View Full Version : The Dreaded Dead Pixel


Pages : [1] 2

Tom Chaney
May 26th, 2006, 03:52 AM
Hi Gang,

I was out shooting some of the very cool clouds yesterday, and when I brought the camera in and reviewed the footage on my 42" Plasma, I saw the dreaded dead pixel.

It was not there just days ago.

Can someone help me out? I would like to shoot some more this weekend.

Thanks,

Tom Chaney

email me at tom@tomchaney.com

Scott Jaco
May 26th, 2006, 07:01 AM
Make sure it's the camera and not the TV that has the dead pixel.

John Mitchell
May 26th, 2006, 09:12 AM
Take it in for service - there is a pixel masking facility in the camera's firmware. Apparently it doesn't take too long, they can generally do it while u wait.

Craig Roblewsky
May 26th, 2006, 10:13 AM
There is a fix you can perform on your own. You can contact JVC and they can send you the instructions. I just went through this last week and the built in correction worked great.

- Craig

Ken Freed JVC
May 26th, 2006, 10:53 AM
Tom, try this:

CCD cameras can develop CCD blemishes where a pixel or two can appear white in the image. These may be caused by atmospheric radiation and also by manufacturing impurities that begin to show after years of use.

This is so common on CCD cameras by all manufacturers that we all build into our pro cameras the ability to detect and correct these blemishes.

On the GY-HV100U camera this process is provided in an advanced Menu accessed while in 24p mode by holding the Focus Assist button (the one on the camera body) when you press and hold the Menu/Status button for at least 5 seconds. You must be in 24p mode for this operation.

You see an advanced menu where the second line says CAMERA1. Move the cursor down and push the shutter wheel to enter that menu and now you have a list with the sixth item being PIXEL COMPEN. Move down to it and select that item, the word CANCEL blinks. Change it to EXECUTE and push again. It takes about 10 seconds or so to perform the operation and then tells you to power off the camera. You are done. Make sure Focus Assist is now OFF and exit 24p mode if desired.

However there are a few things for you to note about this process:
1. The process only conceals blemishes that show at 0db of gain. Blemishes that only show at 3dB, 6dB, 9dB or 18dB will not be corrected. Also there is a threshold that the blemish must be above to be detected. Over time those blemishes may become greater and then show at 0db and then will conceal.
2. Because of a good reason that would take too long to explain, the process can “miss” a blemish. It is absolutely likely that you might have to run the procedure a few times to find the blemish and conceal it.
3. Warming the camera makes the blemishes brighter and easier to correct. Some people wrap the camera in a blanket or coat to help warm it up. We DO NOT recommend warming your GY-HD100U in a microwave oven.

Chad Terpstra
May 26th, 2006, 02:20 PM
I had a dead pixel right out the box. Very bright and easy-to-spot. Thankfully I fixed in using the service menu before I shot anything important.

However, I still have about 5-6 dead pixels at 12db and even 6db gain, though they are much less pronouced. Are these fixable at all? At what point does it qualify for service? I also have some SSE, but nothing that's been detrimental or noticeable too badly. I see it when I look for it at even 0db.

Antony Michael Wilson
May 26th, 2006, 02:31 PM
We had two dead pixels after 20 hours' use. I know others who also had dead pixels within the first 50 hours...

Tom Chaney
May 26th, 2006, 05:22 PM
Hey Everyone,

Thanks for the overwhelming response to my problem.

I love this place!

I hope that I can return the favors.

Tom Chaney

Tim Dashwood
May 26th, 2006, 09:13 PM
Tom, try this:

CCD cameras can develop CCD blemishes where a pixel or two can appear white in the image. These may be caused by atmospheric radiation and also by manufacturing impurities that begin to show after years of use.

This is so common on CCD cameras by all manufacturers that we all build into our pro cameras the ability to detect and correct these blemishes.

On the GY-HV100U camera this process is provided in a Service Menu accessed while in 24p mode by holding the Focus Assist button (the one on the camera body) when you press and hold the Menu/Status button for at least 5 seconds. You must be in 24p mode for this operation.

