On 1/23/2017 7:57 PM, Jim Nichols wrote:
Moose,
As a potential X-T2 purchaser, I appreciate your taking time to present these results.
Then I'm glad I did!
I had downloaded and examined the images in my own clumsy way, and had come to the same conclusions. I am not at all
happy with the orange peel affect that can come from LRCC sharpening.
At least in part ham handed application, I suspect. The technique is actually quite old, deriving from a mechanical
technique using duplicate negs in film days. It's proper name, from the film process, is UnSharp Masking, USM. What it
does is increase contrast at edges, making them more obvious to human vision. Perhaps unfortunately, it is symmetrical.
So if it makes a line or lines of pixels darker on one side, it also makes the same number of pixels on the other side
lighter. This later causes things from slightly off looking edges to obvious looking halos.
PS some time ago added some 'smart' sharpening options that work better; are potentially more subtle than the original.
I don't know much about that, as I was already using other solutions. I'm guessing LR doesn't have that?
In any case, NeatImage works in a different way. In fact, it probably shouldn't be called sharpening at all, but we're
stuck with it, probably because 'deconvolution' is such an awkward, maybe slightly suspicious sounding, word and a
process that no one understands*
Suffice it to say that NI and other generalized deconvolution tools generally do a better job of recovering visible
detail without artifacts than USM based tools.
The Clarity slider in LR may also help in making details more visible.
Edgy Moose
* You could call it 'guess what the image looked like before being messed up by
lens and camera.' :-)
Yes, Mike, you are part of the tiny minority who test the rule. ;-)
I usually minimize sharpening in LR and end up using Focus Magic, and, in extreme cases, a touch of Neat Image. LR
does a lot of things well, but sharpening does not appear to be its strong suit.
Jim Nichols
Tullahoma, TN USA
On 1/23/2017 6:30 PM, Moose wrote:
On 1/22/2017 10:31 PM, SwissPace wrote:
Hi List, but mainly directed at any Fuji X-T2, X-Pro2 or X100F owners.
Am I the only one who is having difficulty getting good results from the raw X-T2 files in LR 6.8, I am not seeing
the sharpness and they seem to be very grainy when compared to the X-T1.
Color me befuddelated. I just looked at the IQ page of dpreview's test of the X-T2.
<https://www.dpreview.com/reviews/fujifilm-x-t2/6>
Their comparison tool continues to improve, and provides an excellent way to compare things like resolution, DR,etc.
between cameras, without relying on different subjects, lighting and other circumstances, memory, etc.
At 800 ISO, no matter what Lighting or Image size options I choose, JPEG or Raw, on any part of the target with
detail, the X-T2 image is sharper, clearer, more detailed and at least no noisier than the X-T1 image. At a couple of
other high-ish ISOs, the same thing holds. At 200, neither has much noise, but the sharpness/detail advantage holds
for the X-T2.
They do say "Raw files are manually corrected.", without any detail, but the differences are still quite clear in the
JPEGs.
I have found following this advice gave improvements
<http://petebridgwood.com/wp/2016/02/x-pro2-render-my-raw/>
I then tried getting rid of the noise/grain with dfine2 which does clear it up but reduces the sharpness slightly,
Ok I admit that may be down to technique or software how are the rest of you dealing with these new generation Fuji
raw files in LR
Mr Bridgwood is leading you astray.
<http://www.moosemystic.net/Gallery/tech/Fuji_X-T2/X-T2_proc2.htm>
I didn't do the NR and sharpening in LR, although I'm sure very similar results are possible with LR's tools, and
identical ones with the same plug-ins that I used. I'm more adept in PS and it's WAY easier to set up roll-overs with
layers.
One could argue that these portraits are interesting with the effect of his settings, but I don't think it's what you
are looking for.
On to something more subtle.
<http://www.moosemystic.net/Gallery/tech/Fuji_X-T2/X-T2_proc1.htm>
I don't know if it's sensor limitations, lens (probably not) or imperfect X-Trans sensor demosaicing, but there's a
point in the finest feathery parts where details just sort of mush away. It's not nearly as obvious, troublesome in
the hard edged, B&W test parts of the image, but also very obvious in the breast feathers of your woodpecker. It is,
in any case, there in both JPEG and Raw files.
Bridgwood's post is about printed results, and it's possible that such artificially crunchy files give him prints
that he likes. It might, for example, use entirely artificial texture to give a sense of detail in the mushy feathers.
here is an example of the grain that I am seeing at iso 800, it has only been processed using the advice above and
cropped - the Woodpecker
As you can see above, that grain is entirely artificial. all you have to do to get rid of it is not to create it in
the first place.
<https://thattimeoflife.smugmug.com/Other/OMlist/i-wLmq587/L> you may need to
view it at original size
Yup, has the Bridgwood grit look, all right.
Subtlety Moose
--
What if the Hokey Pokey *IS* what it's all about?
--
_________________________________________________________________
Options: http://lists.thomasclausen.net/mailman/listinfo/olympus
Archives: http://lists.thomasclausen.net/mailman/private/olympus/
Themed Olympus Photo Exhibition: http://www.tope.nl/
|