Golden Gate JVC vs. CineAlta

[Stephen van Vuuren]

I posted some quick grabs:

http://www.sevensmilingsharks.com/golden_gate_test.htm

As a "indie moviemaker" DVX100 owner and former XL1 owner, I looked at these clips to see if it would be worth switching since 16:9 is compelling and the Panasonic anamorphic has been getting very mixed reviews.

However, no 24p is big loss for me and the footage does not impress me. I fail to see the resolution argument. Does a thousand bad pixels make a better image than 500 good pixels?

The footage, so far, looks flat, extremely narrow VHS like latitude that robs the picture of all depth (especially zero modeling in shadow areas) and crazy edges.

It's not HD, it's not miniDV. It some kind of freak, lots of pixels, lots of noise and not much else.

Plus, the control issues would drive me bonkers. I fail to see indie guys as having hours to tweak lighting, filters and camera settings to compensate for glaring flaws. It's the opposite. I got rid on my XL1 to a DVX100 that grabs a quality image with ease.

Fully manual and flexible.

PS. Any San Fran DVX100 shooter available to grab a shot of the Golden Gate to match here. Would be interesting.

[Chris Hurd]

Well, Stephen, I'd have to say that this isn't the right camera for you.

[Yang Wen]

Ahh yes, that's the HD i'm used to seeing and expect. I don' t know why you guys are such adamant proponents of the JVC's colors? How is that more life-like than the CineAlta's colors right below it?

[Boyd Ostroff]

Just a random observation, the CineAlta frame was shot either early morning or late afternoon - look at the long shadows and stong sidelight effect. The JVC seems to have been shot closer to noon - the shadows are underneath the cars. Much flatter lighting which gives a much more harsh effect. There also appears to be more mist in the air in the CineAlta shot which softens the effect.

But aside from that, wouldn't one expect a huge difference between cameras that are (maybe) $100,000 apart in price?... ;-)

[Stephen van Vuuren]

Well, I'm still puzzled at the market for this camera. My posting the screen grab was to clearly point out the image does not resemble HD that I've seen from any other HD cam other than the frame size.

Sure the CineAlta is way more money, but then again, I wonder how well this image would fare shot with any number of DV cams. From this grab, I think most DV cams, all the way down to micro-DV & Hi-8 cams show nearly as well, many better than the JVC.

Obviously, the CineAlta image looks much better, but if you look more objectively items like latitude, color rendition, etc. it seems much more like the digital still camera market. There are plenty of new cheap 5 megapixel cams that don't take as nice images as older 2 megapixel cams.

Yes, pixels count when you blow up an image. But if the image looks crappy to begin with, do you really want to blow it up?

My judge of any camera is regardless of size, cost, still, motion, market, it must take a nice image in it's intended format. JVC says this is a HD cam.

Well, it's HD image is awful by HD standards. However, the cam shoots other formats. Has anyone even tried the DV settting?

That would at least allow us to compare DV to DV and see how the CCD and lens stacks up.

If it shoot really nice DV and the HD is a bonus feature, then I see this cam as having a market. But if shoots poor HD, mediocre DV and nothing else, all I can see it for is a status symbol (I'm shooting HD, look at my big HD screen) than a useful camera.

[Paul Mogg]

Boyd, you are completely correct. I shot the Bridge shot at around noon time, and the Cinealta shot was made in the late afternoon, making my shot a much flatter image. As you also said, there is a lot more mist in the air in the CineAlta shot. Steven, you are perfectly entitled to your opinion, but I don't think many would agree that it's quite as black and white as you're making it out to be. To judge these images one against the other, you really have to view them all, in motion, on an HD monitor or projected on a large screen, at HD resolutions, not as still grabbed from video and viewed on a computer monitor.
By the way, have you ever seen DVX100 footage blown up to a 40' screen? It's not a pretty sight IMHO, the blocky pixelation of 720*480 image, and the terrible motion artifacts and ghosting really come through on a 35mm blowup. Looks great on an NTSC TV though. Just my personal opinion on what I've personaally seen, and something I think any person intending to make a movie with that camera needs to consider for themselves before spending their hard-earned money on a project.

