joe 1008
Veteran
Yes. You'll have to find the right place to stick an Atomos Niña Blade or something similar.
AG-DVX200 4K Large-Sensor Handheld Camcorder
- Thread starter Screenforme
- Start date
Yes. You'll have to find the right place to stick an Atomos Niña Blade or something similar.
filmfan222
Well-known member
That's kind of my point. I currently own a Sony FS7 with metabones and set of lenses. I also own a BMPCC, BMC4K and all the rigging associated with all those kits. I just wanted to get rid of all these and rely on one piece of equipment. The FS7 produces a beautiful image, but the menus are awful. The BMD cameras have brilliant menus and 10bit, but require a fair amount of rigging. I just was really hoping Panasonic would produce something that kind of hit all the right sweet spots for our work so we could replace those cameras with 1-2 DVX200s.
Barry_Green
Moderator
Nice overview by Jonathan Lawrence. I agree with him on most everything, although there are a few things they said that I think might need some clarification or even correction...
1) They're right about the LCD, it's gorgeous, but it's quite reflective and that does make it harder to see in daylight. You may need a hood, or to use the EVF.
2) Focus -- I reported this to Panasonic. Prior Panasonics were unique among free-spinning focus ring cameras; the Panasonics always had precise repeatable focus rings that you could actually set marks on, and they would hold. In the preproduction DVX200, they don't always hit the exact spot. That may or may not change; I'm hoping it does. The DVX200 has three ways to use the focus ring, it's got a "fine" mode where it's extremely long throw for excellent precision, it's got a "fast" mode for ENG-style work where the travel from minimum to maximum is a lot less, and it has a "speed" setting, where the focus distance moves further the faster you move the ring, and less the slower you move it (so it's like a combination of "fine" and "fast"). I don't see how the "speed" mode could be made repeatable, but I'm hoping the "fine" and "fast" modes can. In any case, there's another much better solution, which is that you don't have to mark the focus ring at all, you can just use the camera's "focus transition" function. You program in a start point and an end point, and the speed at which you want the transition to happen, and the camera will handle shifting focus for you. With that function, it will never ever miss the focus mark.
3) They said the lens is "28 to 167". That's incorrect, they're mixing the actual millimeters with the "full frame equivalent". The actual lens is 12.8 to 167, and the FF equivalent is 28 to 365mm.
4) Mr. Lawrence praises how lightweight it is. Compared to, say, a Red One, yes it is. Compared to a GH4, no, it's more like a block of iron. So keep in mind the weight of the camera, it's designed for handheld operation and it is very well balanced, but whether you consider it "light" or "heavy" is entirely relative to your frame of reference. It's a little heavier than a PX270/HPX250/AC160 cameras. They mentioned it as being lighter than the Sony EX1; I think that was probably incorrect; the EX1 specs I can find say that it was 2.8kg with battery, memory card and lens hood; the Panasonic specs say the DVX200 is 2.7kg without battery, lens hood or accessories. So I would say it's about the same weight as an EX1, perhaps a tiny bit more.
5) at 5:10 in the interview he says that it's 8-bit 4:2:2. That's not correct for internally-recorded footage. It's always 8-bit 4:2:2 output (unless you choose 10-bit 4:2:2 output). But the recorded footage is always 8-bit 4:2:0. The fact that they did a whole preproduction trip, evaluation, short film and other footage, and never noticed the limitations of it being 4:2:0 probably seem to imply that its 4:2:0 ain't that bad, right? But for technical accuracy, what he stated is incorrect, it's recorded as 4:2:0.
6) Dynamic range, noise, and V-Log. Jonathan said "between 10 and 12 stops, it boasts 12 stops". That's all correct. The camera in V-Log delivers a full 12 stops. In the other gammas, it's less; they said specifically that they shot in Cine-V and Cine-D; I don't have the numberrs in front of me but I'd be surprised if Cine-V is holding more than 10 stops, so I'm saying that their observations, taken in whole, are correct; depending on the gamma you use, it will affect the total dynamic range. And yes, VLOG is noisier than the other gammas. VLOG is as close to "raw" images as you're going to get from this camera; it bypasses nearly all image processing (such as sharpening, color adjustments, etc). So you're going to have to do your own noise reduction in post, as a standard part of your workflow. The grading process for VLOG is going to include gamma adjustment, color correction, color saturation, noise reduction, and sharpening, just like you'd have to do in a RAW workflow.
