FAQ FAQ  Forum Search   Events   Register Register  Login Login

TP: What to do about Hot Pixels

Page  12>
Author
keesey View Drop Down
Groupie
Groupie

Joined: 03 March 2007
Location: United Kingdom
Status: Offline
Posts: 89
Post Options Post Options   Quote keesey Quote  Post ReplyReply Direct Link To This Post Topic: TP: What to do about Hot Pixels
    Posted: 26 June 2007 at 03:16
I have just carried out a 30 second exposure this evening and have found there to be 27 hot pixels, as far as I know the camera should reset these every month and so that means all these have been created in the last 26 days?

How many hot pixels should we expect before a replacement CCD?

Any details on this issue are appreciated

Mark


Edited by brettania - 03 July 2007 at 12:37
A100 | Minolta 28-85/3.5-4.5 | Minolta 50/1.7 | Minolta 70-210/4 | Minolta 75-300/4.5-5.6 | Sigma 170-500 APO | Sony 18-70/3.5-5.6 | Teleplus MC4
 



Back to Top
brettania View Drop Down
Admin Group
Admin Group
Dyxum factotum

Joined: 17 July 2005
Country: New Zealand
Location: Auckland
Status: Offline
Posts: 20650
Post Options Post Options   Quote brettania Quote  Post ReplyReply Direct Link To This Post Posted: 26 June 2007 at 04:24
Originally posted by keesey keesey wrote:

I have just carried out a 30 second exposure this evening and have found there to be 27 hot pixels, as far as I know the camera should reset these every month and so that means all these have been created in the last 26 days?


You can try resetting the date to next month, bang off a couple of pics, then reset the date.

Originally posted by keesey keesey wrote:


How many hot pixels should we expect before a replacement CCD?


Never seen any info on that -- presume the problem has not yet arisen with even the 7D.

Edited by brettania - 26 June 2007 at 04:25
Back to Top
davesmith View Drop Down
Groupie
Groupie

Joined: 19 September 2006
Location: Australia
Status: Offline
Posts: 139
Post Options Post Options   Quote davesmith Quote  Post ReplyReply Direct Link To This Post Posted: 26 June 2007 at 06:59
The fact there is a built in reset at the beginning of the month would suggest it's pretty common, but manageable.
Cheers,
Dave
My Galleries
Back to Top
harveyzone View Drop Down
Senior Member
Senior Member

Joined: 15 September 2006
Country: United Kingdom
Location: Worcestershire
Status: Offline
Posts: 918
Post Options Post Options   Quote harveyzone Quote  Post ReplyReply Direct Link To This Post Posted: 26 June 2007 at 11:19
Originally posted by brettania brettania wrote:

Originally posted by keesey keesey wrote:


How many hot pixels should we expect before a replacement CCD?


Never seen any info on that -- presume the problem has not yet arisen with even the 7D.


How would anyone know how many hot pixels they have in total if they are mapped out every month? One would have to do a check on the last day of every month and keep a note. I am guessing that not many people do that.

--
Tom.

Edited by harveyzone - 26 June 2007 at 11:20
Back to Top
madcat207 View Drop Down
Senior Member
Senior Member

Joined: 26 January 2007
Country: United States
Location: NM
Status: Offline
Posts: 1716
Post Options Post Options   Quote madcat207 Quote  Post ReplyReply Direct Link To This Post Posted: 01 July 2007 at 19:27
Originally posted by harveyzone harveyzone wrote:

Originally posted by brettania brettania wrote:

Originally posted by keesey keesey wrote:


How many hot pixels should we expect before a replacement CCD?


Never seen any info on that -- presume the problem has not yet arisen with even the 7D.


How would anyone know how many hot pixels they have in total if they are mapped out every month? One would have to do a check on the last day of every month and keep a note. I am guessing that not many people do that.

--
Tom.


Well, I can be a test subject, in a sense. I was looking at some birthday pictures i too just last night, and noticed 5 hot pixels, in ONE spot (reallllly annoying, beacuase they are in the middle too), in a + shape.

Do you have to do anything special to have the camera map these pixels (i mean, how would it know they are hot anyway?) Given that i am photographing a wedding this next weekend, i would like to not have shots with a big white + on em (i know, i can do a dark frame subtraction, but that will get real old after 300+ pictures...)

I am going to go power up my camera, and see if this is true. To be honest, i don't believe the camera does do that kind of mapping; i will report back either way..
Back to Top
madcat207 View Drop Down
Senior Member
Senior Member

Joined: 26 January 2007
Country: United States
Location: NM
Status: Offline
Posts: 1716
Post Options Post Options   Quote madcat207 Quote  Post ReplyReply Direct Link To This Post Posted: 01 July 2007 at 20:23
Well, when I am wrong, I am willing to admit it....








