HD100U Split Screen Will Not Be Fixed?
 

I spoke to a JVC representative at a seminar today and she told me that JVC will NOT fix the split screen and it is the trade off for getting the HD resolution. Besides very few people will ever have that problem with the camera and the ones that do, can easily fix it.

I agree with you that JVC does not consider SSE a "problem" that can be fixed.

But, I assume what she also means that JVC will "fix" or "replace" units that don't meet their QC requirements.

Of all the NTSC reports I've read, there are clearly a few that are obviously BAD. Much like most, or all, the PAL units.

THe problem is not, IMHO, that JVC has not admitted that SSE exists, it's that they have not yet published a simple test that can be used to determine if they should send it back to JVC. Of course, such a test may not be possible without test equipment -- in which case they at least need to publish a Guideline on the return procedure.

And you can hear the sound of cards going back into wallets...

That's quite a surprising statement considering the number of HDV cameras currently on the market without this effect. While JVC may have there reasons, you would think such a statement would be preceded or at least quickly followed by detailed standands for exchange repair as Steve suggested.

While no camera is without it's quirks and artifacts, the evidence posted so far says that this exceeds the definition of "quirk" or "artifact" and is clearly a "flaw".

The only solution is to either obtain a camera where the level of the flaw is minimal and not impacting your shooting style or buy another camera.

Though I love 24p, I'm going to skip this camera and wait for the HVX and see if it meets my HD production needs.

I spent 3 hours with the JVC sales manager here in Chicago today. We had the camera and tried everything to get it to split screen. It would not. I asked him what the policy was and he told me if there was a split under 9db, the camera would be replaced by JVC. NO questions asked. He also explained to me in detail what is causing the problem. It is a comparitor circuit that is reacting too slow and that by design the comparitor met the spec set by the engineers. The CCD's are scanned from the center "line" out to the edge (per CCD block). Some cameras came out initially that were not QCed.

As I wrote, I spent a few hour in many different scenarios and no split on the camera we were working with. The camera was tethered to a laptop and it's 1394 output was monitored in Liquid's live capture as well as DVrack HDV's monitor. No split on that unit at all in any scenario.

I think the slogon "if you can by a better true progressive camcorder at the same price -- do so" is one which most folks will listen to. Clearly, there are no other choices -- as Canon's isn't 24p (and I wouldn't buy such an ergonomic kludge even if it did) and any P2-based camcorder will not be cost-effective until the end of the decade.

There's no evidence that professionals now shooting HDCAM, CineAlta, and DVCPRO HD will have issues with this camcorder. Nor, will those who now shoot 35mm film. These folks know how to use lighting and know how to use the menu controls to get maximum quality.

The fact is, if "DV shooters" are willing to learn and adapt, they too will be able to get great results. We all know what happens when a major technology shift occurs. Those that master the new CRAFT will thrive. Those that don't, ... .

It's clear that "DV camcorders" have enabled lots of folks to make a living shooting video without actually learning the fundamentals of photpgraphy, film, and video technology. (The so-called "democritization" of video.) Technology has allowed them to never have to learn their craft. (And, it also matches the huge drop in America turning-out students who have more than a minimal knowedge of science and math -- plus reading/writing. All of which, are the basis for learning a craft.)

Conversly, if there is one characteristic of the HD100 I love most -- it is that it calls on everything I've ever learned -- starting with 8mm and a PortaPk. In fact, it's forcing me to get out books on "film lighting" that I never got into as video arrived. Now, that we have access to near 35mm film quality -- it demands we respect that capability. Now I need to buy a light meter. :)

If you want to stay in the "DV world" but simply want an increase in resolution -- why not buy a Sony FX1/Z1? It really does offer what it seems many folks want. (Except it is almost 2-stops less sensitive that the PD170/VX2100.) And, there is a huge experience base on moving both 50i and 60i to film. The FX1 is particularly cost effective! I would not have spent 4 months writing a book on the two Sony HDV camcorders if I felt they did not offer what many folks want.

Stephen, do you remember some of the most extreme scenarios you tested? Can you share those with us, please?

Absolutely,

White wall across a gamut of settings scenarios.
Single light off screen shooting a dark backdrop across a gamut of settings scenarios
Gain in dark scenarios up to 18db.
White balanced against blue, grey, yellow, green and then shot.

etc. etc. 3 hours of trying to make it split at the Museum of Science and Industry, indoor and outdoor. No split. I proposed to go out and shoot the HD-100 at night in downtown Chicago next week. I think that will happen and I'll try to force it to split screen again.

