updated 14 Oct 2004

 

JVC HD10 compared to Panasonic HDC-27F, Part Two
an article by Jon Fordham, Director of Photography

Jon is a contributing member of our DV Info Net Community.

I carefully did the classic video critical focus zoom. The zoom felt a little funky, but I didn't pay too much attention to it since I don't do zooming when shooting a traditional style piece. I usually do my best to adhere to the traditional cinematographic techniques of dollys and such when the camera needs to push in or out. The focus was difficult to tell if it was hitting or not and the even more frustrating part was the typical passing by it and trying to get it to stop at the right focus position with a servo assisted lens such as this. This however isn't unusual for those experienced with these type of lenses and after a few hours with this focusing system, you can usually get use to the speed and quirks of it.

After we wrapped, I had a few minutes while discussing the day with the director at his residence. The director has a Samsung 1080i High Definition Television set with component inputs. So I asked if I could test the 720 to 1080 output option and take a peek at the HD10's footage on a true High Definition set. I connected the adapter cable to the HD10, the components into the set and pressed play. Just as promised, the image came up on the screen in 1080i, correct aspect ratio and all! I was impressed. Not to mention that I was relieved to see what the image really looked like. I let out a sigh of relief as I realized that the LCD was a horrible representation of actual quality of the image.

Fortunately, my instincts on the exposure were as spot on as they could be for the expected limited latitude of the CCD. The lack of full manual control of all aspects of the camera, coupled with the "low end" specs of the camera itself produced the image that one would expect. However, the image was by no means lacking in image quality. The image had a very narrow contrast range producing a very contrasty image. The highlights were very bright and almost washed out and the shadows were very dark to the point of almost going ink black. However, the highlights did still hold plenty of detail and the shadows were by no means murky or indiscernible. The color was noticeably desaturated as expected from a single CCD. And I believe that the single CCD's lack of color definition is what made the highlights look more washed out and almost overexposed than they really were. I don't think the color looked bad, or overly washed out. But, I was surprised by a close up shot on a cigar being lit by one of the lawyers. The flame from the lighter as it ignited the cigar was green! Weird UFO green. The brighter levels in the flame were almost white. But the blue and orange areas of the flame were somehow green.

The 30P motion signature produced nice movement, but still had that smoother more "video like" look to it. Certainly better looking movement than 60i, but not as nice as 24P. The image was very clean and clear. No obvious noise. And as with all HD images, it was razor sharp. Too sharp, in my opinion. Sharpness is the major complaint among many who work in HD. And when it comes to filmmakers and Cinematographers, sharpness is usually the first thing they notice and dislike. In my opinion, the whole point of High Definition imaging is increased resolution. The ability to capture and resolve the many subtle nuances of an image. Unfortunately, that increased resolution and ability to resolve those subtle nuances often translates into a very sharp picture. For lack of a better way of putting it, unnaturally sharp. But even though the sharpness of the HD10 was a bit noticeably edge enhanced, it wasn't overly edge enhanced to the point of being a major hindrance. After a few minutes of checking it out, I disconnected the HD10 from the 1080i set and headed home to try and get four hours sleep in the unacceptable eight hour turnaround time. Non-union and independent. Need I say more?

The following evening, I had the fortune of a proper ten hour turn around time. So, I foolishly spent my valuable time letting my curiosity get the best of me! I connected the HD10 via Component to a Sony DSR-80 DVCAM studio VTR. I connected the DSR-80 to my home television set (a 27" Sony Trinitron) via the Y/C (S-Video) output, and set the HD10 to output the 720/30P signal as a standard 480/60i NTSC signal. Again the image came up no problem and everything worked as it should. And again, I was impressed. Upon downconversion, the image maintained the same qualities as it had on the 1080i television. Of course, with the obvious lower res reproduction symptoms that any consumer Standard Definition television set has.

Then to see what we've all been waiting for, I connected a Panasonic HDC130 DVCPRO-HD deck to the Sony DSR-80 DVCAM deck via SDI. The HDC130 downconverted the 720/24P DVCPRO-HD signal flawlessly. No fuss, no issues, no problems at all. Man, it's nice when things work the way they should! I popped in the Varicam's DVCPRO-HD in camera original tape from the bar and spent a few minutes scanning through it. The difference in image quality between the HD10 and the 27F was immediately noticeable. The image from the 27F contained beautiful and accurate color with just the perfect level of saturation for my taste. The contrast ratio was a nice and even level of highlight and shadow. Easily reproducing the highlights at a level that maintained the color and detail accurately as well as clearly reproducing the fine details in the shadows. The contrast provided plenty of range of deep blacks and bright whites while still producing a natural and even tone. Of course, the 24P motion signature produced the expected "cinematic" movement. And in true High Definition fashion, the image was razor sharp. The difference though, between the "sharpness" of the HD10's image and the 27F's image, is in the way the image looks sharp. The HD10 doesn't show the subtle nuances of detail and has an "electronic" edge enhanced quality to it. The 27F's sharpness has a quality of extreme resolution. Capturing those subtleties and resolving detail without having an unnatural electronic feel.

Skipping ahead to check out the close-up of the cigar being lit that the HD10 showed such unusual color on, I find that the color on the 27F is absolutely perfect. The levels aren't at all over exposed, the oranges are beautiful and deep in color, the blues are solid, and the flame of the cigar burning comes across clearly and accurately. Even the tiny bits of ash as the tobacco burns away is beautifully detailed.

Fortunately for the HD10, my eye was able to get the proper exposure for the footage that I have had the opportunity to sit down and evaluate. A combination of experience and a hell of a lot of LCD usage on many many cameras! I agree that it would've have been far superior to have a monitor available that would display the actual 720/30P image that the camera was acquiring. However, since the camera does not output a full res image of what it's getting, attempting to accurately evaluate the exposure using the "downconverted" image would've (in my opinion) been just as safe or dangerous as using the LCD. Whenever working as a D.I.T. and the budget allows, I often insist that a downconverted NTSC image be available when shooting 1080. I've worked with many DP's who are unfamiliar with 1080 HD. Those unfamiliar with the medium and the format tend to get too trusting of the monitor and forget that 99.9% of the audience will never actually see the original full res 1080 image that they are seeing on set. This usually leads to them wanting to crush the blacks more, or push the color higher, or any number of other image possibilities that might look fine on a Sony 24" professional HD display. But once it hits NTSC, or film out, or color correction can become loss of shadow detail, bleeding and artifact ridden colors, etc.

By that same rationale, using the HD10's live "camera donwconverted" image as a point of reference may be just as justified as any other method given the exact same reasoning. How much of the audience (especially given the particulars of this new HDV format) are actually going to ever see the image in any other format other than SD. So sure, with more experience using this camera, I'm certain I'll be able to evaluate the difference (if any) in the full resolution image vs the live "camera downconverted" image to be able to rely on a properly calibrated CRT monitor for exposure verification as I usually do.

By the way, with the "exposure, shutter speed, f/stop" situation: What I've been doing is setting the "exposure" using the little dial, then turning on the "shutter speed" control and setting it to 1/60. The exposure seems to hold even though the readout disappears. But I don't know that this is the case.

Written by Jon Fordham, Director of Photography.
Thrown together by Chris Hurd.