.. i was wrong! At first, when i did a picture with the cap on, the hot pixels were still there. However, I remembered this thread also suggested setting the date ahead manually, and low and behold, it worked!

Here is a black frame before pushing it ahead:


And here is a black frame after pushing it ahead.


Well, in the end, i am frustrated that some of my pictures have a big bright dot on them, but I am happy to know that the camera actually does fix itself. In fact, i even was able to see it - after setting the date forward, powering off and then back on, the camera actually took longer to boot, which does tell me it was actually remapping! Thanks for the advice guys!
 



Back to Top
brettania View Drop Down
Admin Group
Admin Group
Dyxum factotum

Joined: 17 July 2005
Country: New Zealand
Location: Auckland
Status: Offline
Posts: 20650
Post Options Post Options   Quote brettania Quote  Post ReplyReply Direct Link To This Post Posted: 02 July 2007 at 07:36

You're welcome.

-
Back to Top
keesey View Drop Down
Groupie
Groupie

Joined: 03 March 2007
Location: United Kingdom
Status: Offline
Posts: 89
Post Options Post Options   Quote keesey Quote  Post ReplyReply Direct Link To This Post Posted: 02 July 2007 at 11:27
I have finally got round to testing this by putting the date forward on 2 occasions now so the first shot is taken in July, the second in August and the 3rd in September. I have noticed that upon moving the date forward the camera takes several times longer to turn itself off which I am assuming is the camera recalibrating.

The screen shot below shows the hot pixels taken over the 3 months. Sorry about the width but on resizing the pixels disappear.



Also is noise reduction meant to work in all modes? As it only seems to do anything in full automatic. When in full automatic the total number of hot pixels does reduce slightly to 22. There are also strange red blobs in my images which are 12 pixels (4x3) in size and there are 6 of these present in each frame. The red sections look like this

A100 | Minolta 28-85/3.5-4.5 | Minolta 50/1.7 | Minolta 70-210/4 | Minolta 75-300/4.5-5.6 | Sigma 170-500 APO | Sony 18-70/3.5-5.6 | Teleplus MC4
Back to Top
mikethelaserman View Drop Down
Senior Member
Senior Member

Joined: 23 November 2005
Country: United Kingdom
Location: United Kingdom
Status: Offline
Posts: 292
Post Options Post Options   Quote mikethelaserman Quote  Post ReplyReply Direct Link To This Post Posted: 03 July 2007 at 00:44
As each month goes by and the hot pixels are mapped out, eventually the sensor must run out of useable pixels.

If around 20 pixels per month was typical, then loosing 1000 or so over four years might be OK if they were well spread out. Bad news if they were grouped in the centre of the frame as in Madcap's case above.

In theory, people that fly their cameras long haul frequently should loose pixels faster because of the increased solar radiation (gamma rays is it?) which affects semiconductors (and fogs acrylic eye glass lenses).

I have never seen any info on expected sensor life and the normal failure rate calculations for semiconductors are not appropriate for individual pixels. I'd be very interested to find out.
Back to Top
ClevaTreva View Drop Down
Groupie
Groupie

Joined: 02 July 2007
Location: United Kingdom
Status: Offline
Posts: 60
Post Options Post Options   Quote ClevaTreva Quote  Post ReplyReply Direct Link To This Post Posted: 03 July 2007 at 01:21
What kills semi conductors faster are the following:

A faulty circuit at birth (narrow path becomes narrower and breaks). Should show itself up on a CCD by being in clusters or lines.

Static. Not often a problem for things like computers because they are inside a shielded box. If your camera body were metal ... but it isn't. You are the main source of static (unless you are into doing work in thunder storms or humid weather). Obviously, any metal part you touch will take the static, unless you ground yourself first.

Overheating. Not a likely problem for a camera, but cutting down the load on the camera's processor is a good thing (like switching off as much as the auto stuff as you are not using).

Rapid on/off power ups.

Personally, I have most of the goodies off (ssr, nr, af etc). I find it better to shoot raw and let my software (I use Paintshop Pro X and Artizen) do the file conversion and do the nr by hand.
Back to Top
mikethelaserman View Drop Down
Senior Member
Senior Member

Joined: 23 November 2005
Country: United Kingdom
Location: United Kingdom
Status: Offline
Posts: 292
Post Options Post Options   Quote mikethelaserman Quote  Post ReplyReply Direct Link To This Post Posted: 03 July 2007 at 12:06
ClevaTreva - Agreed that the normal life factors for semiconductor devices will affect CCD sensors too.
The camera sensor is a special case however - the sensor cells are working close to the theoretical limits for leakage current. (Dark current leakage in the 6MP KMD7 and KMD5 sensor is claimed to be 0.02 electrons per second - about 1 electron per minute!). Anything that causes a permanent and significant increase in this leakage current will generate a hot pixel.
Gamma rays will punch through any material and in semiconductors, may leave behind a conductive trail. In a "normal" semicoductor devices (CPU, DRAM etc.) this would not normally alter the functionality, though it is usual for electronics in satellites etc. to be "rad-hardened".
Not possible to rad-harden a photo sensor because you want light to get at it!

