Original compared with ACR and Photo Ninja

[John Geerts]

First of all, being not a huge fan of 'dramatic' post-processing, the motivation is to shoot as to be published. But reality is different. Possible cropping, downsizing, exposure-correction (shooting with  underexposure), noise problems and some other minor things. 

For a number of reasons now entered the trial-period with Photo Ninja and getting to learn the software, which looks more logical than ACR in a number of ways.

For a bit of testing, especially on geometric changes,  a streetview  with the Nikkor AFS  F/2.8   17-35     at    17mm in F/8 
First the original (not processed), Second the Photo processed in ACR and the third one in PN.                       - love to hear the comments.

[Akira]

John, I'm not big fan of exaggerated post-processing (PP) either, so far as snap shots or land- / cityscapes that are meant to look natural are concerned.  (Of course, this doesn't apply to any artistic expressions.)  For the natural looking scenes, I'm mostly bothered by the exaggerated contrast, clarity (micro contrast), sharpening and saturation.

In this particular case, I would find the clouds of the third example are a bit too "dramatized".  On the other hand, I'm not really bothered by the enhanced perspective of the super-wideangle lens.

[pluton]

Are you asking about the cropping or the distortion correction, or both?

[Frank Fremerey]

What is the "Original"? JPEG out of cam and which settings?

I see the biggest difference between the 3 versions is the distortion correction & crop as Keith put it.

Why do you aim to do these in the RAW processor anyway?

For me RAW processing is to extract maximum tonality for the goal I am aming at. A technical pic needs different treatment from a highkey....

distortion I care about later in Photoshop. Crop too.

[Airy]

No opinion here - it looks like you processed the pic in ACR and PN with different goals in mind. The PN pic looks like having "excessive" contrast and saturation, something you also can achieve with ACR.

It is only recently that I became aware that I was mistaking software presets for software "signature". Presets are quite arbitrary and relate, probably, with marketing thoughts, a bit like the warmish-yellowish cast imposed on printed photos by the shop around the corner - shouldn't all those summer snapshots be filled with sunshine ?

I am pretty sure that, as long as you stay far from extremes, you will be able to produce similar results with both pieces of software. I'd expect biggest differences coming from highlight recovery, shadow recovery, sharpening, and noise reduction (especially in conjunction: de-noising + sharpening). Your sample picture does not however test any of these *probably* discriminating factors.

The Fierce Bear of the North mentioned highlight recovery as one strength of PN. I'd be curious about the other criteria. I have not tested these for myself yet. In my D700 times, I was more interested in de-noising, and at the time, DxO was the clear leader. Meanwhile, sensors have improved (D800, Df...) and I do not care much about the remaining differences between DxO, ACR and RawTherapee. Right now I am more concerned by proper rendering of reds, as I am shooting lots of small flowers and not seeing any details in their oversaturated petals. Here the problem seems to be 1) an intrinsic weakness of Bayer matrix sensors (foveons seem better); 2) a software default setting issue (the "Adobe standard" profile sucks).

[John Geerts]

Thanks for the comments.

Akira: Yes, it seems the colour-changes are more powerful in PN (that's a point of learning, I guess), but the skies here are  quite like this in July with lots of rain, and sunshine. I would rather say the original ones are too flat....

Pluton: Well both. But there is no real cropping done, PN does do different changes in the size of the image, within lens-corrections, which needs to be corrected to have a full frame again.

Frank: Original is RAW but opened in ACR with no changes applied, but it takes standard D700 settings and cameraprofile ACR 4.6.  For me the Raw Processing should give the same treatment to all the RAW-files in reducing Workflow.

Airy: Thanks,  yes food for thought. I think it just needs some time in setting - up the desired profiles and settings in PN, and / or  reconsider - ACR profile.  Running two different systems is not really an option, of course.   De-noising and a good Highlight Recovery are important items, but this example was just about handling of distortion actually.

[Frank Fremerey]

I can speak for PN because I use it regularly.

The exposure settings are what PN messes with automatically. But. All of these have their Zero point.

When PN opens a picture to absurd values lieke -2.5 exp plus 36 lighting and -anything highlights. I reset the values to 0 and get a better starting point. This does not explain the distortion  correction.

I do not know the right words because the computer is busy installing WinX/64.

[Airy]

I do not care about distortion correction in general, because no current software stores non-AI / AI / AIS lens profiles, as far as I know.
Correcting "regular" (barrel or pincushion) distortion by hand is very simple. Now I am used to my lenses and know by rote what values to input to get a result that pleases the eye (not necessarily the charts; it is seldom useful to be precise to the third decimal place).

The only auto distortion correction I am grateful for is, "moustache" distortion. Here, the only lens I have that badly needs it is the 20/2.8 AIS. There is no dedicated profile, but fortunately the AF version has about the same optical formula, and there is a profile available. I have not investigated the way to build a profile in ACR or whatever other environment. I know that Photoshop will allow any distortion, as complex as it may be, to be corrected "by hand", but that's not a profile in ACR.

[simsurace]

Some vertical lines on your processed shots look curved. There might be some uncorrected distortion going on.

[Frank Fremerey]

Airy: you can submit samples of any lens to Tom Niemann. This is why I bought PTlens

[pluton]

John, the distortion correction in these shots is, in my mind, inadequate...at least for "realistic" commercial or documentary photography.  It seems partially corrected in the second two, of course.  It could be that the positioning of the buildings in this shot places them at the worst possible locations vis-a-vis this lens' wave distortion at 17mm.
I have had a copy of the Nikon AFS 17-35 for 8 years, and I *feel* that I have never seen this much barrel distortion at 17mm.  The Adobe profile that comes with Lightroom seems to overcorrect it into positive(pincushion) distortion, but I can pull that back with the manual controls, and get very low, or not perceptable, distortion with straight-sided buildings and such, with this lens at 17mm.
Note: I have never played with the distortion correction in Photo Ninja, only Adobe.
 

[simsurace]

I frequently encounter this problem myself and decided to use this shot in order to find out how far I can go.

With a bit of work, one can perfectly correct this shot. I used PTGui with the Panorama Tools optimizer. Using the Line control points, I can align all windows of a given column instead of only the top and bottom ones, resulting in a more accurate distortion correction. This method does not rely on any lens profiles, and Panorama Tools are free, even though the PTGui interface makes it somewhat easier to set the control points. I had to fill in and stretch the corners a bit in Photoshop though.

[Andrea B.]

Simone's distortion corrections are the best.

This photo seems to have some crazy distortion which is very difficult to correct in any app I own. When I attempted correcting the distortion in Photo Ninja, I could get the left side looking straight, but then the right side would look bad. And vice versa. In Photo Ninja's auto distortion tool (placement of lines), the corrections were made on both the Pincushion/Barrel slider and the Mustache slider.

Well, here is my attempt anyway. I did brighten up the photo a bit. Perhaps too much for such a dark day? But it is just my interpretation.

[simsurace]

This photo seems to have some crazy distortion which is very difficult to correct in any app I own.

Indeed, PTGui finds distortion parameters which look nothing like simple barrel distortion (exaggerated about ten-fold):

[pluton]

Just for amusement, here's the 17-35, at 17, handheld on a target at about 1.5-2 meters; the distortion is complex.