happens after ~50min.
So, my idea is to replicate different scenarios with different ambient temperature; then start for example a 3h simulated session, and discard the photos taken during the first hour.

Hope it makes sense

Thanks!

[Burly]

Depending on your setup try dithering and do away with darks if possible i dither between frames and use APT connecting to mount via Eqmod basically the mount moves image approx 10pixels between subs does away for the need for darks and saves time and for dslr imaging is a better way more info here https://skyandtelescope.org/astronomy-b ... ro-images/

[mgutierrez]

Depending on your setup try dithering and do away with darks if possible i dither between frames and use APT connecting to mount via Eqmod basically the mount moves image approx 10pixels between subs does away for the need for darks and saves time and for dslr imaging is a better way more info here https://skyandtelescope.org/astronomy-b ... ro-images/

Thanks for the reply.
Yes, dithering is also another improvement I'm planning to implement. I'm waiting my new qhy5 cam for autoguiding. I'll make some tests once received. I have an eq5 mount with the corresponding ar and dec motors, and an autoguiding port.
So, are not darks needed anymore when using dithering? I guess having also darks will improve the snr. Or maybe barely makes difference. Well, I'll try.

Regards

[Burly]

No darks not required all I do is every couple of months I take 50 bias frames , and in DSS or Siril during stacking I load the bias as darks , I was told to do that with dithering by a respected Astrophotographers on a local forum and seems to work ok so lights and flats on the night and saves time .

[mgutierrez]

No darks not required all I do is every couple of months I take 50 bias frames , and in DSS or Siril during stacking I load the bias as darks , I was told to do that with dithering by a respected Astrophotographers on a local forum and seems to work ok so lights and flats on the night and saves time .

Well, it seems it makes sense to subtract bias in the same way than they were darks, since the lights frames don't have thermal noise (or almost) and only should contain read noise, that need to be subtracted. Makes sense, I think. Thanks for the tip!

M