Film vs Digital – the truth

There is an ongoing debate at Rangefinders Forum that has been running for years titled Film vs Digital. There are countless “experts” (I am not one of them) claiming one way or the other as to which is better, film or digital. With the ever increasing sensor megapixel counts of the recent years, the debate of which has a higher resolution (digital?) would appear to have been put to bed. Appearances aren’t always everything.

I ran across an interesting thread on Photrio, the old APUG page. Instead of pontificators and pseudo experts, a real expert whom does scientific testing of films for major film manufactures chimed in. His name is Henning Serger. I found his posts original, credible and delightfully insightful. So I am doing something different, I am cutting and pasting his comments regarding film resolution (in 35mm no doubt!) to share with you here, to expand the audience as well as preserve his comments for future reference without having to sift through the entire thread. Understand, this is not me, this is Henning Serger’s comments, to be found at https://www.photrio.com/forum/threads/rollei-rpx-25-grain-and-resolution.115244/

Dear Karl-Gustaf,

thanks for mentioning my test report about the Adox CHS 100 II and Rollei RPX in the edition III.2014 of “PhotoKlassik”.
And also thanks to all the others here for their positive comments about PhotoKlassik.

I’ve just discovered this thread incidentally. Due to lots of work my presence here is quite rare.

To answer the questions of the OP concerning resolution and grain of the RPX 25:
Well, as I am running a little test lab for lenses, film, developers, sensors I can give you test results of this film compared to all other films in this group of very fine grained, high resolution films.
Here we go:

First, some information about my test method: I am using several test patterns, and a resolution chart with an object contrast of 1:4 (two stops). That is an object contrast you can find in lots of details in your daily, normal photography.
With higher object contrast details you will get higher resolution values, and with lower object contrast you will get lower values.

Test camera is my F6, with MLU on, MC-30 cable release, 1/250s shutter speed, focus bracketing, Berlebach 3032 tripod.
Test lens: Zeiss Makro-Planar 2/50 ZF at f 5,6.
So the following resolution values give you the system resolution of the film and this lens in combination. With better lenses, you get higher values. And with worse lenses, you get lower values.
The first, lower value represent clearly separated linepairs per millimetre, the second, higher number is the resolution limit at which a contrast difference is still visible.
My standard test developer for the resolution tests is Spur HRX, because it gives very good sharpness combined with very fine grain, resulting in very good resolution.

Rollei Pan 25: 80 – 90 Lp/mm

Rollei RPX 25: 120 – 130 Lp/mm

Rollei Retro 80S; ISO 25/15°: 135 – 145 Lp/mm

Adox CHS 50: 90 – 100 Lp/mm

Ilford PAN F+: 110 – 130 Lp/mm

Adox CHS 100: 80 – 90 Lp/mm

Adox CHS 100 II : 100 – 110 Lp/mm

Adox Silvermax: 105 – 115 Lp/mm

Fomapan 100 : 75 – 90 Lp/mm

Ilford FP4+: 75 – 90 Lp/mm

Ilford Delta 100: 130 – 140 Lp/mm

Kodak Plus-X: 80 – 100 Lp/mm

Kodak T-Max 100: 135 – 150 Lp/mm

Fuji Acros 100: 115 – 130 Lp/mm

If you want even significantly higher resolution and better sharpness, grain, then go for:

Agfa Copex Rapid; ISO 25/15° – 32/16°; Spur Modular UR New developer: 165 – 180 Lp/mm

Adox CMS 20 II, Spur Modular UR New or Adotech II developer: 240 – 260 Lp/mm.

Grain:
Finest grain of all by a huge margin has Adox CMS 20 II.
Then Retro 80S follows.
A little bit behind the 80S on a similar level Rollei RPX 25, TMX, Acros, Pan F+, Agfa Copex Rapid are following. The differences between these 5 films are very small and not really field relevant. And the ranking can be a little bit different when different developers are used.
Then the discontinued Pan 25 and Delta 100 follow.
After that Adox Silvermax, which is the finest grained of the traditional ISO 100 films with classic cubic crystals.

And just for comparison:
Under identical test conditions, same test chart, same lens and aperture, the
Nikon D800 delivers 80 – 85 Lp/mm,
and the D800E delivers 90 – 95 Lp/mm.

And 
Provia 100F: 120 – 135 Lp/mm
Velvia 100: 125 – 140 Lp/mm
Ektar 100: 90 – 105 Lp/mm (Provia and Velvia also have a bit finer grain than Ektar).

I hope these test results are helpful for you.

Have a nice weekend,
best regards,
Henning

Further reading….

I think we have to be precise here: Sharpness (edge sharpness / contour sharpness) ist not a problem with modern digital cameras. With a bit PP you can get brutally sharp results. Personally I don’t like it, because it looks very unnatural. Our eyes work different, create a different sharpness character.
And film in its sharpness characteristics is much more similar to our eyes, it looks more natural.