You see an advanced menu where the second line says CAMERA1. Move the cursor down and push the shutter wheel to enter that menu and now you have a list with the sixth item being PIXEL COMPEN. Move down to it and select that item, the word CANCEL blinks. Change it to EXECUTE and push again. It takes about 10 seconds or so to perform the operation and then tells you to power off the camera. You are done. Make sure Focus Assist is now OFF and exit 24p mode if desired.

However there are a few things for you to note about this process:
1. The process only conceals blemishes that show at 0db of gain. Blemishes that only show at 3dB, 6dB, 9dB or 18dB will not be corrected. Also there is a threshold that the blemish must be above to be detected. Over time those blemishes may become greater and then show at 0db and then will conceal.
2. Because of a good reason that would take too long to explain, the process can “miss” a blemish. It is absolutely likely that you might have to run the procedure a few times to find the blemish and conceal it.
3. Warming the camera makes the blemishes brighter and easier to correct. Some people wrap the camera in a blanket or coat to help warm it up. We do not recommend warming your GY-HD100U in a microwave oven.

Thank you Ken. We have been keeping a lid on this procedure at hdvinfo.net for the past 9 months. Mostly because the JVC engineer who fixed my pixels asked me not to share the information on the net. I'm glad someone from JVC has finally shared the procedure - and we will let it stick from now on.

However, we should warn with a disclaimer that this procedure performed by the end user is at his/her own risk, and any damages caused (filling the pixel map memory) will most likely not be covered under warranty.

With that said, I'd like to make this procedure part of a sticky.


BTW, Atmospheric radiation can cause it, but I think the x-rays at airports are also to blame. I seem to get a new one everytime one of my cameras travels.

Scott Jaco
September 7th, 2006, 03:46 AM
I just ran this process. It worked quite well. I had to run it twice to fix the pixel. The first time only partially fixed the problem, the second time seemed to do the trick.

I gave my camera a good workout afterwards, cranked the gain to +18db and pointed it in dark places, put the lens cap on to see if I could get the pixel to show but so far no problems!

I would recommend plugging the camera into an TV when performing this process so you can do a decent comparison.

David Scattergood
September 7th, 2006, 04:03 AM
^^ Indeed - my dead pixels were not initially evident through the viewfinder - composite out to the TV showed them up quick enough mind.
Worked first time for me...saved me a lot of hassle sending the camera back.
Thanks folks.

Keith Nealy
September 7th, 2006, 02:55 PM
Carl, doesn't this seem like this should be a quality control issue at JVC?

Why don't they check for this before they send them out the door and then fix it before the consumer opens the box.

This just puts a lot of bad press out there that this camera has troubles from day one.

I for one hope they get this all resolved with the 250 as I plan to get one as soon as it ships.

Aloha,

Keith

PS - any news on that? it's been too quiet.

John Vincent
September 7th, 2006, 03:14 PM
Well, not really. As Tim and others have said, it can be caused by x-rays, et al. This can (and will) happen post factory shipping.

john
evilgeniusentertainment.com

Carl Hicks
September 7th, 2006, 10:34 PM
Carl, doesn't this seem like this should be a quality control issue at JVC?

Why don't they check for this before they send them out the door and then fix it before the consumer opens the box.

This just puts a lot of bad press out there that this camera has troubles from day one.

I for one hope they get this all resolved with the 250 as I plan to get one as soon as it ships.

Aloha,

Keith

PS - any news on that? it's been too quiet.


Keith,

Stuck pixels are a fact of life in the technology of CCD cameras, and that's been the case since the late 80's, when CCD cameras first hit the market, regardless of brand or model. In a former position, I sold cameras made by JVC, Panasonic, Sony, and Ikegami, and I encountered pixel issues with all of them from time to time.

All manufactures check for stuck pixels as part of the QC process. Many times, pixel problems don't show up until weeks, months, or years after the camera is built. As Tim and others have correctly noted, the issue is often brought on by extended exposure to atmospheric radiation - gamma rays I think. For example, when you fly with a CCD camera at high altitudes for long periods of time - like an overseas flight for example - you expose the camera to a substantial increase of exposure to atmospheric radiation.

JVC Pro cameras have a service feature that allows the owner to mask stuck pixels. Keep in mind as Tim said, you don't want to over use this feature, as there is a limit to what it can do. Also remember that a stuck pixel might not show up in the viewfinder, but will usually show up on a HD monitor. So, use a good, real HD display to check out the camera.

