Chasing "softness" in small formats ~ troisieme part ~ Showing My Homework

I recently wrote a blog entry where I talked about lenses that seem to fill the gap between Full Blown Soft Focus for 35mm format and "normal" sharp/clinical lenses.  What I shared were mainly Lens Porn, er, sorry, Portraits illustrations because I find these cheap lenses beautiful to take photos of.  I did not share my homework, even though I wrote at length about differences between the optics.  This blog entry corrects this omission.

Keep in mind that the trick to these first element focusing lenses is that the greatest optical imperfection effect is generally found in subjects closer/closest to the camera.  Of course it depends on lens design, but this is the tendency I've seen thus far.

Often these first element focusing lenses for 35mm format that are easily adaptable to digital derive from or are implementations of the early Cooke Triplet.  The design is two positive elements on each end with one negative element in the middle.  This is pretty simple, easy to manufacture, and can be very low cost.  Lenses on the used market can be nearly Give Away cheap.  I recently picked up a digital adaptable first element focusing Cooke triplet design lens for less than 9Euro.

Homework ~

 

Steinheil Auto-Cassaron
50mm f/2.8 at f/4
processed using a film sim
that I like just to see how
the low contrast of the lens
might play out against a
contrast-inducing LUT

 

Comments ~ (borrowed in large part from an earlier blog post)

In reverse order, from bottom to top... 

Ricoh 55mm f/2.2 Riconar - Optical imperfections galore - the kinds of imperfections vary depending on subject distance.  While I don't show this here, it's easily seems comparing close to distance focused subject at f/2.2 and f/4.  It delivers rather decent contrast, actually.  

Interesting highlight "glow" effects.  Controllable by aperture and subject distance, which is potentially useful.  One would have to map out distance/aperture to know which settings to use.  

f/11 can make a pretty sharp image of distant subjects.  Regarding the effects I'm going after, between f/2.2 and f/5.6 the optical imperfections at all distances can play well toward generating a decent "Pictorialist" style enlarged to "normal" viewing size/distance.

ISOC Iscotar 50mm f/2.8 - Optical imperfects somewhere between the Ricoh and Steinheil.  Good contrast.  

Interesting highlight "glow" effects.  Controllable, as with the Ricoh, by aperture and subject distance.  f/8 and f/11 can make a decently sharp image at greater subject distances if desired.  

Between f/2.8 and f/5.6 the optical imperfections can play well toward generating a decent "Pictorialist" style enlarged to "normal" viewing size/distance.  I think of the ISCO as a slightly more rational German Riconar.

Steinheil Auto-Cassaron 50mm f/2.8 - Subtle optical imperfections, spherical aberration at all apertures and all subject distances. Low contrast.    Using the "haze" removal control during processing cleans up a scene, but why use it if I'm looking for "soft focus?"  Have I mentioned this is a low contrast lens?  There must be an echo in here.  Either that or it's strongly evident from looking at the results.

To me this lens is like using a Heliar large format lens from Voigtlander.  Back in the day I owned two of these, a 15cm and 21cm f/4.5, both in Compound shutter.  These lenses had similar underlying detail to what I see with the 50mm.  Missing the rendering of those old lenses I'm happy to discover the Steinheil.  The more I stare at Steinheil images the more I wonder if this isn't a basic trait to how Steinheil designed their optics?  As with the Heliars I find this a really interesting way to make an image.  There seems to me to be a lot of potential for processing unique small format digital images.  I have at least one more Steinheil optic coming to try to confirm/deny this line of thinking.

As the lens is stopped down the underlying detail begins to extend from the center toward the edges of the field.  The effect is common to how triplets behave and I saw this most particularly in a Meyer Domiplan 50mm I once had.  It was sharper from wide open than any Zeiss Tessar I ever saw (and I had more than a few of these over the years).  The ISCO behaves this way too.  That is, wide open the center of the field can be surprisingly sharp and the mid to edge of the field showing subtle/not-so-subtle optical defects of various kinds.  These clean up as the aperture is stopped down and the sharpness spreads out.

