BMPCC Hot Pixels

For what it's worth, here's how my camera looked. Both images are ISO 1600 and a 360-degree shutter, 30 fps and with peaking on. The first was taken with the 1:1 zoom engaged, and the second with the zoom off. It takes about 7 minutes for the camera to warm up and show you a nice star field. Not every pixel that shows as green with peaking on is so bright in the recorded image, but some of them are light enough grey to appear over any dark area of the picture.

I had one of the earliest cameras with the white orb problem (manufactured 2013-08-15 according to the RMA). I sent it back for white orb calibration and added a note asking about the hot pixels. They sent me a brand new camera with no explanation, and the new camera has many more hot pixels than the original one shown here. I should have kept my mouth shut.

I don't understand why the hot pixels get worse with longer exposure times. Anyway, it's just another in a long series of disappointments from this camera: lack of audio meters and a histogram; nearly unusable ProRes recording in a proprietary "BMDFilm" color space; delayed support for raw recording. But I'm not going to sweat the small stuff. Pixels are small. Now I just need to figure out if raw recording is actually usable.

Tony Rivera said they were investigating this a couple of weeks ago. Tony, could you tell us how going research or something?, in my case took from the 23/10 I sent my 3rd unit to LondonBM waiting for a solution, before they send me the replacement unit untested in the same conditions of Hot Pixels, apparently are waiting for an answer too, what I implied that in case that you find the problem, you need to send to repair our BMPCC or repair by firmware will be? I would like a quick response, and please, inform too to BMLondon.
This is my situation, i say to other Europe people that if are thinking in send they unit for Hots, probably get the same situation.

My BMPCC has been in repair for the Hot Pixel issue for about three weeks now, and apparently had been shipped to London for repairs (I'm located in the Netherlands).

I've just received word from my supplier that Blackmagic in London had communicated to them that repairs
were taking longer than expected, as the problem was more difficult than first imagined. They were
communicating with Australia about the source of the issue and apologized for the time it is taking to
solve the issue. Apparently in Australia they are doing tests and research about the hot pixel issue.

When my supplier asked why they didn't just send another camera to me as a replacement they said the
issue might be present in a multitude of cameras, and thus they wanted to be sure to pinpoint the issue
and not send back another defective unit.

So clearly Blackmagic is aware of the issue and recognizes it as being something more than just 'a part of the deal because it's not a low-light-camera'. It's good to hear the issue is being worked on, but I hope it won't take too long to make an official statement about the hot-pixel issue and present a plan to fix it.

Yes, have a similar response to me in BM London. Should be congratulated for a perfect communication between user and BM London, have never taken more than two, three days to respond to any of my questions. Not all post-sale service can boast that. . So wait, hope no more than hours or some days

Hey,
I was wondering if anyone who had issues with the hot pixels in the BMPCC and sent it in for repairs has heard any news yet. My camera is in repair for almost two months now, and the last news I heard was a few weeks ago, when Blackmagic communicated that the issue was 'more severe than expected' and they were doing tests to determine the cause and solution.

Are there any similar experiences from people, or maybe some other recent news or info about the issue?

I unfortunately do not have any further update at this time.

So are these Hot pixels, an issue or a known side effect of the camera settings?

When I spoke to BMD they let me know that hot pixels are to be expected at such high ISO, and that it is normal..

I can replicate the very same scenario with my BMCC, as shown above...

So is this a fault/issue or just a "side effect"?

I am also experiencing the problem with hot pixels, plus I seem to have the highlight grid issue as well.

Unfortunately the guy from the place I bought it thinks I am lying and don't know anything about images cameras (I work 60+ hrs a week as a color grader, editor and videographer and this is the first I've seen of this kind on any image that's crossed my monitor so far) and I am having a hard time making him take back the camera and send it in.

I get "hot pixels" at iso 400, 800 and 1600. While they are less noticeable at 400 and 800 I can push the grade in resolve to make them appear. And the hot pixels match throughout the clips and different iso settings.
I find it hard to believe this is normal.
As is, the camera is 100% useless to me, seeing as this issue is across the iso ranges.

Personally, I will have to get a confirmation by mail from Black Magic to show my dealer, before he will acknowledge the issue and send it back for repairs.

Man am I stressed out by this.

If you want to see the static pixels you have to view the images at full res.

graded cap test @ iso 400

graded cap test @ iso 400

iso_400.jpg (251.25 KiB) Viewed 19012 times

graded cap test @ iso 800

graded cap test @ iso 800

iso_800.jpg (600.1 KiB) Viewed 19012 times

graded cap test @ iso 1600

graded cap test @ iso 1600

iso_1600.jpg (196.37 KiB) Viewed 19012 times

Also, I am getting a real white christmas on my display when do the cap test and set the pocket to iso 1600 and turn on focus peaking.

