Software Intervention ~ Takumar 28mm f/3.5 SMC vs Sigma 24mm f/3.5 DG DN

One of the basic assumptions about lenses for photography that I have accepted unchallenged is that new lenses would be "obviously better" than old lenses.  My thinking was that modern optical design tools improve optical products in meaningful/useful ways leading to technically "better" photographs.

Images loaded into image processing software undergo many adjustments _before_ a starting image is displayed.  Some of these adjustments make optical defect "corrections."  Which made me wonder about the aforementioned optical design tools.

Turning software automation off often reveals an interesting, unexpected truth.  Modern lenses may _require_ software intervention to look good.  In many ways old manual focus lenses look better.  Old lenses often exhibit less geometric distortion, less chromatic aberration, and display certain level of "sharpness", even at wide apertures, the modern lenses appear to need help with.  

This little "discovery" lit up my brain cells and got me to thinking.  What if I applied lens corrections to manual focus lenses?  Specifically, what if I corrected for spherical aberration and geometric distortion?  Not that old lenses need much geometric distortion correction, that is.  What effect might software intervention have on old lenses?

Post-Pandemic I found myself once again exploring manual focus optics.  Previously I'd made a decision to move to all modern AF but, it appears, an Old Itch still needed to be scratched.  I now have several beautiful Takumar, ever more Nikkors (including one particularly early SLR F-mount lens), Pentax-M, and third party Kiron.

Recently, I found myself in a favorite location and took two similar focal length lenses off to a car show. One lens was a new Sigma 24mm f/3.5 DG DN.  I really like this lens.  It's light, sharp, and the AF is blazing fast.  It has never let me down and is a jewel of an optic.

The other lens was an old c.1970's Asahi Takumar 28mm f/3.5 SMC (version 2).  It is a multi-coated 49mm filter thread later version of that focal length.  It, too, is very light, compact, very smooth focusing, and relatively quick and easy to use.  In fact, it, too, is a little jewel of a lens.

Setup ~ 

• Sony A7RII set on a tripod 
• 2 second self timer
• ISO100 
• Images shot at f/11

Image processing ~

Deployed four RawTherapee tools, starting from a "neutral" profile (thereby avoiding  automated software intervention) -

• Demosaic 
• Color management (Sony Camera "Standard" .dcp)
• Automated Chromatic Aberration correction ~ RAW -> chromatic aberration
• Capture Sharpen image sharpening ~ RAW -> Capture Sharpen (note: not USM)

Here is an example of what I see -

 

Comments ~

I once again confirm for myself that software intervention can be quite useful, regardless of the vintage of a lens.  In fact, old lenses tend to require less software intervention.

If I didn't already know which lens was which, there's no way I would be able to tell the difference between the old Takumar and the new Sigma.  Can you see any meaningful difference?  Maybe my eyes are getting old, but I can't.

While this is just one very small, minor example of what software intervention can do for images taken with old manual focus lens compared with modern AF .lcp enhanced optics, I see similar things with nearly all my old lenses. 

Software Intervention ~ a few musings

Back when dinosaurs roamed the earth I photographed using film and printed my own black and white images.  For a short time (four years) I printed for various labs around southern California.  My colleagues and I were involved in various projects, including making limited edition runs of various things for different galleries and producing prints for exhibitions and portfolios of famous photographers.

I felt I had a good working knowledge of the requirements that met customer expectations.  To meet those requirements, our equipment and chemical processes were carefully controlled.  Our enlargers were carefully aligned (Omega D-series were easy compared with Beseler).  The enlarging lenses we used were the best we could find (EL-Nikkor, Schneider Componon-S). The chemicals were mixed in a consistent way and temperatures were carefully monitored.  We used densitometers to verify everything was as expected.  Lamps were left on to ensure constant temperatures in condenser enlarger heads and we waited a moment after putting a negative in carrier into the enlarger so the negative would snap into flatness (if it matters and you don't understand what I just said, ask and I'll try and explain a little further).  Global image contrast was controlled by selecting paper types and/or filters.  Local corrections were made using dodging/burning techniques.