With my Auto-Cassaron it's as if the lens designers kept/allowed the spherical aberration to gain consistency in other areas of optical design.  Resolution, field distortion, chromatic aberration and coma are better controlled than in the Ricoh and ISCO.  While more subtle than many large format soft focus lenses from Back in the Day, the Steinheil for small format might make for a decent "Pictorialist" style lens where image viewing sizes can vary depending on the electronic display system. 

 

Images made with a Steinheil
Auto-Cassaron 50mm f/2.8 at f/2.8 or f/4 

Chasing "softness" in small formats ~ deuxieme part

I'm chasing pixies again.  Or still.  Or some little "corner case" like that.

The topic of soft focus photography on small formats is interesting to me.  I'd like to reliably replicate the "look" and "feel" of Photo Secessionist/Pictorialist era prints in digital.  This means looking for a way to get "just the right amount of softness" in an image using digital sensors.

 

Sony A7RII
Ricoh Riconar 55mm f/2.2 @f/2.8 

All of the made for purpose adaptable to digital 35mm SLR Soft Focus lenses I've looked give results that are too "strong."  That is, the level of veiling "softness" (spherical aberration) does not stand up to looking at an image at "normal" viewing distances.

I'm convinced that soft focus lenses from the late 1800's/early 1900's "work" because they were made by contact printing and lens designers could allow "just the right amount" of optical imperfections to be pleasing to viewers.  4x5 inch.  5x7 inch.  Whole plate.  8x10 inch.  Any attempt to enlarge a negative made with a special purpose soft focus lens meets with disaster.  So contact printing it was.  Period.

Adaptable 35mm/Full Frame Soft Focus lenses to my way of thinking fail where large format film lenses could succeed.  Fujifilm 85mm f/4 SF.  Pentax 85mm f/2.2 and f/2.8.  Minolta 85mm f/4 Varisoft.  Canon 135mm SF.  These lenses are nearly impossible to work with if trying to come close to what the original Pictorialists were able to create. 

Though, it must be said, I have it on good PhD thesis-level on this topic authority that Pictorialists in the 1970's and 1980's were able to successfully use the Minolta Varisoft with the softness ring set between "0" and "1" (ie: very very gentle application of spherical aberration).  The photographer had to know what they were doing.

Looking at pixies, odd pieces of fluff, and photographic lens histories I found that something potentially useful.   Wollensak's Velostigmat Series II tessar formula is a normally sharp lens that came with soft focus capability.  The soft focus effect is achieved by unscrewing the first element moving it away from the two optical groups placed on either side of the aperture.  By extending the distance between the first and second element a pleasing soft focus effect is achievable.

This got me to thinking. Perhaps pixies do exist, in spite of the lack of empirical evidence.

What if I took a cheap Soviet tessar and extended the distance between the first and second elements?  Would the soft focus effect seen in large format film photography be achievable in smaller formats?  Perhaps even in some controllable or easily understood manner?

Being occupied with the Rigors of Retirement, time passed and I casually let the subject rest.

One day I was thumbing through a Pentax on-line forum and discovered that the Ricoh 55mm f/2.2 Riconar is a first element focuser.  It's not a tessar, nor a Cooke triplet.  It's a four element air space design of some kind, of which I've never seen the likes of before.  So I thought I'd give it a try as they are nicely inexpensive.  What a strange and interesting lens the Ricoh turned out to be. 

I then learned the Germans made a LOT of Cooke triplet first element focusing lenses... because they were cheap... because they are easy to manufacture... and because they were (barely?) sufficient to the task.  The first element has, in concept, enough power to control but not contain many of the optical imperfections down stream.  

Enter a Steinheil Auto-Cassaron 50mm f/2.8 Cooke triplet.  Such a beautiful lens.  Oh my. My many thanks to Bonzo Din or turning me onto this one.

Enter an ISCO Iscotar 50mm f/2.8 Cooke triplet. Such a gorgeous lens.  Oh boy.

Avoiding the Meyer Domiplan 50mm f/2.8.  Been there.  Done that.  Soap bubble ain't my thang.  