@f1.8, iso 1600

display @ f1.8 iso 1600

displ_001b.jpg (900.39 KiB) Viewed 19012 times

@f14, iso 1600

display @ f14 iso 1600

displ_002b.jpg (916.71 KiB) Viewed 19012 times

I am also experiencing the problem with hot pixels, plus I seem to have the highlight grid issue as well.

What lens do you use when you see the grid issue?

Anyone with a hot-pixel free BMPCC or is this an issue currently present on all of them?

By the way, you can shoot a blackframe with the lens cap on and subtract it from your shots in postproduction to get rid of the hot pixels.
Note that this does only work in linear color space and theoretically you would have to compensate the camera's respone curve to get exact results. It does the job without the latter, though.

I guess an on the fly darkframe subtraction would be useful in firmware as you would not loose the pixels by mapping them out, but correct their values instead.

By the way, you can shoot a blackframe with the lens cap on and subtract it from your shots in postproduction to get rid of the hot pixels.

But only if they are "true" (stuck) hot pixels. From earlier postings i get the impression that the hot pixels in this case appear randomly and fluctuate, depending on temperature? I static mask wouldn't help then...

I am also experiencing the problem with hot pixels, plus I seem to have the highlight grid issue as well.

What lens do you use when you see the grid issue?

I used the Olympus Pen 17/1.8 which is my only lens at the moment.

Ok, so another fast lens produces the grid... Thx for the info!

By the way, you can shoot a blackframe with the lens cap on and subtract it from your shots in postproduction to get rid of the hot pixels.

But only if they are "true" (stuck) hot pixels. From earlier postings i get the impression that the hot pixels in this case appear randomly and fluctuate, depending on temperature? I static mask wouldn't help then...

Nope , they stick once the camera is warmed up. The fluctuations are minor, like stars twinkling.

But are these the same pixels every time you start the camera?

Yes, same pixels every time.

Ok, then a hotpixel subtraction might work, or even better an algorithm to interpolate the hot pixels.

I really only get those "twinkling" hotpixel when the lenscap is on. Otherwise no noticeable hotpixels @1600 ASA and 180

I feel like I'm being a bit repetitious, but I do wish Blackmagic would communicate something substantial about this issue soon. My camera has been in repair for weeks (since the start of November) and I've received no timeline on which I can expect my camera to return.

Consider the fact that my first BMPCC also had to be returned because of a loose HMDI-port, and you end up with a situation where my BMPCC has been away in repair longer than I've held it in my hands since I've received it in october.

As Blackmagic doesn't want to replace my BMPCC with another one because the issue with the Hot Pixels was present in multiple camera's, I'm left with not owning my purchase for weeks on end and no indication if, or when, I will receive my camera, and if the issue has been resolved.

I imagine it's tough to officially communicate about such an issue beyond saying 'they're looking into it' because of all the legal context of every statement you make. Do you officially call it a defect? Do you admit a production error? etc.

But it's so painful, as a customer to be left in the dark about this. More than two months since I've received my first BMPCC and I've yet to pleasantly use it and really film and experiment with it cause it's been away in repairs for so long. And now that we wrap up 2013, I have no clue whatsoever what will happen with my BMPCC.

When the eventual politics and legal stuff is resolved and Blackmagic is ready to communicate fully about the defects as I've experienced them, I really, at some point, would like to hear something about the cause of the issue and their eventually presented solution. Heck; for all I care; replace it with a BMCC and charge me the difference, just throw up a solution of some kind for those who are affected.

Seriously, spare yourself the heartache and long letters and just get your money back (if you can) and get another camera until they sort this out. Don't mean that in a harsh way, but no reason to be a couple months without a cam. Or do you already have another cam?

Mine is free. I cannot seem to find any way of making them show up on my camera, not even one.

Can you give us a some kind of frame pulled of your footage, or some short test on vimeo?!
That'd be great and would make a few of us feel better.

cheers

Could you please share with me the Support e-mail address of BMD?
I have a fixed white hot pixel on my footage, even visible in broad daylight and I'd like to contact them.
Thanks in advance.

30fps 360 shutter with zero light, lifted from 800 to 1600 = not normal conditions.

You will bring out hot pixels in any raw camera by doing this.


It is only a problem if you are seeing them in nominal filming conditions.

Here is a quick example I just did:

30fps 360 shutter = 1/30's (if my maths is correct), shot at 800, lifted 1 stop in post - without any focus peaking on -a 100% crop from a Canon 5d Mkii

http://tommajerski.com/publicimages/IMG_8152.jpg

As you can see, many hot pixels.

Do I see these in properly lit photographs? no.

Unless you are seeing these in properly exposed shots, not pushed several stops above 800 in post - its normal. It is not a fault. You will see it in any raw capable camera with the equivalent settings.

