How many Hot and Cold pixels reported by cr2hdr.exe for Dual-ISO on your camera?

[l_d_allan]

I'm curious if my 5d2 and T3i / 600d have a relatively low, normal, or high amount of Hot and Cold pixels, as reported by cr2hdr.exe when processing Dual-ISO files.

So for apples and apples:

• 30 seconds
• at ISO 400
• with lens cap on (or body cap)
• Dual-ISO 100/1600
• Room temperature, and exposure not done in LiveView, so as to not increase sensor heating
• Also, I think it would be appropriate to have recently done the remapping of hot/stuck pixels by taking a 30 to 60 second exposure with the body cap on. Also, have ML disabled by pressing SET while camera booting.
• How much use has the camera had ... reported shutter activations (including in/out of LiveView)
• How much of the usage has been video? (which I believe can be more likely to overheat the sensor ... or not?)
• other issues I'm leaving out?

• 5d2 with about 201,000 shutter activations, almost no video: 1477 cold pixels, no hot pixels reported
• T3i with about 30,000 shutter activations, no video: 374 cold pixels, no hot pixels reported

Edit: Did some Dual-ISO images this morning, with very different Hot/Cold pixel counts reported:
Hot: 25 to 660
Cold: 25,636 to 46,030 (but several had zero)

Seems odd ... they weren't particularly long exposures ... 1 to 10 seconds. However, one or more may have been with LiveView ON ... don't recall specifically. I suppose that could make a difference. Or not?

[l_d_allan]

FYI:

• Using 6d for star-trails and dual-iso (perhaps not a good combination?)
• Temp about 30 degrees F ... 120 sec Bulb exposures ... ISO 100 (quite a bit cooler than room temperature, so not "apples and apples")
• Hot pixels: from 2266 to 6256
• Cold pixels: from 173 to 776
• Dark frames:
• Hot pixels: 0 to 0 (bit surprised that the "dark frames" have zero hot pixels)
• Cold pixels: 83 to 127

[a1ex]

I guess the converter just doesn't have enough data to match the exposures in a dark frame, so the result is probably bogus.

May I see some of these raw files?

[l_d_allan]

>> bit surprised that the "dark frames" have zero hot pixels

> May I see some of these raw files?

OK ... thanks for checking on this possible anomaly.

• Just RAW cr2's: http://www.canprint.org/Misc/140111a_DeckStarTrails_JustCr2s.zip
• Both .cr2's and .dng's: http://www.canprint.org/Misc/140111a_DeckStarTrails_CheckHotColdPixels.zip

• includes 140111a_192889_DeckStarTrails_DI_DarkFrame.CR2 and 140111a_192889_DeckStarTrails_DI_DarkFrame.dng
• 192886 and 192887 are actual StarTrails, not dark-frames
• Rather large zips ... ~ 114MB for .zip that includes both cr2's and dng's ... ~ 63MB for .zip that only has .cr2's
• FWIW: you can see the "Big Dipper" towards the top-middle, and from there find the North Star.
• FYI: these are poor StarTrails, intended for testing with Dual-ISO. They are with a Samyang 8mm FishEye on full-frame, from my deck within city limits of Colorado Springs, so the ambient light is rather "polluted". For real StarTrails, I go about 50 to 100 miles east from where this image was taken: http://berean.zenfolio.com/zf_slides/h54be6206#h54bea6d4