DEAD PIXELS

Hey guys,

Very excited as I just received my camera yesterday. Just noticing a lot of dead pixels, red and blue in footage (Prores Film 800 at night). They are very noticeable when colour corrected, I count 10-15 red and blue spots, there could be more. I can also even see 1 dead pixel on the camera's LCD. Has anyone else noticed this problem. I am not sure what firmware I have, whatever was shipped with the camera, my Canon 24-70mm 2.8 L aperture works if that is any indicator.

Any help would be greatly appreciated as I don't want to have to send my new toy back to the factory especially after such a long wait till Christmas!

Cheers,

Sam

I have a dead pixel on the LCD monitor too.
A strobe effect on the screen and on the record in low light condition.
À flange effect that make the infinite on my Lens , out of focus.
À so noisy fan, that I almost can't ear the voice of someone speaking in front of the camera. (Never eard that even in the 70's camescope)
AND for finish I should have handle with the cam that I Never get With any excuses.
(

A strobe effect on the screen and on the record in low light condition.

Is this strobe effect present in your ProRes footage as well or only the RAW footage? Also, what do you grade your footage with?

I have a dead pixel on the LCD monitor too.
A strobe effect on the screen and on the record in low light condition.
(

change the Shutter speed


also, i had dead pixels on lcd, but not in footage. I say i had cos when i turned on the camera (same scene, same light) the dead pixel gone

Uploaded with ImageShack.us

You might have to click on the image and look at it full screen in ImageShack to see the pixels

Grrrr....

This camera is very young, unavailable and has o lot of different problems!!!

Hope that Blackmagic will answer us very quickly for all "lucky" customer who received their camera with problems:

___DEAD PIXELS on sensor or Screen___
___INFINITY focus lenses___
___VERY NOISI FAN___
___STROBE_in low light___

Thanks

Uploaded with ImageShack.us

You might have to click on the image and look at it full screen in ImageShack to see the pixels

I must be blind?? is it just above the white remote?

Here a sample with the cap on the camera, shutter angle 180°.
Settings in Camera RAW:
Export from PS with Camera RAW settings:
RAW-Frame from the camera:http://file.blaufilm.de/bmcc_forum_frame.dng

It's not realy visible in normal conditions... Looks much better than the noise in EOS 5d MKII RAW footage. Nice grain instead of colored noise and lines.
But the sensor noise on the upper half is different from the bottom half. Can someone please share some footage with lens cap on?

Having the same issue. The dead pixels are not on the LCD as they don't appear when displaying the menus. However, they appear when shooting. The problem pixels would move depending on Gamma, ISO, and Record Formats. Only had a short time to play with the BMCC last night, and didn't have media available at the time, but plan to shoot tests today. Also going to try a firmware upgrade to see if that clears up the issue.

Due to the fact that the pixels change position, and appear to flash on and off at times, I'm hoping it's merely a software issue that will be addressed with firmware. I'm worried that it could be voltage leaks between photosites, or another issue with the sensor itself. If so, Blackmagic has a serious problem.

Is there any method of black balancing this camera?

Thanks for the post. You should circle the problem pixels when you upload a still.

I think these are more likely to be hot pixels than dead pixels.

Without seeing the footage it sounds like you are under exposing. The camera needs to be "exposed to the right" to get the best image out of the sensor. Under exposing will bring out issues like hot pixels and flickering.

Thanks for the post. You should circle the problem pixels when you upload a still.

I think these are more likely to be hot pixels than dead pixels.

Yes, i saw the img and he has blue/red pixels in the darker zones. Maybe im wrong but these hot pixels can be fixed with firmware.. Right?

A strobe effect on the screen and on the record in low light condition.

Is this strobe effect present in your ProRes footage as well or only the RAW footage? Also, what do you grade your footage with?

Yes on the two of them. My cam goes in RMA very soon.

