sensor remapping...

gjl711 wrote in post #13966971
Can you post a sample? 100% crop would help as well. ISO 1600 should be pretty clean and white dots doesn't sound like sensor noise. Sensor noise are random colored specks.

lannes wrote in post #13967032
Interesting blog on sensor remapping and experiences at Canon service

http://www.hdcamteam.c​om …er-experience-hot-pixels/

you could try the method, where you do a manual self clean for at least a minute with the body cap on, before you send it to Canon

Canon Hot(dead) pixel fix (do it in dark room):

Step 1 - Detach the lens from your Canon camera
Step 2 - Put on the camera cap (do not leave your camera open).

Step 3 - Turn your camera on. Go to the menu and select "Sensor cleaning" under the utilities menu.

Step 4 - Turn the dial to select "Clean manually" under the sensor cleaning option.

Step 5 - Turn the dial to select "OK" and then press the "Set" button. The mirror will lock up and the shutter will open. Allow the shutter to remain open for at least 30 seconds. In your case I'd leave it sitting like that for 60 seconds.
Step 6 - Turn your camera off and reattach the lens.

Take a few test shots to see if the problem has been corrected.

Background Info:
---------------
There is really no mystery about sensors having bad pixels- they all do but the cameras map the true dead and stuck pixels out. The confusion comes from newbies that don't understand how hot pixels behave. As explained by others above hot pixels are just pixels that have greater than average dark current. The visibility of dark current scales by exposure time and ISO, and dark current itself increases very rapidly with increased sensor temperature. This means that for long exposures especially at higher ISO settings "hot" pixels will appear to be stuck on: even though they may work OK for short duration low ISO exposures.

It is true that current Canon DSLR cameras create a new bad pixel map whenever the firmware is updated or the manual sensor clean function is selected (I have personally verified this for my 5D2 which has two hot pixels). Newbies may report that this doesn't fix their hot pixels because the function only maps out pixels that are bad enough to appear stuck for exposures of about one second at ISO 100. Pixels that work at this setting are not mapped out because they work fine for typical photographs. If you care most about long exposures then you can increase the number of pixels mapped out by first heating up the sensor by using live view. If the manual sensor clean is selected right after using live view for at least two minutes then the increased sensor temperature will result in more of the marginal pixels appearing as stuck to (and therefore mapped out by) the mapping out routine. Of course this will need to be repeated after every time the sensor is cleaned or the firmware is updated or the marginal pixels will reappear.

If the number of hot pixels is still too much for your liking then you can turn on the long exposure noise reduction function or better yet just use Lightroom or Photoshop to perform the raw conversion. These programs use Adobe's ACR that does a good job of removing hot pixels. This is much preferable to the Nikon method which always applies a crude hot pixel removal routine to the raw data of images with exposures of about one second or longer. This forced Nikon hot pixel suppression is probably why there are less complaints about hot pixels in the Nikon forums but it also removes real small image details like faint stars in the night sky. This is one reason why Canon cameras are often preferred for amateur astronomy.

No I am not a sensor designer- but I doubt that you will get someone like Eric Fossum to answer such a basic question. A great deal of information about sensors is freely available though. There are many papers available from the past workshops that Fossum organizes:
http://www.imagesensor​s.org …hops/Past%20Wor​kshops.htm

Here are two papers about modeling dark current from a Google search:

http://www.ee.bgu.ac.i​l …APS_DC_Modeling​_Final.pdf

http://www.imagesensor​s.org …%20Folkerts%20e​t%20al.pdf

Here is an article by the author of the IRIS astronomical imaging software about Nikon's hot pixel removal processing:
http://www.astrosurf.c​om/~buil/nikon_test/te​st.htm

swetsastonic wrote in post #13967327
I'm actually not sure what it is...noise is the only word I could think to describe it. here's a sample...

lannes wrote in post #13967442
That looks more like noise, are you using Canon's DPP ?