[Boyd Ostroff]

<<<-- Originally posted by Paul Mogg: By the way, have you ever seen DVX100 footage blown up to a 40' screen? It's not a pretty sight -->>>

I'm doing a project in September where we'll project DV on a 44'x25' screen, and am feeling a little nervous about it. We'll be using a big 12,000 lumen DLP projector however, not film. We did a test last winter in the theatre, and some things looked good, others didn't. Lots of people were impressed but I was more critical, having shot the footage. For one thing, during that test we were using a 16:9 aspect ratio but it was just cropped to 720x360 in a 4:3 frame. The actual show will use 720x480 anamorphic which should produce a noticeable improvement (combination of real video shot with a PDX-10 and computer animation). DV itself is a big limitation. I have animated sequences (done in Bryce) where the individual frames rendered as JPEG's look terrific, but after compressing as DV you get banding, blockiness, ragged edges, etc. In this case it has nothing to do with optics or CCD's, just the limitations of the DV codec.

However I just saw "28 days later" last night in a movie theatre and must say that I was impressed. As I understand, it was shot on stock PAL XL-1s for the most part. The scenes with fine detail had the usual DV problems (especially noticeable in cityscapes, landscapes with trees, etc), but I was surprised that there was no real evidence of "blocky pixelation of 720*480 image"; I really expected to see that myself. Perhaps it has to do with how it was processed in post and printed to film? The "ghosting" was there, but not as prominent as I would have expected either. All in all, it encouraged me that DV really can play on a large screen. OF COURSE there are compromises though and it won't ever look like film (or HD).

I see this JVC cam as sort of a "grand experiment". Obviously it has its problems, but hopefully it will inspire the other companies to follow suit and the software companies to develop low cost mainstream tools for editing. Not ready to buy one for myself, but it's fascinating to follow its development.

[Stephen van Vuuren]

<<<-- Steven, you are perfectly entitled to your opinion, but I don't think many would agree that it's quite as black and white as you're making it out to be. To judge these images one against the other, you really have to view them all, in motion, on an HD monitor or projected on a large screen, at HD resolutions, not as still grabbed from video and viewed on a computer monitor.->>>

Why is a HD monitor necessary to make the JVC footage look good? The CineAlta footage, in motion and still, looks outstanding, as it should. The JVC looks pretty bad, despite the frame size. Vegas 4 fully supports HD footage, so I think this argument is baseless. I've seen CineAlta footage through a viewfinder, on small and large monitors and projected digitally and blow up to film. I've seen all manner of DV footage the same.


<<<By the way, have you ever seen DVX100 footage blown up to a 40' screen? It's not a pretty sight IMHO, the blocky pixelation of 720*480 image, and the terrible motion artifacts and ghosting really come through on a 35mm blowup. Looks great on an NTSC TV though. -->>>

Something is clearly wrong with this statement. DVX100 shoots 24p. If the footage is shot 24p advanced, then edited and transfer properly there will be zero motion artifacts, ghosting as you are transferring one frame to one frame. This would not be the case with this JVC. If fact, 30p to 24p is one of the process that creates the most artifacts (check FAQs at film transfer houses).

I'm sure the JVC suffers less resolution loss on a 40 ft screen, but again, resolution is hardly everything. For 720*480, the DVX100 produces the best blowups of any miniDV camera:

http://www.dvinfo.net/conf/showthread.php?s=&threadid=7879&highlight=DVX100+Blowup+to+35mm
http://www.theasc.com/magazine/product.htm (If you can get the actual issue, the images are better).

My impression would be completely the opposite. Shooting Cinegamma with the vertical resolution thin setting looks much better blown up with the DVX100 than viewed on NTSC because it captures much more vertical resolution than NTSC can display.

I have not seen the Duart blow-ups myself, but I intend to project my footage digital via progressive scan projectors.

