Pixel white dots issue explained here...

Gibby

Gibby

Red Savant
Page 9 of the new HD100U manual, available to everyone online at the JVC web site as of 1/10/05, clearly explains the "mysterious" white pixel phenomena that Barrty Green and countless others are discussing. Read Page 9, Characteristic CCD Phenomena, White Dots:

Quote:

"High temperatures can cause CCD sensor panels to produce the effect of white dots in the image. This condition is conspicuous especially when gain is applied. This is characteristic of the charge-coupled device (CCD). As far as possible, use the unit under conditions where the temperature of the unit does not increase."

Mystery solved. So it is not "burned out pixels", "defective CCD's", "dropouts", etc. Indeed, in almost every post I've seen on this board and others about "dead pixels", the user has boosted the gain to 18db to even see them. In 30 years of professional shooting in every analog and digital format available, I can count the times I've used 18db gain for my professional productions on one hand. When you boost the gain to 18db on most cameras you will find "white spots", snow, and artifacts. Why would any professional shoot 720p24 or 720p30 footage at 18db anyway?

Perhaps the camera Barry used at NAB had been powered up in the indoor environment for hours, and thus it had accumulated some heat. Even then, if Barry had to go to 18db to see the "white dots" that's not especially troubling to me.

Before everyone does a "tecnno-lemming" thing, and rushes over the "I told you so" cliff, I'd suggest that we give this camera a chance. When it has been shot using the extensive menus (check the manual), by seasoned DPs, in every condition possible, then professionals can BEGIN to draw educated opinions on the camera's performance envelope.

On another thread on this board, a poster from Scotland reported that the JVC representative stated that there is a self-repair set up in menu for "dead pixels". I'd be surprised it the HD100/101 doesn't have such an auto-fix function in conjunction with the balck balance function.

I should have my HD100U this week, and as promised to Barry, I'll make a trip to Las Vegas so we can jointly put it through its paces, indoors and out, using all the menu options, and accessories. If Jared would like, that footage can then be posted on this board.

Again - the mysterious white spots are not dead pixels

HD100U manual link: http://books.jvcservice.com/booklist.asp?Model=gy-hd100u

Gibby
www.cut4.tv
 
..did you do know that barry has ordered his own HD100U as well? hows that for giving it a chance. hes just reporting on what hes seeing and experiencing. you can push +18 on other cams and not get white pixels... it likely does have a way to fix them.. but lets just call it like we see it as it comes. and thank you for the info on the pixels.. good to know. :)
 
Yeah, I'm well aware that Barry has ordered an HD100. We're friends that go back to when my business was based in Las vegas. Me and Barry both sat at the same table and recieved Emmy Awards together a few years ago. My offer to him to take time from my busy television schedule and travel to Las Vegas with my HD100 is based on my friendship with Barry, and my curiousity as a professional shooter to define the capabilities of new equipment. If he gets his HD100 before me, there will be no need of me taking a trip to Las Vegas.

Gibby
www.cut4.tv
 
Just used another HD100 last night, and yep - dead pixel in the lower left. Similar position to other cameras.

Note: you don't NEED to go to 18dB of gain to see it, it's just that that's what makes it easiest to see. Just like the odd brightness pattern on the CCD -- there's a split-screen effect that happens, one side of the image is darker than the other, split right down the middle. It's a lot easier to see it with 18dB of gain on, but the D.P. on the shoot yesterday saw the effect in the 0dB gain shoots we were doing. And it's not necessarily related to high temperatures -- we were shooting in Los Angeles, 10:00 at night, it was probably 70 degrees, when we finally got around to checking for dead pixels.

The dead-pixel thing is an issue. It's a dead pixel -- it shows up in every shot in every frame. The color isn't always white, it depends on what else is around it -- sometimes the dead pixels are green, or red, or whatever. Now, the good thing is that they're small, and can be easily missed (on high-def, all the pixels are tiny). The bad thing is, if you've got one or more, you'll see it in every frame you shoot. And that's irritating. You'll see it easier under 18dB gain conditions, but you'll see it in 0dB as well. I have yet to see an HD100 that didn't have dead pixels on it, or hear a report from someone who has an HD100, saying that they don't have at least one dead pixel. Someone on DVInfo said he had no dead pixels; I told him how to test for it and sure enough he reported that he had nine! Once you know they're there, you'll see they're in every shot, every frame. If you didn't notice them before, that's just because you didn't notice them; it doesn't mean they're not there and that someone else won't notice 'em.

Now, here's the potential good news: I have been reliably informed that there is a masking procedure in the camera's service menus that can hide the dead pixels. Perhaps the early cameras out there haven't been properly "masked" as part of the release to the public? I don't know if this process is something the end user can do, or if it's something that requires a trip to the service department. Either way I'm not too overly worried about it, because as long as it can be fixed, then that relieves a lot of worry. I'd still like to hear official confirmation on the policy from JVC, but I'm less concerned.
 
