Home · Register · Join Upload & Sell

Moderated by: Fred Miranda
Username  

  New fredmiranda.com Mobile Site
  New Feature: SMS Notification alert
  New Feature: Buy & Sell Watchlist
  

FM Forums | Canon Forum | Join Upload & Sell

1
       2       3       4       end
  

Archive 2015 · 5DS-R sharpening.

  
 
ben egbert
Offline
• • • • • •
Upload & Sell: Off
p.1 #1 · p.1 #1 · 5DS-R sharpening.


A new camera a new workflow. Please share your sharpening workflow. For me this will be an ongoing refinement.

I typically sharpen in three steps.

1. Presharp during conversion.
2. Sharpen after creative work is finished. (in my case HDR which applies some clarity)
3. Output sharpen.

Presharpen.

I found that my old ACR settings 50,0.7,7 with mask 30 were almost impossible to see any change. That was, by moving the amount slider to full it hardly made a change. The key was the detail setting. After some experimenting, I settled on 40,1.0,40, mask 30.

I am still setting Luminance at 50% and cannot see any changes in detail sharpness at this setting. This is to avoid sky artifacts after HDR. The masking slider does not help me here.

Final sharpen.

A large radius small amount sharpening of 12,50,1

Then USM 100, 0.7, 0 darken mode.

Output sharpen.

Prints: Qimage at default settings, but I have not printed yet.

Web 850 wide: Bicubic smoother in one step followed by Topaz 333. (no change from old settings yet)



Jun 26, 2015 at 10:53 AM
ben egbert
Offline
• • • • • •
Upload & Sell: Off
p.1 #2 · p.1 #2 · 5DS-R sharpening.


While at Butterfly Lake awaiting sunset, I shot alternate brackets sets focused on the end of the log and then on the mountain. The difference in sharpness is very hard to see with a slight nod to infinity for the mountain and slightly sharper left side trees. But the right side corner is less sharp on the infinity shot.

I have taken these through the entire process including HDR and my sharpening methods. I don’t see any artifacts or halos.

Here are 100% crops of both versions. No output sharpening applied because these are not downsized other than the reference full image.

The left side looks pretty soft at this presentation, but will look pretty good at 20x30 print size. I do need to work on my technique here though. These are 17mm, f8 for near and f9 for infinity. I stopped down as a focus marker.

Edit:

I had a bad action that upsized the images before saving, I deleted all images and will repost later. Sorry.

Edited on Jun 26, 2015 at 12:44 PM · View previous versions



Jun 26, 2015 at 10:54 AM
ben egbert
Offline
• • • • • •
Upload & Sell: Off
p.1 #3 · p.1 #3 · 5DS-R sharpening.


Here is the first group of corrected images.





Full image







near focus, showing target focus area







log, with infinity focus







lower right corner near focus







lower right corner at infinity focus




Jun 26, 2015 at 01:16 PM
ben egbert
Offline
• • • • • •
Upload & Sell: Off
p.1 #4 · p.1 #4 · 5DS-R sharpening.


The rest.





Mountain with near focus







Mountain, infinity focus, this is the target







Left side near focus







Left side infinity focus




Jun 26, 2015 at 01:21 PM
newhaven
Offline
• • •
Upload & Sell: Off
p.1 #5 · p.1 #5 · 5DS-R sharpening.


After reading How to Focus a Landscape Scene, I began using a CoC value of 2.5* pixel size for depth of field calculations. For a quick read, scroll down to "A smaller CoC size for the digital era".


Jun 26, 2015 at 02:27 PM
ben egbert
Offline
• • • • • •
Upload & Sell: Off
p.1 #6 · p.1 #6 · 5DS-R sharpening.


newhaven wrote:
After reading How to Focus a Landscape Scene, I began using a CoC value of 2.5* pixel size for depth of field calculations. For a quick read, scroll down to "A smaller CoC size for the digital era".


But consider this, I show two focus points here, one very near and one at infinity (neither at hyperfocal) , Surely ideal focus is inside that range. But the fact is, both are within my tolerance for acceptable focus.

The near and far are so close it is hard for me to see any difference. Only at the left side and the corner do I see differences. I suspect a focus at the left side would have been nearly perfect .

The fact is, each set represents three images blended and I don't see any sign of motion blur. I suppose stopping down some might help, I need to test that.

I have been burned too many times with manual focus, but this camera with 16X live view and fine detail may prove the exception. I could try that, but it is hard for me to see how that would help.'

The point here, it is more than focus that is causing left and corner softness. It is no doubt the lens which is the 11-24. As super sharp lens, but edges and corners are always softer.



Jun 26, 2015 at 04:06 PM
Peter Figen
Online
• • • • •
Upload & Sell: On
p.1 #7 · p.1 #7 · 5DS-R sharpening.


Ben,

These all look pretty crunchy and over sharpened. Visible white halos and oversharpened noise. Are these 100% crops? It's weird, but you mountain near focus almost looks better than the far focus but maybe it's been sharpened more judging by the intensity of the white halo. And on the log too. The near focus just feels too crunchy and the two are shot at slight different apertures as well. There's visible stairstepping too, indicating an abundance of sharpening, but really, you shouldn't be seeing ANY stairstepping at 100 percent. I'd be curious to try processing these in both C1 and Iridient.