I noticed this issue on my camera, but only in dark areas that are overexposed. I spent the afternoon freaking out about it and sent tech support an email. They responded saying I can send it in and that it would take 5-10 days to either send a new one or repair.

At this point i'm too in love with this camera (despite the hot pixels in bad light, which I managed to fix by CC) to send it back. I think I will hold on to it now until a more concrete reason/solution is given by blackmagic and just make sure I shoot in proper lighting conditions with the right settings.

A short update from my case. I contacted Blackmagic directly and received an update. Not much new information but good to hear, '...As previously mentioned we are investigating this behaviour and are trying to address it as quickly as possible..'

So it seems clear again; this is not normal behavior that is always present in underexposed images, not in my (and probably many) case, and is deemed a real issue also by Blackmagic.

Anyway, they (Blackmagic) offered a BMCC MFT as a loan until the BMPCC is fixed, which I really appreciated (even though I need to invest in SSD's and batteries for that to use) and so for me, it's better news than before. I might be able to shoot with a Cinema Camera again soon, and can even experience the larger size to try it out.

I hope that a fix can be presented soon to all who suffer from it.

I wonder why they don't simply send you another BMPCC that has been tested.

As far as I understood, they seem to believe the issue is too wide spread among the BMPCC's and that another pocket camera could exhibit the same issue. That's why I didn't get a replacement immediately back in November. Though I would assume that there are tons of Pockets out there that don't exhibit the hot-pixel issue as severe, or else the complaints would be more widespread.
The testing at Blackmagic might then be also aimed at figuring out why some Pocket's exhibit the Hot pixel issue more or less than other models.

So I'm noticing this as well, at higher ASA. It reminds me of static grain, like we use to see in static 35mm adapters. I prefer to shoot at the lowest ASA to avoid noise, but it does bother me to know I can't shoot at higher ASA if I needed to. Curious do you see any hot pixels at 200, 400 ?

So I'm noticing this as well, at higher ASA. It reminds me of static grain, like we use to see in static 35mm adapters. I prefer to shoot at the lowest ASA to avoid noise, but it does bother me to know I can't shoot at higher ASA if I needed to. Curious do you see any hot pixels at 200, 400 ?

No, do you? If so, can you show us?

I get stuck pixels @ iso 400 but they are hard/impossible to see without squeezing them out using curves in resolve - but then they are there, and are the exact same stuck pixels as in iso 800 and 1600

I get stuck pixels @ iso 400 but they are hard/impossible to see without squeezing them out using curves in resolve - but then they are there, and are the exact same stuck pixels as in iso 800 and 1600

In "squeezing" them out, are you sure you are not effectively lifting the footage up beyond 400 then?

So what can you do, besides returning the Pockets to BM?

Apparently BM is aware of the problem. For me as long as I expose correctly and don't push the ASA too high, I won't notice the pixels.

So i'm just going to keep my camera for now until there are any definite fixes or announcements

"No, do you? If so, can you show us?"


At the moment I've only tested it through the camera. I'll shoot a few more cap test and see. I fab products for a living and understand the difficulty in offering a "perfect" product. Honestly I didn't expect the camera to be perfect and that goes for every other camera we've used. I'm very happy with the camera so far and got nothing but love for BM !

We've had the same issue with two Pocket Cameras we purchased. The only response we got was we need to light and expose the shots better but you can't always control that if you're filming a live concert. The footage that was shot with both pocket cameras has white dots all over it. We've never had this issue on any camera before regardless of the lighting or exposure. We wish somebody would address this issue and let us know what's going to be done to fix it.

My first camera had about 40 hot pixels, some flashing some static. I received my new camera and it's got about 5. I'm not sure if I should get it replaced or wait for a fix...

I just noticed that the longer we filmed a live concert the worse the pixel issue got. We started off with a clean image and as the minutes kept going by the footage kept getting worse and worse ending up with what looks like a White Christmas. My best guess is that the longer the camera is recording the more stuck pixels appear due to the heat.

I haven't seen white dots that STAY on my image, but i haev one shot in particular that I get two white dots, reasonably close to each other, for a SINGLE frame, and then they vanish... Anyone encounter this before?

I've made an interesting discovery today (why didn't I checked it before and why nobody noticed it?). The thing is that I can see some "hot pixels" when BMPCC is operating in ASA1600 and when focus picking assist is on. So basically when I focus on a subject I can see a green outline around it's shape... and some hot pixels. If I go out of focus the green outline is gone because obviously this subject is no longer in focus but sadly the hot pixels are still there and are indeed recorded on card.

