Originally Posted by Film-Niko
Film-Niko: Well sorry but your methodology must be poor, because that is simply incorrect. Portra 400 has better resolving power than Provia 400X. The MTF curve has a better response. Portra 400 also has whole orders of magnitude more dynamic range. Provia 400X will clip both highlights and shadows at the same time before pushing, pushing it to 1600.. it's just absolutely nasty in that respect.
Also have a look at the characteristic curves, look where the Provia 400X "shoulders" (or toe I guess) the highlights before completely flattening out (not to mention clipping the highlights let alone being flat).
Portra 400 remains a steep straight line for 6 stops past this point where Provia 400X begins to flatten out, with no sign of flattening out into a shoulder yet. Not to mention the highlights aren't going to be clipped.
I haven't shot chromes for a while, but I used to push and pull up to 3/4 of a stop. When doing commercial work in the old days, I'd shoot 3 sheets of EPP, run a test. From the test, I'd do my "Balancing" based on the test by pushing or pulling the remaining sheets. Some of the work was critical color work and processing that wasn't "Normal" caused color shifts. I'd avoid push or pull processing because of that and the special handling of the film would cost extra. From my experience, it seems that Fuji RDP or RTP pushed and pulled well.
My methodology isn't poor at all, because I use the test methods used and published by Zeiss.
Originally Posted by Athiril
The only difference is that I am using less contrast for my tests (because in real photography situations there is less contrast compared to the five stops contrast Zeiss is using), and my tripod is not a heavy Sachtler one which is used by Zeiss.
I've got 105 linepairs per millimeter with Provia 400X.
And 75 lp/mm with new Portra 400.
Did you have tested the resolution of both films in real photography, using your lens?
Which values have you got?
Or are you only looking at MTF curves?
All photographers who have did detailed tests by themselves know that MTF data has a very limited relevance to real life shots.
Originally Posted by Athiril
MTF tests are done under conditions quite different to our real shooting conditions.
If you want to know how much resolution a film and a lens can deliver, take photographs under the same conditions and compare these photographs.
I've done that. And Provia 400X surpasses Portra 400 in resolution and sharpness.
Fuji Pro 400H is a bit better in this respect compared to Portra 400, too, but not as good as 400X.
Some reasons were already mentioned. One was to manage contrast and to adopt to high contrast scenes.
Originally Posted by Ektagraphic
There is another method to adopt to higher contrast scenes, with which you can increase the dynamic range by one stop.
This method is used for decades by photographers working with the zone system.
Mostly used with BW film, but it also works well with color film.
It is based on the physical fact that when you double the light intensity for zone I, it doesn't effect the highlight zones in a visible way.
You measure your exposure, and the value for light intensity for zone I is 1.
Then the value for highlight zone VIII is 128 (128x or seven stops more light intensity).
If you double the light intensity for zone I, the value for light intensity will be 1 + 1 = 2.
Then the light intensity for highlight zone VIII will be 128 + 1 = 129.
The difference between 128 and 129 is less than 1%, and that is not visible in the picture.
Therefore you can increase shadow detail by giving extra exposure to the shadows without affecting the highlights.
A trick successfully used for decades by old school photographers knowing the zone system.
It works best with using a tripod and multiple exposure setting: Measure the right exposure, then stop down for the first exposure increasing the shadow detail.
Stopping down will be 4 stops (for zone I) with BW film in general, and because of the steeper characteristic curve of color slide film in the 3 1/3, 3 1/2 or 3 2/3 stops range with slide film. It depends on the film type how much you have to stop down for best results. With more contrastier films with less dynamic range like Velvia 50 stop down less, with films with higher dynamic range like Sensia 100 or Astia 100F stop down a bit more for best results.
Make the first shot with this stopped down value.
Then you make on the same negative / positive a second shot with normal (in case of slide film highlight oriented) exposure.
Your shadow detail will be increased by the first extra exposure, but it will not visible affect the highlights, which are determined by the second, normal exposure.
