Dead Pixels On Ccd's...

[DownUnderCamMan]

G'day Guys...

Just wondered if one of you manfacturing camera guru's could explain to me why cosmic rays & gamma rays have the effect of causing dead pixels in camera CCD's?

So does this mean that when receiving cameras from suppliers abroad you are better off waiting for the slow boat, rather than having them despatched by air?

(As more cosmic rays up there!!!)...

Sony apparently state they are not able to produce CCD's free from white/dead pixels.

Is this the same for the few other CCD manufacturers?

[Service Engineer]

All CCD's have dead pixels when manufactured, but the manufacturers usually map them out so there are ignored by ccd image processors. After time all camera ccd's develop more dead pixels and these may show up on images as most camera's do not have dead (hot) pixel mappout procedure built in.

A pixel of the CCD sensor that doesn't react to light is always black (dead pixel), while one with a short circuit always shows white (hot or stuck pixel), with the possibility that one of the 3 base colors only is dead or hot.

[boldrx]

All CCD's have dead pixels when manufactured, but the manufacturers usually map them out so there are ignored by ccd image processors. After time all camera ccd's develop more dead pixels and these may show up on images as most camera's do not have dead (hot) pixel mappout procedure built in.

All cameras have dead pixels.

The more sensitive the imaging device, the more obvious they are, particularly in low light when they are am,plified

[CameraGimp]

You are asking about blemishes that appear post manufacture and it is getting mixed it up with blemishes that occur during manufacture.

First there are the pixel failures that happen during fabrication of the CCD. These come in different flavours, mainly dark or light, as mentioned in someone else's reply and they are measured in two ways, the signal value and the position on the imager. In the white pixel failures the higher the value the more visible it is and if it is in the middle of the imager it is more likely to be seen than a dead pixel on the edge of the imager. So you allow x blemishes of y intensity in a region Z on the imager. Sony has a specification for CCD blemishes. All CCD's have blemishes or dead pixels or whatever you want to call them. Fortunately modern DSP chipsets allow camera manufacturers to mask a certain number of pixel failures. The pixel is still "dead" but the dsp inputs a value taken from the pixels around it to make it less obvious. Temperature also effects blemishes. This is why some blemishes only become visible after the camera has warmed up.

Now you ask about the effect of cosmic rays on CCD's and I want hate you for it. Yes there is proof that CCD's can develop blemishes after fabrication and it is thought that cosmic rays are the cause. This is not a joke however I'd be lying if I said I understood it. I say there is proof as I have been given a presentation on the subject by some much more wise men than me from Japan.

So a CCD manufacturer will grade his CCD's prior to shipment but there is no guarantee that they will be the same when they arrive at the customer. Is the slow boat better than a fast plane? Depends which way you go I have seen information showing routes from Japan that avoid cosmic ray hotspots! I can't believe this as I type it. I also know that research was carried out where CCDs where transported in heated containers as this was thought to reduce the possibility of post manufacture blemishes, but probably not the cheapest or the safest way to transport CCD's when part of a camera. I have never understood why a camera manufactured 5 years ago hasn't developed 100's of blemishes where as a CCD left on a shelf for two months seems to attract them. I question if sensitive imagers are more susceptible to blemishes than less sensitive ones and wonder if it is a case that blemishes are more noticeable on sensitive cameras as they are used in low light applications?

I have experienced CCD's from Sony and Sharp both had blemishes. Sharp had a considerable amount more than the Sony CCD's and as a manufacturer we reject a higher % of Sharp chips however I expect that was because we had a tighter specification limit on allowable blemishes with Sony. If we had the same spec with Sharp they would have been as good.

CG.

[DownUnderCamMan]

Thanks Camera Gimp...

Can you imagine booking a flight & asking for the route with the least cosmic rays please...

One thing...

Who would do the clever DSP chip masking of dead pixels.

Sony the chip manufacturer?

Or would this be done by the camera manufacturer?

[CameraGimp]

The masking would be done by the camera manufacturer and not the chip manufacturer. If you think about it you would have to pair a CCD with a dsp chip and keep them together through out the whole supply chain for that to work.

In my day blemish correction was done at the very end of the manufacturing process. The setup operator would drive a dot around the screen, like a simple video game, until it masked the blemishes and then save the position in the dsp. This may have changed, the SS2 DSP which came out after I left the world of CCTV was going to support automatic blemish correction. The process would have been cap the lens, tell the dsp to detect blemishes and then mask them.