Home
SearchSearch MemberlistMemberlist RegisterRegister ProfileProfile Log in to check your private messagesLog in to check your private messages Log inLog in
About post processing
View previous topic :: View next topic  

Do you post process your images?
I just resize and sharpen.
3%
 3%  [ 1 ]
Yes, but only a few basic settings, curves, levels, sharpening++
42%
 42%  [ 14 ]
I sometimes do alot of PP
33%
 33%  [ 11 ]
I shoot RAW and just resize and develop without adjustments
0%
 0%  [ 0 ]
No, I shoot jpg and just resize them for web
6%
 6%  [ 2 ]
I try to do minimum of PP on this forum to present a honest description of lens abilities og character
15%
 15%  [ 5 ]
Total Votes : 33



PostPosted: Fri May 01, 2015 4:14 am    Post subject: Reply with quote

Look, in that EOS, there is NO zero sharpening, OK? You must choose a value. It defaults to 3, but it has two values lesser than that and four above it. Is 3 zero sharpness? No. It's just the default value that Canon has chosen, which is in fact a quite muddy value. At least with my XS it is. So I choose not to use it. And I don't subscribe to your theory that somehow this sharpness level will work more on one image than another. That's the equivalent to choosing a rubber yardstick to make your measurements, which renders pointless any result. Your conjecture is meaningless unless you can back it up with hard evidence.


PostPosted: Fri May 01, 2015 2:37 pm    Post subject: Reply with quote

cooltouch wrote:
Look, in that EOS, there is NO zero sharpening, OK? You must choose a value. It defaults to 3, but it has two values lesser than that and four above it. Is 3 zero sharpness? No. It's just the default value that Canon has chosen, which is in fact a quite muddy value. At least with my XS it is. So I choose not to use it. And I don't subscribe to your theory that somehow this sharpness level will work more on one image than another. That's the equivalent to choosing a rubber yardstick to make your measurements, which renders pointless any result. Your conjecture is meaningless unless you can back it up with hard evidence.


Wow, on the web there is plenty of "hard evidence" -- explanations of how sharpening works -- not my "conjecture." Perhaps they can explain better so you can understand. I don't enjoy abuse when trying to help somebody -- I'm done with this.


PostPosted: Sat May 02, 2015 2:03 am    Post subject: Reply with quote

cooltouch wrote:
Look, in that EOS, there is NO zero sharpening, OK? You must choose a value. It defaults to 3, but it has two values lesser than that and four above it. Is 3 zero sharpness? No. It's just the default value that Canon has chosen, which is in fact a quite muddy value. At least with my XS it is. So I choose not to use it. And I don't subscribe to your theory that somehow this sharpness level will work more on one image than another. That's the equivalent to choosing a rubber yardstick to make your measurements, which renders pointless any result. Your conjecture is meaningless unless you can back it up with hard evidence.


Sharpening is always a positive value -- I would say on the XS a sharpening value of "1" would equate to "no sharpening", which is the goal. However, in-camera sharpening is not utilized when using raw image and conversion software.

It makes sense to me that feeding the sharpen algorithm different data will result in different output, i.e., the sharpness routines can possibly work better on one test image than another.

Would you, as you say, "subscribe" to the notion that the sharpened images get more similar as more sharpening is used. i.e. any sharpening makes the lens differences less apparent? That would make lens comparison more difficult, yes? Another reason not to sharpen images use to compare lens' sharpness...

Another factor is computer sharpening algorithms are typical better than in-camera algorithm, simple because more code can be fit on a computer, and, a computer has much more computing power, thus the algorithm can be more complex/accurate/featured, etc.. I would re-develop any images made using both in-camera and computer sharpening, to use only the computer algorithms, to get best results...


