HVX 1/3 HD vs SDX 800/900 2/3 SD?

Couple of things have been on my mind lately and sorry if using my friend Bart Boge as an example here, but a few of the things he said are things I've been thinking about lately, and so really this post is to present my current views on some things that are now relevant to this discussion, the increased dynamic range of bigger sensors and high perceived value in HD lenses. These things aren't as a big of deal as I used to think.

Yes, bigger sensors will reproduce a broader dynamic range, but, how much? I mean, does it DOUBLE? I don't believe so. Which leads me to this: In my experience, most of "dynamic range" issues on any format, big or little, film or electronic, can be "made up for" with a underexposure, an optical polarizer filter, and some DSP. Whether the processing is built into the camera (cinegamma) or it's done in post, doesn't make much difference either. On the camera it's handy to have as you can see it right then, right there. Post is probably better, though, as there are more options and more time to think about them.

The DVX100, as shown by Reel Stream, underexposes nearly everything by around two f/stops. Look at the sample pictures pre and post DSP if you think I'm up to no good by saying this. What this tells us is that most of the "magic" of the DVX100 is simply DSP. We all knew that already, right?! The clever thing is that Panasonic *designed* the camera to underexpose by default, cranking up the shadows and mids to a "normal" level, while compressing the highlights in DSP. This is the same technology as used on VariCam, SDX900, etc. They call it CineGamma. To me, it's underexposing. Whatever.

The point is, underexpose everything by two stops, then crank the shadows and mids in post, leave the highlights alone. Big chips = big latitude? Latitude, smatitude.

This shot came from a Canon GL2, a 1/4" CCD camcorder:
groundbreaking.jpg
groundbreakingGL2.jpg


I don't personally think it's lacking in much latitude, yet look at the extreme range of tones represented in this lame little 8-bit camcorder! It's pretty good if I say so myself. Of course, I shot it, so naturally I would think that. :)

Additionally, the lenses on even a cheap HD camera appears to 'resolve' fine. It's mainly tolerances for chromatic aberration, so far as I can tell, that differentiate a cheap lens from an expensive lens. But, I'm not the least bit convinced they are worth the multiple times price point as their cheaper, almost as good, cousins.

The point of this rant: I don't need a 2/3" chip camera or an HD lens.
Good results are possible even with 'cheap' camera equipment. The same care as taken on bigger formats to control exposure contrast can also be taken when use smaller, less expensive formats with more than acceptable results. I believe a lot of what I used to perceive as 'increased latitude' merely came down to experienced cinematographers controlling contrasts well. I don't really believe the difference is *as*, key word, significant as I used to think.

Brian Wells
 
I somewhat agree wells. Dynamic range is a function of aparature setting and alot of people shoot with the aparature open too wide to really get black. Instead the focus has been the low light abilities of the camera which force people to shoot with an un-needed extra stop or two just for the sake of 'lightness' in the picture. This is a trap and does not show off a camera's true dynamic range.

NBC are you sure you're setting up your 1/3 CCD camera correctly? When you're shooting are you sure your not overexposing for the sake of a brighter picture and thus a noise issue?

Please be advised that these are just general statements and observations from content I've seen posted over the years.
 
Brian,

I agree with 100% of what you stated in your last post, although until the unit actually comes out we can't know what we will actually be getting.

Perhaps the "latitude" issue has been overstated. It might be our way to recalibrate our expectations for this camera. When we hear "1080p" in a sub $6k camera, we think "VariCam for the masses!" It will NOT be a VariCam, just like a DVX100A is not 16mm film. But in certain contexts and applications, a DVX can be more or less indistinguishable from a 16mm shoot (well-lit, color corrected, and straight to DVD, for example). Perhaps, once Barry puts it through its paces, we'll learn about the situations and set-ups that deliver "VariCam-like" quality footage, and which ones make the HVX look like a crappy prosumer camera.

Then it will be up to clever DPs to shoot for the former and avoid the latter.
 
I can find lots of shots from my old GL or my DVX that show great range. I know that with experience you can get very decent range out of a 1/3" chip. And sure you can do lots of post work. But my point was, that there obviously must be a reason that bigger chips exsist, and that from my experience, you don't have to work quite as hard when shooting with a bigger chip.

I understand that there will always be trade-offs at different price points. I was just disscussing the possibility of moving up the food-chain. I may not know everything, but I do have lots of experience with 1/3" cameras. Sometimes I feel that we have just about reached the limit of what can be done. Some guys from our circle are now working at higher levels in the industry, and spend hours and hours with full HD or film. They often talk about never wanting to work with DV (1/3" prosumer) ever again. Especially when piling on the FXs.

I am an indie guy at heart, and love what that entails, but I do daydream of selling a project where I am not making excuses for production value right out of the gate. I have been thru the wringer on selling a DV feature, and it's not pretty. Trust me, until 24p DV arrived it was next to impossible. I myself used to hang on those rare examples of a DV feature being sold, but hey I don't have the Aliens chick in my movie! (TadPole, i think)

24p may help only because the buyer may not automatically equate it with just DV, it has something else to offer.

