Dead pixel...?!

[axisvideo]

Dear Alex and developers,
I had problems with the latest version of ML:
during some timelapse night in all the frames appear several pixels white and blue ... before the time lapse, in the test exposures, there were not. After i removed ML menu, test shots with Canon menu in high ISO values ??and long times 10 seconds, there are no dead pixels, what happen?
One suggestion if you let me: in the next release it would not be better to divide the functions of ML kind of interest, meaning you have to do a time lapse loads only the functions of the time lapse, if you have to do a HDR load functions for HDR, if you work with the video loads only the video features. May be this we could avoid the possible conflicts that seem to make but not completely stable ML. I do not know ...
Thank you,

Massimo

Config file
# Magic Lantern v2.3.RC2.2012Jul15.5D2212. (d5735cf30260+ (unified) tip)
# Build on 2012-07-15 10:10:32 by alex@thinkpad
# Configuration saved on 2012/07/28 09:30:51
menu.upside.down = 0
menu.first = 9542382
debug.timed-dump = 0
audio.monitoring = 1
audio.draw-meters = 2
audio.filters = 1
audio.input-choice = 4
audio.alc-enable = 0
audio.o2gain = 0
audio.lovl = 0
audio.mic-power = 1
audio.dgain.r = 0
audio.dgain.l = 0
audio.mgain = 4
buffer.warning.level = 70
bitrate.indicator = 0
time.indicator = 3
h264.bitrate-factor = 10
h264.bitrate-mode = 1
h264.qscale.plus16 = 8
battery.drain.rate.rev = 88
lcd.adjust.position = 0
bmp.color.scheme = 0
preview.saturation = 1
preview.contrast = 3
arrows.bright_sat = 0
arrows.tv_av = 0
arrows.iso_kelvin = 0
arrows.audio = 0
arrows.set = 1
arrows.mode = 0
digital.zoom.shortcut = 1
swap.menu = 0
quick.review.allow.zoom = 0
halfshutter.sticky = 0
play.lv.button = 0
quick.delete = 0
play.set.wheel = 2
play.quick.zoom = 2
af.frame.autohide = 1
burst.auto.picquality = 0
dof.preview.sticky = 0
bv.needed.by.av = 0
bv.needed.by.tv = 0
bv.needed.by.iso = 0
movie.log = 0
shutter.display.degrees = 0
display.dont.mirror = 1
lcdsensor.wakeup = 1
electronic.level = 0
defish.projection = 0
defish.preview = 0
idle.rec = 0
idle.display.gdraw_off.after = 0
idle.display.dim.after = 0
idle.display.turn_off.after = 0
spotmeter.position = 1
spotmeter.formula = 0
spotmeter.draw = 1
spotmeter.size = 5
clear.preview.mode = 0
clear.preview.delay = 1000
clear.preview = 0
vectorscope.draw = 0
waveform.bg = 38
waveform.size = 0
waveform.draw = 1
hist.log = 1
hist.warn = 3
hist.colorspace = 1
hist.draw = 1
focus.peaking.grayscale = 0
focus.peaking.color = 7
focus.peaking.thr = 5
focus.peaking.method = 1
focus.peaking = 1
zoom.overlay.split = 0
zoom.overlay.pos = 1
zoom.overlay.x = 1
zoom.overlay.size = 1
zoom.overlay.trig = 3
zoom.overlay = 0
falsecolor.palette = 0
falsecolor.draw = 0
crop.playback = 0
crop.movieonly = 0
crop.index = 0
crop.enable = 0
zebra.rec = 1
zebra.thr.lo = 0
zebra.thr.hi = 95
zebra.colorspace = 2
zebra.draw = 1
global.draw = 1
transparent.overlay.y = 0
transparent.overlay.x = 0
transparent.overlay = 0
disp.mode.x = 343
disp.mode.c = 1
disp.mode.b = 1
disp.mode.a = 1
disp.mode = 0
lv.disp.profiles = 0
bramp.calib.5 = 0
bramp.calib.4 = 0
bramp.calib.3 = 0
bramp.calib.2 = 0
bramp.calib.1 = 0
bramp.calib.0 = 0
bramp.calib.m1 = 0
bramp.calib.m2 = 0
bramp.calib.m3 = 0
bramp.calib.m4 = 0
bramp.calib.m5 = 0
bramp.calib.sig = 0
picstyle.rec = 0
picstyle.rec.sub = 1
bulb.ramping.manual.focus = 1000
bulb.ramping.manual.expo = 1000
bulb.ramping.percentile = 50
bulb.ramping.auto.speed = 100
bulb.ramping.auto = 1
bulb.ramping = 0
motion.trigger = 0
motion.release-level = 8
mlu.auto = 1
bulb.duration.index = 15
bulb.timer = 0
zoom.auto.exposure = 0
zoom.focus_ring = 0
zoom.halfshutter = 0
zoom.sharpen = 0
zoom.disable.x10 = 0
zoom.disable.x5 = 0
silent.pic.sweepdelay = 350
silent.pic.highres = 0
silent.pic.mode = 0
silent.pic = 0
lv_3rd_party_flash = 0
flash_and_no_flash = 0
audio.release-level = 10
focus.trap = 0
interval.stop_after = 2
interval.movie.duration.index = 2
interval.start.timer.index = 3
interval.timer.index = 16
hdr.scripts = 2
hdr.iso = 0
hdr.seq = 1
hdr.delay = 1
hdr.ev_spacing = 16
hdr.frames = 1
hdr.enabled = 0
iso_selection = 0
lvae.disp.gain = 0
lvae.iso.spd = 10
lvae.iso.max = 104
lvae.iso.min = 72
bv.av = 48
bv.tv = 111
bv.iso = 88
bv.auto = 0
rec.notify = 3
screen_layout.ext = 1
screen_layout.lcd = 0
shutter.lock.value = 0
shutter.lock = 0
enable-liveview = 0
movie.rec-key = 0
movie.mode-remap = 0
movie.restart = 0
wbs.ba = 100
wbs.gm = 100
wb.kelvin = 6500
hdmi.force.vga = 0
focus.patterns = 0
movie.af = 0
focus.follow.rev.v = 0
focus.follow.rev.h = 0
focus.follow.mode = 0
focus.follow = 0
focus.step = 5
focus.stack = 0
focus.rack.delay = 2
focus.wait = 1
focus.delay = 1
focus.stepsize = 2
fps.sound.disable = 1
fps.preset = 0
fps.timer.b.off = 1000
fps.timer.a.off = 1000
fps.override.idx = 10
fps.override = 0
hdrv.iso.b = 101
hdrv.iso.a = 72
hdrv.en = 0
lcd.release = 0
digic.cartoon = 0
digic.swap-uv = 0
digic.negative = 0
digic.desaturate = 0
digic.black = 100
digic.iso.gain.photo = 1024
digic.iso.gain.movie = 1024

