Ironically though this just goes to prove my point, because I can still screen shot the f/4L center and make it as sharp in PS as the f/2.8L II center @ f/4.
I hate sharpening if it can be avoided. Sharpening has consequences that often I don't like
Then you're missing out.
I do it at iso 100 and 200, then only sparingly going upward in iso
Sharpening can certainly be over done, but that doesn't mean it has to be over done. You can buy the very best sensor and lens small format can offer and end up with an image that has a lower IQ to human observers then a properly processed image from someone with mid-level equipment.
or you can buy sharp glass and a body with weak aa in the first place...
my 100 L on my 5dc required zero sharpening - if you sharpened, the images were oversharpened
In some respects hardware still dominates. A plugin cannot restore detail never resolved by the sensor+lens in the first place. NR can help but sensor size still dominates at high ISO.
it's a multi-variable function - sharpness AND NR, and they are not mutually exclusive
In others, software and post processing can represent a larger gain in IQ then hardware offers, sometimes even from the entry to the pro level.
I don't think of IQ being post processing
try to sharpen mushy line pairs - you get even greater mush
Have you tried it? ;-) One of the early articles on digital camera MTF50 resolution using Imatest points out that you can get very different results simply by sharpening prior to the test. (If I can find it I'll post the link.)
Here is an image from a review site where the reviewer claimed they could noise redux the 50d image and get better resolution than the 40d. Mush...I kept my 40d, the 40d colors were better also
I'm on board that most cannot tell difference between 18 mpxl and 36 mpxl on prints below 16 x 24 -- as you and fuzzy have proclaimed over the years - when at iso 100 and sharpening. But I shoot mostly much higher than iso 100 and I hate what sharpening does to damage high iso images.
Past a certain ISO...depending on both scene and taste...sharpening is no longer useful due to the way it impacts noise. You and I are in complete agreement there.
correct
IMHO those ISOs are not conducive to really large prints any way. Crop can definitely get mushy at higher ISOs but it's not the format size so much as the aggressive NR. FF also gets mushy as ISOs go up compared to itself at ISO 100. It just does so later then crop.
I just shot bball - ss 1/400 f2.8 with my 70-200 L at iso 12500 and the 7d2 - cleaned up nicely in DPP - no sharpening - 8x10's are not a problem
Therefore, it may just be me, but I'm using TDP and dxo as starting points for potential hints on what issues I might have with lenses and combos
The very nature of TDP's test will exaggerate differences, which may be perfect as a pointer, though it can also introduce differences that aren't really there (errors; shooting distances that are different from average for the lens). Don't get me wrong, I do like TDP a lot and will cite them. But you have to be on guard even there.
correct - Bryan messed over my 70-200 f2.8 L at the ISO site also
DxO's tests...even where they point you in the right direction, their gross misuse of terminology leaves a very false impression, especially between formats.
comparing between formats is wrong to do - agreed
And if you ignore the Mpix label entirely and treat it simply as a synthetic score, it still often doesn't make any sense.
agree - grain of salt - but can be an ooc sharpness indicator
and if other sites show the same - then a flag
I ask the question - is the lens camera combo appear to be sharp enough ooc so I will not ruin the image at hi iso - since I don't want to sharpen at hi iso
talk to the guy who couldn't get sharp feathers on the 7d + 100-400 I combo.
I didn't like the mark I even when I had the opportunity to shoot a friend's on FF. That's how I ended up with the 300 f/4L IS instead. But I'm glad you brought that up.
the 100-400 II with 1.4 III ext is on my wish list - 7d2 handles f8 with ext
DxO says 6 Mpix for the 100-400 and 7 Mpix for the 300 f/4L IS, both on the 7D. I have zero problem getting sharp anything out of the 300 on my 7D.
glad it works for you
And the time I did get to play with both on two different bodies (7D and 5D2) the 7D+300 combo still pulled way ahead of the 5D2+100-400 combo at 300mm, before any processing on my part. Yet according to DxO, both are apparently worthless on the 7D.
well, 6 MPix or 7MPix can be processed at low iso - to appear really good - I have great images from my 6 mpixl 10d - but my point goes back to ooc sharpness at high iso - my 100 L, my 35 f2 IS and my 70-200 L all meet my ooc sharpness expectations not to have to sharpen at high iso.
If I had listened to DxO (well...if DxO had been publishing Mpix back then) I would have taken out a small loan for a 300 f/2.8 I guess.
yeah - I wish to hit the lotto also
my point is this:
in order to come close to the FF IQ, the crop cameras must have sharp glass.
The 7d2 is only 1 stop away from the 5d3 in high iso and only 1/2 stop away from the 5d2 at high iso
but if you put weak glass on the 7d2, then you're going to have to sharpen more at hi iso and the noise differential between the FF and the 7d2 gets wider
I like to look at it this way
My 7d2 + 35 F2 IS will score 14 or 15 MPix at dxo (based on the way the 70d scored with this lens)
This is the FF equivalent of 56 f3.2 with 4 stop IS with a MPix of 14-15 and only 1/2 stop in high iso from the 5d2
The 5D2 with the great 24-70 f2.8 II only scores a MPix of 15 (call it ooc sharpness if you want vs resolution) AND DOESN'T have IS and is QUITE HEAVY and costs at least $1K more
BOTH are sharp enough not to require sharpening at high ISO
And I know it's a prime vs a zoom, but I'll just move my feet.
This is a crop setup that IMO closes the gap on FF advantage.
just my 2 cents on how to study and think about and potentially use and deploy the data sets from multiple sites