Maybe the HVX will blow our minds. I really hope so. I am already on the list to buy one. This was never a comment on that camera. I have been a Panasonic fan for longer than most.

I was shooting stuff with the DVX when it first hit the street.

I'm just day dreaming here. :)

Cheers,
Chris
 
thisiswells said:
I believe a lot of what I used to perceive as 'increased latitude' merely came down to experienced cinematographers controlling contrasts well. I don't really believe the difference is *as*, key word, significant as I used to think.
Click to expand...
First of all, I just assume as a given when approaching these kinds of discussions, that I'm lensing a scene with a brightness ratio that I've already controlled to a manageable level through lighting and lighting control tools. Although I'm not an engineer, at least in my experience, I've noticed that it's not so much the size of the sensor, but the quality of the sensor and the sophistication of the electronics processing the sensor data which determines the camera's contrast handling ability, i.e., "exposure latitude."

I disagree with that view, Wells, in that I 've seen significant differences in contrast-handling capability among various cameras depending on the electronics in the camera. I'm speaking from direct experience derived from direct observations under practical shooting conditions. For example, broadcast Sony cameras' (and others') DCC circuitry demonstrably and visibly increase your practical exposure latitude for a given aperture value by approximately 1.5 stops by softly compressing the upper shoulder of the curve, as Sony claims, by "600%."

I once did a side-by-side comparison between a Sony DXC637, BVW400, and a BVW600. The monitor of the live composite output was A-B switched so that instantaneous visual comparisons could be made. A high-contrast, studio interior was lit with a single HMI Fresnel. As you switched from camera to camera head, a significant difference in resolution, colorimetry and exposure latitude was clearly visible. The differences among the three cameras was substantial and quite obvious to even the casual obvserver. Most notable was the increased amount of detail, especially in the highlights, of the BVW600's image over the 637 and 400's image. The BVW600 clearly had significantly superior dynamic range than the less expensive cameras. Believe me, if you all could have witnessed that monitor that day, I'm sure you would all agree, the differences were dramatic.
 
NBC, It's a nice thought to compare in that fasion but not very scientific. I admit to comparing things in that fasion as well but the reality is that the R & D departments have the real information about the lattitude of a camera and the circuitry's designed lattitude.

Case in point with Panasonics own equipment. (I'm about to get beat up!)

DVC30 has better saturation than the DVX100 and yet the DVC30 has smaller chips. The fact that the DVC30 was developed after the DVX100 is testament to the fact that some things were learned from the DVX that were put into the DVC30's design. Same thing goes for the HVX when it comes out, it will be superior just by virtue of R & D from cameras before it's time. Panasonic is very good at it. The HVX will be not slouch. I'd be willing to bet a fair amount on it and as a matter of fact I will when the time comes and it will be because of Panasonics reputation, mainly.

I think you're on the right track for you. The sony is worthy....
 
I will also add that the 2/3 cameras such as the Varicam and the F900 are adjustable to huge degrees. It isn't just a case of what range they have out of the box, but what can be done with tweaking. One of the ex-BBC high def consultants and camera set-up gurus managed to get an 11 stop range on the Varicam and F900, the same kind of latitude that many films are capable of. The cameras were measured with instruments and were not just a guessed.

Its not just the size of the CCD that gives the larger cameras more sensivity either. Its also the simple fact that they have much larger lenses and therefore let in far more light. Combined with Super HAD CCDs and there is a huuuuge difference between the 2/3 cameras and the smaller ones.

For a real test try the GL2 in a forest with the huge difference between light and dark areas on the ground
where the sun manages to get through the trees.

But regardless, the HVX will be good. I wish people would stop being so finnicky about the technical details, especially ones we don't know about because nobody has seen the camera working. When it does arrive it WONT be as good as a Varicam. It WILL be good however so just get on and make something with it!

I could have stuck with an XM1 for all eternity. It wasn't ideal, but it wasn't stopping me from making the best videos I could. The 2/3" gave me the flexibility to do more exactly what I wanted. In other words when I watch someones video I don't give a rats a** whether it was shot on HD, SD, an HVX or a DVX. As a viewer all I care about is whether it engages my attention or not and is well put together.
 
stephenlnoe said:
NBC, It's a nice thought to compare in that fasion but not very scientific. I admit to comparing things in that fasion as well but the reality is that the R & D departments have the real information about the lattitude of a camera and the circuitry's designed lattitude.
Click to expand...
That comment simply doesn't make any sense. I don't care what the manufacturer's spec sheet says (which marketing re-interprets from engineering) . . . there's no substitute for field testing. It's like the old joke . . . "Are going to believe [the picture] or your lying eyes?"

If you believe the initial Sony brochure (marketing-ese) copy for their Digital Betacam camcorder when it first came out, you were told that a DVW700 has "11 stops of latitude." NOTHING has 11 stops--not even film negative.
 