I have a dead pixel on the LCD monitor too.
A strobe effect on the screen and on the record in low light condition.
(

change the Shutter speed


also, i had dead pixels on lcd, but not in footage. I say i had cos when i turned on the camera (same scene, same light) the dead pixel gone

I dis but still the same.

A strobe effect on the screen and on the record in low light condition.

Is this strobe effect present in your ProRes footage as well or only the RAW footage? Also, what do you grade your footage with?

Yes on the two of them. My cam goes in RMA very soon.

Did you use Lightroom or Photoshop/CameraRAW to grade the frames by any chance, or did you use Davinci Resolve?

My biggest problem right now is that i don't have my mac with me but just an old, ugly pc.

Anyway, I can see i got this same problem for sure ! :

But the sensor noise on the upper half is different from the bottom half. Can someone please share some footage with lens cap on?

[quote="alvanh"]A strobe effect on the screen and on the record in low light condition.

Is this strobe effect present in your ProRes footage as well or only the RAW footage? Also, what do you grade your footage with?

Yes on the two of them. My cam goes in RMA very soon.

Did you use Lightroom or Photoshop/CameraRAW to grade the frames by any chance, or did you use Davinci Resolve?[/quote]

I used Resolve.

Hey guys,

Very excited as I just received my camera yesterday. Just noticing a lot of dead pixels, red and blue in footage (Prores Film 800 at night). They are very noticeable when colour corrected, I count 10-15 red and blue spots, there could be more. I can also even see 1 dead pixel on the camera's LCD. Has anyone else noticed this problem. I am not sure what firmware I have, whatever was shipped with the camera, my Canon 24-70mm 2.8 L aperture works if that is any indicator.

Any help would be greatly appreciated as I don't want to have to send my new toy back to the factory especially after such a long wait till Christmas!

Cheers,

Sam

I noticed on mine, its the same but if you are indeed in low light and at iso 800 and up, you see a bunch on blue, red and yellow pixels. If you adjust the ISO to 200 they are faint but still there. I am just lighting whatever i have to shoot with lots of light.

While looking at the picture posted. I define them as hot pixels. I see these things on my HM700 with high gain too.

It is only notisable in the darker area's were hot pixels appear if the chip get's a higher light rate then the other pixels around it.

The HM700 has a build in system to calibrate / ignore those pix's

Guys PLEASE help me :

I am experiencing, like many of you, the damn LINE problem, the famous line that devides the screen in a darker (greenish) part above and one brighter down. It appears clearly when shooting on low light conditions.
My problem is: I left my MAC in south america, I still couldn't watch no footage on a pc to see if the problem persists on the footage once downloaded to the computer.

did any of you find a solution to this problem? I mean, many people say "just don't shoot in low light" but it seems such a ridicolous answer, since the line appears even in "normal" artificial light situations.

Does the LINE persist in the footage once downloaded? Do you think it's just a software problem or something more serious?

Please, without my mac i'm in panic, i can't make no tests, I really hope I'll get some answers from you guys!

Thanks!

Yes after downloading you will still see the line!!

Read this topic and you know why. viewtopic.php?f=2&t=4604

that's an answer from a user...

Let's not take it as gold, this line becomes really clear even in normal light condition, i don't think it's normal at all.

Since i know many of you have noticed it, i'd like to know if you got an answer from blackmagic about this issue.

tnx

Well the thing i know is that JVC had similiar problems with there HM 700 schoulder cams. This was later solved by a firmware update.

Problem there was that they had 2 processors to proces the sensor. Otherwise it was not fast enough.

But that problem also was due in low light situations. Just remember it is a 3K cam instead of 25K cam. A 25K cam like sony will be field tested by sony for a year before releasing anything onto the market.

It's like the new update from Canon 5D mk3 were you can get uncompressed HDMI out. This software update will be released in April. But was announced half a year ago. At that time they already had the firmware working. This is the testing fase!! And let's fase it. These guys are Camera builders pro!!
A lot of engineers with a lot of knowhow and they have dealed with al the small "child decieces" many times before.

