4 ms·
>M-CAM 1 took this long-exposure photograph of Mercury's north pole I'm curious how this works. The dynamic range between the sunlit parts and the dark portion
by bragr 2y ago
>M-CAM 1 took this long-exposure photograph of Mercury's north pole
I'm curious how this works. The dynamic range between the sunlit parts and the dark portions must be huge at that distance from the Sun. Anyone have the technical details on the camera or post processing they use to achieve this? Is it really a long exposure or is it a series of photos at different exposures stitched together?
Edit: details at the bottom seem to imply a single photo, but that "long exposure" really isn't that long
>This image of Mercury's surface was taken by M-CAM 1 [...] using an integration time of 40 milliseconds.
- therealfiona 2y agoIntegration implies multiple pictures taken and stacked. This is how we do deep sky astrophotography on our back yards.
- fecal_henge 2y agoI think this is their terminology for exposure time. The sensor is integrating charge from incident photons during this period. Of course the image could be stacked also! Edit: the sensor is integrating CURRENT. Charge is the integral!
- tejtm 2y agoIntegration over time and integration over bit fields are both ways we do deep sky astrophotography. Integration over time is used to collect more photons without saturating sensor wells and integrating bit fields (stacking) is used to increase signal to noise
- tempodox 2y agoWhat does that mean exactly? What's the relation between integration time and exposure time?
- car 2y agoWondering same. If my math is right, this was a 1/25 sec. exposure. (40/1000)
- dylan604 2y agoIf you're thinking that long exposure automatically means something longer than 1 second, it might not mean long exposure to you. I'd guess this is a fixed aperture system where the main way to control the exposure is with shutter speed. But for images taken in bright sunlight, you can use shutter speeds 1/250, 1/1000, 1/2000, or even higher type numbers. In those terms, 40 milliseconds is 10 times slower/longer than 1/250. So for the M-CAM 1 system, 40 milliseconds could be an extremely long exposure