|
|||||||||
|
Thread Tools | Search this Thread |
June 10th, 2006, 05:10 PM | #1 |
Major Player
Join Date: Apr 2005
Posts: 512
|
For the CA complainers: CA from a $14k broadcast lens
Today I freelanced as a studio camera operator for a home-shopping type network that broadcasts out of town. I was noticing throughout the entire day that the lenses on our cameras exhibited terrible CA at full tele. It was especially bad because we were shooting a lot of jewelry, with lots of highlights which really tend to bring CA out. A piece of silver jewelry could end up looking almost purple on air. These were $14k 2/3" EFP lenses, and it's not like they were being run wide open or anything, about f4.5 through a 2x zoom extender. The only real explanation I can think of is that the designers were expecting that if you were at full tele the focus would be racked on objects much further away.
Anyway, my point here is that the JVC stock lens is actually beating out much more expensive lenses in terms of picture quality. |
June 11th, 2006, 05:50 AM | #2 |
Inner Circle
Join Date: Feb 2006
Location: Belfast, UK
Posts: 6,152
|
You'd expect to get increased CA when using a X 2 extender: the lens should be stopped down when using one. Working close up at the tele end of the zoom will push the optics to the limit, then putting in the x 2 extender just exaggerates the problem. The designers have to put in compromises to keep the price down.
You wouldn't see the levels of CA in normal highlights on SD broadcast quality lenses that the JVC examples have. HD has really pushed the lens manufacturers in overcoming CA, even with high end lenses. |
June 11th, 2006, 04:42 PM | #3 |
Major Player
Join Date: Aug 2005
Location: Belgium
Posts: 497
|
I have to say that the amount of CA is limited enough in my standard Fujinon lens on the JVC as long as there is just enough light to go of the "open" spot for the aperture. If there isn't quite enough light (e.g. accidentally having an extra ND filter on) the CA is horrifying....
Of course you'd want the iris to be fully opened for short focal depth, but you can always check the color screen id the CA is a little to much.... |
June 11th, 2006, 07:00 PM | #4 | |
Regular Crew
Join Date: Jun 2004
Location: Hollywood, CA and Roma, Italia
Posts: 155
|
Quote:
When I initially set up a new video cam, I first use the DSC color and GS charts (from DSC Labs) to set up a "broadcast legal" color profile (or sets of profiles). Then, to check resolution/registration/CA, I use a set of 8 cards I had made (based on the old Ike 79 chart) which are composed of black/white, blue/white, green/white, and red/white registration charts (both positive and negative). I then shoot tests at full wide, mid zoom and full-on (at the lens' widest aperture, f/5.6, and f/11), and if it has a 2x, I drop that in at the full-on setting to test it. No video interchangeable or built in modular lens I have ever tested (HD or SD) has been entirely free of some CA (except a Canon GL-2, go figure). In general, the HD lenses have the most CA issues. Of the current crop of HD DV cams I have tested, the Sony Z1 had the least CA (but it also has the softest image which disguises CA to some extent), the Canon H1 the most. The JVC stock lens is about average, the Fuji 13mm JVC option is better, but still not entirely free of CA. At this point in the evolution of HD lenses, some CA appears to be the nature of the beast. It's difficult to make a 100% definitive assumption on any of the current lenses (or modular camera/lens combos) because the biggest factor is the individual lens or modular combo being tested. In general, CA is most evident at the full-on zoom setting at wide open or deep stops. Most CA can be minimized by shooting at less than full-on zoom in the f/4 to f/5.6 range. Upping the gain also helps to disguise some CA problems. I only do CA registration tests to reject or accept a particular cam for inclusion into our inventory or if someone is reporting a problem with it, not for comparison purposes. So (except for my general observations of the CA issue being more prevalent in HD lenses), the above should in no way be construed as any type of brand/model comparison test. |
|
| ||||||
|
|