AG-DVX200 4K Large-Sensor Handheld Camcorder

do you suppose you could post sole footage that demonstrates once and for all how sharp that lens is? some uncompressed stuff? since there was so much grumbling over that? thanks.
I can't think of anything Barry could post that is more demonstrative than a resolution chart - and what was seen previously shows resolution up to the system limit. Why uncompressed video? What is that going to show you (about resolution) that a still won't?

I suppose it could be argued that was from the centre of the chart, so for a fuller test maybe something to show corner sharpness at max aperture? And at different zoom settings? But whilst such is inevitably likely to look worse, then, well, same applies to pretty well all such lenses. It's also likely to use in-camera processing to reduce such as chromatic aberration at the edges. (As such as the RX10 do.) I just don't see sharpness (or lack of it) as any issue with this camera, and there's no reason it should be if it fully reads and deBayers the whole chip.
 
wow. all i was asking was to see some uncompressed footage (vs the youtube stuff weve been seeing), as there have been some folks complaining about the "lack of real detail" in the images posted this far. if the yt compression is contributing to the softness, uncompressed might help show what the cam can really do. make sense?
 
Would that be a fair assessment-- noise on the DVX200 at 3,200-4,000 ISO is comparable to the AF100 at 1,600-2,000 ISO (as of right now, on a pre-production model, with v0.32 firmware which may or may not improve)?
Well, as always, the answer is complex. The AF100 had chroma noise that the DVX200 just doesn't have. So even if there were to be comparable amounts of overall grain, the DVX200 would still look cleaner because it wouldn't have any of the colored chroma noise.

Then, there's lens considerations. I don't have the 12-35 or 35-100, so I have to make do with the 14-140. I just did a very low-light shot (one 40w fluorescent bounced into the ceiling), and aimed the AF100 at a scene, put it on 3200 ISO and zoomed in wherever it needed to be, opened the iris up, and shot it. I then put the DVX200 in the same position, zoomed to the same spot, opened up the iris, put it on "high sensitivity" mode (because, why wouldn't you?) and then set the ISO to match the AF100's exposure. It only needed 1250 ISO to match what the AF100 needed 3200 ISO to deliver.

So it's definitely better. And yes, the noise was cleaner, and the footage way sharper. We all know the AF100 was never the sharpest camera; Panasonic rated it at 800 TVL. I believe they rate the DVX200 in excess of 1000 TVL when in 1080 mode. In any case, it's sharper and clearer in 1080p, obviously it's insanely sharper when in 4K vs. the AF100's 1080p.

But it is not as easy as you might think to make such a noise comparison, because the gamma curves and color are not quite the same between the two cameras yet. Even using the same settings, the images just look different, the AF100 has more saturated colors, the gamma curves are different, if you expose to get the highlights to match brightness then the shadows are at different levels, etc. I could spend a while trying to tweak them to match better, but that's wasted effort at this point because the next version of firmware might change the color science and invalidate that work.

There is some noise in a low-light scene at 1250 ISO, if you pixel-peep for it and magnify it you will see it. But I can't imagine as to how anyone would view it as a problem, it looks plenty clean; it's certainly notably less than on the AF100 when at equivalent brightness. And, furthermore, this was with zero effort taken towards minimizing noise. This was with detail level 0, v-detail 0, in a medium-noise gamma, etc. If I were to tweak the settings towards minimizing noise I'm pretty sure it would be notably cleaner.

I'm not saying it's a miracle of absolutely zero noise anywhere. There is some noise in the image, about the same level at 800 ISO as the HPX250 has when the HPX250 is at 320 ISO. But I find the level of noise notably lower, and much more pleasing, than the AF100 has when at equivalent exposure.
 
Last edited:
wow. all i was asking was to see some uncompressed footage (vs the youtube stuff weve been seeing), as there have been some folks complaining about the "lack of real detail" in the images posted this far. if the yt compression is contributing to the softness, uncompressed might help show what the cam can really do. make sense?
Sorry if my previous remarks came across as more harsh than intended! But I really don't see what uncompressed footage is going to tell you that a high quality frame grab won't - at least in so far as sharpness goes. And I doubt if you'd see any perceivable sharpness difference between uncompressed and normal in-camera compression anyway.