We did all these things to make certain the field of film grain would be absolutely sharp from extreme corner to extreme corner across the entire print while working to reveal the subject/scene in all its potential glory.  Making good prints was a mixture of alchemy, art, and craft.

Digital image processing comes with a few more tools than what were available to us "back in the day."  Using these tools helps us to go beyond what was done in the Old Dinosaur Film Days. 

Mulling such things over in my mind, I thought it might be interesting to comment a little on the topic.  Here's a list of things we can do in digital to "improve" a basic image beyond film era global contrast and local dodging/burning.  We can now correct/control:

• Optical distortion (pincushion, barrel)
• Chromatic aberration
• Optical vignetting (illumination fall-off around the edges)
• "Sharpness"
• "Local contrast" 

Here's a short list of the kinds of things that won't be corrected/controlled with neither film nor digital. 

• Optical coma effects
• Optical field curvature (field flatness)

There is (at least) one thing printing from film could do that is more than a little different in digital, and that is image enlarging.  Non-contact printed film sizes obviously require enlargement.  Digital is a different beast, though digital tools exist that enable enlargement possibilities.  I've found G'Mic's DCCI2X and CNN2X upscaling tools to be fairly convincing in taking a 42mpixel image and turning it into a 160+mpixel monster.

I've said a few things about how I feel I can make an old film-era lens look like a new digital-age one.  Using automated chromatic aberration correction and Capture Sharpen in RawTherapee takes my images a long ways toward matching the performance of my new digital AF Zeiss.  If an old lens suffers from lower contrast, a gentle Local Contrast and/or USM application finishes the job.

Taking the Insanity a little further, yesterday I downloaded a Leica Q2 .dng file and looked at it long and hard.  Given Leica's reputation I thought I might see some magic that's not commonly seen in other camera systems.  However, I wasn't "grokking" it, so I turned on/off the various software interventions... and... oh!... sure enough, image processing corrects/changes the outcome, even with hugely expensive Leica.

It's all rather exciting, actually.  Digital image processing software intervention can "democratize" image quality.  

Old lenses?  New lenses?  Apparent image "sharpness?"  On some level it's pretty much all the same.  Which is, I'm sure, partially the point some people try to make when saying things like "equipment doesn't matter."  Of course "equipment matters", but maybe not just in ways I usually think.

One Last Thought: There's something of a community of old lens user-commenters ("influencers?" - perish the thought) on various sites/platforms spouting/touting/sprouting the benefits of using old lenses.  They tend to say things like "old lenses have so much "character"..." and "modern lenses are so "clinically" sharp..." If what they want is "character" maybe they should try turning off software intervention on modern lenses and see what happens.  It might boggle/confuse/re-inform their world view.  Not that there's anything wrong with using old manual focus lenses, right?

Software Intervention ~ chromatic aberration, capture sharpen, local contrast, and USM corrections on a Super Cheap Kiron 28mm f/2

In the prior post I looked at chromatic aberration and capture sharpen corrections on a beautiful old Nikon Nikkor-S 50mm f/1.4 c.1972 lens.  I then took a small step to see what controlling local contrast and adding a very light USM might have and found the old Nikkor could match (under many circumstances) the performance of a Sony 55mm f/1.8 FE ZA that was managed by the common and now customary lens correction profile (.lcp) file.

For this post I want to consider similar software interventions on an incredibly inexpensive Kiron 28mm f/2 as well as a cheap, widely available Pentax-M 28mm f/2.8.  Nobody seems to like these lenses and prices on the open market tend to reflect this thought. 