Regarding the GY-HD250U: It has been discussed many times on this forum. Just do a search. It is targeted for Oct. - Nov. delivery.

Regards, Carl

Lee Alford
September 8th, 2006, 10:53 AM
it just seems to be of an issue with the hd1 and not the other cameras. they never talk about having problems with it while the hd1 owners seems to discuss a bad pixel on a weekly basis.

Antony Michael Wilson
September 8th, 2006, 11:20 AM
I appreciate that dead pixels can happen to all CCDs on any video camera and that it is just a limitation of the technology and that there are things one can do to avoid it. I also appreciate that it is an easy fix on the HD100 (as long as you never need to use gain). However, I'd have to agree that dead pixels do seem to happen with great frequency and regularity on this camera, from what I can tell from this board. I've already had two dead pixels on the HD100 and never had any on any other camera I have ever owned or rented in the last 10 years. Maybe I've just been lucky...

All in all, I think it might actually be a fair criticism of an otherwise great camera.

Scott Jaco
September 8th, 2006, 01:20 PM
I also appreciate that it is an easy fix on the HD100 (as long as you never need to use gain).

Actually I've cranked the gain to +18db without seeing the bad pixel. I think what they are saying is that the bad pixel has to be visable at 0db in order for the software to "mask" it. Once the pixel is masked, it shouldn't be visable at any gain setting, since it's been disabled. At least thats what it would seem.

I have a consumer JVC with a couple of bad pixels also. It's a high-end consumer handycam but it doesn't have the "mask" feature.

I do think you are right, I never see anyone complain about the Sony Z1u's having bad pixels.

Antony Michael Wilson
September 8th, 2006, 03:48 PM
That's interesting. I masked our bad pixel in the well-known fashion and it is always visible again with any gain added. Maybe I have a lemon?

John Vincent
September 9th, 2006, 04:24 AM
It could be another pixel that was next to the original dead one. Try the procedure one more time after the camera is heated up. If it remains, I'd ship it in for inspection/repair, with a list of all the procedures you've tried up to now. Obviously, you can't use the camera w/ a dead pixel...

john
evilgeniusentertainment.com

Antony Michael Wilson
September 9th, 2006, 05:04 AM
I've tried the procedure several times. It's masked fine without gain but the gain brings it back. I was sure that this is normal and I seem to remember Carl Hicks saying that this was the case. Either way, I'm not that bothered because I use gain very rarely...

Carl, if you're reading this, what is the official line on masked pixels with gain up?

Bill Edmunds
September 10th, 2006, 08:42 AM
1. The process only conceals blemishes that show at 0db of gain. Blemishes that only show at 3dB, 6dB, 9dB or 18dB will not be corrected. Also there is a threshold that the blemish must be above to be detected. Over time those blemishes may become greater and then show at 0db and then will conceal.
2. Because of a good reason that would take too long to explain, the process can “miss” a blemish. It is absolutely likely that you might have to run the procedure a few times to find the blemish and conceal it.
3. Warming the camera makes the blemishes brighter and easier to correct. Some people wrap the camera in a blanket or coat to help warm it up. We DO NOT recommend warming your GY-HD100U in a microwave oven.
If there is a blemish that cannot be fixed with this solution, what are we supposed to do? Send it back to JVC? What solutions are there for blemishes that show up under gain settings of 6db and over?

Scott Jaco
September 10th, 2006, 02:05 PM
Well, have you tried the "mask" function yet?

John Mitchell
September 10th, 2006, 09:45 PM
BTW, Atmospheric radiation can cause it, but I think the x-rays at airports are also to blame. I seem to get a new one everytime one of my cameras travels.

Not sure about the X-rays but someone told me that travelling in a jet you are exposed to far higher levels of gamma radiation than standard which could also explain the dead pixels...

Richard Hunter
September 10th, 2006, 11:35 PM
Not sure about the X-rays but someone told me that travelling in a jet you are exposed to far higher levels of gamma radiation than standard which could also explain the dead pixels...