PostPosted: Sat May 02, 2015 2:49 am    Post subject: Reply with quote

visualopsins wrote:
cooltouch wrote:
Look, in that EOS, there is NO zero sharpening, OK? You must choose a value. It defaults to 3, but it has two values lesser than that and four above it. Is 3 zero sharpness? No. It's just the default value that Canon has chosen, which is in fact a quite muddy value. At least with my XS it is. So I choose not to use it. And I don't subscribe to your theory that somehow this sharpness level will work more on one image than another. That's the equivalent to choosing a rubber yardstick to make your measurements, which renders pointless any result. Your conjecture is meaningless unless you can back it up with hard evidence.


Sharpening is always a positive value -- I would say on the XS a sharpening value of "1" would equate to "no sharpening", which is the goal. However, in-camera sharpening is not utilized when using raw image and conversion software.


This is not true. Canon's DPP brings the camera's sharpness setting into its environment. But, just like the camera, this setting can be changed -- with the same results. And the value of "1" is not sharpening at all, it is the opposite. It blurs the image.

To quote Adobe on sharpening: "Sharpening enhances the definition of edges in an image." Really, that's all sharpening does. It enhances the edges in an image. It does not somehow draw forth dormant resolution that hasn't been expressed.

Quote:
It makes sense to me that feeding the sharpen algorithm different data will result in different output, i.e., the sharpness routines can possibly work better on one test image than another.


This may be the case if one of the images has no edges to enhance. Please recall the above definition. But this whole business of comparing different images is a false dichotomy because that's not what I'm doing. When I'm doing comparisons, I'm not comparing different images -- or at least different scenes. I'm' comparing the same identical scene with different lenses. So your -- I dunno -- thought experiment? -- it doesn't apply to what I'm doing. By comparing the same scene with different lenses, I make sure that everything, including the sharpness setting, is kept the same so that the only differences are due to the lenses being used.

Quote:
Would you, as you say, "subscribe" to the notion that the sharpened images get more similar as more sharpening is used. i.e. any sharpening makes the lens difference less apparent? That would make lens comparison more difficult, yes? Another reason not to sharpen images use to compare lens' sharpness...


It makes NO difference because I don't change the amount of sharpening when I'm doing lens comparisons. Let me fill you in on a little something I've noticed over the years about applying sharpness to images. If the image is already perfectly in focus, sharpening routines add nothing but noise. Even enhancing edge detail adds an insignificant amount of change -- except noise -- to an image. So if I've managed to achieve perfect focus with my test lenses, no sharpening -- beyond what the camera is set to -- is even necessary. And with my NEX, I don't even have a sharpening setting, so there's your zero sharpness right there.

Quote:

Another factor is computer sharpening algorithms are typical better than in-camera algorithm, simple because more code can be fit on a computer, and, a computer has much more computing power, thus the algorithm can be more complex/accurate/featured, etc.. I would re-develop any images made using both in-camera and computer sharpening, to use only the computer algorithms, to get best results...


But not if you're doing lens tests. Well, maybe when you are, but not when I am. I do NOT add any sharpening routines AT ALL to the images of my lens tests -- beyond where the camera was set, which is always 5 out of 7, in the case of my EOS XS. Have I made my point yet? Incidentally, I arrived at the "5" figure after doing many comparisons of images, studying them for artifacts, stair-steps, and other annoyances. "5" is the limit. Anything greater than "5" and artifacts appear. So I've never seen anything wrong with getting the most out of that camera's sensor, given that its anti-aliasing filter probably removed most of the sharpness to begin with. Oh, and when I mentioned above a perfectly focused image not benefiting from any further sharpening, I wasn't talking about in-camera sharpening. I was talking about my computer software's sharpening. Even it doesn't really make much of a difference -- except for adding noise.


PostPosted: Sat May 02, 2015 4:23 am    Post subject: Reply with quote

cooltouch wrote:
visualopsins wrote:
cooltouch wrote:
Look, in that EOS, there is NO zero sharpening, OK? You must choose a value. It defaults to 3, but it has two values lesser than that and four above it. Is 3 zero sharpness? No. It's just the default value that Canon has chosen, which is in fact a quite muddy value. At least with my XS it is. So I choose not to use it. And I don't subscribe to your theory that somehow this sharpness level will work more on one image than another. That's the equivalent to choosing a rubber yardstick to make your measurements, which renders pointless any result. Your conjecture is meaningless unless you can back it up with hard evidence.