Setup ~ 

• Sony A7 set on a tripod 
• +1EV (because of the strong whites and knowing whites saturate at EV+3.5) 
• 2 second self timer
• ISO100 focusing on the central flower
  
• Images made at f/2.8 and f/5.6
• Processed in RawTherapee 

Voila! encore a mundane scene of my Rescue Orchid - 

 

 

 Taking five processing steps and sharing the results in the following image -

• Demosaic and only color management as the starting point
• Add RAW -> chromatic aberration correction to the starting point
• Add Capture Sharpen to chromatic aberration, demosaic, and color managed image
• Add Rawtherapee -> Detail -> Local Contrast very light amplitude = 0.05
• Add Rawtherapee -> Detail -> Sharpening USM

Comments ~

Well, well, well... would you have a look at that, will 'ya?  Incredible, isn't it?  Or, if you've become somewhat jaded to the whole exercise, like me: Huh.

Identifying the tools .lcp files use to correct modern AF lenses and then applying that knowledge to isolating and using similar tools on old manual focus lenses has shown me several things.

First, in surprisingly many ways, modern lenses require software intervention to look as good as they do.  Second, designers of earlier lenses, using by today's standards rather rudimentary calculations, were able to achieve decent performance in terms of field flatness, field distortions, and chromatic aberration.

While there may be little to nothing software can do to correct for field flatness, coma at wide apertures in high speed lenses, or resolution fall-off toward the edges of a field (such is commonly seen in early wide angle lenses at the extreme corners), software intervention seems to work wonders on old lenses in just about ever other way.

From my perspective there is nothing to fear from using old lenses.  To prove this point I may post a few images I've recently made to see if people can tell which lens made which image.  I know, I've played this game before, so we'll see.

If a person can't make a decent image, it's likely not the fault of the lens, regardless of the age of the optic.

Chromatic Aberration ~ software intervention Nikon Nikkor-S 50mm f/1.4 vs Sony 55mm f/1.8 FE ZA

Now that I was on a roll looking at chromatic aberration and sharpness software tools I hauled out an old Nikon Nikkor 50mm f/1.4 c.1972 and a currently manufactured Sony 55mm f/1.8 FE ZA. 

What follows is actually the second pass that I made at this specific comparison.  I tried photographing the Rescue Orchid against strong backlight and found the Nikkor-S flares badly.  So I flipped the scene around and shot with the light coming from behind me.  Chromatic aberration was still present, so I was able to process the way I've learned works well.

Then I took two additional software tools and tried to match the output of the Nikkor-S to that of the Sony 55mm.  I wanted to confirm what I already suspected, which is to say, with a little help, old lenses image might be made to look every bit as good as new.

Setup ~ 

• Sony A6300 set on a tripod 
• +2EV (because of the strong whites and knowing whites saturate at EV+3.5) 
• 2 second self timer
• ISO100 focusing on the stick 
• Images made at f/5.6 and f/11
• Processed in RawTherapee 

Voila! encore a mundane scene of my Rescue Orchid - 

 

 

Three processing steps and sharing the results in the following image -

• Demosaic and only color management as the starting point
• Add RAW -> chromatic aberration correction to the starting point
• Add Capture Sharpen to chromatic aberration, demosaic, and color managed image

Using  

• Rawtherapee -> Detail -> Local Contrast 
• Rawtherapee -> Detail -> Sharpening 

 I took a quick look at matching the Nikkor-S to the Sony FE.

 

Comments ~

As in prior comparisons I see that the automated chromatic aberration correction does a fine job here.  There's not much CA in the Sony FE, so this lens is easy on the CA tool.  There is a bit more CA in the Nikkor, but this, too, cleans up quickly and nicely.

The Nikkor-S has less local contrast and is a touch softer than the Sony FE after CA correction and Capture Sharpen.  Adding two tools, then gently nudging the local contrast and USM sharpness I was able to show myself that at f/5.6 and f/11 the Nikkor-S and Sony FE have little useful/important/visible difference between them.

 

Chromatic Aberration ~ software intervention Nikon Nikkor-P 10.5cm f/2.5 c.1959/1960

Once I sorted out how to apply chromatic aberration software corrections to old manual focus lenses I set about to have a look at a few of my oldest lenses, beginning with a well used, mostly clean glass Nikon Nikkor-P 10.5cm f/2.5 9 aperture blade "tick-mark" transition c.1959/1960 lens.