Hi John. It's usually cosmic rays that cause problems at high altitudes. Alpha particles can penetrate aircraft skin and equipment housings and take out individual logic gates in memory circuits or CCD cells in cameras. If the damage is temorary (soft error) it is called Single Event Upset, if it's permanent (hard error) then it's Single Event Burnout.

Xrays and gamma rays are not as localised. If they were powerful enough to damage the CCD, they would be more likely to wipe the whole sensor rather than just kill single pixels.

Marc Brewer
October 8th, 2006, 04:56 PM
Helo all,

I've tried "the procedure" but the cursor skips over the pixel masking funtion strait to page back. I can see the pixel in 0 db but of course it shows more in 3 db gain to 6 and above...

ideas anyone?

MB

Antony Michael Wilson
October 8th, 2006, 05:04 PM
You need to be in 24p HDV mode for the pixel mask function to work. Have you checked this?

Carl Hicks
October 9th, 2006, 08:14 AM
If there is a blemish that cannot be fixed with this solution, what are we supposed to do? Send it back to JVC? What solutions are there for blemishes that show up under gain settings of 6db and over?


Bill,

Have you discussed this with your reseller and / or JVC District Sales Manager?

Carl

Miko Kirchoff
October 20th, 2006, 05:59 PM
I found I had the so called,"DEAD PIXELS" in my brand new JVCGYHD100U and believe me, wasn't too thrilled about it. The insult to my injury was the Express Video Supply Guys not giving me much compassion after spending lots of dough with them over the years. I won't purchase from them anymore. No service after the sale.

Meanwhile, I shot a local car commercial with new camera and when I went to post, found this crazy little white dot on the monitor. I called JVC and after a long, circuitous trip thru an endless maze of recorded message prompts, got a guy in Jersey who showed me the super secret menu allowing the camera to mask out the problem pixels. It worked, or so I thought, 'til I shot a lovely lady against a green screen background on a high res monitor. The dot came back and almost ruined the shoot. I had to frame her in such a way as to avoid having it dance across her face when she moved. Good thing it was Green Screen project or I would have been in some sh**!

I went back to the JVC expert and tried the masking process again. So far, it's been working. I wrote down everything the Tech said step by step (just in case) in my owners manual. If you'd like the info, I'll relay it via e-mail to you.

Meanwhile, I've had this rig over 4 months now and have been shooting with it almost daily, and really don't think much of it.

Miko Kirchoff
skyrocketvideo@yahoo.com

Mark Silva
October 23rd, 2006, 11:19 AM
I'd like to ask, out of curiosity, Carl, how many times can the masking function be run before it won't work any longer for masking pixels?

I've heard it can only be done a certain number of times before the masking buffer fills and is unable to function.

William Hohauser
November 8th, 2006, 05:34 PM
On the GY-HV100U camera this process is provided in a Service Menu accessed while in 24p mode by holding the Focus Assist button (the one on the camera body) when you press and hold the Menu/Status button for at least 5 seconds. You must be in 24p mode for this operation.

You see an advanced menu where the second line says CAMERA1. Move the cursor down and push the shutter wheel to enter that menu and now you have a list with the sixth item being PIXEL COMPEN. Move down to it and select that item, the word CANCEL blinks. Change it to EXECUTE and push again. It takes about 10 seconds or so to perform the operation and then tells you to power off the camera. You are done. Make sure Focus Assist is now OFF and exit 24p mode if desired.


I am unable to get the advanced menu at all. The camera is in 24p mode, no tape and on AC power. It has the firmware upgrade from JVC.

Are the buttons pressed simultaneously or in succesion? If in succession, is the Focus Assist button first or the Menu button? When does the 5 seconds start?

Any help would be appreciated, a dreaded dead pixel has appeared.

William Hohauser
November 8th, 2006, 08:16 PM
I have answered my own question and have found that the instructions need a few small additions:

1) Make sure the camera is in 24p.

2) Only one Focus Assist button will work in this operation. Use the button next to the Peaking adjust knob on the left side towards the front of the camera. The Focus Assist button on the handle, while technically on the camera, will not work nor will the RET button on the lens if it's changed to Focus Assist.

3) Turn off Focus Assist if it's on. Press and hold the button then press and hold the Menu/status button.

There you go - Service menu!