[DMcG]

I have a similar issue. I've just installed Magic Lantern v2.3 on my 600d on top of firmware 1.0.2. I was playing with the features including focus peaking and fps override. I now have what looks to me very much like a defective pixel which is red and always visible at the same point on the screen both in live view and in playback in video and stills mode. Needless to say I am a bit worried and would be grateful for any reassurance or helpful advice.

I have now formatted the card and removed ML. Red pixel is still clearly visable. Very concerned 

[ilguercio]

Very concerned about ONE red pixel?
Seriously?

[Malcolm Debono]

http://www.slashgear.com/how-to-remove-stuck-pixel-on-your-digital-slr-2227392/

[DMcG]

Thanks very much Malcolm, that worked a treat! Panic over 

@Ilguercio yes I was very concerned having never heard of this issue before and being completely new to Magic Lantern. This forum is somewhere to report problems and get solutions, not to come and mock those experiencing difficulties. If you are so familiar with the problem, why didn't you simply share your knowledge? Not a great first impression of community spirit here.

Thanks again for providing the solution so quickly though Malcolm. I'll be reinstalling ML today, happy to finally be using this great tool.

[Malcolm Debono]

You're welcome!

While dead pixels aren't really an issue for photos since they are barely noticeable and can easily be cloned out, they are much worse when present in videos.

[ilguercio]

They should not be a concern because they're not caused by ML but they are present in a sensor anyway.
You can't expect to have 18.000.000 working pixels/18.000.000. Chances are you will find dead or just hot pixels from time to time and the best way to try and solve this issue is the sensor cleaning method suggested by Malcolm. It's not mocking around but since it is a common problem of photography in general it was a bit strange to me that someone has never noticed them.
Raw converters already map them out and alter them by looking at the surrounding pixels. So, it's not something bad, don't be concerned.

[screamer]

Ops, i don't know if it's magic lantern related, but the strange thing is that me too noticed yesterday, for the first time, a bunch of red or white pixels in a star timelapse. and they are still there. I've tried the manual cleaning procedure, and not only one time. But seems that these pixels are burned out :\ the strange thing is that i'm shure that pixels wasn't there only some day ago. I compared the shots of yesterday with the shots of 1 week ago (quite same shooting conditions, with black sky and some stars). And there was'nt.
I know in post one can remove them, but it's strange that they happens in the same time (they are at least 5 hot pixels really visible, and other less visible random distributed.
I'm a little worried about this, because if (and i hope no) magic lantern caused this, for example with some function that warm up so much the sensor or something like this, i'm scared that continuing to use it i will have an unusable camera in a couple of months.