Other observations: The lens does have CA but not nearly as bad (on the one we are using) as originally posted. The lens does breath, alot! It's like a second zoom, but, it is very sharp and we got between 700-800 lines resolution. Ergonomically it is the best of the price range followed by the Z1 and HVX (tie) and then the XL-H1 (last). You'll have to be Popeye to hold up the H1 for very long. I HAVE HANDLED ALL OF THEM PERSONALLY and I like them all.

She basically said they were NOT going to do anything and itm was a choice between them and very expensive cameras. The bottom line that she said, and I agree is that you won't have the SSE proplem if you are doing it right. She said that this camera may not be for all. That it is really geared towards cinematographers and pros who will spend the time to make the settings correct. They had to try really hard to get the SSE on the test, and even when they got it, it was very difficult to see. You have to set the gain real high, whcih most people won't do, unless they are trying to shoot night exterior with very little light, or some other instances that have been pointed out here. From what is aw and heard, and what i have read here, the people that have shot with this camera are VERY happy and the footage I saw today was absolutely AMAZING!! I could not tell it was video. Note: this was without any Prime Lense Adapter. I can't wait till I get mine and share some of the footage I am going to test shoot.

I found that the biggest issue on the streets at night -- is what do you WB on?

If you use a white card and MWB using steeet lights you can get a nice "white" but I note that most movies want a bit of the orange look to add "mood." Preset (3200) might be OK.


If the scene is mostly lit by light from within stores, then that is a very similar issue. You may want the "Natural Born Killer" blue-green look -- not true white. Preset (3200) might be OK.

It is an open issue of how the HD100 will handle scenes that are NOT truly white balanced.

I have also tried FAW. This is a very neat function! Give it about 10-seconds and it moves quite close to the correct WB value. But, I've never tried it at night in a situation where there is almost no white. (FAW assumes there is a "typical" amount of white in the scene.)

I raise these questions, because WB seems to play a role once one has sufficient light.

Speaking of light -- streets have a statistically non "normal" distribution of illumination levels -- some very bright and huge amounts of dark.Thus an F4 AVERAGE that I keep recommending may not be correct. So if you want to avoid a split on well lit pavement (for example) that looks dark gray, you may want to zoom into it and set its exposure to F2. Now zoom back and re-frame.

Looking forward to your report -- thank you for your extensive testing.

I am reading some justifcation of a clear design flaw in JVCs camera (should a courier have to know how to drive an indy car...).

Nothing too difficult to understand, if they can't fix it, JVC were clearly not up to the job, they bit off more than they could chew. They should now remove the promotional material trying to sell this camera to any market that can't control the lighting environment or add a disclaimer.

If Sony, Panasonic or Canon bring out a camera with the same problem I'll gladly eat my words, but all I see is JVC making excuses for poor engineering. It is not the cost of HD, it is the cost of depending on 3rd parties for all your CCD and integrated circuits. The Z1 has 12% more pixels than JVCs camera, so clearly it is not a problem with getting out these pixels in the required time frame. Its the same number of pixels in the same time frame. 24p adds even more time to get the pixels out...

I wonder if the HD7000 will also have this problem.

Guy,

Have you actually had your hands on the camera?

Yes I have.

Did it split for you?

I agree with you Guy. I got this camera a few days ago and saw split screen in low light messing around with it right out of the box in my living room. Ive played around with it more in better lit situations and the split screen dissapears. I can live with that because pretty much all the work I will do will have controlled lighting. But it would really suck for someone who has to tape in low light. It is a serious flaw, and its bad business to continue to sell it without correcting the problem. This issue has definately tarnished the image (no pun intended) of a camera that could have been so great

Shes very wrong. There are other choices. Z1, FX1, HC1, XL-H1, HVX, and many excellent SD cameras in the same prosumer price range. So far, of the ones that are available, JVC is the only one with a serious flaw.

No I didn't see SSE, but it was before anyone had ever raised the split issue so I didn't even think such a fault could exist and so didn't light to expose the problem, it was in a studio type of environment where lighting was soft, even and ample. The lighting was nothing like the lighting experienced in my line of work in the real world, but never before in any camera has this made any difference.

I spent at least about two hours with the camera, and at that stage was totally impressed.

