HVX200 Resolution Question

Arman

Member
I have searched this forum and read Adam Wilts shootout article and I am confused.

What is the resolution of the image captured by the HVX200 in both 720p and 1080i modes?

I realize that the output resolution is standard 1280x720 and something like 1920x1080 respectively. This is achieved through pixel shifting a smaller 3 CCD array. From Adam wilts article, it would seem that the HVX lacks considerably in the resolution department. What am I missing here?

I have scoured the net and have found different guesses at what the HVX is really grabbing. Something like 1440 x 810. Why does Adam Wilt's article suggest it has roughly half the res of the other low cost HD cams?

?
 
Panasonic published a white paper on this very topic several months ago, which has been posted somewhere on DVXUser. A search of Panasonic's site or of the forums here should find it. This topic has been beaten to death here, so I wouldn't anticipate very enthusiastic reponses on the issue.
 
Arman - the whole subject is so complex that it really isn't possible to quote a simple figure such as 1440x810 and say - "this is the resolution of the camera". For starters, resolution is never 100% up to a number, then vanishes to zero, in practice it tails off gradually - as the detail gets finer, it becomes less distinct. (You may like to do a serch for "mtf" or "modulation transfer function")

Secondly, the basic situation is complicated by the whole theory behind pixel shifting. It's a good technique, but it's important to realise that it doesn't provide any enhancement for an image comprised of saturated colour. The resolution figures you may get from a black and white chart will be better than if the chart was green and white, say. Statements such as "pixel shifting boosts resolution by 30%" are simplistic, and don't tell the full truth.
 
Thanks for the responses guys. After reading the Wilt article, I was surprised that the HVX appeared to resolve about half the lines of resolution of the other cameras.

I shot my last short with the DVX100 and was happy with it. I am primarily looking for an inexpensive solution for feature work that will yield a sharper image than the DVX when blown up for a film out. I am interested in the workflow benifits of the HVX over the competition and the codec being superior than the HDV options out there as well.

Am I right in thinking that the HVX200 is (under the right conditions) capable of giving me more detail than the DVX100? Yes - I understand that it is an HD camera that records x amount of pixels. The Wilt article had me thinking that if the camera only resolves to 540 lines, will it be much better than the DVX???
 
Quit reading articles and start looking at footage. The HVX shows way hella more detail than a DVX does.

Let me put it this way: yesterday we spent about six hours calibrating HVXs against the new HPX2000, Panasonic's new top-of-the-line CineAlta fighter; 2/3" chips and all. With a $60,000 lens on it. Was the HVX softer? Yeah, some. Maybe 10-15%. Could you intercut seamlessly between the two? Absolutely.

The HPX was 1.5 stops more sensitive, and could see about 1 stop further into the shadows, and was a bit cleaner. But for overall "high def"ness? It was probably 85% as good as the $27,000 camera. This is real-world footage, test charts, scopes, etc., and the guy I'm working with is C.R. Caillouet, one of the most brilliant HD technicians on Earth who's been doing HD since the 80's, works for NASA, and knows more about HD than Rosie O'Donnell knows about twinkies. His right-hand-man and drinking buddy is Charles Poynton, if that helps spell it out. I mean, here's a snippet from his resume: http://www.vu-la.com/resume/past.html

The point is: the HVX is a real high-def camera that delivers a real high-def picture. At least 85% of the picture you get from Panasonic's newest top-of-the-line HD camcorder on the market. You can see a similar report from another of our members who screened F900 footage against the HVX:
http://www.dvxuser.com/V6/showpost.php?p=837039&postcount=1

So I say, start looking at actual footage, rather than at res chart numbers (unless, of course, all you want to buy a cam for is to shoot res charts; if that's the case I can strongly recommend the XLH1, as it performed superbly on our res chart tests.)
 
Last edited:
Barry -

Yes, I should spend some more time looking at footage. It's just just hard to find any that is not Quicktimed to death, resized or compressed.

Screen grabs have been helpful too, but don't tell the whole story.

I went to a local video trade show last week to check out the Panasonic booth. They had a HVX200 "Tips and Tricks" seminar but they didn't show any footage! Uggg.

I live in Portland OR, so maybe I'll try and contact someone local so I can get my hands on some real footage.


Again, thanks for comments.
 
Barry can you say what the settings were on the hvx that got you to HPX calibration and in what environment the tests were done?
 
Take a look at Barry's article "Canon XH-A1 vs. HVX200". While the Canon HX-A1 has a higher resolution than the HVX-200, a side by side look at the photos shows the HVX-200 as having much better color. The HVX's colors look rich and the Canon's look a bit muddy to me.

http://www.dvxuser.com/articles/xha1/


Bob Diaz
 
After buying the HVX I was concerned about the resolution being enough of an improvement to justify the price. Not having the P2 card, I shot some SD footage and was even more concerned. While it looked good in a small window, when blown up on my 32" Toshiba monitor, it looked pretty
bad. Then I got the P2 and shot some 720P. MAJOR improvement. I downloaded some "Red" footage for comparison (slightly compressed of course). I found the HVX holds its own, especially in color rendition and when you shoot 1080. You absolutely must know what you are doing when shooting
with the camera to get optimum footage, which is why I am out daily shooting
and practicing with the settings and scene files.

Then after days and days of cloudy weather since buying the cam, we had a couple of days of gorgeous shooting weather, like magic hour all day. This footage really shines and makes me glad I made the decision to go with the HVX.
 
I did read Canon vs HVX200 article here. I don't think I'd go the HDV compression route.

Thanks for sharing mainstreetprod - I think I will have to get my hands on one and see what I can do!
 
when you shoot 1080. You absolutely must know what you are doing when shooting
with the camera to get optimum footage,

mainstreet, could you elaborate on that? Are u saying 1080 is more difficult to shoot...why? because its harder to get the focus sharp? Please elaborate. Thanks.
 
Of the footage for the production? Not in my domain to say; it's all owned by the Iditarod organization. But here's some stills of our test chart of the DSC CamBelles, one shot from the HVX200 and one from the HPX2000. I won't tell you which is which, but see how easy it is to tell which came from which...

http://www.icexpo.com/HVX200/P2CamBelles1.bmp
http://www.icexpo.com/HVX200/P2CamBelles2.bmp

Note: there's been absolutely no color correction or post-processing of any type. These shots were captured by DV Rack 2.0 HD, by recording footage and then using the still-frame grabber. We were attempting to match the cameras and this is the result. The two cams were sitting side by side, and we had two laptops running DV Rack and a series of test charts and C.R. had built a custom light box to house the charts. The CamBelles are a skin-tone test chart from the CamBook, made by DSC Labs. So this wasn't a shot of live girls at the beach (we're in Alaska in February, after all!) but it's a shot of a test-chart.
 
Barry_Green said:
I won't tell you which is which, but see how easy it is to tell which came from which...
I've only got a small laptop here, but my vote is Belles2 looks a lot better than Belles1 on simple first viewing.........?
 
Belles2 is a little more saturated and looks like the detail has been cranked a little. More visable noise.

Belles1 is slightly flat and has a softer image. I think I would rather see moving pictures of Belles1. Less noise maybe?
 
Last edited:
Hi Barry,

Belles2 is HPX2000 to my eyes.

I see a little detail filtering (detail coring), too much for my taste in the Belles1 image.

I bet you could get the HVX closer to HPX2000 by lowering the detail coring and by lowering the master ped by one or two clicks.
 
Back
Top