Gibby said:
We're friends that go back to when my business was based in Las vegas. Me and Barry both sat at the same table and recieved Emmy Awards together a few years ago. My offer to him to take time from my busy television schedule and travel to Las Vegas with my HD100 is based on my friendship with Barry, and my curiousity as a professional shooter to define the capabilities of new equipment. If he gets his HD100 before me, there will be no need of me taking a trip to Las Vegas.
Click to expand...
Hey Steve, thanks, and I very much appreciate the offer. I also appreciate the perspective you bring and the experience and another viewpoint! Hopefully these cameras will be out soon in the US. I was talking to another of our members last night and he was told a date of August 25th from his reseller.

I really want to like this camera. As everyone who's read anything I've written knows, I don't think much of the HDV format, I believe it's the weakest part of the camera, but the camera itself is very cool. (And at least the flavor of HDV that the JVC uses is the best version, least artifacty, most bits allocated per pixel, etc.) The menus are extensive and powerful. The Focus Assist really works well. It looks cool -- especially when you bolt a mini35 on it, with a Cooke 100mm S4 lens and a mattebox! ;) I think there's a definite place in the production world for a camera like this. I have hope that it'll be spectacular. I will admit that my first experiences have given me serious concerns, but I'm still going forward with it unless someone brings up a showstopper somewhere.
 
Just like the odd brightness pattern on the CCD -- there's a split-screen effect that happens, one side of the image is darker than the other, split right down the middle.
Click to expand...

When I saw a demo in Rome, the camera has the same defect. But the JVC guy told us it was because it was a pre-production model... Hope he said the truth.

We all want this camera to be a good one!
 
A friend in the UK has one with this split screen effect.

Yes, a couple of the hand built cameras had this issue (in extremis).

A firmware update via SD card is in the works. Apparently.
 
Hey guys, quick question that I hope doesn't sound too dumb.

Is there anyway to resuscitate these dead pixels? Does the masking utility cure the pixels or does it just cure the symptoms? What if more pixels die after that, more masking?

Okay, maybe it wasn't such a quick question. I'm like other people in that I want to like this camera. I love my DVX and I think the JVC camera could be a very useful tool, as well. I just want to know what I'm getting myself into.

Thanx for the help.
 
I am sure JVC will come up with a solution to this problem, because if you are shooting in countries where the temperature is high, then you are doomed.
 
Great, i live in southeast mississippi where the temp all week has been in the upper 90's with heat index in the 100's.

Maybe not a good choice.
 
Don't make decisions on this camera without spending some 1-on-1 time with it. I worked with it a full day, witnessed firsthand all the discussed issues, and still am VERY excited to recieve mine.
 
True. I hate when other people jump to conclusions without having all the info and here i am doing the same thing. I'd like to hear about your experience with the camera, nateweaver.
 
obiezra said:
True. I hate when other people jump to conclusions without having all the info and here i am doing the same thing. I'd like to hear about your experience with the camera, nateweaver.
Click to expand...

My comments would basically mirror Charles's in the article...if fact, indirectly, some of my comments are IN the article!
 
nateweaver said:
Don't make decisions on this camera without spending some 1-on-1 time with it. I worked with it a full day, witnessed firsthand all the discussed issues, and still am VERY excited to recieve mine.
Click to expand...

Nate, I hope you have better luck than we did. In fact I hope the US release is an outstanding success as it will mean the camera can reach the potential you demonstrated in your test.

Our first camera was returned with 2 dead pixels at 0 gain and the split screen effect. The noise on one side of the split was red and I wonder if that is one of the underlying reasons the skin tones from the camera appear 'reddish'?

The second camera is on the way back to the factory. The split screen effect was still noticeable, but not to the same extent as the first. When switching through gain the VF would 'bloom' and then settle to disguise the split. Once you picked the split, you could make it out through the gain switches back to '0'

This wasn't the reason we sent the camera back. We sent the camera back because the chromatic aberration from the 16x is so severe that I don't know how anyone could shoot with it.

I understand that it is a cheap lens and allowances have to be made, but the results we got were unacceptable. I should clarify that it could be a combination of bad lens and bad camera that gave us the results we got. I'm not saying every HD100 will have this problem. This is our experience.

I'm hoping that the 13x HD lens does drop in price and that for us, the third tiime is the charm.
 
In simple terms it's misaligned colors. Usually results in a ghost or shadow color around an object. In crass terms, think of watching a red & blue 3-D movie without the glasses on -- that would be an extreme example of what chromatic aberration looks like.
 
You must log in or register to reply here.
Back
Top