Carl Hicks
November 8th, 2006, 11:02 PM
I am unable to get the advanced menu at all. The camera is in 24p mode, no tape and on AC power. It has the firmware upgrade from JVC.

Are the buttons pressed simultaneously or in succesion? If in succession, is the Focus Assist button first or the Menu button? When does the 5 seconds start?

Any help would be appreciated, a dreaded dead pixel has appeared.

William, you press and hold all three buttons at the same time. Hold all three pressed until you see the service menu come up.

Oh, I see now that you figured it out!

Regards,

Mark Silva
November 9th, 2006, 11:04 AM
Carl how many times can this be performed before the masking buffer fills?

John Vincent
November 9th, 2006, 01:33 PM
Carl how many times can this be performed before the masking buffer fills?

That is a great question - I've already had to do the process twice... And I have not put very many hours on the camera. I also wonder if there is a way to empty/reset the buffer. Not something I'm losing too much sleep about, but the warning to "not do it too many times" is a bit vague.

Carl?

john
evilgeniusentertainment.com

Carl Hicks
November 9th, 2006, 11:31 PM
That is a great question - I've already had to do the process twice... And I have not put very many hours on the camera. I also wonder if there is a way to empty/reset the buffer. Not something I'm losing too much sleep about, but the warning to "not do it too many times" is a bit vague.

Carl?

john
evilgeniusentertainment.com

I would not worry much about it. In the early days of CCD error correction, the buffer was very small, so you did have to be careful not to over do it. But with the current technology, the memory is much larger. A product engineer I spoke to did not have an exact number, but said it's at least 100, maybe more.

Mark Silva
November 9th, 2006, 11:55 PM
I would not worry much about it. In the early days of CCD error correction, the buffer was very small, so you did have to be careful not to over do it. But with the current technology, the memory is much larger. A product engineer I spoke to did not have an exact number, but said it's at least 100, maybe more.


excellent!

I REALLY appreciate that feature btw.

Marc Colemont
November 10th, 2006, 02:07 AM
Just to be sure. I guess the lens cap must be on to perform this function?

Werner Wesp
November 10th, 2006, 05:50 AM
No other warranty will be void after performing this yourself? The european service guy told me here that after a certain number of times a hardware change was needed (don't ask me od what exactly), so I'm a bit weary of doing these things myself...

Carl Hicks
November 10th, 2006, 07:21 AM
Just to be sure. I guess the lens cap must be on to perform this function?

Not necessary to have lens cap on.

The key thing is to make sure the camera isfully warmed up. An hour turned on prior to running the proceedure is a good idea.

Carl Hicks
November 10th, 2006, 07:22 AM
No other warranty will be void after performing this yourself? The european service guy told me here that after a certain number of times a hardware change was needed (don't ask me od what exactly), so I'm a bit weary of doing these things myself...


Here in the U.S. there has been no warranty issues wih this proceedure that I know of. I can't speak for other country's policies.

John Vincent
November 10th, 2006, 12:19 PM
Carl - thanks for the info! An hour, huh? Good to know... Does jacking up the gain help in any way?

john
evilgeniusentertainment.com

Carl Hicks
November 12th, 2006, 08:31 PM
Carl - thanks for the info! An hour, huh? Good to know... Does jacking up the gain help in any way?

john
evilgeniusentertainment.com

I don't think so.

Skye Fitzgerald
April 20th, 2007, 12:25 PM
Hey everyone - just finished up reading this thread...thought I'd share my experience on this issue. I have an HD100 (still under warranty) that I've had repeated, and uncommonly bright bad pixels develop with on a regular basis.

I fly a lot and so have been very careful to keep my eye on this issue and
at first I thought the highly touted pixel compensation menu was the solution. However, this has not been the case...

I have now had three subsequent shoots marred by significant dead pixel issues on this camera in fairly quick succession - one in Antarctica, one in Cambodia, and one in Oregon.

I should note that in each instance prior to pixel loss:

1. The pixel compensation program was run on the camera AFTER the camera was fully warmed up for over two hours.
2. The pixel compensation software was run AFTER each flight I made with the camera.
3. The problem has developed within 3-7 shooting days of the start of each shoot.

