AG-DVX200 4K Large-Sensor Handheld Camcorder

Glad to hear you're testing it Matt- the big question from me is how long does it hold fast aperture through the zoom range? If it drops off to f4 as soon as you zoom in then that would be a real disappointment, but if it holds 2.8 to around 100mm then it's a different story...

Tom, Barry addressed this in a recent thread. The lens leaves f2.8 very quickly an is at about f4.2 at 100mm (FF eq.). Sad to say but the "f2.8" lens specs are more marketing than practical use.
 
Here you go:
f2.8 at 12.8mm
f3 at 13.9mm
f3.2 at 15.9mm
f3.4 at 19.9mm
f3.5 at 23.2mm
f3.6 at 24.9mm
f3.8 at 26.8mm
f4 at 29.9mm
f4.2 at 33.6mm
f4.5 84.6mm-167mm

The lens ramps almost instantly and doesn't really settle down until 33.6mm and above. The above figures are measurements straight off the camera.
 
Hello Matt. I have a very clear idear about the test conditions I would be interested in:

Extreme run and gun. Exactly: set the lens to constant f/4 as this will make it possible to shoot with a constant aperture for much of the zoom range. Show examples of the possible DOF in different conditions.

Shoot in low light conditions at ISO 6400 as in many occasions you will need such an ISO to compensate for the aperture of f/4.

Shoot handheld in HD while walking: how good will the 5-axis stabilization be?

Do an AF stress test.

Shoot HD in V-log and try different grades. What look can you achieve?

If the camera passes all this, I'm in.

I second this extreme run and gun question. This seems to be what the camera is designed for. I actually was able to handle the camera at EVS. Nicely balanced although bigger than expected.
 
Can "color science" and its finality to a released product be explained in a paragraph or less?
Cameras have digital signal processors in them that take the raw sensor data and translate it into color and brightness data. How that translation happens, how the colors are rendered, and all the other image processing, is what I'm lumping under the label "color science."

So at this point, the final image rendering is not complete. I know for a fact of certain aspects which simply aren't finalized yet. So the images it's making now, are not necessarily representative of what the final camera will make. They can change all sorts of parameters, from the intensity, hue, shade, and phase of the colors, or the aggressiveness of the noise reduction system, or -- well, anything they need to, until they get it delivering the images the way that they want it to.

That's why I say, any judgements made on the prerelease cameras are simply foolhardy. You cannot buy the prerelease cameras. The final camera may make images that are different from what you've seen from the prerelease cameras. There is literally zero point in pixel-peeping over preproduction camera footage, because things can, may, and potentially will change.

What is the engineering reasoning for having a higher bit rate - even if it's variable - at lower resolutions? To me, it should be the opposite, as it is on many products.
Well, in the case of 1080p at 200 megabits, the reasoning is that it's intraframe-only. You need more data in an intraframe codec, in a long-GoP codec it requires less bitrate to make comparable images.

But if you're asking why does 1080p get 200 megabits whereas 4K gets only 150, I can't give you a definitive reason but I can say that my guess is that there's only so much processing power to go around. These cameras don't have infinite processing power, there's only so much power that can be derived from a 7.2v battery, and so they try to make the most of what they can. If all the power is being absorbed in processing UHD video at 60 fps, then there may not be a ton of power left over for encoding a massive bitrate into a long-GoP codec. But that's speculation; I don't know the intimate details of what the limitations of the processors are, I only know what the results are, so I can tell you the results but I can only speculate as to how they get to those results.
 
barry or noel,

I'm curious how the servo zoom feels as to ease into and ease out when starting.
Really good.

is it possible to start a zoom and not notice it slightly jolt into action?
I say yes. I say it with the qualification that there certainly seem to be a whole lot of folks a lot more picky than me, so I'm sure someone somewhere will complain, but to me it has the smoothest, silkiest zoom of any Panasonic handheld camera ever.

my old EX1 was decent at this where my HMC150 was not as finessed when starting or stopping a zoom.
That's being generous. The HMC150 lurched noticeably when it started a zoom. The older series of cameras only had three servo speeds. I haven't documented how many the new camera has, but the slowest speed is incredibly slow, I've used the power zoom to adjust framing by 0.1mm before. It's much nicer than any prior Panasonic.
 