To get a good and working product you need a lot of field testing and adjustment before a product like this is finnished and working in upper most condition. I have been a software developer and engineer for 10 years. Designed a lot of products (still do in my spare time) some stuff is just hard to get working correctly.

Just Remember it's there first camera!! It takes time.

this line becomes really clear even in normal light condition, i don't think it's normal at all.

tnx

So far, the only examples shown where the line is visible have not been in normal light conditions.

Please share a DNG with "normal" light conditions where it is visible.

@Xtreemec :

this is the right answer, I noticed that many people tend to justify the problems of this camera.

everyone talks about the famous line as if it were a normal thing, but it is not. Unfortunately, as I wrote before, I have problems with my computer and I can not save the dng file. That said, the line appears on the screen in discrete artificial light situations.

this line is not normal, it is much worse than the noise. A video with NOISE can be recovered, a video with this ugly LIGHT is unusable.

When I write a Blackmagic no one answers me "everything is normal", I don't want them to take advantage by the fact that few complain about this and not solve the problem.

Let 's focus on the problems of this camcorder more than on its merits, so that they will improve its characteristics.

I have sent a DNG file of the line in the footage to BMD support a while ago. This is the reply I got:

Thank you for sending that through.

I'm getting in contact with one of our product managers about this to see if there is any explanation to this behaviour.

I opened the DNG file that you sent through and opened it in Photoshop. I then increased the exposure to +5 and was able to see the 50/50 issue that you reported.

At this present time, the issue needs to be investigated further as to see whether this is normal, or a firmware or software fault.

I have set a follow up for this email and I will get back to you as soon as I have some further information for you.

Regards,
Paul Wilson
Technical Support Consultant
Blackmagic Design EMEA

That was about two weeks ago. I didn't receive a follow up eMail yet.

@Xtreemec :

I noticed that many people tend to justify the problems of this camera.

everyone talks about the famous line as if it were a normal thing, but it is not. Unfortunately, as I wrote before, I have problems with my computer and I can not save the dng file. That said, the line appears on the screen in discrete artificial light situations.

this line is not normal, it is much worse than the noise. A video with NOISE can be recovered, a video with this ugly LIGHT is unusable.

When I write a Blackmagic no one answers me "everything is normal", I don't want them to take advantage by the fact that few complain about this and not solve the problem.

Let 's focus on the problems of this camcorder more than on its merits, so that they will improve its characteristics.

Hang on a second, so far, the only people who have reported the problem and provided examples of it have all been in abnormal/very low light situations and or shots that have been pushed too far in post production. It's not that people are trying to "justify" the problems, they are trying to get to the bottom of what the real issue is. So far, it is fairly clear that this artefact is ONLY visible in situations which are not well lit - shots that are outside the normal operating conditions of a "cinema" camera. A major aspect is that this camera shoots RAW - so you are able to see the footage without any in camera noise reduction and you can also boost the footage way more to the point at which on another camera you would just see compression artefacts - whereas on this camera you see the "signature" of the sensor.

Here is an example I posted previously - this was taken on a 5d Mkii without noise reduction - using the exact same lens and settings as someones example of the horizontal split effect.

Look at the clear horizontal and vertical lines present in this image. Taken on the "low light king".

This kind of artefact is completely normal when shooting raw with insufficient light or when you boost exposure too much in post.

Several people have reported the same effect, each time they have posted an example dng - each dng was severally under lit.


When you are able to -if you can demonstrate this problem in a scene with "normal" lighting - you will be the first.

It is easy to make the claim that people are eager to "defend" the camera - but it could be equally said that some people are eager to "attack" the camera. I wish to do neither - I want to ascertain if there IS actually a problem with the camera in normal conditions - if it it just the same affect one should expect from any other raw camera in the same circumstances.

You cannot expect anyone to look into a problem that you are unable to show evidence for.

I have sent a DNG file of the line in the footage to BMD support a while ago. This is the reply I got:

Thank you for sending that through.

I'm getting in contact with one of our product managers about this to see if there is any explanation to this behaviour.