To review, here's how I enabled automated chromatic aberration correction in RawTherapee.

 Rawtherapee -> RAW -> Chromatic Aberration Correction -> select Auto-correction

Pretty simple, right?

Setup ~ 

• Sony A6300 set on a tripod 
• +1EV (because of the strong white back-light) 
• 2 second self timer
• ISO100 focusing on the stick 
• Images made at f/2.5, f/4, f/5.6, and f/8
  
• Processed in RawTherapee 

Voila! encore the mundane scene of my Rescue Orchid - 

 

 

I took three processing steps and share the results in the following image.

• Demosaic and color management only as the starting point
• Add RAW -> chromatic aberration correction to the starting point
• Add Capture Sharpen to chromatic aberration, demosaic, and color managed image

Comments ~ 

There's not much to say.  It's obvious there is little CA in this lens.  It renders beautifully at all apertures.  Using chromatic aberration and "sharpness" software intervention on this optic feels like it brings this ancient lens right into line with modern products quite nicely.

Chromatic Aberration ~ software intervention Nikkor, Pentax-M, Takumar

Looking at software intervention for correcting chromatic aberration in a new Sigma 24mm f/3.5 i-Contemporary lens and learning just how important (reliant?) lens correction profiles can be to image processing, I turned my attention to three old manual focus lenses to see what might be done for them.  The challenge is, of course, that there are no lens correction profiles (.lcp) files for old lenses.

Fortunately, RawTherapee provides a surprisingly complete set of tools to work with.  Opening this software I see a tool that might be useful for correcting chromatic aberration.

Rawtherapee -> Transform -> Chromatic Aberration Correction 

There are two sliders to work with and I can look at contrasty edges of a subject and use the tool to remove CA color tints.  I can then save the settings as a recipe and recall it later as needed.

Looking around the software a bit further I stumbled upon an automated method.

 Rawtherapee -> RAW -> Chromatic Aberration Correction -> select Auto-correction

This tool automatically detects and removes CA shortly after demosaicing an image.  This tool appears to operate earlier in image processing than the Transform CA tool.  All I need to do is save a single recipe with the RAW CA correction enabled and I no longer have to store individual lens recipes.  This is the tool I used in the following comparison.

Setup ~ 

• Sony A6300 set on a tripod 
• +1EV (because of the strong white back-light) 
• 2 second self timer
• ISO100 focusing on the stick 
• Processed in RawTherapee 

I compared three old manual focus lenses ~

• Nikon Nikkor 24mm f/2.8 Ai
• Pentax-M 28mm f/2.8 
• Takumar 28mm f/3.5 SMC (second version)
  

Voila! a mundane scene of my Rescue Orchid - 

 

 

In RawTherapee and for each of the three lenses I took three processing steps and share the results in the following image.

• Demosaic and color management only as the starting point
• Add RAW -> chromatic aberration correction to the starting point
• Add Capture Sharpen to chromatic aberration, demosaic, and color managed image

 

Comments ~

 

RawTherapee's automated chromatic aberration correction tool works the treat.  It cleans up the CA tints and my impression is that an image becomes sharper in just this one simple step.  Then, with the addition of Capture Sharpen, images from old manual focus lenses can match their modern optical counterparts.

 

The old lenses I used here showed less CA than the Sigma 24mm I compared them against.  These lenses must be lighter on the tool to achieve similar results. Modern versus old?  At this level it doesn't seem to make much difference.  

 

My Nikon Nikkor 24mm f/2.8 Ai produces images that have felt "fat" to me.  I've often wondered about that.  Now I understand.  The "fatness" comes from un-corrected CA.  When the CA is cleaned up I doubt anyone could tell the difference between the old Nikkor and the new Sigma.  Images are "clinically" sharp after a little processing.

 

All hail software intervention.