I've said it before, my only real issue with this cam is the possibility of ending up with the SSE when shooting a scene in which I have limited control over lighting, and will not ever have more than one chance to shoot it, and it is live.

I have never disputed the suitability of this camera in other shooting environments.

I wonder if the dealer would be kind enough to let you take the camera for a couple of days and test it out? That is par for the course over here (JVC and Sony).

I'd be happy to try one for a wedding, I'd have to keep my DV cam handy though incase I need to switch back suddenly. I think I'd actually have someone shooting backup on DV for that job anyway. I'm up for the work and cost of paying someone for the day to shoot backup, but I doubt JVC are going to lend me a camera. My next video job of my own is on the 26 Nov.

I agree with you 110%. Its ridiculous for JVC, cause I don 't really think she is speaking with all facts and data, to come out with a camera that has split screen in low light. They should fix it. I Still think the fcamera is great and I will make the most of it, and from what I've seen it really looks amazing.

I really can't believe this, the camera has a serious flaw that no other camera on the market has and JVC aren't going to fix it. It's OK to say that this camera will only be used by serious professionals who know how to light, but I'm a professional, I know how to light and there is no way I am going to buy a camera with such a flaw. Fact is sometimes you have to use gain, particularly in corporate work, where you aren't always in control of the lighting and I wouldn't consider going on a shoot knowing that there are some shots I couldn't get because of a fault on the camera.

I'm very disappointed as we are starting to shoot more drama at the moment and the JVC would have been a great B roll camera as it seems much more suitable than the Z1.

I know JVC are the "Rebels" against Sony's "Empire", but no matter how many people say this camera is acceptable, the fact is it's not.

If Sony had released a camera with such a problem the forums would be spewing hate as we speak.

One sale lost.

So what's the problem? When shooting some serious drama, you have plenty of light AND plenty of time. You only need to know how to operate the cam then....

Try to do the math right... The SONY is INTERLACED, so it has to process (and 'read out') only half the pixels (i.e. data) per time-frame....

It is on average the same, but with twice the precision, you can easily calculate the JVC has to work twice as fast. (per 2 fields - 1 frame it is the same (1/25th of a second), but it is not quite the same at 1/50th of a second.)

The thing is when I'm shooting serious Drama I use higher end cameras! HDV is pitched at the lower end of the filmmaking spectrum where correct lighting setups and time are not always available.

It would also have to do corporate work to pay its way, I agree if all you are shooting is well lit material then I'm sure the camera is very good and certainly the footage I've seen is amazing. But my experience over 15 years of shooting Corporate, Broadcast and film tells me there are times when you NEED gain, whether you like it or not and sometimes in situations where you did not expect it.

Having the restriction of not be able to use gain at all is a deal-breaker for me personally as far as buying one goes. However I'm sure I'll still be renting one when I know the shoot is fully controlled like studio based work.

I just don't think the whole situation is going to do JVC's reputation much good at all.

Perhaps something was lost in translation due to I'm not
being a native English reader or writer. But isn't this just ONE
person that told ONE person that the camera will not be replaced?
I've heard others say it will be replaced if it shows under 9db gain.
Others got theirs replaced no questions asked. And a lot
of people don't even have the problem.

I think it's unfair to base a decition on what one rep told
one person when there is contradicting statements from
other reps.

Perhaps I misread the thread or missed some information?
But I generally don't base my facts on what one rep said
when there a bunch saying otherwise.

Kind and friendly regards
Andreas

Fair point - I hope that rep was wrong!

"Try to do the math right... The SONY is INTERLACED, so it has to process (and 'read out') only half the pixels (i.e. data) per time-frame...."

Interlaced is just half the pixels at twice the rate of progressive, same number of pixels per second though.

So again:

It's twice as much in one period (and doing nothing in the other period). It is more problematic. What you are saying is that they are ON AVERAGE the same (in 1/25 of a second). On 1/50th of a second the JVC reads out the whole CCD or reads out nothing. And it is even faster then that, look at the shutter speeds:

e.g. shutter speed 1/1000th of a second: the sony needs to read out only half the pixels in that time-interval as the JVC does. You can understand that the JVC's method is a lot more straining.

You were talking about a great B-camcorder. During well lit serious drama it should be fine then...

I do not have the problem
 

I do not have a split screen problem. When I purchased the camera from my dealer, we agreed that if there was one, the camera could be exchanged or returned. That pretty much covers it for me.

