Barry Goyette
lvl.4
Flight distance : 14928 ft
United States
Offline
|
Skyris Posted at 2017-12-9 21:40
Sorry i didnt realise that every sample dng someone had sent in hadnt focused properly.
All I am saying is that every sample I have seen has been soft and mushy at 100% I am only talking about sharpness here, my $400 lens suffers from poor vignetting at f 2.8-f4 but can be corrected in post the sharpness cant (well to a certain degree)...
The x5s was an expensive camera and the stills werent that great and had been fighting to get sharp stills on every shoot for my clients.
ok ...first...you're that guy? with the crazy bracket hanging off his i2? Let me just say its an honor to meet you. Any one who goes to those lengths to up his image Q deserves our respect and admiration.
But.....let's get real. There is a distinct difference between your $400 lens sitting on a full frame 42mp sensor that HAS NO OLPF....and another lens that is sitting on a 24mp sensor that has a cinema grade (read: strong) OLPF.
I get your argument that at 200% one looks sharp and the other doesn't. But 99% of the time when see this type of difference, it's not the lens...it's the OLPF (which believe it or not, has a purpose!). If you took your fancy $400 lens and married it to the x7 ( (or vice versa with the DL lens), you would see what I mean. As it too, would look blurry compared to what you see on the a7r2, (which has no filter designed to remove high frequency detail like the x7 does.)
Your screenshot on my computer is displaying at 200%, and looks blurry. A camera with an OLPF is never going to look sharp at 200%, as the detail is being purposely cancelled out so that motion related aliasing doesn't occur at 100% (this type of aliasing is a hallmark of dii's processing since the beginning and is one of the reasons its cameras have generally not been accepted by the film industry --even the x5s aliases a bit, and this would appear to be something DJI has tried to fix with the x7. (Take a look at this screenshot at 100%, sharpened properly...and take your eyes off that overexposed car for a second, and look at the long grass at the top of the frame. (ie...stop looking for aliased hard edges as proof of resolution...they aren't... they are proof of aliasing, which isn't detail.)
Something not shown by ray's tests is the increase in color resolution and depth in the x7 over the x5s, which is substantial, which delivers an image with more apparent clarity and depth, and to me makes the camera worth every penny, but I'm getting off-message).
Personally, if I were in your situation, given your goals of maximum resolution, I'd take every dollar I just spent on an x7 and lens and apply it to a copter that can actually carry your a7r2. You'll be happier. The x7 wasn't designed for what you want it to do. My guess is that that camera will show up sooner or later on dii's website, but it's not the camera that the x7 is now.
If you're interested in looking at the tests I referred to, they are here https://www.dropbox.com/sh/iym7r ... 03jbz_2hjWzE4a?dl=0
There are two things to look at here...one is the 9.5 rings, which are being resolved, even in the corners, on all the lenses, including the 16mm in the corners, which is impressive. Secondly, and it's the most important thing, is that it's doing it with absolutely zero moire or aliasing. This is super important in cinematography, and I would argue, that DJI got this tuned almost exactly right given that's the market this camera is targeted at. If you want to get a deeper understanding, you'll need to search up Ian's charts for the X5s (look in december/january on the I2 forum at rcgroups. You'll see how much better the x7 performs in a true "side by side" than the x5s did on both fronts (overall image resolution and aliasing/moire).
|
-
|