There's no way in hell I'm buying a pre-production version of this camera. No way. Sorry.
Nobody should be buying a pre-production version of anything, agreed.

And as long as I'm being brutally honest...I have a confession to make...have a guilty pleasure...I own an HC-X1000. And am actually very pleased with it. It's a great little camera once you learn how to properly use it. There - I've said it. It takes an investment of some time to get the best results from a new camera. And whether one is satisfied or not depends a lot upon how realistic one's expectations were when the camera was purchased.
Exactly.

Am going to make a bold prediction - even after production versions of the DVX-200 are into consumer hands, there will initially be lots of sub-par footage posted from those anxious to be among the first with "real" DVX-200 footage - which will be endlessly criticized and dissected on forums as examples of what the camera is not capable of. Haha...bold prediction? No, entirely predictable. And the current nay-sayers will have plenty of fodder to say, "I told you so."
Yep. This is exactly what will happen. And then one day MacGregor will get his hands on one, shoot something good, and the whole world will be stood on its head.

And one last thing I can say with very high confidence; Barry Green will tell it like it is.
I do try. There are things I can't say yet, due to nondisclosure agreements, but I am hinting when I say that I've documented specific things in the color science that aren't working properly yet, and I have every confidence they will be fixed long before anyone can buy a production model. That's just the fact -- there are unfinished elements, even at this point in the game. Will it change "everything"? No, of course not, but it could certainly explain some of the complaints.

My fondest wish is that the engineers and marketing team will learn from this, and shift to a more Apple-style product introduction. No loaners to people who post pre-production footage, no rumors, none of that. Just one day, launch a finished product, with a proper introduction and proper footage. Say "This is what it does, and you can buy it at your local dealer today." It would save endless months of hand-wringing and pixel-peeping and sniping and all that. That would be a much better way to introduce a product, given today's hyper-short-attention-span world.
 
Here you go:
f2.8 at 12.8mm
f3 at 13.9mm
f3.2 at 15.9mm
f3.4 at 19.9mm
f3.5 at 23.2mm
f3.6 at 24.9mm
f3.8 at 26.8mm
f4 at 29.9mm
f4.2 at 33.6mm
f4.5 84.6mm-167mm

The lens ramps almost instantly and doesn't really settle down until 33.6mm and above. The above figures are measurements straight off the camera.

And here in lies the Achilles heel of the camera. It will will still be very useful for many functions but most of the camera's limitations will centre around this lens. I do think it would be worthwhile for Panasonic to make a version of this camera with a more robust and flexible lens, even if it means hiking up the price by $3000-$4000. I know this is not exactly the same sensor as the GH4 but I think its close enough to make some comparisons as far as light sensitivity and the only way to make to the GH4 usable in low light is with lenses that very wide apertures ( F1.2, 1.4 etc..).

Thanks for the detailed breakdown Matthew.

Any sort of run and gun indoor lighting based testing would be nice to see. Not necessarily footage of your cat beside a candle but maybe a person (to get skin tones) being interviewed in less than ideal light. Indoor wedding type scenarios. Banquet hall lighting, speeches etc... you know that kind of thing.

Thanks
 
I also find it hard to believe the image from the final camera will be a full change from what is now.
Of course it won't be, and nobody ever said it would. I've said that it is "not necessarily representative", because things can change.

The problem is that we just don't know if we are seeing the best the camera can do now.
That is absolutely guaranteed.