7) Glad to see them confirm my observation that the noise (in full color) is "virtually undetectable". It really is rather clean at 500 ISO, or even 1000 ISO. It gets noisier as you go higher, but in high sensitivity at 2000 ISO it's an effective 4000 ISO and it's still reasonably clean. I know that there are going to be pixel-peepers who see some shadow noise and complain about it, but -- looking on a monitor from a reasonably proper viewing distance, the image is quite clean.
8) at about 10:20, they mention lag. There is lag on the LCD, but there's less lag on the EVF, and even less lag on the external monitor. So their concern over the external monitor issue may not be really warranted, the external monitor is the least laggy of all the display options. I don't know if they were basing their concerns on the pull-out LCD, or on an actual external monitor, but it does make a difference of a few frames.
9) Jello. There is definitely some jello in the UHD 24p and 30p. Much less in UHD 60p, and extremely little in 1080p, 1080i, 720p, or 480i or 480p. And yes, it's obviously much exaggerated on the long end of the lens, just like it is on any camera. From my limited experience with the GH4, I'd say the rolling shutter was about comparable between the GH4 and the DVX200; they're certainly in the same ballpark. Now, if you go zooming in to 167mm and wave the camera around, you'll see lots of skew and/or jello... try that with a GH4 and you'll see the same thing. Rolling shutter is more pronounced at slower frame rates and higher image magnification -- again, just like on most rolling-shutter cameras. And, I may point out, this is not a finished product, and there is the possibility that this may be improved or changed before the final release version. I'd also add -- the LCD panel exaggerates the perception of skew and jello -- presumably because it scans differently. The EVF is a much more accurate gauge of what the recorded image will look like, when it comes to skew or jello.
10) at 12:28, they're talking about putting a Letus 35mm adapter on the front. Really? I mean, yes, if you need to eke out another stop or two of DOF then maybe, but I can't imagine anyone would go back to that. The DVX200 eliminates the need for that for almost everything... if I was needing a super-shallow DOF shot for some purpose or other, I'd probably use a GH4 with a Voigtlander rather than try to work with some spinning-ground-glass 35mm adapter. The whole concept of using one of those runs counter to the whole point of the DVX200... so... yeah, I don't get that. But to each his own I guess.
11) at around 14:40 they're talking about the ports, and he says that it has TC in/out. Just some clarification, that's not a full-time jammed sync signal; it's used for synchronizing the timecode together on two cameras (or an audio recorder or slate or some other device) but then you have to disconnect them and each maintains its own timecode from then on. It's labeled TC PRESET IN/OUT, and it can transmit and sync to other devices, but not in a continuously-cabled mode.
12) At 15:02 they're talking about the HDMI port. That's an HDMI 2.0 port, so it does support full UHD/59.94p. Many monitors (and even many recorders) don't yet support HDMI 2.0, they are frequently HDMI 1.4 (which limits UHD or 4K to 29.97). The DVX200 is compatible with those ports; it plugs right into them and works properly, you can set a menu option that says to limit the HDMI output to 29.97 to make it fully compatible with HDMI 1.4 devices.
1) They're right about the LCD, it's gorgeous, but it's quite reflective and that does make it harder to see in daylight. You may need a hood, or to use the EVF.
2) Focus -- I reported this to Panasonic. Prior Panasonics were unique among free-spinning focus ring cameras; the Panasonics always had precise repeatable focus rings that you could actually set marks on, and they would hold. In the preproduction DVX200, they don't always hit the exact spot. That may or may not change; I'm hoping it does. The DVX200 has three ways to use the focus ring, it's got a "fine" mode where it's extremely long throw for excellent precision, it's got a "fast" mode for ENG-style work where the travel from minimum to maximum is a lot less, and it has a "speed" setting, where the focus distance moves further the faster you move the ring, and less the slower you move it (so it's like a combination of "fine" and "fast"). I don't see how the "speed" mode could be made repeatable, but I'm hoping the "fine" and "fast" modes can. In any case, there's another much better solution, which is that you don't have to mark the focus ring at all, you can just use the camera's "focus transition" function. You program in a start point and an end point, and the speed at which you want the transition to happen, and the camera will handle shifting focus for you. With that function, it will never ever miss the focus mark.