There is a guy in the US who recons he can judge a pilot's or astronaught's flying hours by the ammount of fogging in acrylic eye glass lenses. Each gamma particle leaves a trail in the acrylic and the cumulative effect is a visible fogging of the lens. We do get gamma rays at ground level but at far less density/frequency.

I did fly my KM5D to Italy and back with no adverse effects, but it would be really interesting if anyone who regularly flies long-haul (i.e. HIGH - not light aircraft stuff) has seen an increase in the number of hot pixels.
Back to Top
ClevaTreva View Drop Down
Groupie
Groupie

Joined: 02 July 2007
Location: United Kingdom
Status: Offline
Posts: 60
Post Options Post Options   Quote ClevaTreva Quote  Post ReplyReply Direct Link To This Post Posted: 03 July 2007 at 12:29
Originally posted by mikethelaserman mikethelaserman wrote:

ClevaTreva - Agreed that the normal life factors for semiconductor devices will affect CCD sensors too.
The camera sensor is a special case however - the sensor cells are working close to the theoretical limits for leakage current. (Dark current leakage in the 6MP KMD7 and KMD5 sensor is claimed to be 0.02 electrons per second - about 1 electron per minute!). Anything that causes a permanent and significant increase in this leakage current will generate a hot pixel.
Gamma rays will punch through any material and in semiconductors, may leave behind a conductive trail. In a "normal" semicoductor devices (CPU, DRAM etc.) this would not normally alter the functionality, though it is usual for electronics in satellites etc. to be "rad-hardened".
Not possible to rad-harden a photo sensor because you want light to get at it!

There is a guy in the US who recons he can judge a pilot's or astronaught's flying hours by the ammount of fogging in acrylic eye glass lenses. Each gamma particle leaves a trail in the acrylic and the cumulative effect is a visible fogging of the lens. We do get gamma rays at ground level but at far less density/frequency.

I did fly my KM5D to Italy and back with no adverse effects, but it would be really interesting if anyone who regularly flies long-haul (i.e. HIGH - not light aircraft stuff) has seen an increase in the number of hot pixels.


And I learn more :)
Back to Top
tomaspinall View Drop Down
Newbie
Newbie

Joined: 03 July 2007
Location: United Kingdom
Status: Offline
Posts: 15
Post Options Post Options   Quote tomaspinall Quote  Post ReplyReply Direct Link To This Post Posted: 03 July 2007 at 21:30
Hey guys, I am completely new here and this is now just my second post after posting the obligatory Hello!

I just want to say a massive thankyou to the guys in this thread that helped me get rid of the horrible white + on my CCD! I was worrying about it for the last week and just happened to find the thread on here the other day. Just tried forwarding the time on and back and lo and behold the hot pixel has gone so massive thanks for your advice!

(I have been reading the forums for a while now but never joined, had to join to say thanks tho, the advice is much appreciated)
Back to Top
Sharpshooter View Drop Down
Newbie
Newbie

Joined: 19 January 2007
Country: United States
Location: Iowa
Status: Offline
Posts: 15
Post Options Post Options   Quote Sharpshooter Quote  Post ReplyReply Direct Link To This Post Posted: 06 July 2007 at 17:00
I also have a problem with hot pixels, which I noticed very prominently in the fireworks pictures I shot the other night. I tried setting the date ahead to August 1, then Sep 1, taking a test shot after each, with no effect. Because I'm using a 5D, is it possible that this 'trick' doesn't work? Would that feature be left out of the 5D and reincorporated into the A100, or did I just do it wrong?

On the other hand, noise reduction removed ALL the hot pixels. But NR takes extra time, which is often inconvenient, and doesn't work in continuous advance drive mode which I needed for fireworks.

Any thoughts or suggestions?

Thanks,

Sharpshooter
Back to Top
Dyxum main page >  Forum Home > Dyxum Community > Knowledge Base Page  12>

Forum Jump Forum Permissions View Drop Down



This page was generated in 0.078 seconds.

Monitor calibration strip

Dyxum.com - Home of the alpha system photographer

In memory of Cameron Hill - brettania

Feel free to contact us if needed.