Making a habit of avoiding cheap lenses all my life colored my sense of what is possible.  Cheap lenses are bad lenses, right?  The trick question is in which ways is a lens "bad?"  Mike Johnson from Darkroom Magazine/The Online Photographer said something to the effect that all lenses have their gifts.  And so it is with first element focusing triplets and quadruplets.  I've found they can provide a controllable sense of softness, particularly for subjects nearer to the camera than not.   There live the pixies I've been searching for.

 

Status: Here's what I've experienced using these lenses thus far.

Ricoh 55mm f/2.2 Riconar - Optical imperfections galore - the kinds of imperfections vary depending on subject distance. Decent contrast.  Interesting highlight "glow" effects.  Controllable by aperture and subject distance.  f/11 can make a fairly sharp image of distant subjects.  Between f/2.2 and f/5.6 the optical imperfections at all distances can play well toward generating a decent "Pictorialist" style enlarged to "normal" viewing size/distance.

 

Steinheil Auto-Cassaron 50mm f/2.8 - Subtle optical imperfections, spherical aberration at all apertures and all subject distances. Low contrast.    Using the "haze" removal control during processing cleans up a scene, but why use it if I'm looking for "soft focus?"  

This lens is like using a Heliar large format lens from Voigtlander.  Back in the day I owned two of these, a 15cm and 21cm f/4.5, both in Compound shutter.  These lenses had similar underlying detail to what I see with the 50mm.  Missing the rendering of those old lenses I'm happy to discover the Steinheil.

As the Auto-Cassaron is stopped down the underlying detail begins to extend toward the edges of the field.  The effect is really interesting to see.  It's as if the lens designers kept the spherical aberration to gain consistency in other areas of optical design, such as resolution, field distortion, chromatic aberration and coma.  The Steinheil obviously makes for a decent "Pictorialist" style lens for small formats, though of a different characteristics compared to the Ricoh. 

 

ISOC Iscotar 50mm f/2.8 - Optical imperfects somewhere between the Ricoh and Steinheil.  Good contrast.  Interesting highlight "glow" effects.  Controllable, as with the Ricoh, by aperture and subject distance.  f/8 and f/11 can make a decently sharp image at greater subject distances if needed.  Between f/2.8 and f/4.5 the optical imperfections can play well toward generating a decent "Pictorialist" style enlarged to "normal" viewing size/distance.  I think of the ISCO as a slightly more rational German Riconar.

 

 

Are there more German first element focusing triplets in my future?  Stay tuned.  

Digital Zone System ~ Sony A6300 in-camera jpg generation

I continue to pick at a question about where Sony places tone values for its in-camera jpg generated images.  My question this time is how manufacturer default settings record luminance.

In the Sony BIONZ and BIONZ X universe I've noted that Creative Style Black and White jpgs at its out of the box manufacturer default setting clips the shadow areas hard compared to sensor RAW.  The Creative Style BW range of tone expression is compressed into a narrow range.  It's a very steep curve into darkness.

It's this tonal compression that sent me in search of ways of expanding the in-camera generated Black and White jpg tonal range and found using DRO at specific values useful.  DRO2 or DRO3 raise the shadow values in a controlled way, though the bottom of the tonal range is still -5EV/Zone 0.  By comparison, RAW files using very early NEX reach pure black at -7EV and later sensors from A6000 onward reach pure black at -8EV or -9EV.

That's on the stills Creative Style side of the Sony systems I have.  

On the video side in the BIONZ X systems I have there are additional imaging functions under the heading of Picture Profile.  I originally thought these were an extension of Creative Style.  Only recently did it occur to me to confirm/deny this.

Et voila! here I am with ever more words and illustrations of what I've found. 

For this post I share the Sister of the Mother of All Digital Zone System Charts for a Sony A6300 that I have.  Illustrating the effects of parameter changes (contrast, knee, etc.) are not covered here.  That would be the Mother of All Digital Zone System Charts had I taken the time to do all that.  However, I feel a pretty clear understanding of Sony is doing can be developed.  Additional details and refinements are left to the reader to pursue.