3) They said the lens is "28 to 167". That's incorrect, they're mixing the actual millimeters with the "full frame equivalent". The actual lens is 12.8 to 167, and the FF equivalent is 28 to 365mm.
4) Mr. Lawrence praises how lightweight it is. Compared to, say, a Red One, yes it is. Compared to a GH4, no, it's more like a block of iron. So keep in mind the weight of the camera, it's designed for handheld operation and it is very well balanced, but whether you consider it "light" or "heavy" is entirely relative to your frame of reference. It's a little heavier than a PX270/HPX250/AC160 cameras. They mentioned it as being lighter than the Sony EX1; I think that was probably incorrect; the EX1 specs I can find say that it was 2.8kg with battery, memory card and lens hood; the Panasonic specs say the DVX200 is 2.7kg without battery, lens hood or accessories. So I would say it's about the same weight as an EX1, perhaps a tiny bit more.
5) at 5:10 in the interview he says that it's 8-bit 4:2:2. That's not correct for internally-recorded footage. It's always 8-bit 4:2:2 output (unless you choose 10-bit 4:2:2 output). But the recorded footage is always 8-bit 4:2:0. The fact that they did a whole preproduction trip, evaluation, short film and other footage, and never noticed the limitations of it being 4:2:0 probably seem to imply that its 4:2:0 ain't that bad, right? But for technical accuracy, what he stated is incorrect, it's recorded as 4:2:0.
6) Dynamic range, noise, and V-Log. Jonathan said "between 10 and 12 stops, it boasts 12 stops". That's all correct. The camera in V-Log delivers a full 12 stops. In the other gammas, it's less; they said specifically that they shot in Cine-V and Cine-D; I don't have the numberrs in front of me but I'd be surprised if Cine-V is holding more than 10 stops, so I'm saying that their observations, taken in whole, are correct; depending on the gamma you use, it will affect the total dynamic range. And yes, VLOG is noisier than the other gammas. VLOG is as close to "raw" images as you're going to get from this camera; it bypasses nearly all image processing (such as sharpening, color adjustments, etc). So you're going to have to do your own noise reduction in post, as a standard part of your workflow. The grading process for VLOG is going to include gamma adjustment, color correction, color saturation, noise reduction, and sharpening, just like you'd have to do in a RAW workflow.
7) Glad to see them confirm my observation that the noise (in full color) is "virtually undetectable". It really is rather clean at 500 ISO, or even 1000 ISO. It gets noisier as you go higher, but in high sensitivity at 2000 ISO it's an effective 4000 ISO and it's still reasonably clean. I know that there are going to be pixel-peepers who see some shadow noise and complain about it, but -- looking on a monitor from a reasonably proper viewing distance, the image is quite clean.
8) at about 10:20, they mention lag. There is lag on the LCD, but there's less lag on the EVF, and even less lag on the external monitor. So their concern over the external monitor issue may not be really warranted, the external monitor is the least laggy of all the display options. I don't know if they were basing their concerns on the pull-out LCD, or on an actual external monitor, but it does make a difference of a few frames.
9) Jello. There is definitely some jello in the UHD 24p and 30p. Much less in UHD 60p, and extremely little in 1080p, 1080i, 720p, or 480i or 480p. And yes, it's obviously much exaggerated on the long end of the lens, just like it is on any camera. From my limited experience with the GH4, I'd say the rolling shutter was about comparable between the GH4 and the DVX200; they're certainly in the same ballpark. Now, if you go zooming in to 167mm and wave the camera around, you'll see lots of skew and/or jello... try that with a GH4 and you'll see the same thing. Rolling shutter is more pronounced at slower frame rates and higher image magnification -- again, just like on most rolling-shutter cameras. And, I may point out, this is not a finished product, and there is the possibility that this may be improved or changed before the final release version. I'd also add -- the LCD panel exaggerates the perception of skew and jello -- presumably because it scans differently. The EVF is a much more accurate gauge of what the recorded image will look like, when it comes to skew or jello.
