Friday, February 16, 2024

True Focal Length ~ Sony 35mm f/2.8 ZA, Sigma 24mm f/3.5 DG DN, Tamron 20mm f/2.8 Di III

Casually reading something on the internet, I stumbled across a comment about how the field of view narrows when applying lens corrections to the Tamron 20mm f/2.8 Di III.  The writer claimed the lens went from 20mm uncorrected to 22mm corrected.

 

Bugatti ~ Retromobile ~ 2024

Image taken using at
Tamron 20mm f/2.8 Di III

 

I instantly thought that a gap from 24mm to 22mm might not be worth the effort to carry two lenses, the Tamron 20mm and a Sigma 24mm f/3.5 DG DN.

I've caught myself out in the past in reading something, not fully understanding the situation, selling a lens or camera, only to find out later that I was wrong and that the "problem" I was trying to solve was actually something completely different.

Case in point: There was a wonderfully sharp little Sigma 60mm f/2.8 DN EX E that suddenly developed a problem focusing.  I read on the internet that the AF mechanism was not up to snuff, so I sold the optic.  Only to find out a few weeks later that there was some grease stuck around the AF circuit on the shutter release of my Sony A6000.  I cleaned the A6000 and instantly everything was good again and the AF worked correctly.  All was not lost, however.  I picked up a brand new Sony 50mm f/1.8 SEL OSS that has proven to be great jewel-like little lens.  Still, I learned something from the experience.

Thinking that the Tamron corrected and Sigma lenses were too close together in focal length, my first impulse was to sell the 20mm Tamron and buy a Sigma 17mm f/4 DG DN or a Sigma 20mm f/2 DG DN.

But hold the horses.  Maybe I should check the actual focal lengths of these and then make a better informed decision.

Taking a ruler, I measured 1100mm from the tripod mounted camera to a spot on the floor where I stretched the ruler parallel to the film plane of the camera.  Then I took a photo using three lenses, the Tamron 20mm f/2.8 Di III, the Sigma 24mm f/3.5 DG DN, and the Sony 35mm f/2.8 ZA.  

I then noted the numbers on the ruler in the horizontal direction at the very edges of the scene uncorrected and then corrected, calculated the base of the isosceles triangle, calculated the angle at the film plane of the camera, then calculated the actual focal length of the lens at 1100mm distance from the subject.  Easy peasey.  Right.  Here's what I found.

Tamron 20mm f/2.8 Di III uncorrected - 18mm

Tamron 20mm f/2.8 Di III corrected - 21mm

Sigma 24mm f/3.5 DG DN uncorrected - 27mm

Sigma 24mm f/3.5 DG DN corrected - 27mm

Sony 35mm f/2.8 ZA uncorrected - 39mm

Sony 35mm f/2.8 ZA corrected - 39mm

Well, lookee there, will ya?  

The Sigma and Sony lenses have longer focal lengths than marked when focused on a subject 1100mm away.  I would have expected that with an old manual focus lenses were the entire lens group moved forward when focused on close subjects.  Extending the lens distance to the image plane increases focal length.  But since the Sony and Sigma are internally focusing lenses, I thought there might be a lot less of what cinematographers call "focus breathing."  That's where the size of a subject changes with changes in focus.

OK.  So I learned that the Sigma and Sony lenses are longer than expected.  

What about the Tamron?  It starts wider than marked on the front of the lens when uncorrected.  The lens suffers from a large amount of barrel distortion and I can see quite a shift around the edges when a correction is applied.  Once applied, the lens measures 21mm.  This is much closer to what is marked on the lens than either the Sigma or the Sony.

A gap from 21mm to 27mm between two lenses is actually quite enough for me.  I don't feel the need to find something different.  Buying and selling lenses is becoming a chore and is something of a thrash, so I'll stick with what I have for now.  I'm glad I checked.

Saturday, February 10, 2024

RawTherapee dcp and lcp files...

I need to remember something detailed.

 

Retromobile ~ 2024

 

In using RawTherapee to process RAW files I've found it has sophisticated color and lens management systems.

Interestingly, RawTherapee accepts industry standard dcp and lcp files.  dcp files are for color management and lcp are for lens corrections.

Rawtherapee comes with some dcp and lcp files on installation which appear to be updated from time to time.  I'm not sure who generates these files, but they seem to have done a good job for the lenses and camera models covered by the software distribution.

However, I found some of my cameras (Sony NEX, A6000, A5000, A7) and lenses (Sigma 24mm and a few others) are not supported by RawTherapee automation for the version I'm running (the latest), so I set off in search of good dcp and lcp files to fill in the gaps.