Simon Wyndham said:
One of the ex-BBC high def consultants and camera set-up gurus managed to get an 11 stop range on the Varicam and F900, the same kind of latitude that many films are capable of.
Click to expand...
It's funny you mentioned the 11 stops. I hadn't even read your second post util after I wrote the 11-stop thing about the DVW700! I would be very interested to hear the details of the methodology used in the f-stop calculation. If an F900 can be tweaked to actually hold detail in an 11-stop range, I'll eat my hat!

I've performed exposure tests with an F900 in cooperation with Plus 8 Digital engineers in Burbank using a footcandle meter and a waveform monitor--in my experience, the camera has no where near that exposure range (and neither does film).
 
The VariCam absolutely can be tweaked to 11 stops. I've seen it with my own eyes in person at a demonstration performed by the HD Expo. The methodology is pretty simple. Two small tabletop sets are setup next to each other, one with primarily dark objects on black duvatyne and the other primarily light shade or white objects on a white cloth. They are lit separately with quartz and Kino's with a divider placed in between, and the light level is precisely measured with a Sekonic. Or, multiple Sekonics if you bring your own. Point the camera at the two sets, look at the HD CRT Monitor and the Leader scope and voila, 11 stops of latitude. It's fab.
 
Exactly wells. NBC shooter, that stop range is measured precisely. Its not guess work, and I can tell you that the likes of Alan Roberts who devised the settings knows everything about everything to do with HD. He can be found over at the http://forums.dvdoctor.net/forumdisplay.php?s=&daysprune=&f=10 forums.

He's the main BBC high def consultant and has worked with high def since the very beginning in the early 80's. He's always ready to explain stuff (although be warned that he will probably blow your mind!)

Alans exact quote to me was "HD cameras routinely pull in 11-12 stops, occasionally
13"
 
Now that's just funny. I have no idea who that comment is directed towards, but it doesn't make any sense. I haven't witnessed any trolling in this thread. Just guys talking 'bout cameras, that's all.
 
Stephen, in what way is pointing out that an extremely highly respected and knowledgable member of the BBC, a person who is acknowledged as a world authority on high definition systems, has adjusted and measured the cameras as having this range of latitude trolling?

If you wish to read the white papers by Alan you can read them here;
http://www.bbc.co.uk/rd/pubs/whp/whp053.shtml
http://www.bbc.co.uk/rd/pubs/whp/whp034.shtml

I have pointed these out, and Alans general work, many times to doubters. Yet despite the fact that he works very closely with the likes of Panasonic (to the point of testing the prototype of the Varicam and having input into how it could be improved) people still insist that he is somehow wrong, and his work flawed. So be it. But I have pointed out the forum where Alan hangs out, so if you doubt his research you are perfectly free to go over there and ask him yourself!
 
A "troll" is someone who either throws down argumentative "red herring" posts, or who shills for a position clearly contrary to the constituency of the forum--and in either case is motivated chiefly by the desire to touch off a flame war.

I have not read a single post here that would qualify a poster as a "troll." The BBC info is actually quite fascinating, IMHO.

..I'm sure Jan and her team have massaged the engineering of the HVX to have 11-12 stops of usable latitude. In fact, Panasonic will discontinue the VariCam line as soon as the HVX is released. The new 1/3 CCDs will be so much better than anything before them that they'll completely blow all extant 2/3 chip cameras away. The Leica lenses will be able to resolve to sub-pixel focus resolution, and DOF will be exactly the same as a 35mm still camera. In fact, Fuji and Kodak are already discussing the retirement of their film stock product lines, which, again, will coincide with the HVX release date...

<yawn>..oops..! ..Just nodded off, there. Must have been dreaming...

This post is the closest one to "troll" status thus far.
 
Some guy I talked with claimed the dvx's CCDs probably had 7-8 stops of lattitude, but because of NTSC could only display five or so. Looking at the real stream andromeda, this seems to make sense.

Hopefully panny will be able to make a CCD that retains that kind of dynamic range and improve the DSP (they're working with a far better format now after all) so that there's even more dynamic range than the dvx. More likely than not, though, this will be cancelled out by the tiny photosites or the massive amounts of noise generated by built in gain used to get a decent sensetivity rating.
 
@ Simon Wyndham- My humble apology to you and try not to take offense. I was responding to NBC's remarks and not yours. It seems he's trying to justify his purchase (to me). I should have replied to his response and not used the 'quick reply'. I'm happy for the user if he got what he wanted, but is seems like it has been an ongoing discussion to justify a Sony DSR500.

can you forgive me hombre????
 
stephenlnoe said:
Dynamic range is a function of aparature setting . . .
Click to expand...
If this is the kind of "information" the forum thinks is valid and not worth any clarification or discussion (or god forbid, any disagreement), then suspend my account right now, ban my posting priviledges, and call me any names you like . . . sheesh!
 
You must log in or register to reply here.
Back
Top