From this point forward, my company will be shooting all of our video projects with the JVC. We are now selling all of the other rigs we own. I hope that FCP and AVID will support the 720 24P mode in the very near future. I also, look forward to sharing information here other who are interested in getting the most from the camera.

-Dave Dessel

Steve Connor,

Go try out the camera. The chatter on the forums is not sufficient to your own eye. I'd be surprised if you didn't like the camera right out of the gate and on into a relationship. Do you have a dealer nearby?

Guy the point is threefold - first of all Sony restricted the size of their sensor to 960 pixels wide x 1080. JVC claims that in one third inch CCDs, 1000 is the magic number where you need to scan the chip in two halves and they speculate that this is why Sony have restricted the width to 960 (I know you assumed that the chip was full 1920 x 1080 - no such luck amigo).

Secondly an interlaced camera scans the same number of pixels on AVERAGE as the progressive camera, BUT it doesn't have to scan the whole chip at one time (which is the killer according to JVC), only each alternate line (although it does have to do it twice as fast - but not in this case as I'll explain next).

Thirdly the JVC actually scans the CCD @ 50/60Hz which is why they include a 50/60 progressive mode for standard definition.

With so many sensors crammed on to such a small chip, heat is believed to be the major problem.

And the irony here is that Sony make the GY-HD100 CCD.

While technically JVC may have made a huge error here, I think it's patently unfair to accuse them of being technically incompetent. Out of all the Japanese electronic firms, JVC have on the engineering front been the most innovative. This is what got them into trouble here - being innovative. Sony on the other hand have a reputation for design stodginess that has justifiably built them an enviable field reputation.

I'm also of the opinion that what one rep said to one person is not a company position - to date there has been no official announcement from JVC on SSE.

The HD100 does actually process the full 50p/60p even though it doesn't record it on tape.

1280x720x60=55296000
1440x1080x30=46656000

1280x720=921600
1440x540=777600

That gives the JVC an almost 19% increase.

The Z1 also only takes 960 pixels instead of 1440.

1280x720x60=55296000
960x1080x30=31104000

That gives the JVC a 78% increase.

The HC1 is only a single chip so maybe it only has 1/3 of the data to deal with.

1280x720x60x3=165888000
1920x1080x30x1=62208000

That gives the JVC a 267% increase.

The JVC HD1 and HD10 were single chip as well as only 30p.

1280x720x60x3=165888000
1280x720x30x1=27648000

This gives the HD100 a 600% increase.


We including JVC are very interested to see how Canaon has dealt with this issue since they claim to use full 1440x1080 chips. Even though this would still place it 19% under the JVC's datarate I don't think that would be enough to prevent this issue.

It was established 2 weeks ago that gain has almost nothing to do with SSE.

The facts:

1) You do not want to use gain above +12dB because of noise, not because of SSE. For film work, I would limit gain to +6dB for max. image quality. Not because of SSE.

2) If you need to use more than +12dB gain the color will be desaturate and the color balance will be off. (JVC should have limited gain to +12dB, as the gain above this point is useless.)

3) When you use gain, there must be enough light so that WITH gain you get an AVERAGE exposure of at least F4. Given that one often uses gain in situations that have an lot of dark areas and a few bright areas -- you want to expose the dark areas for F2. In short, you expose just as you would any negative film.

The only times you'll get SSE is if you are in a very dark situation where even using +12dB gain the iris still reports it is OPEN. That is your warning sign. If the iris is open at +12dB gain, you clearly do not have enough light.

REDEFINING SSE:

Every camera -- when not getting suficient light does something bad to the image. Some lose chroma saturation. Some go magenta or green. Some have fixed pattern noise. Some have a picture full of AM chroma noise. Some have a picture full of PM chroma noise. Some only have noise on reds. Some have luma noise. The JVC has SSE. All of these "bads" make the video unusable -- unless of course you capture some rare event -- in which case we know that even cell phone video will be bought.

With every camera -- one doesn't shoot in situations that provoke the "bad." Or, one adds light!

It's time to simply lump SSE in with all the other artifacts that occur from under-exposure. Just as we understand smear to be an artifact of point source over-exposure. (Something we also avoid if we want our video to look like film.)

BOTTOM LINE: the FX1/Z1 is about 2-stops less sensitive than the PD170/VX2100. The HD100 is about 2-stops less senstive than the FX1/Z1. We expect ANY progressive camcorder to be 1-stop less sensitive -- so the JVC may be about 1-stop less sensitive than we might have hoped.

