Nathan67 wrote:
The original was not sharpened, the hasselblad files are not like dlsr files... comnpared to my canon files they need a whole lot less... and this was sharp as.... to be honest I am not seeing the "wierd artifacts" usm applied here was minimal...
It is basically the left edge of the left mountain. There is a dark line all the way down, grey near the top, blue near the bottom. The sky looks slightly brighter in the same places, along the same edge. Compare to the right mountain and you will see that they look different.
I had a Sinar 22MP eMotion 54LV for my Contax 645 for some time, and found I couldn't really sharpen at all sometimes. Artifacts instantly. Have you tried without USM?
Ok.. here is a fun one... part of a series on Lego Star Wars... ( I have two small sons in the house and I find lego figures everywhere...) The title Stue Wars is a play on words... Stue is the Norwegian for living room... enjoy.
this with canon 1DSmkIII and 851.2L I read at the start that pano with any equipment was OK here....
http://www.lediardfoto.com/stuewarspanoweb.jpg
And a making of pano... taken in-camera with my fuji x-100 in pano mode....
carstenw wrote:
It is basically the left edge of the left mountain. There is a dark line all the way down, grey near the top, blue near the bottom. The sky looks slightly brighter in the same places, along the same edge. Compare to the right mountain and you will see that they look different.
I had a Sinar 22MP eMotion 54LV for my Contax 645 for some time, and found I couldn't really sharpen at all sometimes. Artifacts instantly. Have you tried without USM?
Actually that line you are seeing is there in the full rez file too.. but there you can of course see that the snow does not go all the way up on that side of the mountain.. there is a sharp ridge... as for the variations in sky colour.. at 100% you can see small clouds in that area...
Is the sky also slightly lighter there in real life? I guess the blue line is a limitation of the lens you used then. Maybe run a lens correction on the file?
alundeb wrote:
Sure you mean 0.1 and not 1.0? Radius 0.1 pixels is almost invisible in already sharpened images, even at 500%.
0.1 radius is easily visible with PS smart sharpen, especially if you use it in "remove lens blur" mode, even at 10% strength. I usually do my final sharpening (before converting to srgb and saving to jpeg) at 0.1 pixels radius at 20% and then repeat once more at 10%.
AhamB wrote:
0.1 radius is easily visible with PS smart sharpen, especially if you use it in "remove lens blur" mode, even at 10% strength. I usually do my final sharpening (before converting to srgb and saving to jpeg) at 0.1 pixels radius at 20% and then repeat once more at 10%.
I know, but he said "usm". That is a specific function in PS, not a description that fits all sharpening methods. Maybe he meant "smart sharpen" instead then.
I see sharpening haloes in both of Nathan's shots, especially the village one when viewed at full size. Still, much improved from the original images he posted.
Ok guys, i am not stupid... Wien i say 0.1 radius usm, i mean usm ok,.. Not smart sharpen, if I had meant smart sharpen i would have said Smart Sharpen
As for what you are seeing carsten, I see no blue line... Trust me, the lens is no problem ok... I'm not into measurebating really... I could post 100% crops but really, no point...
Nathan, thanks for you contribution, it was interesting and very nice images. Welcome back soon!
It must be something wrong with my photoshop CS5 then, since 30% USM on the resized Bud pano is completely invisible until I increase the radius to 0.4 pixels. USM with 0.1 pixels radius is actually invisible at all strengths on my computer. I run CS5.1 32 bit on Windows.
I didn't assume that you used Photoshop. It is just that USM with a specified radius and strength is supposed to behave the same in different programs. If you use another USM engine, and radius 0.1 pixels and more than 30% gives haols, then your engine is much rougher with USM than photoshop. It could even be that it rounds the radius up to a larger effective quantity. It should be possible to sharpen an already sharp image more than USM 0.2 pixel and 50% without getting problems with halos.
Well... I after downsizing using that better algorithm sharpened using o.1 and increased the % until I could see artefacts and then backed off a bit.. This was in graphic converter for mac which has the lanczos3... Felt it was better to sharpen and save there than save, openn in ps sharpen there and so save again... But really, you guys are really picking at images... And to be fair really at the size they are being presented, is that really worth it?
Great, I got reminded by this thread that there are better algorithms for downsizing... After years of aperture, then Lightroom workflow things get kinda forgotten.... But really, I never asked for ths kind of nitpicking, nor is it really in the spirit of the thread in my opinion... I just can't be arsed with it all... To put it bluntly
(seems I wasn't quite all the way OUT)
Nathan, in the end I guess the reason a couple of people chimed in with comments is that the sharpening/resizing didn't look that great, compared to many other shots, and people wanted to help out. If you don't want to bother with it, fine, I guess no one will comment again.