As I leap off into the deep end of things I should note that the process used in all my Digital Zone System work is easily implemented on all non-Sony systems.  All this is simply noting how cameras place luminance values given the various imaging functions a manufacturer offers.  In fact, I've spot checked this entire approach against Panasonic Lumix S9 and Fuji GFX100RF systems and found the basis of understanding holds true across at least three manufacturers cameras.

Setup ~

• RAW processed 1EV separation values as reference
• In-camera generated 1EV separation jpgs for...
• Creative Style Black and White with settings at manufacturer default
• Picture Profiles selecting...
• Gamma
• Color Mode Black and White
• All other settings at manufacturer default for each Profile

Results ~

 

Comments ~

Looking at the result in columns from left to right...

RAW 0EV -  This is my tone value reference.  Though not shown here, -8EV is pure black.  1EV tone separation is achieved from -8EV to +2EV.  From +2EV to +4EV is the "shoulder" of the curve (ala film).  Film stopped being 1EV separated at -2EV, which explains why digital imaging systems offer a TON more shadow detail than film when 0EV (the standard) is set at Zone 5.

Creative Style Black and White - Pure black is -4EV Zone 1.  Pure white is +4EV.  Zones 4 through 1 show a steep drop in tone compared to the RAW reference.  Visually, images made using this setting are contrasty with what I would call "inky" blacks.

Sills - The first Picture Profile "style" I come to shows what appears to me to be a very similar result to Creative Style Black and White.

Movie - This Picture Profile is the first illustration I come to of Sony moving tonal values compared with Creative Style Black and White and Picture Profile Stills.  Notably, the shadow values are "opening up" with Zone 0 being properly pure black.  This setting behaves similarly to what Creative Style with DRO enabled can do.  If there was nothing further, at its default setting I feel Picture Profile Movie could be a satisfactory in-camera jpg imager. 

Cine 1 - Now we're cooking with gas.  Tonal values are expressed through EV -6 Zone -1, though the highlights are a little "hot" with how close +3EV is to pure white.  To correct for the highlights being "hot", under-exposing by -0.3EV moves the entire curve appropriately (to more closely match the Zone System definition).  For the way I was trained to "see" BW images from all the way back into the Film Days I could stop right here.  

A very useful BW Sony recipe:

• Picture Profile Gamma Cine 1
• Color Mode Black and White
• -0.3EV
• All other Picture Profile Cine 1 settings factory defaults

Cine 2 - Wackiness ensues.  Highlights are strongly compressed.  Perhaps if someone is "color grading" video, having tone value in the highlights could be interesting.  For BW stills work?  I'm not so sure.  YMWV.

Cine 3 - Hah!  Another potentially useful manufacture default Picture Profile.  This is very similar to Cine 1, except for the exposure as the tones are expressed.  There's no need to -0.3EV with this setting.  It seems usable straight out of the box: 

• Picture Profile Gamma Cine 3
• Color Mode Black and White
• All other Picture Profile Cine 1 settings factory defaults

Cine 4 - This Picture Profile setting sets -5EV as pure black, where Cine 1 and Cine 3 set pure black at -6EV.  Said another way, Cine 4 appears to be using less of the native dynamic range of the sensor that Cine 1 and 3.  Depending on the environment a person shoots in, this could be sufficient as an out of the box Color Mode BW setting.

SLog 2 - All Holy Hell breaks loose and the FULL dynamic range of the sensor is expressed. +4EV down through -8EV have tone.  Images using SLog 2 "look" flat.  Referencing back to how my eyes were trained on film prints, SLog 2 "feels un-natural."  It's a strange sensation.

Of course SLog 2 is meant for video work and the output of this setting is expected to be color graded after the fact.  That's where the Cool Kids get to apply their oh so special color LUTs to achieve certain "cinematic effects."

For BW stills work I envision this setting being useful in extremely high contrast situations, then off-loading a jpg from the camera to a mobile device for processing where an 8bit color space is all that's available.  Otherwise we'd just shot RAW and process it in a 16 or 32 bit color space, right? 

