The GetDPI Photography Forum

Great to see you here. Join our insightful photographic forum today and start tapping into a huge wealth of photographic knowledge. Completing our simple registration process will allow you to gain access to exclusive content, add your own topics and posts, share your work and connect with other members through your own private inbox! And don’t forget to say hi!

"Orange Peel" Texture Enhanced Tangerine

k-hawinkler

Well-known member
In diglloyd: Sony A7R: ?Orange Peel? Texture Lloyd Chambers speculates about the nature of the “orange peel” texture of some of his Sony A7R images. Well, I thought it may be educational to take a single picture of tangerines, process the Sony raw file with Capture One Pro (C1), and show 2 resulting images, one without and one with “orange peel” texture.

I only modified the 3 Sharpening parameters Amount, Radius, and Threshold.
Setting these 3 parameters to minimum or maximum values like these

1000/0.2/0.0 or
1000/0.2/12.0 or
1000/2.5/12.0

results in an image like the following without the dreaded “orange peel” texture.



The 3 different images are basically indistinguishable.


However with this set of parameters

1000/2.5/0.0

the generated image shows the so called “orange peel” texture.



My take away from this is to keep the Radius smaller than 1, and Threshold larger than 0.
A good starting set for sharpening parameters in C1 that doesn't generate the “orange peel” texture might be something like this:

less than 300/0.6/1.0

or C1's default settings of:

120/0.8/1.0

I tend to look at an image at 200% to 400% magnification and aggressively move the sliders.
Then it's easy to spot the onset of “orange peel” texture and back off the sliders.

I shot the raw file of the tangerines with A7r2 and 55/1.8 FE.
 

ErikKaffehr

Well-known member
Hi,

Thanks for sharing!

I am new to the A7rII and I have not seen that kinds of artefacts on any of my Sony cameras. I do think I have a very basic understanding of the theory involved and I don't think it can explain what Lloyd claims he can see. None of his postings is really clear on which parts of the image should have the claimed artefacts. I guess that you are expected to see…

But, I don't think brain works that way. Once you have observed some kind of artefact you can be conscious about it. Personally, I am likely to react to axial chromatic aberrations and colour aliasing, those are the artefacts I have learned to hate.

Now, Lloyd Chambers complains a lot about tonal compression scheme, which I would say is basically sound. Now that shifts focus from the "Delta compression" which have been proven to cause artefacts under some conditions. Regarding that Delta compression I would really suggest that Sony needs to reconsider it.

Regarding the tonal compression, I don't think it is major issue, but I would guess that Sony needs a new image processing pipeline to implement linear 14-bit coding. They may need to do it sooner or later.

Best regards
Erik


In diglloyd: Sony A7R: ?Orange Peel? Texture Lloyd Chambers speculates about the nature of the “orange peel” texture of some of his Sony A7R images. Well, I thought it may be educational to take a single picture of tangerines, process the Sony raw file with Capture One Pro (C1), and show 2 resulting images, one without and one with “orange peel” texture.

I only modified the 3 Sharpening parameters Amount, Radius, and Threshold.
Setting these 3 parameters to minimum or maximum values like these

1000/0.2/0.0 or
1000/0.2/12.0 or
1000/2.5/12.0

results in an image like the following without the dreaded “orange peel” texture.



The 3 different images are basically indistinguishable.


However with this set of parameters

1000/2.5/0.0

the generated image shows the so called “orange peel” texture.



My take away from this is to keep the Radius smaller than 1, and Threshold larger than 0.
A good starting set for sharpening parameters in C1 that doesn't generate the “orange peel” texture might be something like this:

less than 300/0.6/1.0

or C1's default settings of:

120/0.8/1.0

I tend to look at an image at 200% to 400% magnification and aggressively move the sliders.
Then it's easy to spot the onset of “orange peel” texture and back off the sliders.

I shot the raw file of the tangerines with A7r2 and 55/1.8 FE.
 

fmueller

Active member
I see orange peel, lots of it in both versions.
:ROTFL:

Somebody should send this to DigLloyd.
 
Last edited:

dmward

Member
My experience with sharpening is that its always possible to go too far.
Several people steeped in sharpening using Lightroom, C1, and/or Photoshop suggest that one should evaluate sharpening at 1:2 since most often printing or publishing to the internet masks sharpening artifacts to some extent if one knows what they are doing while doing creative sharpening. One is never going to be looking at an image as closely as it appears on screen at 1:1 let alone 2:1 or greater. After all, an A7RII image, printed at 300 pixels per inch, the standard for C prints suggests looking at sharpening at 1:3 on a 90 pixel per inch monitor. According to my math that's a 29" print on the long side. 240 pixels per inch is a 33 inch print. And that's still more than 1:2.

Looking at the two versions posted, what would cause me more concern that the orange peel that probably won't be seen at 1:2 is the black halo that has appeared around the central tangerine when the sharpening was applied.
 

Lars

Active member
Interesting - thanks for posting. The first image is clearly more appealing.

Any chance you can post an unsharpened image for reference?
 

k-hawinkler

Well-known member
Interesting - thanks for posting. The first image is clearly more appealing.

Any chance you can post an unsharpened image for reference?

Thanks, sure, no problem. Here goes.

Amount/Radius/Threshold 0/0.2/0.0



It's an ISO 500 shot, so I used Define 2 in Nik.


I think we can all agree that the original parameters I used were completely off the wall in order to make a point.
My conclusion is to be careful so that not inadvertently low amplitude high frequency noise gets amplified through post processing.
If one chooses to work with a very large Detail setting for certain images, then one also has to pay attention in particular to Masking, please see previous sentence.
In this case a too large value for Amount and Radius is also not good.

I certainly think 14 bit raw data with lossless compression will be more robust than the current implementation.
However, I would not attribute the items discussed here to Sony's current implementation of their raw file.
 

Lars

Active member
It's an ISO 500 shot, so I used Define 2 in Nik.

I think we can all agree that the original parameters I used were completely off the wall in order to make a point.
My conclusion is to be careful so that not inadvertently low amplitude high frequency noise gets amplified through post processing.
If one chooses to work with a very large Detail setting for certain images, then one also has to pay attention in particular to Masking, please see previous sentence.
In this case a too large value for Amount and Radius is also not good.

I certainly think 14 bit raw data with lossless compression will be more robust than the current implementation.
However, I would not attribute the items discussed here to Sony's current implementation of their raw file.
Thanks. Yeah in a well exposed image like this 12 or 14 bits is so far below the signal you'll probably never see any difference in sharpening. You might be able to measure it but it won't show to the eye. IMHO.
 
Top