That's really the whole point of charts and test images. They test individual attributes of performance in a controlled fashion. In the case of resolution charts they even distinguish between "real" detail, and detail enhancement, or aliasing etc etc. "Real" images make for much more interesting "real" films - but are much less use for camera evaluation, it's way more difficult to tell what's causing what. I wouldn't shoot charts for a real film, but trying to evaluate camera performance on "real images" isn't sensible either.

Let's say you did get some uncompressed footage. But what conclusions could you draw - resolution results would likely vary with zoom angle, aperture it was shot at, let alone whether it was centre or edge of frame? If that's all there is, then OK, better than nothing, but when a (very good) chart result image has already been presented, then what else can other footage tell you?
 
Second test, I put the DVX200 in 3200 ISO, so it was 3200 vs. 3200. The DVX's image was much brighter, as you'd expect. In some places it was a lot cleaner in noise, in some not as much. Like all the Panasonics, the noise affects certain brightness regions; by pushing much of the image up out of that darker area, it became cleaner in noise even though it was at a higher ISO. But the areas that were subject to noise became noisier because of the higher ISO. Never noisier than the AF100, but noisier than it was at 1250 (of course).

once I start getting into the 1,600-2,000 ISO range, the noise starts to make the footage unusable
Can you share a clip that you consider unusable?
 
Second test, I put the DVX200 in 3200 ISO, so it was 3200 vs. 3200. The DVX's image was much brighter, as you'd expect. In some places it was a lot cleaner in noise, in some not as much. Like all the Panasonics, the noise affects certain brightness regions; by pushing much of the image up out of that darker area, it became cleaner in noise even though it was at a higher ISO. But the areas that were subject to noise became noisier because of the higher ISO. Never noisier than the AF100, but noisier than it was at 1250 (of course).


Can you share a clip that you consider unusable?


This was the first thing I could find digging through some old stuff (2012). Uploaded the .MTS file straight to Vimeo. I hadn't bought the 35-100mm yet, so not quite sure what lens I was using ([EDIT] Nikon 28-70 f2.8). ISO would have been between 2,000 and 3,200.

[EDIT] inserted a different clip
 
Last edited:
oh sorry were HQ stills posted? guess i missed them. when i say "uncompressed" i mean in terms of camera-original quality footage vs something thats been chewed on by youtube or vimeo.
 
This was the first thing I could find digging through some old stuff (2012). Uploaded the .MTS file straight to Vimeo. I hadn't bought the 35-100mm yet, so not quite sure what lens I was using ([EDIT] Nikon 28-70 f2.8). ISO would have been between 2,000 and 3,200.

[EDIT] inserted a different clip
Yeah.... Those crawly blues in the blacks. I dealt with that heavily on a few AF100 shoots. FCPX color picker helped a lot in taking the color saturation out of that noise for me.

But we are talking about the DVX200 here. I'm VERY glad to hear from Barry that the chroma noise is much less on the DVX200. For me, if the DVX200 also has better over-exposure roll off (especially in the chroma (yellow)), I'd be a VERY happy man!
 
Yeah.... Those crawly blues in the blacks. I dealt with that heavily on a few AF100 shoots. FCPX color picker helped a lot in taking the color saturation out of that noise for me.

But we are talking about the DVX200 here. I'm VERY glad to hear from Barry that the chroma noise is much less on the DVX200. For me, if the DVX200 also has better over-exposure roll off (especially in the chroma (yellow)), I'd be a VERY happy man!

There is enough picture information that we were able lighten it up in post... but it also enhanced the "crawly blues" (I like that description :) ). Neat got rid of a lot of it, but at the cost of messing up the video in other ways. We had to go back and shoot another operation with the room lights on for a little while (which is now our standard procedure for this sort of thing)... which sucks, because the effect from the lights from the instruments and monitors looks really cool.