SLog 3 - Ack!  Here we go again with crushed highlights.  I see zero value in this setting for BW stills work. 

Chasing "softness" in small formats...

I was rather excited when stumbling upon the Ricoh Riconar 55mm f/2.2.  It's a soft focus lens in cheap kit lens drag and I thought I'd skinned a Fat Calf.  

I've been looking for a lens with gentle optical incorrections.  The made to purpose "soft focus" lenses for Minolta, Pentax, Canon, Leica, etc, etc, etc, are all way too soft for me.  The effect doesn't whisper it's presence, it hits me over the head with a hammer and announces "I am a Soft Fekk'n Focus Lens, fer Gawds! sake!!"  The Ricoh, on the other hand, appears to offer a controllable level of softness, much like early large format portrait and pictorialist lenses.

 

 

 

Having struck gold once, the experience motivated me to look for other old first optical element/group focusing lenses for 35mm format cameras.

Bonzo Din shared the fact the a German made Steinheil comes with a front element focusing lens.  It's called the Steinheil 50mm f/2.8 Auto-Cassaron.  Oh boy! quick as a bunny I found and purchased one.  The aperture and focusing ring were sticky, so I disassembled the simple lens as far as I could, applied denatured ETOH to everything trying to unbind the bound bits (kinky! you say).  Cleaned and lubed the focusing threads.  The Steinheil became usable.

 

 

 

What I found is The lens is an implementation of the classic Cooke triplet.  This Steinheil is indeed first element focusing.  The second and third elements are fixed, one on each side of the aperture.  The Steinheil's aperture control mechanism is as simple as can be.  Lens coatings are applied only to the outer surfaces front and back.  Surfaces on the inside of the lens are uncoated.   The lens is obviously designed for lowest cost manufacturing, just like the Ricoh 55mm f/2.2 Riconar.  Cheap, cheap, cheap.

Once my aging mind was coming unstuck, I remembered that Saint Ansel used a Turner Reich 12-1/4" triple convertible as well as a Cooke 12-1/4" series XV triple convertible early on.  I've inspected prints made from negatives exposed using those lenses and the results are quite good. 

 

 

 

In use I've found the softness generated to by the Steinheil to be "gentler" than the Ricoh's.  The German lens is very well behaved across the field.  The extreme edges can be a little soft until stopped way down, but nothing compared with the Ricoh.  The field is as flat as can be.

My tail was wagging like one on a medium sized dog.

Trying to better understand what I was looking at, I found I've forgotten more than a fair bit of triple element three group history.  Wikipedia to the rescue.  Cooke designed the earliest examples of the triplet. See - https://en.wikipedia.org/wiki/Cooke_triplet

Now I was really happy.  I felt I'd stumbled upon something.  Something secret.  Something delicious.  Something...

 

Except.  Except.  Except.  I didn't.  All too often late to the party, I find everyone who's anyone already knows about three element three group Cooke design lenses for small formats.   

It turns out the Germans, French, English, and Americans were cranking out Cooke triplets by the boats load! for smaller format 620, 120, and 35mm cameras.  The Germans continued this well and deep into the 35mm SLR times (1950's through 1980's).

Ricoh's 55mm that I was first smitten with is actually a four element four group lens.  So I'll not count it in the list that follows.  Simply remember it's cheap, widely available, and can be fun for use in soft focus work.

From Germany, I see the following Pentax M42 thread mount Cooke lenses - 

• E. Ludwig Meritar 50mm f/2.9 - East Germany
• ISCO Göttingen Iscotar 50 mm f/2.8 - Edixa M42 spec
• ISCO Göttingen Iscovitar 50 mm f/2.8 - Pracktica/Pentax M42 spec
• Meyer-Optik Görlitz Domiplan 50mm f/2.8 <- Soap Bubble Bokeh!!!  Ick.  Ack.
• Steinheil 50mm f/2.8 Auto-Cassaron

Browsing old catalogs shows these lenses were available under different model namings and different mounts.  Some lenses outside the M42 list above are more easily adaptable than others and I'm sure anyone with a keen interest in such things can quickly sort out what to do and which lenses merit attention.