All these people posting footage from the DVX200 seem to be totally amateurs.
I wouldn't say that, some are quite skilled, but I'd dare say few if any have had enough time with the camera to learn its particulars and fine-tune it towards its sweet spot. If someone gave you a camera with the level of complexity of this camera, and said "you have it for two days, then I need it back" -- would you be confident you'd extract the best that the camera could do? I sure wouldn't.

I think you have unrealistic expectations though. All that from a $4200 camera just won't happen. There will always be catches. I have a FS7 which costs much more without a lens. It has a good share of catches and pisses me off on a daily basis. But I know it's not a F55 or F5 and costs much less. So I live with it.
Exactly what I've been saying. This DVX200 is almost 1/3 the cost of an FS7 with lens. Guess what -- the FS7 is more camera. I mean, really -- should we even have to explain that? The DVX200 isn't going to out-film a dedicated cinema camera, and there's no way in hell I'd go shooting a football came on a BlackMagic Cine camera. The DVX is an events/live production/reality TV/documentary camera that can also do film-style shooting (within the limitation of its iris). It's unlike anything else out there at its price point. It should be evaluated based on what it does and what it costs, and whether or not what it does is what any given user wants or needs.
 
Exactly what I've been saying. This DVX200 is almost 1/3 the cost of an FS7 with lens. Guess what -- the FS7 is more camera. I mean, really -- should we even have to explain that? The DVX200 isn't going to out-film a dedicated cinema camera, and there's no way in hell I'd go shooting a football came on a BlackMagic Cine camera.

Thanks for all your contributions Barry.

This is exactly my frustration with Panasonic's treatment of their 4/3's sensor products- DSLR's and lower end prosumer camcorders ala the DVX200. I want a Panny that DOES try to compete with the FS7, with less compromises albeit at a higher price than the DVX200. I love 4/3's, but Panasonic keeps on serving up salads when I want a meal... Their current 4/3's product lineup leads me to believe that 4/3 is eventually going to be a dead end format. I hope I'm wrong, I'm a huge fan of the sensor size.

The irony of all this is their extensive lineup of great still lenses and very competent still cameras- and yet no higher end implementation of the 4/3s sensor....? They absolutely need to introduce a constant aperture 14mm-300mm ish range video lens. F2.8 would be great sure- but I'd settle for a constant F3.5 or even a 2.8-4 variable.

Maybe they suffer from the delusion that we are all going to go out and buy a Varicam. Not happening. Panasonic, how's the 2/3 inch market going for you? ;-\ If this is indeed their strategy ( I hope not) , it's extremely short sighted and certainly leads me to pessimism regarding the format for motion imaging. This may end up being the only viable application of the sensor size anyway. Unless their is some dramatic progress on the sensor size and low light capability, its days as a viable still sensor format may be coming to an end as well- sadly.

... Dear Panasonic, I'm not buying a Varicam- but I want something more than a DVX200... or at least a DVX200 that does not have some of these compromises. Money and budgets are tighter these days, and I want a system that has some legs, a system that I can invest into. I'm a huge fan of the 4/3s sensor for motion work, but your commitment to it appears tepid and unsure.

Sincerely,
A current HPX300, GH3,GH4 and GX1 owner. A former DVX100, DVX100A, HVX200 and HPX170 owner.
 
If Panasonic can not beat the LX100 in image quality with the DVX200 they need to fold up their tent! I think you are not giving the DVX200 a fair shake. Until you can put the camera next to the LX100, setup the camera properly with its pro menu options and shoot a side by side you are just guessing. I like the look of the LX100 and would own one if they did not put the stupid 30min recording restriction on it. But I think the DVX200 will have a great image, especially with all of the menu controls.

I think I'm fair shaking the DVX200 as I'm looking at the footage at hand. What else can I judge IQ by?