10) at 12:28, they're talking about putting a Letus 35mm adapter on the front. Really? I mean, yes, if you need to eke out another stop or two of DOF then maybe, but I can't imagine anyone would go back to that. The DVX200 eliminates the need for that for almost everything... if I was needing a super-shallow DOF shot for some purpose or other, I'd probably use a GH4 with a Voigtlander rather than try to work with some spinning-ground-glass 35mm adapter. The whole concept of using one of those runs counter to the whole point of the DVX200... so... yeah, I don't get that. But to each his own I guess.
11) at around 14:40 they're talking about the ports, and he says that it has TC in/out. Just some clarification, that's not a full-time jammed sync signal; it's used for synchronizing the timecode together on two cameras (or an audio recorder or slate or some other device) but then you have to disconnect them and each maintains its own timecode from then on. It's labeled TC PRESET IN/OUT, and it can transmit and sync to other devices, but not in a continuously-cabled mode.
12) At 15:02 they're talking about the HDMI port. That's an HDMI 2.0 port, so it does support full UHD/59.94p. Many monitors (and even many recorders) don't yet support HDMI 2.0, they are frequently HDMI 1.4 (which limits UHD or 4K to 29.97). The DVX200 is compatible with those ports; it plugs right into them and works properly, you can set a menu option that says to limit the HDMI output to 29.97 to make it fully compatible with HDMI 1.4 devices.
Last edited:
Matthew Allard ACS
Member
Part 2 of my review of the DVX 200 is now up.
http://www.newsshooter.com/2015/08/25/panasonic-dvx-200-review-part-2-v-log-l-scene-profiles-and-colour-science/
http://www.newsshooter.com/2015/08/25/panasonic-dvx-200-review-part-2-v-log-l-scene-profiles-and-colour-science/
philiplipetz
Veteran
"I found that while the image is nice and flat, the way it rolls off highlights on the DVX 200 is not pretty. On the built-in waveform monitor, or when using false colour, the V-Log L image may well show no evidence of highlight clipping, but if you look at the image closely the brightest parts of the image have almost no highlight roll-off and appear clipped, even though they aren’t. There are no options to adjust parameters when using V-Log L. I suspect that V-Log L isn’t the culprit here; it is likely due to how the image is being processed by the camera itself...
The colour rendition the camera produces is generally quite pleasing and more to the point fairly accurate. Where I find the camera has limitations is in the highlights – regardless, it seems, of whether you’re using regular Scene Files or V-Log L. I would have expected that a Log curve would provide more highlight protection on the DVX 200, but this doesn’t seem to be the case. You really need to be very careful when exposing to avoid getting rather nasty blown-out highlights. I think Panasonic have done a pretty good job of getting the most out of the micro 4/3 sensor in this camera, despite the obvious problems with noise and highlight roll off."
note that Matt used a LUT originally designed for the Varicam's 14 stops so it may not have handled highlights well since the LUT expected 14 stops DR and so the roll off was to be applied long after the 12 stop DR DVX200 ran out of image. Not sure as I have zero experience with V Log. Barry what is your experience?
Last edited:
I'm afraid that in the UK the AF100 was never considered acceptable to broadcasters such as the BBC *as an HD camera* (any material shot with such would be considered as part of the non-HD quota within a programme), and that was due to front end performance - not just because of codec. It was tested for them by Alan Roberts and their R&D department - http://thebrownings.name/WHP034/pdf/WHP034-ADD66_rev_2_Panasonic_AF101.pdf - and the conclusion was:
The EX1 with internal codec was (and still is) rated as fully eligible for HD usage within the journalism tier (for news and factual programming), but (as you say) required such as a nanoflash for unrestricted acceptance and general programming. (In practice, I understand the BBC mainly bought Canon XF300's to avoid needing any external recorder.)
For what is being talked about now, then I believe current EBU guidelines don't have any issue with 4:2:0 acquisition for 4K material - the big factor being that it's progressive, the 4:2:2 requirement was because of interlace.