It turns out a certain RentWare implements these two file formats for their own color and lens management systems.  I wondered if I might be able to borrow them?  I run Linux, but have a computer that can boot into Windoze.  Here's what I do.

Boot into Windoze and...

  • In a browser search for "Adobe Camera Raw download" and locate the Adobe site (there are other sites that may offer downloads, but they are highly suspicious and I avoid them like the plague)
  • Download the latest "Camera Raw" plugin (I do not want LR or PS, just the RAW converter part)
  • Execute the "...exe" file to unpack the plugin
  • Descend the "c:\ProgramData\Adobe..." folder structure to locate "CameraProfiles" and "LensProfiles"
  • Copy the contents of these two folder structures into a Linux readable location/media

Boot into Linux and... 

  • Copy the "CameraProfiles" and "LensProfiles" directories and their contents somewhere under $HOME where I can easily find them
  • Open RawTherapee
  • Open an image and...
  • Under Color Management 
    • Select "custom"
    • Open the directory box
    • Locate the "camera profiles" directory
    • Descend the directory to...
    • Locate the camera model
  • Under "Profile Lens Correction" 
    • Select "LCP file"
    • Open the directory box
    • Locate the "lens profiles" directory
    • Descend the directory to...
    • Locate the right lens
  •  The base image is now configured using good dcp and lcp configurations.

Under "Lens corrections" there are selections for "Geometric distortion", "Vignetting", and "Chromatic aberration."  I turn off "Vignetting" because I've found that correction to be too strong for my taste.  But I do turn on "Geometric distortion" and "Chromatic aberration."

Under "Color management" there are selections for "Tone Curve", "Base Table", "Look Table", and "Baseline Exposure."  These are defined here.

On a practical level here's what I do.

  • Open an image in Rawtherapee
  • Let Rawtherapee select the demosaic algorithm (long topic for another time)
  • Set the lens profile
  • Select the color management dcp to be used (more on this in a moment)
  • Select "Tone Curve"
  • Select "Base Table" if selectable (this is not always implemented in some of the dcp files I've seen)
  • Select "Look Table" to get the dcp files color grading (which can be glorious, BTW)
  • Unselect "Baseline Exposure" since there is no jpg reference (read the definitions linked to above)

If I want to make further changes to "Curves", I go to the...

  • Eexposure panel
  • Find the curve function
  • Select "Luminance" 
  • _Then_ make adjustments to the curve
This keeps the colors from shifting.  Remembering, of course, that standard curves modify RGB curves at the same time the luminance curve is changed.  This modifies the color of the image, which I find I do not want to happen since  I like the "Look Table" results and do not want them to change.

From the list of practical things that I do I said I would comment further on selecting a <specific camera model> dcp.  The RentWare distribution is a little complex in how they've implemented their dcp directory structure.  Basically, it comes down to this.  Under "CameraProfiles" we have two ways of further descending the directory structure.

  • "...CameraProfiles/AdobeStandard/<specific camera model>.dcp"
  • "...CameraProfiles/Camera/<specific camera model>/<several dcp to choose from>

I'll start with the ".../Camera/<specific camera model>..." profiles.   From what I can tell these are the RentWares attempt to match specific image style selections offered by the manufacturer.  For instance, with the Sony A7 there are vivid, neutral, standard, landscape, and other dcp selections found in this directory.  If I want an image to look similar to the in-camera style selection, this is a good place to start.

Looking in AdobeStandard/... directory I see the RentWare has offered something a little different.  This is appears to be their own interpretation of what a "good image" would start with.  I find in the case of the Sony A7 that the ".../AdobeStandard/Sony ILCE-7 AdobeStandard.dcp"offers a more muted yellow, for instance, color starting point than the Camera/<specific camera model>/... "standard" dcp.  It pays to experiment and experience these various dcp options.

You notice that I've said nothing about using the RawTherapee "Processing Profiles."  This is because I've found the automated selections to be too strong for my quickly evolving image processing tastes.

OK.  There it is.  Lots and lots of detail.  But if I save a base processing configuration, the processing workflow can collapse to a single button push.  It's pretty sophisticated stuff, but I'm learning it's well worth the while to understand what's going on.

Thursday, February 08, 2024

Sony FE 55mm f/1.8 ZA, Nikon Nikkor-P 105mm f/2.5 ~ considering software magic

In a prior blog entry I commented on how amazing image making is in light of technologies, materials, and manufacturing.  I would like to continue along these lines and add something we never had in the old film/chemistry days.

Software.

To me this is as amazing as the physical implementation of sensors and cameras and lenses.  In software we can implement just about any standard or process or idea we want.