Sharpening is always a positive value -- I would say on the XS a sharpening value of "1" would equate to "no sharpening", which is the goal. However, in-camera sharpening is not utilized when using raw image and conversion software.


This is not true. Canon's DPP brings the camera's sharpness setting into its environment. But, just like the camera, this setting can be changed -- with the same results. And the value of "1" is not sharpening at all, it is the opposite. It blurs the image.

To quote Adobe on sharpening: "Sharpening enhances the definition of edges in an image." Really, that's all sharpening does. It enhances the edges in an image. It does not somehow draw forth dormant resolution that hasn't been expressed.


DPP brings in the camera's sharpness setting, yes, but the sharpness scale in my version of DPP goes from 0 to 10 -- yours does not?! Sharpmess is always a positive value.

Would Adobe say that identical-scene/from different lenses sharpened comparison images are more similar than the unsharpened images? Comparing lens sharpmess would be easier using unsharpened images, yes?


PostPosted: Sat May 02, 2015 4:43 am    Post subject: Reply with quote

cooltouch wrote:
visualopsins wrote:
It makes sense to me that feeding the sharpen algorithm different data will result in different output, i.e., the sharpness routines can possibly work better on one test image than another.


This may be the case if one of the images has no edges to enhance. Please recall the above definition. But this whole business of comparing different images is a false dichotomy because that's not what I'm doing. When I'm doing comparisons, I'm not comparing different images -- or at least different scenes. I'm' comparing the same identical scene with different lenses. So your -- I dunno -- thought experiment? -- it doesn't apply to what I'm doing. By comparing the same scene with different lenses, I make sure that everything, including the sharpness setting, is kept the same so that the only differences are due to the lenses being used.


Of course I was speaking of comparing identical scene with different lenses...


PostPosted: Sat May 02, 2015 4:50 am    Post subject: Reply with quote

cooltouch wrote:
visualopsins wrote:
Would you, as you say, "subscribe" to the notion that the sharpened images get more similar as more sharpening is used. i.e. any sharpening makes the lens difference less apparent? That would make lens comparison more difficult, yes? Another reason not to sharpen images use to compare lens' sharpness...


It makes NO difference because I don't change the amount of sharpening when I'm doing lens comparisons. Let me fill you in on a little something I've noticed over the years about applying sharpness to images. If the image is already perfectly in focus, sharpening routines add nothing but noise. Even enhancing edge detail adds an insignificant amount of change -- except noise -- to an image. So if I've managed to achieve perfect focus with my test lenses, no sharpening -- beyond what the camera is set to -- is even necessary. And with my NEX, I don't even have a sharpening setting, so there's your zero sharpness right there.


I don't understand the part about "with my test lenses, no sharpening -- beyond what the camera is set to" -- the part about "what the camera is set to" implies to me there is sharpening in the camera of test lens images, that you keep the same sharpness setting value DPP pulls in instead of changing that to 0 in DPP before saving jpg.


PostPosted: Sat May 02, 2015 6:06 pm    Post subject: Reply with quote

I'm tired of this. It reminds me of beating my head against a wall. You go do what you think you should do and I'll go ahead and do what I think is valid. I will communicate no further on this matter.


PostPosted: Sat May 02, 2015 7:24 pm    Post subject: Reply with quote

cooltouch wrote:
I'm tired of this. It reminds me of beating my head against a wall. You go do what you think you should do and I'll go ahead and do what I think is valid. I will communicate no further on this matter.


Having beat your head against a wall before explains a lot. Laughing (ducking & running; sorry everybody, I could not resist.)

Cambridge in Colour A Learning Community for Photographers: GUIDE TO IMAGE SHARPENING