Barry_Green
Moderator
VLOG-L is the same as VLOG. The only difference is that in VLOG-L, white is reached at about 4.5 stops over middle gray, whereas on a VariCam with full VLOG it's at 6.5 stops over. There is no rolloff in highlights in VLOG-L, it is a straight remapping of the linear sensor data. It's as close to raw as you're going to get. If you want highlight rolloff, that's what the video gammas are for; they incorporate an s-curve to lift the midtones and roll off the highlights. VLOG doesn't do that -- it's a straight line.
As said before, yes VLOG shows noise. Noise reduction has to be part of your post-processing grading chain of events, along with sharpening, just like they would be in a raw workflow. The entire Scene File menu is basically disabled when you turn on VLOG mode; the camera bypasses pretty much all processing and delivers the footage in its rawest state. It still gets de-bayered etc., but no image processing is done on it. It's possible that some general noise reduction is being done, but all the controls in the menu for noise (detail level, coring, knee APE level, etc) are disabled in VLOG. So you'll have to do that work yourself in post, in the grade.
Barry_Green
Moderator
It was, in the "HD Specialty" category, the same category as the Red One fell into. It was deemed acceptable in that category if using a 50mbps 4:2:2 recorder.
Danilo Del Tufo
Well-known member
New footage (Panasonic 120fps Slow Motion Test), link: https://www.youtube.com/watch?v=VZDCUJ_EW5E
Alpha Beta Omega
Veteran
Funny that the BTS shots shows better IQ :laugh:
Danilo Del Tufo
Well-known member
I noticed a little of aliasing and moire into the footage I post (in the background), has anyone noticed?
philiplipetz
Veteran
Terrifying. Must see some better footage than this and Newsshooter samples. Look at the face comparisons in Newsshooter review using different cans. The DVX200 at that stage of development was clearly subpar. And I am a DVX200 fanboy.
Well now wait. . .let's qualify that statement. DVX footage was super over exposed, BTS footage wasn't. Also, BTS footage was not full screen, making it appear sharper. Also may have had more edge enhancement setting on the the BTS cam, who knows.
We also may be the only tech nerds on the planet that could talk about the video and NOT mention the subject matter.
We also may be the only tech nerds on the planet that could talk about the video and NOT mention the subject matter.
philiplipetz
Veteran
Why does a DVX200 fanboy have fear? This video shows clearly that the color science was lacking for faces at this stage of "near final" development. Look at the end where the camera was compared to other cameras exposed under the same lighting conditions. Look especially at the shadows on the side of Matt's face and the highlight on his forehead.
Last edited:
Mike Harvey
Veteran
Seriously? The DVX footage looked overexposed, and the BTS footage looked like, well, camera phone footage. Nothing about the BTS footage looked like it had better IQ
Alpha Beta Omega
Veteran
BTS footage was probably shot with a smartphone and still colours and exposure looks fine especially compared to the DVX200 which by the look of it had a real photographer behind the camera. Go figure.
The DVX200 shows a very similar behaviour to the AF100 in how it handles highlights, clipping and colours. If treated with care all is well but if you go a little bit on the wrong side it starts to sink. It was actually a long time since I saw such poor highlight handling and skin tones as this footage. Sure it's over exposed and by the look of it the DVX200 can't handle that. Just as Matt pointed out that V-log is flat but it clips bad in a way most other LOG curves do not.
The DVX200 shows a very similar behaviour to the AF100 in how it handles highlights, clipping and colours. If treated with care all is well but if you go a little bit on the wrong side it starts to sink. It was actually a long time since I saw such poor highlight handling and skin tones as this footage. Sure it's over exposed and by the look of it the DVX200 can't handle that. Just as Matt pointed out that V-log is flat but it clips bad in a way most other LOG curves do not.
philiplipetz
Veteran
So VLog is not really a log curve with rolloff at extremes of exposure, but linear compression closer to pseudo RAW. That could explain the highlight rolloff issues. I am sure someone will come up with a proper LUT. So, does this mean it has the advantage (especially for 8 bit) of having slightly more information in the midtones than "real" log curves?
Last edited:
Matthew Allard ACS
Member
I have tested the camera in 120fps in HD and the images are very, very soft.
philiplipetz
Veteran
Matt, thank you for your efforts, they were exhaustive and extremely informative. Well beyond the norm.