If the contrast of the scene is one stop more than the dynamic range of your film, you gain one stop more and can adopt dynamic range of the film to scene contrast.
If the scene contrast is two stops more, you loose only one zone in the shadows with this technique instead of two zones.
In this last case with two stops more scene contrast you can combine this technique with pulling one stop to gain the needed two stops.
I always get excellent results with this zone system technique.
Another recommended option for managing high contrast scenes, which is often overlooked by photographers, is modern fill-in flash technology. With fill-in flash you can also work with zone system technique. It is possible to set the shadows exactly in the zones you want them by reducing the power of the flash.
My F 80 / F90X / F6 / SB-800 combinations are doing little wonders in this respect. Precise light adjustment, combining both highlight and shadow detail in high contrast scenes. You don't see it in the picture that fill-in flash was used. Very natural results.
Originally Posted by Athiril
we've tested the current 400 ISO colour films of the market.
Test method: real life photographs with an object contrast of only 1:4 (two stops).
Lenses: AI-S Nikkor 1,8/50 and Zeiss ZF 2/50. Both at f5,6.
With Nikon F6, MLU, Berlebach 3032 tripod, 1/250s, focus bracketing.
Provia 400X: 105 - 115 linepairs per millimeter
Portra 400 new: 80 - 100 lp/mm
Fuji Pro 400H: 90 - 105 lp/mm
(first value represents clear separated lines, second value the limit where still a difference in contrast can be seen).
Provia 400X clearly surpasses Portra 400 in resolution (and Pro 400H). And with contour sharpness as well.
Not surprising, because in most cases slides films deliver better detail rendition compared to colour negative films of the same speed. Especially at low and medium object contrasts.
ISO 100 films (same test method):
Ektar 100: 90 - 105 lp/mm
E100G, Elitechrome 100, Sensia 100, Provia 100F, Astia 100F: 120 - 135 lp/mm
Velvia 100 / 100F: 125 - 140 lp/mm
By the way, pushing Provia 400X: I have often done it and never seen a colour cast. Colours have always been accurate.
You say how you tested it, but not how it was measured! You obviously cannot measure reversal and negative films by the same method! You can't measure prints either for similar reasons.
So, how did you measure it?
Originally Posted by Photo Engineer
we've used the industry standard method, which is used for decades (e.g. by all lens manufacturers like Zeiss, Leica, Schneider etc.) for analysing system resolution (film + lens).
The resolution charts were analysed under a microscope with 40x and 100x enlargement. Under a microscope you really see what the film has captured. It's the most precise method. Usable for negative film and positve film, BW and colour.
We've discussed our methods and results with Dr. H. Nasse, chief optic designer of Zeiss, and got confirmation.
Furthermore we analysed the whole imaging chain: Optical enlarging with APO-lenses, scanning with drum scanners (Imacon X5 and ICG 370) and Nikon Coolscan 5000 scanner, and projection with different high quality projecting lenses.
Very short summary of the results:
Highest system resolution in the whole imaging chain is achieved with optical enlarging and projection. Loss in resolution compared to the results under the microscope in only 5 - 10%.
Same with slide projection. Getting 120 lp/mm on the projection screen with e.g Velvia 100 is possible with high quality projection lens
[the current digital beamers (the most expensive 7000 - 10000€ class) with their extremely low resolution of max. 2 MP deliver 15 - 20 lp/mm on the screen.
Slide projection is the field where quality advantages of film compared to digital technique is really huge].
With drum scanners the reduction in resolution is bigger, 20 - 40% dependent on the film.
Far behind is 4000 ppi scanning with the Coolscan 5000 (effective 3600 ppi). Worst resolution values of all, significant loss in detail. 70 lp/mm is the limit, more is not possible, even with high resolution films or higher object contrast.
Scanners are limited by the Nyquist frequency. An optical imaging chain (lenses + film) is not, here the limiting factors are diffraction and object contrast.
As resolution is dependant on object contrast, the differences are bigger with higher object contrast (more than our 1:4), and they decrease with less object contrast.