[Yang Wen]

<<<-- Originally posted by Paul Mogg : Boyd, you are completely correct. I shot the Bridge shot at around noon time, and the Cinealta shot was made in the late afternoon, making my shot a much flatter image. As you also said, there is a lot more mist in the air in the CineAlta shot. Steven, you are perfectly entitled to your opinion, but I don't think many would agree that it's quite as black and white as you're making it out to be. To judge these images one against the other, you really have to view them all, in motion, on an HD monitor or projected on a large screen, at HD resolutions, not as still grabbed from video and viewed on a computer monitor.
By the way, have you ever seen DVX100 footage blown up to a 40' screen? It's not a pretty sight IMHO, the blocky pixelation of 720*480 image, and the terrible motion artifacts and ghosting really come through on a 35mm blowup. Looks great on an NTSC TV though. Just my personal opinion on what I've personaally seen, and something I think any person intending to make a movie with that camera needs to consider for themselves before spending their hard-earned money on a project. -->>>

Why would there be motion artifacts if you take true 24P material and convert it to 35mm??? The motion artifacts you see on TV is is understandable because of the telecine process, but on the big screen it should be 24 individual frames / second.

[Chris Hurd]

Stephen

<< Well, I'm still puzzled at the market for this camera. >>

JVC clearly states that the HD1 and HD10U camcorders are intended to supplement their D-VHS home theater market. It's very much a niche market and the camcorders are intended as conceptual, experimental "bleeding edge" technology, not as a replacement for or serious competitor to existing DV formats such as DVCAM, etc. If you have a D-VHS deck or other JVC high-definition home theater components, you might be interested in one of these camcorders. That's about it in a nutshell. Hope this helps,

[Paul Mogg]

<<Why would there be motion artifacts if you take true 24P material and convert it to 35mm??? >>

I have no idea why the DVX100 blowup to 35mm that I saw looked as bad I'm describing, but it did, and I think that anyone that was there would have no problem in agreeing with that statement. The proof is in the pudding as they say, not in the marketing materials.
FYI, the blowup that I saw was done by Monaco labs in San Francisco and shown at the Metreon Theatre at MacWorld earlier this year, they also showed footage from Sony DV 60i cameras which looked considerably better than the DVX100, both in color and lack of artifacts. From what I've seen, DV on a big screen seems to especially fall apart when you attempt to show panoramic vistas, close-ups always look much better to the eye. It's in the panoramic shots, and especially pans, that the lack of detail in the 720*480 DV resolutions becomes extremely apparant.
I am not saying that a blowup to 35mm from the DVX100 cannot look better than what I have seen, it could be that that wasn't a particularly good one. I would just reccomend that anyone intending to use footage from that camera for large screen projection purposes, or for transfer to 35mm film, first go and get a blowup test done, and make the judgement call for themselves. Rather than believing marketing materials that might lead you to believe that it is as easy to make a movie as "flipping a switch". Exactly the same goes for the JVC.

[Stephen van Vuuren]

<<<-- Originally posted by Chris Hurd : Stephen

JVC clearly states that the HD1 and HD10U camcorders are intended to supplement their D-VHS home theater market. It's very much a niche market and the camcorders are intended as conceptual, experimental "bleeding edge" technology, not as a replacement for or serious competitor to existing DV formats such as DVCAM, etc. If you have a D-VHS deck or other JVC high-definition home theater components, you might be interested in one of these camcorders. That's about it in a nutshell. Hope this helps, -->>>

Thanks Chris - I had missed this point and that makes far more sense, though the HD10U seems to be appealing to indie filmmakers, much like the ill-fated Cineline which also offered 16:9 and progressive (though not at the same time!).

Paul:

The DVX100 blowup you saw must have been fundamentally flawed if Sony 60i miniDV had less motion artifacts. Even the best 60i blowups I have seen done by the Orphanage still don't compare with the DV 24p blowups, simple due to motion artifacts.

PS. the blockiness from DV Compression can be pretty much erased using Magic Bullet from the Orphanage. Magic Bullet makes an HD version that might help the JVC footage improve greatly as well.