Robert_Niemann
Well-known member
Yeah, the question goes to Shannon too. Of course.
4-cam comparison -- initial response
- Thread starter Barry_Green
- Start date
Well... we got a camera that's comparable with the best new cameras but has both a better codec, variable frame-rates and a color space that allows for good effects work... It has as well a lens with a very good wide angle side (which i need for filmmaking)...
Inly downside for me is latitude and highlight handling... Right now I stay between jvc and panasonic... can't wait for final results
Inly downside for me is latitude and highlight handling... Right now I stay between jvc and panasonic... can't wait for final results
Steve Roark
Member
Steve Shovlar said:
Actually, the Z1u is about $5000 compared to the $9000 H1, the FX1 is about a third of the price of the H1.
I don't think anyone was expecting the H1 test chart to be 4 times as pretty as the Z1's test chart anyway.:evil:
I don't think I disagree with the spirit of your post, though. An FX1 is still a potent camera, and having deep pockets isn't going to make you a better cameraman.
Barry_Green
Moderator
The Sony was only used in 60i mode. We figured that many film transfer places are saying that they'd prefer 60i over CF25, and Graeme has said that he gets better results from his 60i de-interlacer than CF25 does, so we thought we'd just shoot 60i. Shannon voted for just going with CF25 and letting the chips fall where they may; and there's a lot of merit to that viewpoint, but eventually the decision was to let it run at its full res and if someone wanted to convert it to 24 in post, they'd have the 60i to start from.ecking said:
We didn't put them in "optimal" settings, we put them in comparable settings. I wouldn't necessarily say that those settings are the right settings for all shooting circumstances; I've seen the Sony look a lot better than some of the shots we did because it really works well at higher edge-enhancement levels; it really brings its picture to life. At detail off, the sony borders on unusable, at 5 it's comparable to the others but still just looks too soft; I doubt I'd want to go below 8 on the Sony. The Sony's getting picked on because of the company it keeps and because it's 60i, but seriously, when viewed on its own and not in any sort of back-to-back comparison, it really can make a very nice picture!
Barry_Green
Moderator
It took us so long, I figured it would be in the history books by the time we finished... sheesh. Interesting enough, almost the exact same test is scheduled for around a month from now; Mike Curtis of HDForIndies.com is setting up the same kind of thing. Adam Wilt will be participating, hopefully he can take a lot of what we learned today and take that forward to make theirs run a lot smoother and get some more answers!goldyprog said:
Barry_Green
Moderator
Image-wise? Oh heck no -- there's no way anyone would say that the Canon image was 4x as good as the Sony's, no way. Maybe 20% better? Maybe? But not 4x.Steve Shovlar said:
But you have to keep in mind, the Canon does things the Sony doesn't. It's *not* all about the image! The Canon has HD-SDI, it has interchangeable lenses, it has the 24f mode, it has timecode in & out, it has the genlock, it has the ability to be fully remote controlled via the CONSOLE program. And it has way more telephoto reach. There's a lot you get in the Canon that you don't get in the Sony. You have to factor that in. The image is the most similar aspect, it's everything BUT the image that helps you pick which one is right for you.
I know what you're saying, but I wouldn't say it that way. Value is decided as appropriateness for the task. I'd say image-for-money, oh yes the Sony wins, because the FX1 delivers the same image and it's down to $3000. No 24p mode, but if you want low-noise 60i the Sony has a lot to offer.
Indeed, the JVC did quite well indeed. It was handicapped by its limited frame rates and things like that, but man, when it came time to "bring it", the JVC did very well for itself.
I don't know that that's a fair characterization either though. The Canon likes to present itself as an HD-SDI camera, which just so happens to have an HDV deck on it. But if you want to use it as HD-SDI, you'd have to supply something to record it with. The HVX is a DVCPRO-HD camera that you have to supply something to record with (either some cards or a direct connection to a computer). Base price, the Canon is $3,000 more. Street price, the Canon is more like $3500 more. You could get a Panasonic and a Sony FX1 for the cost of the Canon, and have HDV recording capability as well as a high-def camera that records a superior format (akin to the HD-SDI), but you'd have two cameras.