Well, the method you used is excellent and just fine for comparing 2 optical trains with 1 film, or one train with 2 or more slide films with identical contrasts, but in this case you are comparing a slide film with a contrast of about 1.8 and a negative film with a contrast of about 0.6, so how do you normalize the results?
And, even though you get a difference, how do you explain the data presented by Kodak and others which says that the Ektar is sharper than the Ektachrome?
I have personally done lab tests (minus optics) which give negative films the lead when the contrast differences are ironed out. Also, and in any case, given identical optics the degree of degradation should be comparable between the films. Thus the Kodak data should stand on its own merit.
I recently popped a contrast-increase mask on a 6x7 Ektar neg and enlarged it to 20x24 for testing purposes. That's nothing I plan to do routinely because the overall results are still quite inferior to the
what I get with larger sheet film. But it was a real eye-opener in terms of grain, and the now apparent
fact that the limiting factor with Ektar is going to be lens resolution and plane of focus, not grain magnification or film resolution per se. And yes, all other things being equal, it is distinctly sharper than any Ektachrome film at a comparable level of print contrast.
all BW films were developed to the same contrast.
Originally Posted by Photo Engineer
Colour films: All slide films were developed in standard E6 process, professional lab (Kodak Q-Lab).
All CN films in standard C-41, the same professional lab.
The reasons for this workflow:
1. Normalizing of the overall, global contrast would make no sense, because we want results relevant for real, daily photography, real pictures, and not for pure lab comparisons.
In daily photography no photographer would increase the contrast of his color negs from 0.6 to to 1.8, and no photographer would decrease the contrast of his slide film down to 0.6. The films are not designed for this 'abuse'.
Negative film works best with the lower contrast, and slide film with its higher contrast, that's why it is used in that way. Including all pros and cons.
If you now change the contrast dramatically you would loose the strengts of films in both cases, and you would gain nothing.
And our interest was to see the results the films deliver when they are used in the way all photographers use it.
When you print optically you have the differences between neg and positive film, and when you scan as well (of course also when you project it, but that makes seldom sense with negatives; nevertheless we've done it with high resolution films with extremely interesting results, but that is another story).
That is what we are doing in our daily photography, and are confronted with the (different) results.
To analyse these results was our interest.
In general you have the higher global contrast with slide film, and of course resolution benefits from that. That is an essential characteristic of slide film.
But you have to pay for it with less dynamic range because of the higher contrast.
And the other way round with negative film. Less resolution, but higher d.r..
There is no free lunch.
2. Global contrast is not the major factor for resolution of higher spatial frequencies. Here microcontrast is important, the contrast of the extremely fine black and white lines. And the difference in microcontrast between slide film and negative film at these high spatial frequencies is much much lower than the global contrast. You even see this clearly with the eye under the microscope, and you get that confirmed with densitometric tests (what we did).
I guess you refer to the MTF chart, and not to Kodaks marketing statement....;).
Originally Posted by Photo Engineer
Our experience with doing such tests for more than 20 years now is that MTF data has very little relevance for real life shooting under normal photography conditions. MTF data is generated under completely different conditions, e.g. with much higher object contrast.
Our contrast is lower, because most details in scenes have low(er) object contrast. And our interest is daily photography.
And with lower object contrast slide film performs better.
And there are differences in MTF data from manufacturer to manufacturer (different test conditions).
For example we have made the experience that the MTF data of Fuji is quite conservative. It is relatively easy to reach their resolution values.
A friend of mine, using the same method, but only with littler higher contrast (1:8) and a different lens has even achieved slightly higher resolution with Pro 160C than Fuji data gives for 1:1000 contrast.
These differences between MTF data and test results with real shots (higher as expected resolution in real shooting conditions) was also confirmed by Zeiss.
Therefore the result from our experience: If you want to see the real resolution, sharpness, grain, dynamic range, than take photographs with your equipment, your lenses, your films, your needed object contrast etc.
MTF data is not the answer of your questions concerning your equipment and your individual shooting conditions. Look at your pictures, not at curves from someone else.