There are algorithms for many aspects of image processing that directly effect how we view a photograph.  Exposure, contrast, color profile, color depth, color management, image processing spaces (up to 32bit floating point!), tone mapping, film emulations, lens corrections, image alignment, perspective corrections, image stitching, high dynamic range, and... and... and... there are so many things implemented in software in various ways that it can be mind boggling.

It can be easy to be confused or frustrated and overwhelmed by all the options.  What to they mean?  What do they do?  How do I take advantage of this selection compared to that one over there?

Considering one example from Rawtherapee, there are 19(!!!) options for demosaicing a RAW file on import into this software.  Yes, certain selections can be made for users if you let the software automation retain control, but users are also free to select something else.  Each and every option has a reason to exist.  Understanding them can take time, certainly, but what a rich set of possibilities there are in just this one single step.

Looking at another example, there are at least 15(!!!) sharpen operators in the Open Source G'Mic.  15 different direct ways of sharpening an image.  Each one is implemented based on a formula/algorithm that attempts to enhance some condition or another.  If we add high by-pass filtering in a layer, there are even more image sharpening options, and I've not counted the sharpening operators in the Gimp (where I run G'Mic).

Taking this yet another step further, mixing various operators can refine an image processing sequence.  In the case of the subject matter I currently enjoy photographing (automobiles and motorcycles) and taking into consideration digital enlargements (ie: going from 6000x4000pixels up to, say, 12,000x8000pixels) I've found the following sequence to give generally outstanding results.

  • Open TIF RawTherapee output in the Gimp
  • G'Mic DCCI2x upsize the image
  • G'Mic Inverse Diffusion sharpen using just enough iterations to avoid introducing visible artifacts
  • G'Mic Octave sharpen the image

In this case I use _two_ image sharpeners.  One avoids pixelation (Inverse Diffusion) and the other gently sharpens subject edges.

Sometimes Octave is too strong.  In which case I use one of two sharpen operations. One is applying a high by-pass filter layer blended using "Soft Light" over the base DCCI2x'd Inverse Diffusion sharpened image.  This is brilliant for keeping pixelation to a minimum.

If that approach doesn't work and I need something stronger I use...

  • Create a new G'Mic Gradient Norm layer from the base image
  • Duplicate the base DCCI2x'd Inverse Diffusion sharpened image as another layer
  • Copy the Gradient Norm layer into the mask of the duplicated base image
  • Apply a hard G'Mic sharpener, such as Richardson Lucy
  • Duplicate the sharpened layer to add further sharpness to the over all image

You have to see any and all of this to believe it.  Which is another point I'd like to make.  When using various sharpeners, I've found it really helps to watch the action of a sharpener at full pixel-peeping resolution.

Returning to the very start of the image processing pipeline, there is a sharpening operator in RawTherapee that I've found to be very useful.  It's called "Capture Sharpen."  It's best applied to low-ISO images where noise is at a minimum and is used to "correct" the effects of AA filters or image smearing due to slightly missed focus or slight camera movement.

The effect of Capture Sharpen can be dramatic and yields an image that I feel avoids any effects of being "over sharpened." 

In the case of my comparison of a Sony FE 55mm f/1.8 with a c.1973 Nikon Nikkor-P (Xenar) 105mm f/2.5 pre-Ai, I thought I'd confirm these effects of Capture Sharpen.  Here's what I found.

 

Sony 55mm f/1.8, Nikkor 105mm f/2.5 ~ Capture Sharpen

 

A comment about Capture Sharpen: This operator is used in certain RentWare, too, and sets the basis for their digital enlargement process in something I think they call "Smart Sharpen" and "Super Resolution".  We can do the very same thing in Open Source Software.  I described three ways to perform digital enlargements earlier in this post.

From this little comparison I see that Capture Sharpen applied to the Nikkor-P image brings the crispness/sharpness/contrast well into line with the native un-sharpened Sony FE 55mm.  I can take an old lens and turn it into a digital-era Sharpness Monster.  Applying Capture Sharpen to the Sony FE images takes images to another amazing level of sharpness.  

How cool is all this?  And this is just one little step in the image processing pipeline.  

Software.  I love it!  Oh the possibilities...

Wednesday, February 07, 2024

Sony FE 55mm f/1.8 ZA, Nikon Nikkor-P 105mm f/2.5 ~ considering the magic

Sometimes when I'm comparing this and that and poking here and looking there at photography, equipment, and processing I stand back and marvel at the magnitude of science and engineering that go into making any of this possible.  I've written this before but I feel the need to underscore this point.  It's "magical" on some level.  Consider what knowledge, which materials, what kind of manufacturing it takes to make each step in this process.

