1. Stefan B Avatar Stefan B
  2. nina
  3. NINA
  4. Issues

The pattern name $$SENSORTEMP$$ always returns NULL for DSLRs

Issue #590 resolved
Allan Sun created an issue

I use the pattern name $$SENSORTEMP$$ in my image file name, but NINA always gives NULL for my Canon EOS 500D. I know that we can not get sensor temperature from Canon camera, but there is a EXIF tag “Camera Temperature (EOS Only)“ in the RAW cr2 file (from Irfanview software), which is the camera board’s temperature, would you please let $$SENSORTEMP$$ return this EXIT tag value?

If supported, we can match the light and dark frames just from the file names.

NINA is a great software, thanks all NINA’s developers!

Comments (23)

  1. Stefan B repo owner

    Hi,
    as you already said, the EXIF tag is just the body temperature and not the sensor temperature. Therefore the correct sensor temp info is not really available and putting in the body temperature into that would be questionable.

  3. Allan Sun reporter

    Thanks Stefan. Is it possible to add a new pattern such like $$CAMERABODYTEMP$$ as an alternative?

  5. Bill Richards

    Why do you say that this EXIF value is the body temperature as opposed to the sensor temperature? I cannot find anything online that indicates this, but I may not be looking in the right place.

    In any case, it is a legitimate temperature reading and does provide some indication of the sensor temp even if it’s not 100% accurate. I’ve run a lot of tests and found that for my camera, the thermal noise starts to ramp up dramatically once the temperature hits about 27C so I try to keep the camera below that while shooting.

    Is there some reason this data cannot be extracted from the image file and displayed? I assume that’s what BackyardEOS does since the value only changes when a new image is downloaded.

    I was really hoping that I could switch over to NINA bus this is a showstopper for me.

  6. Stefan B repo owner

    Well the main reason is that due to the proprietary format, reading the exif data isn’t straight forward and not provided by their sdk. Nina already needs to go a special route by calling dcraw/freeimage to be able to read the format at all for image processing. Not having the exif data is mostly a thing of - no one did the implementation for it yet, as it is not trivial effort wise and not much benefit.

  7. Bill Richards

    OK, thanks for the quick response and explanation. I guess I’ll have to figure out how to use ASTAP in BackyardEOS (which I just discovered is now supported).

    BTW - what I saw last weekend was very impressive! You have a great tool here.

  8. Stefan B repo owner

    The Temperature Data will still be available for you in the saved RAW files. You can readout the cr2/cr3 files with a tool like “ExifTool” and check the temperature there. Why do you need the temperature to be displayed directly in app, when you could create the dark frame for that temperature at a later point in time?

  9. Bill Richards

    Yes, I can do that but it’s certainly not as convenient as having it displayed and automatically updated with every image taken.

    Last weekend, I was out in the desert taking images of M31 and had to watch the camera temperature very closely. When it got too warm, I placed a small fan on the camera and stopped shooting for a while, then resumed. So a real-time readout on the display is perfect.

  10. OK Take

    Hello everyone.
    Other sites have the following topics:
    https://theskysearchers.com/viewtopic.php?t=1656
    With in this topic there is a very interesting discussion about camera temperature. It was very helpful. If you don't know it yet, I recommend you to read it.

    by the way,
    Did you know that connecting through " ASCOM Driver for DSLR Cameras " will pass the Canon temperature at the time of shooting to $$ SENSORTEMP $$ ?
    I'm a Canon EOS 6D and Ra user, so I've used this ASCOM Driver to compare the two several times. In a temperature-stable indoor environment, cold-start the two units at the same time, then expose for a very short time (exposure to detect the temperature at that time), then leave enough interval time, and many more times. I repeated it. I noticed that the temperature rise of Ra was overwhelmingly fast. And soon, there was a difference of 10 ° C or more. I'm guessing that Ra is related to being a mirrorless camera.

    From such an experience,
    ・ This camera temperature is not the image sensor temperature.
    ・ This temperature sensor is inside the DIGIC chip and may reflect the work of the DIGIC chip.
    ・ This temperature sensor does not pick up the change in heat generation of the image sensor with the exposure time.
    I agree with this view. And
    ・ It is highly possible that using a dark library with a DSLR is not optimal.
    ・ The highest accuracy is achieved by generating dark frames on the spot rather than trying to approximate using unreliable
    I also agree with this opinion.

    Once the DSLR has a command that allows you to expose with the body shutter closed, it will be comfortable to generate a dark frame on the fly. (Because the DSLR has a deep flange back, it can be difficult to pinch the shutter in the middle of the train.) I hope Canon will embed a temperature sensor in the image sensor in the future.

    Please forgive me for using machine translation.

  11. Bill Richards

    I understand everything discussed in this thread and the thread referenced above. Nevertheless, the temperature provided by the Canon DSLRs is still an indicator of the sensor temperature, even if it’s not an accurate reading. It’s better than nothing at all.

    I have done some experiments in trying to match the Canon temperature readout while creating Darks such that they match the temperatures of the Lights. In doing so, I was able to achieve very good noise reduction results in my stack. In fact, when I actively matched the temperatures I got far better results than just using Darks captured at the end of the night (which is what I used to do).

    So I’m in the camp that anything is better than nothing if one isn’t willing to waste precious time during an imaging session to take Darks in between Lights. Since all of my imaging sessions require a long drive out to isolated areas with dark skies, that time is valuable and I prefer to spend it capturing as many Light subs as possible.

  14. OK Take

    Hi, Bill Richards.
    I have to drive for a few hours to get a decent view of the stars too. So I can understand your feelings. I don't believe that individuals should be denied the fact that they have different ideals to aspire to, and I also think that having more options is not a bad thing in principle.
    I'd like to suggest to those of you who are experiencing the same problems that you are experiencing here to try to connect to NINA using the "ASCOM Driver for DSLR Cameras" . I think using this driver together is a handy way to solve the problem that can be done right now. I feel that the development of this driver has been very active over the last few months.
    I also wanted to share with you what I experienced about the difference between a mirrorless camera and not.

    I'm using machine translation. Please forgive my rudeness.

  16. Dale Ghent

    A new metadata field called CameraTemperature should be created and this value stored there, with the corresponding file name pattern/token created. The reason being is that, strictly speaking, this is not the temperature of the sensor and it would be wrong to present it as such. Canon does not specify in public documentation what this temperature is of.

  17. Stefan B repo owner

    sure it’s not strictly the sensor temperature, but cluttering the ui due to this one special case isn’t a great design either. it would be better to just add a note to the description and/or to the docs.

  20. Jeanfi

    Hello, as the creator of a duplicate “issue”, I would like to thank you very much for this evolution ! Thank you guys.

  22. Stefan B repo owner

    it is part of version 1.11

    Currently it is in preview/nightly state to try out new stuff. Once the development for this is considered final it will transition to beta state to find bugs etc. after that it will then be a released version

  24. Log in to comment
Assignee
–
Type
proposal
Priority
minor
Status
resolved
Component
User Interface
Version
1.10
Votes
1
Watchers
9
Jira: the preferred issue tracker for Bitbucket. Join the team!