Jun 26, 2015 at 05:07 PM
newhaven
Offline
• • •
Upload & Sell: Off
p.1 #8 · p.1 #8 · 5DS-R sharpening.


If focused at infinity DoF spans from hyperfocal distance to infinity.

Both the near and far images look very sharp. My guess is the end of the tree is very near the hyperfocal distance.



Jun 26, 2015 at 06:22 PM
ben egbert
Offline
• • • • • •
Upload & Sell: Off
p.1 #9 · p.1 #9 · 5DS-R sharpening.


Peter Figen wrote:
Ben,

These all look pretty crunchy and over sharpened. Visible white halos and oversharpened noise. Are these 100% crops? It's weird, but you mountain near focus almost looks better than the far focus but maybe it's been sharpened more judging by the intensity of the white halo. And on the log too. The near focus just feels too crunchy and the two are shot at slight different apertures as well. There's visible stairstepping too, indicating an abundance of sharpening, but really, you shouldn't be seeing ANY stairstepping at 100 percent. I'd be curious to try processing these in both
...Show more

Man, I sure don't see any white halo, remember, I am using darken which produces a black halo and I do see that, but heck about 1 pixels worth. I don't see stair stepping or crunchy either. I shot at different apertures because there is no way to tell the focus point in live view after the fact, so I used the aperture as an indicator.

I could maybe put one of the mid exposure shots on Dropbox for you to play with, But you would get different results because the HDR adds some clarity.

I can easily step back on sharpening. Just tell me what to send. near or far.



Jun 26, 2015 at 06:29 PM
ben egbert
Offline
• • • • • •
Upload & Sell: Off
p.1 #10 · p.1 #10 · 5DS-R sharpening.


newhaven wrote:
If focused at infinity DoF spans from hyperfocal distance to infinity.

Both the near and far images look very sharp. My guess is the end of the tree is very near the hyperfocal distance.


That is pretty close to where I had the near focus. I could see the lens infinity position mark change slightly between near and far, both were about at the infinity mark.

I found that the 17TSE had to be focused about where that left tree is because of field curvature. I did some experiments with the 11-24 and thought that was not the case,but now I will revisit that decision.

Come to think of it, I have not had the 17TSE mounted yet so I need to give it a try.

By the way, the full size is a later shot and by that time I had moved infinity focus to the far shore, the left and right corner are both slightly better than the samples posted here that are marked infinity.


By the way, hyperfocal for a CoC .005 is 12 feet, about where that log ended.



Jun 26, 2015 at 06:32 PM
ben egbert
Offline
• • • • • •
Upload & Sell: Off
p.1 #11 · p.1 #11 · 5DS-R sharpening.


I just printed these at 300 PPI, these crops are only 3.4 inches wide, much smaller than displayed here. My monitor displays these at 11 inches wide.

We are all going to need to get used to the image size math.



Jun 26, 2015 at 06:52 PM
Peter Figen
Online
• • • • •
Upload & Sell: On
p.1 #12 · p.1 #12 · 5DS-R sharpening.


Ben - I'm just saying what I'm seeing. Here are the two images with markers pointing to what I'm seeing. The artifacts aren't huge, but they are visible and would get amplified with any kind of output sharpening. I'd love to check out one of the raw files, or put a couple in there if you've got exposure brackets. You've got me email, I think.












Jun 26, 2015 at 07:31 PM
ben egbert
Offline
• • • • • •
Upload & Sell: Off
p.1 #13 · p.1 #13 · 5DS-R sharpening.


Here is a dropbox public link to a mid exposure near and mid exposure infinity. Too big to go by my email. By the way, these mid exposures process fine on their own if you don't do HDR.

Those halos are pretty small, and even smaller at print size which like I said is only 3.4 inches wide.

https://dl.dropboxusercontent.com/u/77536572/150623-0338-5DS-%20R.CR2
https://dl.dropboxusercontent.com/u/77536572/150623-0341-5DS-%20R.CR2





Jun 26, 2015 at 07:43 PM
gdanmitchell
Offline
• • • • • • •
Upload & Sell: Off
p.1 #14 · p.1 #14 · 5DS-R sharpening.


I would lower the sharpening "amount" in the case of the jaggies on the log. I wouldn't worry about that little bit of a halo at 100% magnification along the ridge — it won't be visible in a print.

Also, keep in mind that it is often a good idea to slightly "over-sharpen" for printing on inkjet printers.

I also noticed the effect on the rocky section of the mountain face at 100%. It looks like what we might see if we try to slightly over sharpen in order to make a print that pushes the upper print size boundary. It also would not likely be visible in a print, but it is something to keep an eye on.

Dan

Peter Figen wrote:
Ben - I'm just saying what I'm seeing. Here are the two images with markers pointing to what I'm seeing. The artifacts aren't huge, but they are visible and would get amplified with any kind of output sharpening. I'd love to check out one of the raw files, or put a couple in there if you've got exposure brackets. You've got me email, I think.