The funny thing is that if I disable the focus picking assist I do not see any single "hot pixel" on screen regardless if the subject is in focus, out of focus or I am pushing the camera as much as ASA1600 and 360 degree shutter. I do not have a spare memory card at the moment so I'll try to do some tests with recording in a couple of days. If I won't find any "hot pixels" on footage that means that this issue is related somehow to focus picking assist. I will keep everybody informed in the meantime maybe someone could confirm this?

Kind regards
Adam

We've been shooting with these cameras for a while now and it's happening to more or them. When we noticed it we were shooting at 800 ASA and 180 degree shutter. When we first turn on the pocket camera the image looks clean but as you keep it recording we start noticing more and more white pixels start appearing. We filmed a 1.5 hour live concert and by the end the footage looked horrible. If you're reading this please set your camera up and run it for at least 15 mins without stopping it and see if you notice the same thing happening to your camera. We believe this to be an issue with the camera getting hot but would like for more people to confirm this to Blackmagic so they will stop telling us it's normal. It's not normal and it's not happening with our Blackmagic Design Cinema Camera. It only happens with the pocket cameras we've been using.

Made some tests with switching off focus assist. The hot pixels are on the footage although not as many as on the screen with focus assist on so I presume some hot pixels on screen are fake - maybe focus assist firmware error.

We've been shooting with these cameras for a while now and it's happening to more or them. When we noticed it we were shooting at 800 ASA and 180 degree shutter. When we first turn on the pocket camera the image looks clean but as you keep it recording we start noticing more and more white pixels start appearing. We filmed a 1.5 hour live concert and by the end the footage looked horrible. If you're reading this please set your camera up and run it for at least 15 mins without stopping it and see if you notice the same thing happening to your camera. We believe this to be an issue with the camera getting hot but would like for more people to confirm this to Blackmagic so they will stop telling us it's normal. It's not normal and it's not happening with our Blackmagic Design Cinema Camera. It only happens with the pocket cameras we've been using.

Just tested our BMPCC for hot pixels.

When I first turned the camera on it looked clean (lens cap on, 180deg, 1600ASA, ProRes). No visable hot pixels even with focus peaking on (however when viewed full screen on a PC I could make out some very faint hot pixels but they're barely visable).

Then I left the camera recording for 18 minutes (the body became quite warm), at which point hot pixels became clearly visable and quite numerous.

So, I have to agree with Movie Mafia. This seems to be a heat issue. When the camera is cold everything seems fine.

@movie_mafia

Your post possibly explains why some owners are not seeing the problem (length of take).

It makes me wonder if the same noise happens if the camera is completely pre-cooled (say, in a fridge) -- and then run for 15 minutes, or longer, in that kind of cooled environment.

That would be an interesting test. And depending on the results, I guess it might point to a more fundamental problem that might be harder for BM to fix -- at least for existing units.

Turning the shutter angle down a bit seemed to help this problem for me, although I have been shooting in good lighting conditions.

There's a post way back from November or early December from BM saying they are aware of the problem and looking into it. It sure would be nice to have some type of update!

My understanding is that the camera uses a Peltier effect, thermoelectric cooling system.

Not sure what you would do, as a manufacturer, to solve that problem in existing cameras.

Has anyone else tried running their camera using the same settings we used for at least 15 mins or longer yet? We had our two pocket cameras set at 180 degree shutter and 800 ASA when the pixels started appearing. We are very curious if you setup the camera and just let it record without stopping it if they same thing is happening to more cameras. We returned one of our pocket cameras to the reseller already but they will not let us return the second one because we've had it a month to long. If this is happening to more pocket cameras we are going to have to invest in more Blackmagic Design Cinema Cameras and stop using the pocket all together. We just can't afford to disappoint our clients with an image with dead pixels all over it. We had hoped to use our two pockets as a second and third camera when filming interviews and infomercials but it's looking like that may not be possible if this is happening to more and more pocket cameras.

Frame Grab before the concert started.

Wide 1a.jpg (205.6 KiB) Viewed 18906 times

Frame Grab right when concert ended.

Wide 2a.jpg (396.74 KiB) Viewed 18906 times

For legal reasons we can't show the actual concert footage but you can see in our two attached frame grabs above that before the lights came up the frame is clean and at the end when the lights were turned off the image is covered with dead pixels. Make sure to save each still so you can view them in full rez on your computer and see clearly what we are talking about. The dead pixels are visible with the lights on and off so the lighting and exposure were not the problem. If lighting and exposure was the issue then the dead pixels would have been present at the beginning of the concert and would not have increased as the show went on.

I've been shooting several hour-long stage acts with the pocket cam, set to 800/180° ProRes, latest firmware. I have not seen hot pixels in my footage so far! But i must admit that i only previewed the material at 720p, as that was the final output resolution (it had to be intercut with 720p recordings from other cameras), and most of it is high key material with very bright backgrounds where hot pixels could get through unseen. A quick glance at the end of said footage at full resolution didn't show hot spots either.