temperature sensor С°≈((RAW-150)*1,5)+23

[Андрей Басов]

Hi, Alex.
I measured the temperature sensor in the Canon 550d in degrees Celsius
Infrared Thermometer "Centre 350". You can see it here.
https://docs.google.com/spreadsheet/ccc?key=0Ar2tmqEUFi2vdExFZTZfcE04X3VEWTZhOHd4enVwU0E
Formula: С°≈((RAW-150)*1,5)+23
Sorry, google translate.

[Alex]

Great find, thanks!

Is it OK to display it in Celsius degrees for PAL users and in
Fahrenheit degrees for NTSC users? (based on PAL/NTSC setting from the
camera).

> --
> http://magiclantern.wikia.com/
>
> To post to this group, send email to ml-d...@googlegroups.com
> To unsubscribe from this group, send email to ml-devel+u...@googlegroups.com
> For more options, visit this group at http://groups.google.com/group/ml-devel?hl=en

["Sztupák Sz. Zsolt"]

This wouldn't hold for Japanese and some South-American users (NTSC +
°C) I think.

SztupY

[Андрей Басов]

Alex, minor changes.
С°≈1.5*RAW-202
F°≈2.7*RAW-332

[Alex]

It would be great if these measurements could be repeated on other
cameras.

Over night I've left the cameras at room temperature (powered down, of
course). Now, when powering them on, the 550D displayed 20 degrees,
the 60D displayed 17 degrees and the 5D2 displayed 24. Not a big
difference though (maybe the sensor isn't very accurate).

[Андрей Басов]

During the measurement, said if the camera is turned on, the
temperature of the sensor is not uniform. The top of the sensor is
heated more by 5-9 degrees. Probably why such a difference. Accuracy
of ± 3-4 degrees.

[Андрей Басов]

And not always change the data RAW in real time. Need to turn off and
turn on the camera.

[arm.indy]

Hi,

Could you please take high resolution and high quality pictures of
your 550D mainboard ?
The goal is to be able to recognize electronic components?
and improve this page: http://magiclantern.wikia.com/wiki/Datasheets

Indy

[Morgan Look]

Hey.

Not sure if there are any useful images, but this page has some pretty good teardown pics.

http://ghonis2.ho8.com/rebelmod450d2.html

Morgan.

[Андрей Басов]

1.5 °С = 1 RAW suspiciously not precisely
2.7 °F = 1 RAW

RAW. Can this is not the temperature sensor?

[Colin Peart]

Up here in Canada, degrees C would be preferred as well.  Perhaps it needs to be an option?

Colin Peart
cgap...@gmail.com

[M.v.M.]

 Hi Alex and all,
 On a sailing Blog somebody started to dream about using a camera for Celestial Navigation, they are thinking that it should be possible to automatically measure the shortest distance between the horiz on and the sun on a DSLR. If the optics characteristic is known it should be easy to calculate the angle between horizon and the lowest part of the sun above via counting the  umber of pixels between the two objects. If we know that there is just the horizon as a line and the sole other object is the sun do you think this could be done in ML ? This means recognising the line of the horizon and finding the shortest distance to the overexposed sundisk ? Naturally this could be done in an extra programme but I think it would be a very clever and nice add-on for ML.

The idea went even further: as the suns diameter is constant one could measure the discs diameter to compute the optics characteristic and mm, and so use zoom lenses as well, not only primes.

This idea was first published by

Georg.Ki...@gimmel.franken.de

This is his idea, I simply thought that maybe someone here could be interested in realising it inside ML...

[g3gg0]

this is something i would expect in an extra .lua script or maybe a plugin :)
not as ML core functionality. very interesting idea :)

[M.v.M.]

Well this is just to spread the idea, if Lua can do the job ? Perfect, but maybe it should not be automatic detection, simply set two markers on the zoomed in liveview-playmode  and do a pixelcount could do the job, if we have for example detected or determined a 35 mm optic this would be an easy to compute thing. if we  know the numbers of pixels and know that the optic is set at infinity we should easily be able to compute the angle in a 'livemode'  This is exactly what a sextant is doing, and as we have an exact timestamp for each picture , even if it has an offset to an official timesignal of UTC  this could be much better to shoot the sun or the moon with the camera instead of using an very expensive sextant  where you start to count out loud the seconds after each measurement until you are able to check the actual UTC time on your watch or the ships chronometer.  I could test these measurements against my sextant anytime, so if somebody wants to write a script or a plug-in for ML where we simply count the pixels I could calibrate some optics. like Nikon 50mm 1.4  and 1.8 and 35mm1.4 and a couple of Nikon zooms and a canon 28-105 accordingly...

 Its more a for the fun of it thing and maybe easily done in ML ore quite difficult ...

Thanks for reading Matthias

"g3gg0" <g3gg...@gmail.com> schrieb:

this is something i would expect in an extra .lua script or maybe a plugin :)
not as ML core functionality. very interesting idea :)

[3pointedit]

Found this Atsrophotography page in relation to hot cameras...
http://astrophoto-sv.com/index.php?p=1_84

He has a cooling rig on his 550D! Wonder if he has collected stats on
the chip temps?

[arm.indy]

Thanks Morgan,

the goal is to identify chips.
here we can read chip references
http://www.ifixit.com/Teardown/iPhone-4S-Teardown/6610/2

Indy

On Apr 2, 12:33 am, Morgan Look <morgan.l...@gmail.com> wrote:
> Hey.
>

> Not sure if there are any useful images, but this page has some pretty good teardown pics.http://ghonis2.ho8.com/rebelmod450d2.html