This is why I'm asking so much about the DVX200's low light. If it could handle the above situation (both in sensitivity and noise), while allowing me to not worry about having to swap lenses... I'm more than okay with the f/4.5 aperture at the long end (which I would never need as I'd rarely go past about 115mm at most). This will be a dream camera for low-light live events (like presentations and wedding receptions) and docs.
 
oh sorry were HQ stills posted? guess i missed them. when i say "uncompressed" i mean in terms of camera-original quality footage vs something thats been chewed on by youtube or vimeo.
What I was referring to was Barry's post here - http://www.dvxuser.com/V6/showthrea...corder&p=1986564185&viewfull=1#post1986564185

He states that they are pixel-for-pixel extracts, so take that as good as it gets for looking at resolution, basic resolution at any rate. And they undoubtedly show performance which is very good, close to what the system is capable of. OK, they don't say anything about sharpness at extremes of aperture, or corner performance, but ..... where do you draw the line? "Camera-original quality footage" would be no different, and such a chart result is conclusive - measurable - in a way "real life" pictures are not. (Whether Youtube or uncompressed.)

So can "is the camera sharp enough?" be put to bed once and for all?

Before anybody else says it, then yes, sharpness isn't everything, and the DVX200 may have some foibles - but let's wait for a final version. Personally, I suspect the 4K 60p mode may be the biggest single negative for many people, with it's wide angle being so much more restricted than the other modes due to the cropping. (And I don't see much likelihood of that changing before release?) But if you don't want 60p, then so what?
 
New footage, shot in Cinelike-V (not: V-log), no grading added. Looks very good, except too much YouTube compression. Following Barry's observations, I would expect about one stop more of DR if this was shot in V-log. Anyway, already it's quite a good result IMO.
 
New footage, shot in Cinelike-V (not: V-log), no grading added. Looks very good, except too much YouTube compression. Following Barry's observations, I would expect about one stop more of DR if this was shot in V-log. Anyway, already it's quite a good result IMO.
This is baked in footage with no grading? Then I am impressed. I thought the editor just had poor taste in grading style. But to be that good this early with the camera and baked in?

things are looking really good for this camera!
 
New footage, shot in Cinelike-V (not: V-log), no grading added. Looks very good, except too much YouTube compression. Following Barry's observations, I would expect about one stop more of DR if this was shot in V-log. Anyway, already it's quite a good result IMO.
I didn't see where they said this is ungraded. Please tell me as this is a very important point.
 
At the beginning of the video there is a written sum up of the shooting conditions. There (at 0:14) this is explicitly stated.
Ah. I didnt click the vid you linked (the BTS) as I had already seen the release video and that's what I thought you linked. The release video didn't mention this.
 
Why on earth did they shoot the BTS with a DSLR instead of a GH4 or G7???? wtf Panasonic marketing

This wasn't Panasonic Marketing who did this, it was one of the folks Panasonic lent a test camera to to try out who decided to put this together on his own. From the Youtube page--

After testing the Panasonic DVX200 in various environments like ENG and EVENT style shooting I wanted to see how it performed on set, 'Cinema Style'
My friends and I pulled together this little action scene in a day and used some real world cinema tools like a Flextrack Dolly, Cartoni Smarthead and a Jib from Porta-Jib. ( http://www.porta-jib.com )

We also used 3 LED panels from F&V ( http://www.fvlighting.com )

The Panasonic DVX200 gets a very nice image. Even though the fixed lens takes some getting used to in a cinema style workflow the final result is undeniably 'filmic' .

The true strength of this camera might be News, Event, Documentary and Reality TV.

Thanks to Panasonic for letting me test run their beautiful camera.
http://pro-av.panasonic.net/en/dvx4k/
 
New footage, shot in Cinelike-V (not: V-log), no grading added. Looks very good, except too much YouTube compression. Following Barry's observations, I would expect about one stop more of DR if this was shot in V-log. Anyway, already it's quite a good result IMO.

Hopefully they also shot the exact same scenes in V-Log mode as well. Isn't it the same V-Log L as the GH4? I'm sure they could've acquire the LUT for it. Nice scene. Maybe next time how about some realtime zooms. It's really important.
 
Back
Top