Of course, if I could I would simply shoot everything with a field monitor to be sure dead pixels aren't developing....but as we all know, this isn't always possible.

My thoughts are that a shooter shouldn't be expected to have to run pixel compensation software prior to, and during each shoot, never knowing exactly when - if at all - bad pixels have, or are developing. To me - regardless of what manufacturers may say to protect their product - this is a particular weakness in this particular camera. I've NEVER had this problem except with JVC cameras (and I have two at the moment - though at this point I may never buy another).

My current hope is that JVC has been able to address this weakness in subsequent models as it seems an inherent weakness in the HD100...at this point I'm just trying to figure out what to do with a new camera I can't trust.

-Skye

David Vahey
April 21st, 2007, 06:43 PM
I was freaking out about my lit pixel and BOOM that dang procedure fixed it! I owe you guys my career!

Thanks JVC.

dave
PRODUCER

www.mediaflycreative.com

John Mitchell
April 22nd, 2007, 09:30 PM
I should note that in each instance prior to pixel loss:

1. The pixel compensation program was run on the camera AFTER the camera was fully warmed up for over two hours.
2. The pixel compensation software was run AFTER each flight I made with the camera.
3. The problem has developed within 3-7 shooting days of the start of each shoot.

-Skye

(2) I don't think the pixel compensation program was ever designed to be run so regularly - you've probably filled up the memory. You are only EVER supposed to run the program when you discover a dead pixel (and originally they made you send it in to JVC to do it).

The problem you're relating (3) seems different to standard dead pixels - I'd be returning the camera for a replacement as soon as you can, or if that is not possible at least sending it in to JVC for evaluation. Hope all works out for you.

Skye Fitzgerald
April 22nd, 2007, 11:48 PM
(2) I don't think the pixel compensation program was ever designed to be run so regularly - you've probably filled up the memory. You are only EVER supposed to run the program when you discover a dead pixel (and originally they made you send it in to JVC to do it).

The problem you're relating (3) seems different to standard dead pixels - I'd be returning the camera for a replacement as soon as you can, or if that is not possible at least sending it in to JVC for evaluation. Hope all works out for you.

John - thanks for your thoughts on my posting...I agree with your perspective that I shouldn't have to run the pixel compensation program on a regular basis - and in fact, I only began doing so after consultation with JVC about my repeated, and regularly bad pixels that continued to rapidly develop after every flight. (And now, on nearly every shoot regardless of whether a flight is involved or not).

And yes, the camera is currently with JVC, again, for evaluation. I'll let you know what Doug Mullin's team at JVC says is the cause, and what, if any solution they suggest.

-Skye

Skye Fitzgerald
April 25th, 2007, 12:35 AM
Hey John - just wanted to let you know that I heard back from the JVC techs today regarding the (latest) evaluation on my repeated bad pixels on my HD100:

The JC tech said the entire Optic Block needs to be replaced on the camera, which they are doing. I just hope I don't have the same problem in the remaining months on my warranty...

-Skye

Claude Mangold
April 25th, 2007, 05:02 AM
As posted in another thread, we had a dead pixel (or several) during an elaborate short film shoot in March. The camera was rental. I've heard of pixels popping in all HD cameras but mostly in JVC. We returned the camera. The rental firm in Geneva was not supportive, worse, they didn't check the camera before.

John Mitchell
April 25th, 2007, 10:48 PM
Hey John - just wanted to let you know that I heard back from the JVC techs today regarding the (latest) evaluation on my repeated bad pixels on my HD100:

The JC tech said the entire Optic Block needs to be replaced on the camera, which they are doing. I just hope I don't have the same problem in the remaining months on my warranty...

-Skye

Skye - good news. I'm not seeing excessive reports of dead pixels (they are a fact of life on all CCD cameras), and your situation definitely sounded unusual. Fingers crossed this solves the problem for you.

Scott Shuster
June 11th, 2007, 07:34 AM
I'm thinking we should run pixel comp. immediately before every shoot, just to make certain we are 'up to the second' ensuring that a new 'dead pixel' has not occurred. Is there any aspect of the pixel compensation processing that could negatively affect the performance of the camera if done too often? Or is this as benign as white-balancing, with no danger to the camera's performance no matter how often it is done?