Anyways the talk about getting to know the camera and set it up properly before the good stuff comes out sounds a little bit ridiculos! All cameras I've owned that have shown good IQ didn't require any magic settings. They just looked good by the core IQ. Just like the LX100 does. I put the LX100 in auto and it looks good. My FS7 looks great in Cine EI without touching any settings in order to tweak the IQ. My 5D3 looks good on its own without altering settings. My EX3 looked good, my FS100 looked good my F3 looked good. My AF100 really never took off. But I spent to much time trying to get the good stuff out if it.

When I'm looking at the DVX200 footage at hand I can't but feel it's the AF100 all over again...
 
Ha! Thanks.

Strangely, I would have to agree with Osslund's post above. You turn on the EX1, and even without a profile applied, it makes a very nice image. Shouldn't the DVX be the same? I realize their design philosophies are different (I don't think you can have "no scene file" on this cam like you can on the EX1. . .you're ALWAYS on a scene file on the DVX, right? Even if you don't tweak it?)
 
Re: zoom servo on the 200
Panasonic seems to have finally done away with old 3 speed servo control that has to "click into the gear tooth" when zooming out in servo mode. I've always used my 100a in manual zoom mode because of this engineering decision on the 100a, 200, 150, etc.
The smooth pull out/creep in servo mode on the DVX200 is a real blessing, and one of the things I really like about the camera. The zoom is also very well damped in manual mode as well.
I tried it several times on one of the cameras at the EVS demo last Tuesday (Thanks Jim and the guys at EVS).
Also, I found the noise level at ISO 1600 (4K 24p) barely noticeable on the 30" demo monitor...even close up to the screen. This was true in both Vlog and "Film Like #1 settings. Didn't try it in 1080p though.
Hope this helps.
Ken
 
@Tom Lenham: "how long does it hold fast aperture through the zoom range?"
I checked that on the EVS demo camera last week. The iris drops to 3.4 after zooming in only about 10% of the lens range, and drops rather quickly from there.
If needed relatively constant aperture throughout the zoom range, I'd be using f/4.0 all the way. I was hoping for the 2.8 at least 1/2 way through the range, but it just isn't there.
Hope this helps.
Ken
 
Philip,
No, I didn't. I don't believe it was in high sensitivity mode. The available light in the demo room was pretty dim...probably no more than 16-20 footcandles, so I was pleased that the 1600 ISO looked so decent.
Ken
 
My fondest wish is that the engineers and marketing team will learn from this, and shift to a more Apple-style product introduction. No loaners to people who post pre-production footage, no rumors, none of that. Just one day, launch a finished product, with a proper introduction and proper footage. Say "This is what it does, and you can buy it at your local dealer today." It would save endless months of hand-wringing and pixel-peeping and sniping and all that. That would be a much better way to introduce a product, given today's hyper-short-attention-span world.

Amen.
 
I think I'm fair shaking the DVX200 as I'm looking at the footage at hand. What else can I judge IQ by?

Anyways the talk about getting to know the camera and set it up properly before the good stuff comes out sounds a little bit ridiculos! All cameras I've owned that have shown good IQ didn't require any magic settings. They just looked good by the core IQ. Just like the LX100 does. I put the LX100 in auto and it looks good. My FS7 looks great in Cine EI without touching any settings in order to tweak the IQ. My 5D3 looks good on its own without altering settings. My EX3 looked good, my FS100 looked good my F3 looked good. My AF100 really never took off. But I spent to much time trying to get the good stuff out if it.

When I'm looking at the DVX200 footage at hand I can't but feel it's the AF100 all over again...

You may right buy I think you are wrong to judge a camera before it is officially released. What do you have to gain? Bragging rights that you called it first? I never liked the image out of any AF-100 clip, so I know where you are coming from but picking apart Youtube footage from an unleased camera seems like a dead end road.

BTW, I have always had to setup my cameras to get the best out of them. Images always improve as you learn the camera's strengths and weaknesses. If I remember, the out of the box image from the EX-1 was pretty bad imho. But I never loved that camera's image either. But I see your point. If the DVX-200 still needs "help" after the right reviews and footage then blast away!
 
Back
Top