I may post more on the topic after an ISCO arrives. 

Command Line Interface ~ Linux

I never knew it was "punk" to use a Command Line Interface, but it appears to be so. 

Once the idea struck, it became clear that, yes, indeed, I am increasingly anti-GAFAM (Google Amazon Facebook Apple Microsoft).  OK, so this blog is hosted on one of the GAFAM systems, but this can rectified at a time of my choosing.  Cell phone?  Yes, that too can change.  The tablet can change.  The Big Computer was for a very short period of time running a techno-tyrannical operating system, but most of the time for the past 30 years has been liberated.  It's a matter of effort.

Once I understood a little better the history of punk I could re-frame, re-context the contents of the prior paragraph.  Punk started as a youth movement that responded to arrogant "elite" class Thatcher-ism in the UK.  The US version of punk was something a little different.  It dealt with right wing politics and capitalism.  In both cases punk was a reaction to Bad Things perpetrated by "elite" powers far removed from the experiences of everyday people.  

GAFAM power is even greater than what the original punks responded to starting in the 1970's.  Techno-tyranny is extra-political and trans-national.  It determines what exists and what does not, what is remembered and what is not, what is acceptable and what is not, and it imposes a value system easily consumed by the masses.  Significant portions of the system is "free."  Whatsapp, Facebook, Blogger (the site I write this on), Gmail, Yahoo mail, Twitter (yes, I know it's newer name), Amazon, FaceTime, iTunes, etc, etc, etc on the surface and at first blush cost nothing.

The "costs" are carefully hidden from users.  Many (most?) "free services" on the internet are synonymous with loss of privacy, intrusive data collection, buying/selling of data, leading to a state of shockingly efficient, nearly seamless techo-tyranny means these services are anything but free.  Ads tailored just for our eyes are the least of it. We so quickly accept this as "the way things are" that we become numb to this truth.  Of course none of this is for our liberation, rather for our compliance and sheep-like acceptance.

I'm reminded of the early days of what later evolved into the internet.  We dug for information and knowledge using Gopher.  We spoke rather freely with each other via (unscanned for advertising opportunities) email (hosted on small systems), held community conversations on (largely troll-free) discussion forums and bulletin boards (both commonly hosted on small systems), and read news on something called Usenet.

It felt more like we were moving into a future of our shared creation, rather than a narrowly offered present imposed on us.  Freedom and liberty vs corporate tended bubbles of narrowed for our "protection" tailored for our unique, personalized, and therefore oh so special experience.  Such sadness to see things so incredibly controlled these days.

My effort to limit/restrict the influence of GAFAM now includes a review of tools used in photography.  Cameras and lenses are owned outright (this is the easy part, rather like a current day holdover of an earlier, simpler time).  My image processing tools are never rented and come from the Open Source Community (which, BTW, often implements industry standards _better_ than RentWare).  To speed the image processing pipeline up even further I sometimes use the Command Line Interface to invoke tools that do specific jobs quickly and efficiently.   I must, by definition, be punk.  Huh.  Never knew.  Doesn't change a thing, actually.

Here is my evolving kept for memory reasons list of commands, expanding to include the above motivations for their being in my life.

------------------ original post ------------------- 

Notes to self:  A few useful image processing commands for running in Linux.  All these are much faster to run from the CLI than using an app that's having to manage graphics at the same time. - 

convert *.jpg -average <averaged-filename>.jpg – averaging command

convert *.jpg -evaluate-sequence median <output file-name>.jpg  - a different averaging command

mogrify -format jpg *.png - change file format from png to jpg

mogrify -resize 1920 *.jpg – resizing command

mogrify -bordercolor black -border 10x10 *.jpg – adding a thin black edge to images

mogrify -bordercolor white -border 400x400 *.jpg – adding a white border to images

convert <filename>.<file-extension> -colorspace gray <output filename>.<file-extension> – command to convert a single image to black and white

for i in *.jpg; do convert "$i" -colorspace Gray  "BW_$i"; done – Bash script to convert a bunch of files into black and white

exiftool -a -u -s -G1 <file_name> - to read EXIF image file data

gmic -input <filename.file-extension> scale_dcci2x , cut 0,255 round output <theOutputFileName>.tif - command to perform a DCCI2x upsize

LibreOffice : Impress
menu Insert | Media | Photo Album :: Slide Layout 

 

Black and White digital filters in image processing...