They're not really in direct competition, I think. I think the unique features of the Canon make it well worth the money for those who want, need, and will use those features.
Barry_Green
Moderator
About film-like feel? That's subjective, of course, but no question in my mind that the Panasonic delivered the goods. It just looked more filmish than the Canon did. It looked like a high-def DVX, with all the DVX "mojo"; the Canon looked like a very sharp video camera. It has proper motion, but the grain structure and sharpness and highlight handling made it look a little more like video. Very nice, very good video, and I'm sure with some tweaking in post you could make it look very filmic indeed -- I mean, heck, look at that music video that was shot on the FX1 that looks so fantastic; the Canon has a head start on that... but many of us looked at 'em back to back (I'm thinking me, Rush, Aaron, etc) and said "I just like the HVX better."Robert_Niemann said:
No knock against the JVC, the JVC was right there with it; I just didn't like the large sharpened noise effect. The JVC was kind of in-between; it had the full progressive motion signature like the HVX with no vertical-res aliasing like the HVX, but it had this sharper noise texture that was just a little more video-y.
But it was very good.
And this is all subjective opinion, and with any luck we'll get six or seven subjective opinions, and they may all vary.
I really wish we could have done more scenes; after seeing them put all together in context, it makes me wish we hadn't spent 80% of the day on charts and tests! :grin: I wish we could have gotten the greenscreen shoot and the dark low light shoot and the outdoors shoot.
Next time we try something like this, it should be two or three cameras. You can do almost anything on a two-camera test, so we should pick two, put 'em to the test, and then pit the "winner" (if there is such a term) against the other leading contender, etc. Trying to do four or five or six at a time... well, there ain't enough rogaine in the world to make that work!
Barry_Green
Moderator
The HVX lens performed very, very well. It's good glass.Rodrigo said:
Highlight handling on the scene was very close between the two, with the JVC holding the edge by about one chip on a 10-chip chart.
Barry_Green
Moderator
Well, there's two sets of footage (and there's very little of it, too): there's the uncompressed raw stuff captured in the Mac, and there's the camera-recorded stuff (the P2 card and the three HDV tapes and the HDCAM and VariCam tapes). I have digitized copies of the Sony and JVC tapes, and I have the HVX P2 card footage. I have to figure a way to digitize the Canon footage; when we were there I tried hooking up Shannon's Canon to the HDV Rack program but it just went nowhere. So I don't really know how I'll get the Canon footage off the raw tape right yet... and obviously, everything hinges on that right now!rawfa said:
toke lahti
Well-known member
JVC's resolution is really a surprise.
Maybe that's because it's the only one with "real" lens.
Testing its resolving with two as different lenses as possible would be very interesting.
1080 vs. 720 with HVX would also be very needed, because of the huge difference they use storage space (100Mbps vs. 40Mbps with 24fps).
Did you do any resolution tests to discover chroma resolutions?
Overall low horizontal resolution with all these "hd" 1/3" cameras is a big question mark for me.
Why is that only almost same with 1/3" 4:3 sd cameras than with 1/3" 16:9 hd cameras?
Shouldn't the relation between horz & vert resolution be opposite when moving to wider imagers?
Or do those chips have really horizonally long rectangular pixels just to keep large enough pixel area and at the same time giving better vertical resolution (which is the advertising thing; 720 or 1080).
If we are getting imager resolutions like 640(h)x1080(v) or 480(h)x720(v) or something like that with 16:9 imagers, it really contradicts the benefit of having 4:2:2 when you really get out of imager only something like 4:1:1 because the pixel shape.
Even more extreme with hdv when 4:2:0 turns out to be 4:1:0 with this kind of pixel shapes.
