p.26 #1 · Sigma DP2 Merrill: Have any of you tried it?
nandadevieast wrote:
Herb, as you put so nicely, DP2M can't be compared to different cameras on linear resolution basis only.
D800 will resolve more? Yes, if we shoot black and white charts. But DP2M has more color resolution.
The per pixel sharpness that DP2M has is not a result of resolution alone, whether 4mp or 14mp or 30mp all X3F files will have similar pixel level sharpness. When i look at a D800 file at 100 percent, i still see blurred detail, foliage etc...even 36mp won't be able to give me definition in leaves, infinity...the mushy ness is still there...which is not the case with DP2M files.
Its not about the number of megapixels. ...Show more →
nandadevleast,
Thanks for your reply. Most guys trust the numbers more than their senses
If you go to the dpreview site and compare image quality of different camera's, the Sigma Foveon sensor looks nothing special. My Canon 5D Mk2 and Sony Nex 7 easily match it or even exceed the Foveon sensor in these comparisons.
But at home comparing my own images, all my DP2M landscapes just look more real.
I was thinking about this very nice, really sharp and new Canon 24-70 mm. But then I thought, would I take it if I go out for my landscapes only to come back with still mushy files compared to the DP2M?
p.26 #7 · Sigma DP2 Merrill: Have any of you tried it?
Something different.
Two images made through an ExpoDisc.
Both at ISO 100 at f/8, 1EV overexposed.
One shot towards a complete clear blue sky.
The other towards an average landscape with land - (partly clouded) sky almost 50% - 50%.
Both processed the same. Exposure and WB leveled in the centre frame.
Maximized Vibrancy and Saturation. Tone curve at strong contrast.
We can clearly see that the sensor responds more dramatic with most probably more visible artifacts
with strong blue skylight. Of course this might be also true with very direct sunlight when shooting into the sun.
Every tool has it's limitation and should be used accordingly to prevent unwanted surprises. It is good to explore the limits just to take care of them.
p.26 #9 · Sigma DP2 Merrill: Have any of you tried it?
Do not be afraid of sharpening.
Sometimes images might look over sharpened or it might look there are patterns in grassy landscapes.
In that case you can back off a bit. But it is not forbidden to check what even more sharpening might do
on another image.
This image had +2.0 in SPP (and zero noise reduction!) and additional 10 - 0.5 - 25 - 50 in Lightroom.
No sharpening for the web. I am not shure it will be over the edge when posted but here it goes.
p.26 #10 · Sigma DP2 Merrill: Have any of you tried it?
Herb1911 wrote:
Do not be afraid of sharpening.
Sometimes images might look over sharpened or it might look there are patterns in grassy landscapes.
In that case you can back off a bit. But it is not forbidden to check what even more sharpening might do
on another image.
This image had +2.0 in SPP (and zero noise reduction!) and additional 10 - 0.5 - 25 - 50 in Lightroom.
No sharpening for the web. I am not shure it will be over the edge when posted but here it goes.
Herb
In my opinion, this version is not an improvement over the previous, less-sharpened one. The additional sharpening makes the bokeh more harsh and makes edges look less natural, but doesn't add to my perception of detail in the image.
p.26 #14 · Sigma DP2 Merrill: Have any of you tried it?
alundeb wrote:
Here you go, D800E (36 MP Bayer sensor), 100% crop
Thanks for the nice demonstration of best-in-class Bayer-filtered sensor performance. I think this provides a good comparison with Herb's earlier leafy corner crop for the performance of each sensor type.
The D800E does an admirable job of crisply defining the outlines of leaves/twigs/stems. However, for capturing the fine vein structure within each leaf, the D800E seems very hit-or-miss at the pixel level. I see a lot of muddled, spurious-resolution graininess and "checkerboard" artifacts, where the DP2M consistently presents a very pristine rendering of the natural leaf structures. The D800E tree bark detail often looks "painted" compared to the DP2M. Once the 36MP vs. 15MP difference is taken into account, the D800E image would be nice looking when downsized to DP2M sizes, but it certainly doesn't have anywhere near the advantage that raw pixel counts might imply.
p.26 #18 · Sigma DP2 Merrill: Have any of you tried it?
mpmendenhall wrote:
Thanks for the nice demonstration of best-in-class Bayer-filtered sensor performance. I think this provides a good comparison with Herb's earlier leafy corner crop for the performance of each sensor type.
The D800E does an admirable job of crisply defining the outlines of leaves/twigs/stems. However, for capturing the fine vein structure within each leaf, the D800E seems very hit-or-miss at the pixel level. I see a lot of muddled, spurious-resolution graininess and "checkerboard" artifacts, where the DP2M consistently presents a very pristine rendering of the natural leaf structures. The D800E tree bark detail often looks "painted" compared to the DP2M. Once the 36MP vs. 15MP difference is taken into account, the D800E image would be nice looking when downsized to DP2M sizes, but it certainly doesn't have anywhere near the advantage that raw pixel counts might imply....Show more →
I agree with pretty much everything you say.
It is more interesting for me to post the image first and then listen to comments from others.
Anyway, the main reason I posted it, was that I see so many horribly soft sample 100% crops from the D800E on the internet and even in these discussions, and I just wanted to provide a demonstration of the per pixel sharpness that is possible and that I routinely get from the camera.
Here is another 100% crop from the D800E (Voigtländer Nokton 58 mm at f/5.6).
p.26 #20 · Sigma DP2 Merrill: Have any of you tried it?
sculptormic wrote:
The Raw files are around 50 MP big. I do not assume those are full of worthless information.
I presume you mean 50 MB. Don't assume anything either way. The difference between compression algorithms between companies is huge. Just let the images speak for themselves.