Jun 26, 2015 at 08:33 PM
ben egbert
Offline
• • • • • •
Upload & Sell: Off
p.1 #15 · p.1 #15 · 5DS-R sharpening.


Ok, guys, I am going to need to magnify that and see if I can see the jaggies. I will play with the sharpening until I get rid of it.

Looking forward to other formulas.



Jun 26, 2015 at 09:00 PM
ben egbert
Offline
• • • • • •
Upload & Sell: Off
p.1 #16 · p.1 #16 · 5DS-R sharpening.


Ok, at 200% I can see jaggies. But at 200% I am starting to see jaggies everywhere and pre sharpened. But at 50% of my original strength, I think it is better. But I am going to wait for other comments.

I printed these and the only thing noticeable at a 20x30 print size is the soft left side and the lower right corner. And I have looked at plenty of galleries and this is not at all unusual or unacceptable.






Jun 26, 2015 at 09:28 PM
dakel
Offline
• • •
Upload & Sell: On
p.1 #17 · p.1 #17 · 5DS-R sharpening.


Are you sure you didn't accidentally shoot that photo with a Fuji X-Trans sensor? Ok that was a joke, but the crops do look un-photo like to me.

Beautiful scene by the way and I bet you are happy to have captured it with a 50MP camera. Hopefully you get the sharpening routine down!



Jun 26, 2015 at 09:41 PM
Peter Figen
Online
• • • • •
Upload & Sell: On
p.1 #18 · p.1 #18 · 5DS-R sharpening.


I've started to play around with the raw files and have processed them on both CaptureOne v.8.3 and Iridient Developer. I've left the studio for the weekend (plumbing and auto repair chores this weekend) but on initial inspection, I'm not seeing either the halos or the jaggies in either of these two applications, even after applying sharpening in Ps. Even though Dan is correct about it probably not being an issue at this point, the point is, is that this is not your output sharpening, which will only make both worse, and on a large print, you'd be surprised how big that halo looks up close. And just for shits and grins, I tried a version in Iridient using a new interpolation they have that is very slow but very good, and compared to interpolating in Ps using Bicubic Retaining Detail (or whatever their best thing is called) the Iridient version was palpably better, which is good to know if and when you've got to go really big.

Ben - Did you use any tilts on this image? It was with the T/S-E, right? I'm asking because the left side of the image is noticeably softer than the right side, while the distant mountains in the center are quite sharp, plus the left side has a rather large amount of color fringing - more than the right side. Just curious on that front. But a beautiful composition overall.



Jun 26, 2015 at 10:13 PM
ben egbert
Offline
• • • • • •
Upload & Sell: Off
p.1 #19 · p.1 #19 · 5DS-R sharpening.


dakel wrote:
Are you sure you didn't accidentally shoot that photo with a Fuji X-Trans sensor? Ok that was a joke, but the crops do look un-photo like to me.

Beautiful scene by the way and I bet you are happy to have captured it with a 50MP camera. Hopefully you get the sharpening routine down!


Thanks, I plan on printing the sunset version. I am never sure what unphoto like means. Can you elaborate? I work to get my HDR's as realistic as possible. but not photographically realistic, more eyeball realistic.

The sharpening routine I have now is fine for printing, maybe not for 100% web crops though.

I just went outside and tried an experiment with center versus edge focus at 17mm f8. Then with manual focus and then with the 17TSE. Unfortunately the light was changing too fast and they will be hard to compare. But I did learn some stuff. Even at 16X there is not way I can match live view focus.

Live view focus sends the focus to the exact same place every time as judged by the infinity marker. When I manual focus, I get a range that can be seen on the dial and it varies from one attempt to the next. AF is the same way. During MA you can see how variable the AF is.

I also noted that the dial went to a slightly different location when I live viewed the edge versus the center. Very slight.

I need to find a better scene and light.




Jun 26, 2015 at 10:19 PM
ben egbert
Offline
• • • • • •
Upload & Sell: Off
p.1 #20 · p.1 #20 · 5DS-R sharpening.


Peter Figen wrote:
I've started to play around with the raw files and have processed them on both CaptureOne v.8.3 and Iridient Developer. I've left the studio for the weekend (plumbing and auto repair chores this weekend) but on initial inspection, I'm not seeing either the halos or the jaggies in either of these two applications, even after applying sharpening in Ps. Even though Dan is correct about it probably not being an issue at this point, the point is, is that this is not your output sharpening, which will only make both worse, and on a large print, you'd be surprised
...Show more

This was the 11-24 lens no tilt used. The light at the left was very strong as that was the direction it was coming from.

I already printed some 20x30 equivalent samples and I am not seeing any halos on the prints. Remember, these crops are nearly 11 inches wide on my monitor, but only 3.4 wide in a 20x30 print size. I do see the softness however.

My 17TSe needs to be focused on the edge not the center so maybe the 11-24 does as well.



Jun 26, 2015 at 10:25 PM
1
       2       3       4       end




FM Forums | Canon Forum | Join Upload & Sell

1
       2       3       4       end
    
 

You are not logged in. Login or Register

Username       Or Reset password



This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.