Like Steve Mullen said in DVinfo, that with 1/3" cameras going from SD to HD seems to be in resolution wise going from 600x500 lines to 600x600 lines.
So much for the advertised double or quadruple amount of details...
Maybe that's because it's the only one with "real" lens.
Testing its resolving with two as different lenses as possible would be very interesting.
1080 vs. 720 with HVX would also be very needed, because of the huge difference they use storage space (100Mbps vs. 40Mbps with 24fps).
Did you do any resolution tests to discover chroma resolutions?
Overall low horizontal resolution with all these "hd" 1/3" cameras is a big question mark for me.
Why is that only almost same with 1/3" 4:3 sd cameras than with 1/3" 16:9 hd cameras?
Shouldn't the relation between horz & vert resolution be opposite when moving to wider imagers?
Or do those chips have really horizonally long rectangular pixels just to keep large enough pixel area and at the same time giving better vertical resolution (which is the advertising thing; 720 or 1080).
If we are getting imager resolutions like 640(h)x1080(v) or 480(h)x720(v) or something like that with 16:9 imagers, it really contradicts the benefit of having 4:2:2 when you really get out of imager only something like 4:1:1 because the pixel shape.
Even more extreme with hdv when 4:2:0 turns out to be 4:1:0 with this kind of pixel shapes.
Like Steve Mullen said in DVinfo, that with 1/3" cameras going from SD to HD seems to be in resolution wise going from 600x500 lines to 600x600 lines.
So much for the advertised double or quadruple amount of details...
Barry_Green
Moderator
I've got those charts, but that's on recorded resolution. On live imaging they look very similar, but recorded resolution is lower on 720.toke lahti said:
The live images all looked fairly similar, although (I should leave this up to Adam to explain) there seemed to be some chroma subsampling visible in the vectorscope on some of the cameras but not others. It may be possible to tell differences in chroma resolution in the recorded footage, but again that'll have to wait until I can get the Canon footage digitized.
It's nowhere near the same. All of these cameras beat the living snot out of SD cameras. These are all high-def cameras. The resolution chart numbers have little bearing on what you see in the actual image. The Sony had the lowest chart numbers and it can look stunning and completely, legitimately high-def.
Well, keep in mind that resolution is expressed in terms of lines per picture height. In widescreen you'll actually see less horizontal resolution because the wider imager means that you'll be using less of the surface area of the chip, because the height is shorter in proportion to the width; and since the width of the rectangle you measure is tied to the height (again, lines PER PICTURE HEIGHT), you'll see numbers that seem lower but are in fact resolving more than a comparable 4:3 camera would.
But don't get hung up on that -- they all make great high-def pictures with plenty of resolution, way way way way way more than you'd ever get out of a DVX or XL2.
Well, some do -- the Sony uses 2:1 wide rectangles. But the JVC uses square pixels. The Canon is fairly wide rectangles at 1.33:1, but not as wide as the Sony, and we don't know about the HVX.
No, that's completely wrong -- you DO see double or quadruple the amount of detail! I mean, okay, maybe not quadruple, because there are limits as to how much the lens can focus and the chips can see, but it's extremely higher definition than a standard-def camera.
russc
Active member
Thanks for your hard work on this shootout guys. I guess ultimately it does come down to preference of workflow, footage archiving, and how in fact you'll be using these cameras... but if it is this close in image and price right now ...variable frame rates and true 24pHD will ultimately make the desicion for me. If the HVX does what it says it can do in that area, that will be where I go. I look forward to your official test results aftrer "Playback".
Chris Messineo
Veteran
Barry and Friends,
Thanks so much for doing the test. Some fascinating information already and I can't wait to read the final results.
Although, I have to admit, I am already set on my HVX and it would take a major disaster to turn me away.
Chris
Thanks so much for doing the test. Some fascinating information already and I can't wait to read the final results.
Although, I have to admit, I am already set on my HVX and it would take a major disaster to turn me away.
Chris
