Was giving up on a AF100 but decided to reconsider.

lpowell said:
As for "resolution numbers", there's really no objective way to measure the resolution, it's a subjective judgment call. At high spatial resolutions, the image sensor's output is contaminated with ultra-fine aliasing artifacts that are literally folded down into the upper resolution range. As a result, the genuine resolution of the sensor is submerged within spurious alias products and there's no way to separate them. With digital cameras, "sharpness" is a marriage of refined technology and aesthetic preference.
Click to expand...

Yep. That's what makes it so hard. You can define a limit at which aliasing starts, but it's not like an on/off point, it's -- complex. The GH2 definitely can resolve some details that the AF100's OLPF blurs out, but the tradeoff is that the GH2 also lets more obvious aliasing through, which the AF100 effectively filters out. So where does one stop and the next begin? On a top-end camera like an F65, it's pretty easy to tell, as the fine detail just gradually gets finer and finer until it merges into a smooth mush of gray. But with a GH2, AF100, or even an F3 or FS100, it doesn't work that way. On something like a 7D, it's atrocious, the way the lines blur into a big purple/orange mess.

But even with the GH2, you can see the difficulty... Here's a sample chart from the GH2 in ETC mode. This shows what the GH2 looks like when resolving the chart in ETC, and it basically looks like you'd expect -- clean detail until it extinguishes into gray...
View attachment 58851

But here is an example of what the GH2 looks like in full-sensor mode. This isn't actually the GH2, as I can't find my shot of that right now, this is from the 7D, but the same basic pattern holds in the blacks (ignore the purple/green, that's a Canon phenomenon:
7D-1080-Mid-Trumpets-sideways-lines.jpg


So which is the real res limit? The GH2 in ETC (tiny-sensor cropped) mode looks much better on a res chart than the GH2 in full-sensor large-frame mode. Yet I don't think you'll find anyone who would argue that the GH2 in ETC is "sharper" than the GH2 in full-sensor!

These mega-oversampling/downconverting sensors have changed things and it's not that easy to get to the bottom of what's real and what isn't real detail. That's why I like the wringer, it's designed to do exactly that, especially with color res, but my Wringer isn't high enough res to do an accurate comparison because both the GH2 and the AF100 can resolve it all.
 
This thread has amused me for its life. The resolution minutia of $3000 cams are being picked apart when the functionality of the cam will mean much more. Oh, and don't forget the quality of the lens! A decent quality prime is going to cost more than the cam.

Grant
 
ggrantly said:
This thread has amused me for its life. The resolution minutia of $3000 cams are being picked apart when the functionality of the cam will mean much more. Oh, and don't forget the quality of the lens! A decent quality prime is going to cost more than the cam.

Grant
Click to expand...
Totally agree, the functionality is far more important (to me, at least). I couldn't give a flying fig about another 100 pixels of res, if I can't even monitor the footage accurately!

To me there's a baseline image quality standpoint, and a baseline functionality standpoint. Brilliant images from an unworkable camera are pointless. Brilliant functionality and lousy images are pointless. It's gotta meet minimum standards before it can be considered. And those standards are, obviously, subjective -- there are folks who have lots of time and energy and no money, and they will likely be more willing to put up with a less-functional camera. There are those who need a quick turnaround, who simply cannot be bothered with hacks and workarounds and jerry-rigging solutions and frankenrigs. And there's folks in the middle.

Good thing there's so many different camera models on the market!
 
Barry_Green said:
A lot of what you're reading (like that Art Adams article) was done when the AF100 was brand new and nobody really knew how to shoot with it yet.

There is zero noise in the AF100, when you set the settings properly. ZERO. Zero zero zero discernible noise. Look at this vimeo for comparisons of the noise levels at various ISO's:


As far as highlights go, what people are really complaining about there is when different chroma channels clip at different levels, causing a shift in highlights. And yes, the first answer I offer is "don't overexpose". That's been rule #1 with video, ever since video first came about, fer cryin' out loud. But if you want to increase your overexposure latitude without encountering color shifting, that's also easily done -- once again, it's just about knowing the camera and configuring it for what you're trying to do. The AF100 has a very saturated color palette, and the higher the color saturation, the higher the differences between color channels, and therefore the more likelihood that overexposing will cause one of those color channels to clip. So desaturate the image, and use cine-D gamma, and you will vastly expand the range you can shoot under without encountering color shifts.
Click to expand...

Man, I really really wish I could have attended your AF100 class in LA last year Barry! Are you ever going to do another one?
 
Jon Furtado said:
Man, I really really wish I could have attended your AF100 class in LA last year Barry! Are you ever going to do another one?
Click to expand...
The Stratos schedule and other commitments have really cut down on my live seminars recently, so I don't know. But the AF100 DVD exists, which is basically everything I know (or, well, knew at the time!) presented interactively, so it's been what I refer folks to.
 
Another way to circumvent color shifting in highlights is to do a careful WB with your AF. I use the X rite gray chart exposed to 70 units, and that one thing has made a huge difference in the way high lights look even if they do get a bit over exposed. I've made some really great looking images with my AF since I have had it and the Xrite chart was really the one piece of the puzzle that made it all work for me.

I do not have any financial interest in X rite, just buy some of their charts.
 
You must log in or register to reply here.
Back
Top