What I have quantified in my tests is resolution, at a precisely defined object contrast (1:4). And at this object contrast lots of films have a significantly higher resolving power than a D800 and D800E.
At higher object contrasts the advantage of film is even higher. Zeiss for example did quite a lot of resolution tests with object contrasts in the 1:32 to 1:64 range.
They got:
160 Lp/mm with Velvia 50
170 Lp/mm with Velvia 100
180 Lp/mm with TMX
160 Lp/mm with Acros 100
200 Lp/mm with Agfa APX 25
250 Lp/mm with Agfa Ortho 25
400 Lp/mm with Spur Orthopan UR (Agfa HDP microfilm; almost identical to Adox CMS 20 II).

Digital sensors don’t benefit in resolution from such higher object contrast, because their resolution is limited by the Nyquist freqency. It’s impossible to get higher resolution as the Nyquist frequency.
But the D800 and 800E have a very good resolution at low object contrasts (in the 1:1,3 to 1:2 range). In this range they perform better than most films.

Therefore a lot depends on
– which detail is accurately in focus (only in the focus plane you have the full resolution)
– what object contrast have the detail(s) you have focussed on.
Indeed the resolution levels differ quite a lot from detail to detail just in one single shot, one subject.

Best regards,
Henning

Yes.
Concerning image detail, resolution and grain, scanning is the worst thing you can do with film.
You get much better results concerning these parameters with
– optical printing, especially with the excellent APO enlarging lenses
– slide projection with excellent projection lenses.

We’ve done lots of tests about this topic “imaging chain” during the last years.
We’ve tested the best amatuer scanners like the Nikon Coolscan 5000, and the best drumscanners like the Haselblad / Imacon X5 and the real drum scanner ICG 370 HS.
Result:
Even the best drum scanners cannot resolve all the details on the film. You have a significant loss.

But the good news is:
With optical enlarging with APO enlarging lenses and excellent projection lenses like the Leica Super-Colorplan, Zeiss P-Sonnar, Kindermann 2,4/90 MC-B (Docter-Optics), Rollei AV-Apogon 2,8/120, Schneider AV-Xenotar 2,8/150; the Braun Ultralit, Rollei AV-Apogon 90 etc.
you only have a minimal, not significant loss and you can resolve almost all details on film.
So you can transfer these details on the film
– onto paper
– onto the projection screen.

One example:
The 130 Lp/mm of Delta 100 in our test result in a 120 Lp/mm figure on paper, enlarged with the APO-Rodagon 2,8/50 N.
With the Coolscan 5000 it is only about half of that (55-60) at an object contrast of 1:4.

Similar with Provia 100F in projection: 120 Lp/mm on the projection screen, but only about 60 Lp/mm with the Coolscan 5000.

Another example: Adox CMS 20 II:
240 – 260 Lp/mm on the film in our test with the Makro-Planar at object contrast of 1:4.
Scanned with the drum scanners we’ve got only 130-140 Lp/mm of it.
But with classic optical enlargement with the APO-Rodagon more than 200 Lp/mm (!),
and in projection with the Super-Colorplan and the Kindermann MC-B 230 Lp/mm on the scren (!!).
230 Lp/mm: You would need a 183 MP sensor in your camera, and a digital projector with also 183 MP to get that digitally.
We will never see that in digital.
Slide projection delivers an absolutely unsurpassed performance in detail rendition with big enlargements.
It is the gold standard.

Our classic enlargement techniques, optical enlargement and projection, are still the best ways to completely exploit, completely use the full potential of film.

Best regards,
Henning

There was more exchanges during the thread and I invite you to fully read the full thread linked above. However, what I have copied and pasted from Henning’s posts are far more impressive than anything I have read to date about REAL film resolution. What we are seeing is that even traditional B&W films and color films have FAR more resolving power than anything digital is putting out. This is from the tiny 35mm format, no telling what medium format, 4×5 and 8×10 film potential is for resolution. We also see that your high resolving films, especially Adox cmsII, far out resolves anything ANY digital sensor at any size is currently or in the near future can come close to. This is astonishing! For so long, the experts have said “maybe” 6mp resolution from film, other saying the finest grain film “maybe” 20mp. Nope, try 200+megapixels.

However, this does not come easily. In fact, to get to almost these resolution numbers, one has to optically print or project the film. The problem is digitizing film. The problem has always been there. Now with the best scanners, drum or otherwise, one CAN get to higher resolutions that may match or even exceed 35mm digital, but just barely. So in essence, film in constricted to best results in the full analog process – printing or projecting. For those interested in film resolution with scanning involved, there was a popular post here, https://www.onlandscape.co.uk/2014/12/36-megapixels-vs-6×7-velvia/

Film vs Digital – the truth…Time for me to get my darkroom running again and do some printing.