Can lens or camera have a 'clinical' rendering?

Started 1 month ago | Discussions thread
57even Forum Pro • Posts: 14,738
Re: Can lens or camera have a 'clinical' rendering?

Trebor1 wrote:

57even wrote:

Trebor1 wrote:

Erik Kaffehr wrote:

Hi,

Recently I stumbled over poster claiming that a vendor's lenses provided 'clinical rendering'.

I have often seen that statement and I have always been skeptical. Just to say, discussing clinical rendition, I would always consider white balance. But, I think that lenses are worth some discussion.

Objectively speaking, the objective for a lens would probably to render a point as a point. Neither add or subtract. That is what a perfect lens would do.

I have never seen anyone complaining about a Zeiss Otus being to clinical. But, I guess that they are as well corrected as practical.

But, I wouldn't think that a less well corrected lens would render 'more clinical'.

Obviously, bokeh can be different, but how different are the images below?

The first one is classic medium format Hasselblad lens, Macro Planar 120/4 at f/11 and the second one is a Sony 90/2.8G at f/8. Very different lenses, but I would think they are sort of on pair.

Of course, I do realise that some lenses may have 'adorable faults'. But, in the end, I would expect that good lenses deliver good images.

Best regards

Erik

Zeiss used to believe that this was the case (under certain circumstances) and catered for this market with their Softar range of filters. These were also produced under licence by B+W and Heliopan.

The Softar 1, in combination with the 180mm CFi f/4.0 Sonnar, is shown with MTF plots (pages 14 - 17), in the 2009 Zeiss paper How to Read MTF Curves part II by H H Nasse.

https://lenspire.zeiss.com/photo/app/uploads/2018/04/CLN_MTF_Kurven_2_en.pdf

The Softar 1 had the mildest effect, of the three filters in the range and is mentioned in a 2010 Philip Bloom article as being (perhaps not unsurprisingly) effective at eliminating moire and aliasing on (Canon) DSLRs. The explanation by Zeiss, from their advertising literature is interesting and you can see why these were once popular.

https://philipbloom.net/blog/an-interesting-solution-to-eliminating-moire-and-aliasing-from-your-dslr/

Pictorialism, however defined, has a long history and even today (very expensive) Cinema Primes and zooms are being designed to exhibit particular characteristics for artistic effect: Flaring, oval bokeh, as produced by anamorphic lenses but now apparently a desired effect in conventional lenses, according to one manufacturer.

There is an apparent divide between those seeking optical perfection, or as close as possible, from maximum aperture onward and those who want some lens character/desired flaws. Many appear to want it both ways - 'Dreamy' bokeh and/or spherical aberration, perhaps some of that 'glow' around highlights sought by afficionados of pre-aspheric Leica Summiluxes, vintage sonnars etc. when wide open to perhaps one stop down but 'perfection' thereafter? Recently, I have seen multiple threads at DPR, where posters state that they only use their fast lenses at apertures less than f/2.0 so they are probably not chasing ultimate resolution, in the part of the frame that is in focus?

I found a selection of photographs on Flikr, that were taken with Softars or B+W Pro. The degree of highlight diffusion in landscapes, is sometimes excessive IMHO, when using the stronger Softar II filter but others may disagree?

OTOH, the Zeiss Softar I does seem offer something interesting:

The obligatory cat photo - what is the MTF of this cat?

https://www.flickr.com/photos/thedigitalstory/42442775794/

There are some great flower photographs, at the link below, plus black and white shots on 4x5 and portraits that all appear to retain sufficient resolution, at least at these viewing sizes, unlike some other diffusion filters.

https://www.flickr.com/search/?text=zeiss%20softar

The patents have probably expired by now so there could well be a market for a new 'Softar', if it delivers the smooth but defined look that people seek from their expensive ultra-fast lenses, used at near full aperture?

The photos below show the effects well:

https://www.flickr.com/photos/fwwidall/7434679150/

https://www.flickr.com/photos/fwwidall/10076320764/ (probably has his 3rd and 4th images swapped, regarding labelling?)

Get the same effect by turning the clarity slider down in Lightroom.

That may not work. It certainly didn't work several years ago, when the same suggestion was being made in online forums. Making clarity a negative value, plus by the use of layers, brushes and/or intelligent masking might get you there, or close.

Well, getting close would be fine if I wanted to duplicate the effect exactly, which I may not want to. Fact is, I can get any effect I want pretty much with some masking and blur.

In portraiture, given that there are eye detect and smile detect algorithms in camera and content-aware fill in post then this could perhaps be fully automated? For use in landscape photography probably harder.

https://photofocus.com/photography/shooting-photography/tech-corner-soft-effect-lenses-and-filters-still-relevant-after-all-these-years/

If you look at the photo of the Softar II, in the above link, then you can see that the lenslets are not all the same size and have a semi-random distribution. Unless you fully characterise this and the milder Softar I, stronger Softar III then how can you faithfully replicate the effect?

Personally I wouldn't want to. But I may want something similar.

[here are plug-ins that claim to match the rendering but I haven't seen any scientific evaluation, thus far, the effect will be scene dependent and a single visual image comparison, word of mouth, is not fully representative.

The Tiffen diffusion filters have a different physical implementation than the Softars and the effect is different.

https://www.premiumbeat.com/blog/world-diffusion-filters/

In cinematography this type of filter is still widely used, which suggests that doing it in post is not yet an easy automated process and if manual input is required then this could get tedious and expensive.

This subject was discussed 8 months ago, in this forum:

https://www.dpreview.com/forums/post/6352459

Well doing it in post on a movie is rather different.

It's easier to soften detail in post than add it.

About the only argument I could see in favour of spherical aberration is smoother bokeh behind the subject, which is hard to fake, but you also get a less sharp subject.

I could see more of an argument for lenses with character in the film days, when there was little we could do in post.

These filters can modify bokeh. The Zeiss Softar 1 effect is relatively subtle and it is not really soft-focus, so it would be interesting to compare with the original Nikon Soft 1 filter (ion-diffusion, gradient index) which was withdrawn because of the RoHS directive in 2006 and its replacement, that uses a different operating principle

The trouble with filters is that the net effect rather depends on the lens you use it with.

-- hide signature --

"A designer knows he has achieved perfection not when there is nothing left to add, but when there is nothing left to take away." Antoine de Saint-Exupery

 57even's gear list:57even's gear list
Fujifilm X-Pro2
Post (hide subjects) Posted by
Keyboard shortcuts:
FForum PPrevious NNext WNext unread UUpvote SSubscribe RReply QQuote BBookmark MMy threads
Color scheme? Blue / Yellow