Chromatic Aberrations ~ software intervention and a Sigma 24mm f/3.5 i-Contemporary

It was rather surprising to me that my Sigma 24mm f/3.5 i-Contemporary lens shows as much chromatic aberration as it does.

 

If I believed the Marketing Wizards I would've thought computer optimized optical design (such as what I was led to believe is used in current auto-focus lens design) would've been a vast improvement over earlier hand calculations (such as what was generally used to guide manual focus lens design).  Clearly this is not always the case.  There is something more going on.

 

Image processing software comes with lens correction profiles.  They correct for geometric distortion, vignetting, and chromatic aberration.  So I thought I'd have a look at how .lcp files modify the character of a currently manufactured lens.

 

I'm not sure if users can select individual corrections in RentWare products.  I imagine that people don't know nor care about the underlying automation.  In the case of RawTherapee, I _can_ select the corrections I want and I use this ability for this comparison.

 

Setup ~

• Sony A6300 set on a tripod
• +1EV (because of the strong white back-light)
• 2 second self timer, ISO100
• focusing on the stick
• Processed in RawTherapee 
  

Voila! a mundane scene of my Rescue Orchid - 

 

 

In RawTherapee I took three processing steps and share the results in the following image.

• Demosaic and color management only as the starting point
• Add chromatic aberration correction to the starting point
• Add Capture Sharpen to chromatic aberration, demosaic, and color managed image

 

 

Comments ~

 

The effects of software intervention on image processing should be rather obvious.  My Sigma 24mm f/3.5 i-Contemporary lens goes from showing a lot of chromatic aberration to becoming a sharper looking image at the first step of software chromatic aberration control. Then adding Capture Sharpen and the image takes another leap in perceived "sharpness."  After applying just these two software tools the lens behaves very well.

 

Which raises another question: Are there any tools a person can use when shooting with old lenses?  Can software correct chromatic aberrations in old lenses that do not have lens correction profiles?  I will try to answer this in the next blog post.

Chromatic Aberrations ~ new vs old lenses

I'm not sure how it came to me, but I had a question of how chromatic aberrations might effect the appearance of "sharpness" in an image and took five lenses of relatively different vintages to have a look.

Software automation can correct for various optical "defects", including field distortion, chromatic aberrations (CA), and "sharpness."  As a first step for my comparison I would need to turn all of that off to note what just the lens was doing.

You see, I had the sneaking suspicion that modern lenses were being tweaked into fabulous goodness by lens correction profile, or .lcp, files that are often hidden from a user during RAW image processing, but I wasn't sure.  I'd heard over the years many marketing claims as to new lens design technologies improving all manner of traditional optical defects, so maybe the .lcp files weren't doing as much as I thought?

The lenses on hand included:

• Nikon Nikkor 24mm f/2.8 Ai
• Pentax-M 28mm f/2.8
• Sigma 19mm f/2.8 EX DN E
• Sigma 24mm f/3.5 i-Contemporary
• Takumar 28mm f/3.5 SMC (second version)

Using a Sony A6300 set on a tripod, +1EV (because of the strong white back-light), 2 second self timer, ISO100, focusing on the stick, then processed in RawTherapee using only the demosaic and Camera Standard tone-curve enabled color management steps (to keep software processing to an absolute minimum).  

Voila! a mundane scene of my Rescue Orchid - 

 

There were a few surprises awaiting me.  Looking for chromatic aberration, here's what I found  -

Comments -

The Sigma 24mm f/3.5 shows the strongest chromatic aberrations of the five lenses I looked at whereas the old Pentax-M and older Takumar 28mm lenses show the best CA corrections.  Even the Nikkor 25mm appears to show less CA than the Sigma 24mm and is perhaps better at CA correction than the Sigma 19mm f/2.8 APS-C.

Looking at how CA effects the sense of "sharpness" I see that, yes, indeed, images look less "sharp" when CA is stronger.  Makes me wonder how much modern lenses rely on software intervention to make them look as good as they do?