I opened the DNG file that you sent through and opened it in Photoshop. I then increased the exposure to +5 and was able to see the 50/50 issue that you reported.

At this present time, the issue needs to be investigated further as to see whether this is normal, or a firmware or software fault.

I have set a follow up for this email and I will get back to you as soon as I have some further information for you.

Regards,
Paul Wilson
Technical Support Consultant
Blackmagic Design EMEA

That was about two weeks ago. I didn't receive a follow up eMail yet.

This response shows that there is a problem, which is not a simple problem of low light but something more '.

I hope it is a software problem or firmware as they say and not a sensor problem

Look at the clear horizontal and vertical lines present in this image. Taken on the "low light king".

This kind of artefact is completely normal when shooting raw with insufficient light or when you boost exposure too much in post.

Ok. horizontal and vertical lineS. BMCC gives a perfect split screen line that divides the image in 2 parts giving a totally inusuable footage. That's not normal and needs to be fixed. Noise is one thing, a horizontal line in the middle of the video is another.

Look at the clear horizontal and vertical lines present in this image. Taken on the "low light king".

This kind of artefact is completely normal when shooting raw with insufficient light or when you boost exposure too much in post.

Ok. horizontal and vertical lineS. BMCC gives a perfect split screen line that divides the image in 2 parts giving a totally inusuable footage. That's not normal and needs to be fixed. Noise is one thing, a horizontal line in the middle of the video is another.

Again, show me this effect in a DNG file with sufficient lighting and I will agree that there is a problem. If you can - you will be the first to demonstrate such a thing.

Different sensors have different characteristics - the sensor in the 5d has many vertical and horizontal lines - in the BMCC it has a clear split between top and bottom. Perfectly normal in sub-standard lighting.

I just took adam roberts DNG file (http://www.adamroberts.net/blog/blackma ... ight-test/) and pushed it 4 stops. There is no horizontal line dividing the screen in his footage.

ISO800.jpg (510.67 KiB) Viewed 68629 times

Edit: Or this dng from some guy on BMCuser (pushed 4 stops)

2.jpg (519.24 KiB) Viewed 68628 times

And finally the best example that not every BMCC has this line issue (pushed 4 stops)

night.jpg (484.44 KiB) Viewed 68628 times

i will, i hope, when i finally get my mac back

i'm happy for mister roberts, maybe some cameras have this problem, other don't

BTW: On the 3 pics I posted I can´t find any dead or hot pixels so this is not normal, too.

well, that probably means that many of us will have to send the camera back.

BTW: On the 3 pics I posted I can´t find any dead or hot pixels so this is not normal, too.

Depending on what you use for raw comversion can hide hot pixels, some converters automatically remove them for example.

Also, on all those examples, their histograms and waveforms show more light in the shots than in those which have had the artefact present.

Present an example of the artefact in a shot with "normal" lighting - or as i would prefer to describe it - sufficient lighting. If present in normal conditions then yes it sounds like a problem - but so far no one has shared any dng with such a thing.


Just looked on my ipad and i can clearly see the horizontal line in all of those images.

so you see the line?

and don't tell me these are extreme low light situations!

I just hope this problem will be solved someway...

Just looked on my ipad and i can clearly see the horizontal line in all of those images.

Hmmm. That´s weird.
Here´s the dng
https://docs.google.com/file/d/0B8-f-XP ... YyWk0/edit

I wasn´t able to make the line visible in Adobe Camera RAW.
My screen is REC709 Gamma 2.4 calibrated

I took the dng with the lens cap on and had a look at it. It really looks totally different than the other footage. Not only the "splitscreen" but those blotches of noise.

bmcc.jpg (510.95 KiB) Viewed 68608 times

And if you don´t push it all the way up you can see hotpixels, too.

hotpixels.jpg (50.03 KiB) Viewed 68608 times

so you see the line?

and don't tell me these are extreme low light situations!

I just hope this problem will be solved someway...

They are low light shots pushed 4 stops!
Dont tell me that is normal!