I'm working up to a critique that I'll post at some point in the (hopefully) not too distant future.  To get there from here I want to cover Black and White digital filters in image processing.

The topic arose in my mind when musing over a digital recipe/filter that would accurately/correctly match the spectral response curve of early silver nitrate light sensitive materials, including but certainly not limited to wet-plate collodion.

I found a digital filter that I'm very happy with and for the sake of brevity I'll dispense with the steps I took to the final form.  If anyone wants the complete details, ask.

Here's the starting image.  Top and bottom are grayscale step wedges that I developed for the Digital Zone System I've worked on.  The center of the image is a simple color chip chart.   Examples were processed using RawTherapee.

Starting point -

 

Simple color desaturation - 

Relative RGB channel mix -

Luminance human perception modeling - 

Relative RGB channel mix "Ortho" filter -

Silver Nitrate relative RGB channel mix Red=0 Green=10 Blue=90 filter - 

Comments -

Short answer:  

Digital filters for Black and White color conversions seem to do what they're supposed to.

Long answer:  

Simple désaturation sucks. Colors don't translate to the tonality my eyes would expect to see.  Yet this is EXACTLY what old Black and White film does.  Sure, the ends of the color spectrum might be clipped differently on each end.  The meat of the curve behaves just like this simple de-sat.

Relative RGB channel mix is a minor improvement over simple désaturation. This is to be expected since all channels are set to 33 percent.

Luminance human perception modeling gives an accurate translation of colors into Black and White for the way I "see" tonality and luminance.   This is an outstanding foundation from which to build tonal separation in digital Black and White photography.  Further, in-camera Sony, Fuji GFX, and Panasonic Lumix S (the only system I've looked at) all appear to conform to luminance human perception modeling Black and White jpg generation.

Relative RGB channel mix with RawTherapee's "Ortho" filter seems to look very much like modern orthochromatique film response.  If I want early silver nitrate light sensitive material response, this is most definitely not what I'd look for.  Close-ish.  No cigar.

Silver Nitrate relative RGB channel set specifically to Red=0 Green=10 Blue=90 appears to hit the target.  The tonal response curve closely matches that of old silver nitrate light sensitive materials.  Goal!

A little more:

I could spend far too long looking/comparing/evaluating various combinations of channel mixture and digital filters and this and that.  Should I ever find myself in such a state I processed a number of images.  The collection of the Madness is found here, and scroll right.

In practice, I find the luminance formula in RawTherapee to be excellent for general Black and White conversions.  To explore the early pre-panchromatic Black and White photography "look" my little "Silver Nitrate" formula gives me pleasing results.

Processing an Image

Just for fun I thought I'd post a quick comment on how I processed an image I took at Rodin's Atelier in Meudon.

Image -

... and here's what I did to it: 

• Let the in-camera meter do its best
• Ricoh 55mm wide open at f/2.2  
• *click* the shutter 
• Opened RAW image in RawTherapee
• Applied my 0EV Digital Zone System curve
• Adjusted the ends of the curves for pure white and pure black
• Opened the image in the Gimp
• Back in RawTherapee, processed the sky for tonality and contrast
• Opened the sky processed image in the Gimp as a layer over the first image
• Gimp selected the sky of the first image
• Added a black mask to the second layered image
• Filled the selection area in the mask with pure white
• Adjusted the mask "sharpness" to Gauss soften with a 10 pixel radius
• Flattened the image and saved

Done.  That's it.  That's all it took to get these the way I wanted.

The trick, of course, was protecting the whites/highlights, then applying a correct Zone 3 thru 7 1EV step curve, and stripping in the sky.