We don't expect Sony to "fix" the FX1/Z1 to increase it's sensitivity.
Why would we expect JVC to" fix" the HD100 so that it's more sensitive?

If you want a high sensitivity video camera -- don't buy either a Sony or JVC HDV camcorder! Don't go HD -- stay in the DV world where you have 1 lux camcorders.

I didn't see this being established about two weeks ago.

I've now seen the camera, which I liked a lot, seen the SSE at 9db ( I live in PAL land where there seem to be more problems - JVC are swapping this unit out) and made the decision I won't even consider buying one until there is a clear statement from JVC about the issue. Something, for whatever reason, they have failed to do so far.

I like this camera, I don't have an agenda, I'm (still) a potential buyer of the camera.

Steve:

The SSE issue with the HD100 is a real factor for potential buyers. Not all of us, especially those of us shooting creative, narrative, documentary and pro often choose to shoot in less than ideal exposure conditions, often for creative and image effects.

"Correct exposure" is for engineers and technicians only - grain, noise, crushed blacks, blow-out highlights, smear, flare and various other artifacts are used to great creative effect. A camera is only a tool after all.

However, the SSE is not in a category of effect - it's unpredictable from unit to unit, firmware to firmware and unless you need a split exposure effect, not creatively useful.

SSE is a flaw, bottom line. Potential buyers can either live with it and deal with it via JVC or like me pass on the camera for now.

EDIT - I'm also unsubscribing from HD100 threads. Too much heat for just a pile of circuits (pun intended :) plus I'm going to pass on this cam for now.

Imagine a car, that has got a 5-speed, and after a lot of users have made crashes by using only the 3rd gear, the factory states, that the 3rd and 4th und 5th gear is for professionals only, otherwise You have to drive this car slowly only.

Steve, you just do not know, what we can expect from P2 during the next year, do You? And a FireStore solution for the HVX is already on the run. I am always interested in Your posts, if they are constructive, as they are for the most time.

Heads up, folks -- I have just deleted some personal attacks from several posts (leaving the technical discourse in place as much as possible). Just a reminder that flaming other members is never allowed at DV Info Net. If you want to harangue each other, you'll have to go elsewhere! Let's not get personal here please -- thanks in advance,

Over here in Europe 95% of all cars have manual gearboxes. No-one likes an automatic as it's less sporty, allows lower control, etc, etc....

Now, I've had a few cars already (crashed none of them, by the way), and NOT ONE car had a disclaimer in its manual that you HAVE to use the clutch to change gear...

I know you americans have the whole lawyer-culture, but do you think any user of a car that doesn't use the clutch (and his/her gearbox inevitably dies on him/her) stands a chance to have costs compensated? Or have a right to complain the car is of low quality?

... THAT seems a better analogy.

Most of the car accidents happen in low light situations, do they not?

:-)

that's a good one.

Question remains... you can't control the streets illumination, but you're the one to blame if you don't use your headlights.

Anyhow, I'm only trying to say I'm with steve on this. YOU DON'T HAVE TO break for corners, YOU DON'T HAVE TO have your oil and tyres checked, YOU DON'T HAVE TO stop for read lights, YOU DON'T HAVE TO whitebalance mannually, YOU DON'T HAVE TO go for the proper set-up, YOU DON'T HAVE TO ....

But if it goes (horrebly) wrong, who d'you have to blame?

(robert, this is by no means a personal attack to you... just so you know....)

"I know you assumed that the chip was full 1920 x 1080 - no such luck amigo"
Actually I know the CCD specs very well, which is why I detailed Sony having only 12% extra pixels, not over 100% extra (as 1920x1080 would have calculated)

"And the irony here is that Sony make the GY-HD100 CCD"
This is the first time anyone has indicated who actually makes the CCD. You know this for a fact or just guessing? I'm just curious thats all. I think it is not the CCD iteself that is the problem, it is potentially the circuits just after the CCD, so it could be the CCD itself is just fine.

"I think it's patently unfair to accuse them of being technically incompetent"
I didn't, but I did say they weren't up to this particular task. If being innovative produces cameras with faults like this then perhaps being Innovative isn't such a great thing, and perhaps Sony are onto something.

"to date there has been no official announcement from JVC on SSE"
This is a major concern to many in here, its like publically they completely deny the issue (but in private will discuss it),and that doesn't provide any confidence in them from an end users stand point. I guess it would be worse if they came out publically and denied any problems...