This is not a nightvision camera.

Low light Interior (bar with normal lights and candle)
Low Light Exterior
Low Light Street
Some Sky shots

https://www.dropbox.com/sh/3rkp05ms1z853qs/6jp8HL9aRw
28-105 Sigma AF 2.8

What point are you making?

that some cameras have different respose to low light
for example mine, dont shows the lines that they were talking about.

Also, i had hot pixels (said a few days before) , but as can see not now, just when i turned on the camera was right...

So you mean the hotpixels are sometimes there and sometimes not?

@Kristian Lam: If you´re reading this, can you check if the dead pixel mapping function is not starting everytime you fire up the camera, please?
John Brawley already mentioned the same, he had to turn it off and on again sometimes to activate the mapping function.

that some cameras have different respose to low light
for example mine, dont shows the lines that they were talking about.

Also, i had hot pixels (said a few days before) , but as can see not now, just when i turned on the camera was right...

Please share those as DNG's -or screen shot the scopes before any modification.

Some of them look to be better lit than the other examples.

Tom, here is the dbox folder with the dngs that i have
i deleted some, but i found others with more low light (one street, you can also find a lighted version of that one)
Remember that when converting to jpg the pic usually gets less detail in darks etc etc.. anyway, here you can get darker dngs and lighter dngs to play with the sky.... we usually have greyscale sky...
U have also a screenshot

https://www.dropbox.com/sh/3rkp05ms1z853qs/6jp8HL9aRw

So you mean the hotpixels are sometimes there and sometimes not?.

No, i mean that i saw them once and they never came back.
But well, i maybe i dont have any problem with that, as some of us... but i have a major dropping frames in certified SSD. Nobody is safe

that some cameras have different respose to low light
for example mine, dont shows the lines that they were talking about.

that's the real deal,


many of you are cataloging these problems as normal, but they may be hardware problems involving some cameras

Tom, here is the dbox folder with the dngs that i have
i deleted some, but i found others with more low light (one street, you can also find a lighted version of that one)
Remember that when converting to jpg the pic usually gets less detail in darks etc etc.. anyway, here you can get darker dngs and lighter dngs to play with the sky.... we usually have greyscale sky...
U have also a screenshot

https://www.dropbox.com/sh/3rkp05ms1z853qs/6jp8HL9aRw

Just opened them in Resolve, and just as I expected and have tried to explain - these DNG's are brighter than the dng's with the artefacts. When I boost the darker of these dng's In the darkest and most light starved areas, you can see the artefact.

Let me reiterate - I do not consider the artefact a problem - I assert again that this effect can be brought out on any camera in the same conditions. These conditions are not within the expected limits of a camera such as this one.

Thank you for sharing these DNG's, they are fun and interested to manipulate anyway.

Tom can you post a picture where you brought out the artifact please?

Tom can you post a picture where you brought out the artifact please?

Sure

Its harder to see reduced in size, and I must also reiterate that the quality of light (in both colour temperature and illumination) in the new shot is slightly better than the previous examples of this artefact. Even so, it is clearly visible.
https://docs.google.com/file/d/0BytMyjm ... sp=sharing

I maintain that every bmcc in the same extreme low light conditions, boosted in post, will have the same artefact. The problem here is that "low light" is not a specific thing, people interpret it differently.
For me, low light is an int. night show lit with dim practicals or a room full of candles (Barry Lyndon style, sans the Nasa tech :-p ) Not an external night shot with dim and very very low CRI based street lights. This last point regarding CRI is also a key issue here, I have noticed that different light types tend to bring out the artefact more than others.

In conclusion, it is not a problem, see Kohli's post, it explains it all perfectly - see John Brawley's posts, they confirm the same thing. Notice that one user, whos images are in my example image, worried about the artefacts being a fault with his camera, and then a week or so later released beautiful test shots lit with only a cigarette lighter - without the artefacts. The difference? he wasnt boosting them several stops in post or lifting the blacks.