  • Light hits a light sensitive diode
  • Diode emits a tiny electrical signal
  • Hardware applies gain to amplify the signal (ISO sets the amplification level)
  • Amplified signal enters an analog to digital converter (ADC)
  • Digitized information exits the ADC (may receive additional signal processing as in the cases of extended ISO settings)
  • Digital information can now be written into a RAW file
  • This RAW file can then be loaded into software on a computer/tablette/mobile-phone
  • De-Mosaic function performed on a RAW file
    • Individual "pixels" of RGGB (red, 2x green, blue) information is processed using surrounding "pixel" data to calculate color and luminosity 
    • Such that there is now a different kind of information (de-mosaic calculated) for each "pixel"
    • Changing each and every original red, 2x green and blue locations into full color with brightness information
  • De-Moscaic'd images still look ghastly as the colors are still "off", so...
  • Software "massages" a de-mosaic'd image using ICC, DCP, or other "camera profile" filters
  • The image is now displayed to the user
  • User can now begin performing additional modifications using the tools the software provides

After all this we can now consider various elements of the image making system.  In my case I've looked at lenses.  To me it's absolutely amazing that after passing through the process described above that I can now wonder if one lens is "sharper" than another.  I'm completely and utterly reliant on all that knowledge, understanding, engineering, manufacturing, materials, and software just to reach this potentially interesting point.

Whew!  It sometimes takes my breath away.

With this in mind I wanted to see how the modern Sony FE 55mm f/1.8 ZA compared with an old c.1973 Nikon Nikkor-P (Xenotar) 105mm f/2.5 pre-Ai lens.  Here's what I found.

 

Sony 55mm f/1.8, Nikkor 105mm f/2.5 ~ Comparison

 

It's easy for me to see the newer Sony FE is sharper than the Nikkor-P.  Though I have to admit the old Nikkor-P still looks pretty good.  The differences between the two lenses come down to the Sony FE being the contrastier optic when I use the exact same image processing pipeline applied to both lenses.  

This leaves room for choosing a different process pipeline. Software is flexible in ways I never experienced in the old film/chemistry days when the Nikkor-P was first introduced.  I will consider the role of software sharpening in image processing in the next entry on this topic.
 

Thursday, February 01, 2024

Then and Now ~ part two

Thinking a bit further about the size of the gear we used to have compared to what we have today, I remembered there were some wonderfully compact interchangeable lens cameras.  One in particular attracted my attention, though I never could afford it, even on the used market, and that was the Minolta CL and CLE.

So I thought I'd have a quick look at the lenses and camera body for that small Minolta and compare them to Sony APS-C.  The thing that attracts me, even now, to Sony APS-C is that this series remains smaller than Fuji, Canon, and Nikon devices of equivalent sensor size.  

I noticed years ago just how big Fuji, in particular, is.  Check around and you'll see that Fuji with the small sensor is as big as a Sony full frame.  Makes me wonder what's taking up all that space in those cameras.  Canon makes me wonder the same thing.

Continuing with compactness, it's easy to see just how light and small the Minoltas were.  Then I remembered another camera I never owned and that is the Contax G1.  It's rather modern and the lenses are AF.  It could be interesting to compare AF film lenses with digital.  For this I choose a couple Sigma EX DN and a Sony OSS.

Compact Cameras Weight Length Height Depth
Contax G1 450grams 133mm 77mm 35mm
Minolta CLE 380grams 124mm 77.5mm 32mm
Sony NEX-5T 276grams 111mm 59mm 39mm
Sony A6300 404grams 120mm 67mm 49mm





Lenses



Contax



Biogon 21mm f/2.8 200grams 35mm 59mm
Planar 35mm f/2 160grams 29mm 56mm
Planar 45mm f/2 190grams 39mm 56mm
Zeiss Sonnar 90mm f/2.8 290grams 54mm 59mm





Minolta



28mm f/2.8 135grams 35mm 51mm
40mm f/2 105grams 24.5mm 51mm
90mm f/4 250grams 60mm 51mm





Sony APS-C



Sigma 19mm f/2.8 EX DN 140grams 46mm 61mm
Sigma 30mm f/2.8 EX DN 135grams 39mm 61mm
Sony 50mm f/1.8 OSS 202grams 62mm 62mm

Just as with full frame, the big changes from then to now are the increased capabilities of digital over film.  And in terms of compactness I'm happy to see certain of today's APS-C devices aren't much different than back when we used film.  I wish Sony still offered the NEX series APS-C.  The versions without EVS are truly compact and light.

Onward.

 

Retromobile ~ 2024