I can add as observation that this morning i switch on the camera and the battery was drained out, it's the second time that i have this issue with ml 2.3, never had before. And i understood when it's happen, so from the next time i'll take care of detaching the battery when i know this could be. What i've done was to use intervalometer with behaviour "take pics like a crazy" and with a shutter time  of 30 seconds. So, on a certain point i've switched off the camera (in the middle of a shot). So probably the camera hasn't really turned off. result, the battery was completely discharged this morning (yesterday it was quite full) and i have all this hot pixels. The manual cleaning procedure dosen't work here
I don't know if this things are related, but probably if the sensor was active for all night long this time, and the other time too, i think it could be overheated and the pixels can be burned out for this reason.

here is a shot taken with the cap at 30seconds, as you can see there are a lot of hot pixels :\
http://www.bcaa.it/hotpixels.jpg

[ilguercio]

You must wait 5 seconds between the opening of the card door and the removal of the card itself.
That should solve the issue with the draining.
As for the pixels, every camera can have this kind of problem, just google it.

[Pyrofer]

I think the "concerning" part is that from the sounds of the posts, and my own experience these "stuck" and "dead" pixels were not there before 2.3ML, and suddenly on running this FW they ARE there.

There are 3 options,
1>Coincidence, you just noticed them
2>Something about running ML "wiped" the cameras memory of Stuck pixels and the ones that were always there are now visible again
3>Running ML has cause the camera to heat up/whatever and caused additional pixel issues.

Cases 1 and 2 are complete non-events, Case 3 however is worrying.
I think people just need assurance that it is indeed case 1 or (I expect) 2 and not 3.

Can somebody confirm this in any way? Possibly by comparing old Raw images that may contain a dead pixel they expect has "recently" appeared in ML 2.3? (thus confirming it was always there)

[a1ex]

It may happen when you clear camera settings.

[ilguercio]

ML doesn't heat the camera in any abnormal way, it has been shown many times and the only way to induce overheating is to record for a long time in a hot environment. The camera will then show a thermometer on screen telling you the story is getting quite warm and inviting you to turn off the camera eventually.
Those myths could have existed when ML was at its primitive stage, now though it sound a bit strange that people don't understand that ML doesn't cause the camera to fall into pieces or cook your pizza.
Then again, i appreciate the user feedback about one dead pixel but a quick search on google could have at least cooled your worries as it is a very normal problem in sensors and it is just  ONE pixel anyway.
I want to remember that ML is for ADVANCED users so make sure you don't install it straightaway on your first DSLR and confuse yourself with all the new stuff. Know your camera for a while, understand what's normal and what's not and then install it.
You'll save yourself quite a lot of heart attacks.

[Pyrofer]

These are all very good points, and I am sorry if it seemed I am blaming ML for "heating" or dead pixels.

The fact is, I have used my camera for years and ML on it for as long with no issue. On installing 2.3 latest build I suddenly had a flock of dead pixels all over the screen that were not previously there.

I can pinpoint their appearance exactly at the time I put new ML on by the images I took.

Without a doubt, putting new ML made them appear.

Just telling people to "ignore it" or "google" it does not help.

If ML is clearing something in the camera that had previously mapped out dead pixels, people need to be told BEFORE installing it, something along the lines of,
"ML may cause your camera to forget its dead pixel mapping and show dead pixels you had not previously seen. This is normal and they can be mapped out using this method (...)"

A straight factual answer from somebody who knows what happened to cause this will prevent a wave of "ML broke my Camera " complaints.

I am not saying ML broke my camera, I am saying that people may perceive that and it is bad press ML does not need.

[ilguercio]

But since you should know dead or hot pixels are in every camera (more or less) you should not be worried about this problem anyway.
Nobody can give you a sensor with 100% perfect pixels so that's why our modern cameras can map them and know what to do with them.
I always use google if something is not ML related because it is either something about the technology of our cameras or how something works.
As i said, ML is for advanced users, there was people asking about ML and they didn't even know how to update the firmware in their cameras, as an example.
Nothing wrong with them, sure, but an advanced users should at least know that dead pixels are inevitable. Then, you can talk about what ML did to make them come back.

[rambler]

It may happen when you clear camera settings.

If "it" (clearing some parts of re-map tables) actually ever happens after "clear settings", then that's clearly a bug on canon side. Is there any solid evidence?

[Pyrofer]

Exactly.

I didn't manually clear any settings.

All I did was change the version of ML on my memory card and the dead pixels appeared.
This is obviously sub-optimal.

Whatever the reason for them being there to start with, I want to know why changing ML version made them re-appear.

[Embosser]

I confirm. Hot pixels came during FPS override with slow shutter speed. Them dissapeared later, but footage was spoiled.
Canon 60D.