OT: Resolution of film

A little off-topic, but hoping the experienced film shooters here can

answer. Would like to know what is the attainable resolution in lp/mm

of modern slide film such as Provia 100 under normal conditions (ie

handheld at shutter speeds from 1/15 to 1/500). Does it vary with the

amount of light available? Or is it the same, as long as the correct

exposure is used?

 

I'm interested to compare with digital (admins: not trolling), esp

advanced DSLR's. Dante Stella's website says 35 mm film has effective

resolution of 19 MP or so, but doesn't say if is for ISO 25 film or

for all film. Anyway I'm not interested in resolution from test

charts or tripods, but under real-world shooting conditions.

 

Just to say that I shoot a lot of film now. Yes, I know that there's

a lot more to quality of a picture than just lp/mm, one can talk

about bokeh, contrast, look, colour saturation, etc. And on the

digital side, pixels are not everything, sensor size is far more

important, there are also issues such as chromatic aberration and the

crop factor

 

But the resolution question interests me, because I want to know if

the resolution of digital SLR's has exceeded film-- please, no

digital vs film wars, thanks.

Sounds like you're interested in the theoretical limits of resolution between film uv pixels.

You might find this interesting.

 

http://phot.epfl.ch/workshop/wks99/2_1.htm#titel1

Fuji Provia 100 has a resolution of 145 lpmm. That is 7366 dpi;

for 24x 36 frame, requires 290 x 24 x 290 x36 = 72 MP. No, the resolution of digital SLR has still a long way to catchup, not to mention that there are film with much higher resolution.

This one will run and run as lpm and pixel count don't necessarily match up. That said, Dante's statement sounds far too low for any modern film apart from the really high speed ones.

 

I know that Ilford claimed 200 lpm for Pan F as long ago as the early 'seventies although that would depend on immaculate technique in exposure and development. If you had the technology to do it, I would imagine you could usefully make a 250MByte scan from a 35mm negative which met that specification. That is an order of magnitude greater than the very finest DSLRs currently on sale. This is, however, theoretical stuff for most of us because we simply don't have the equipment or the technique to do it.

Photojournalists' equipment choice has long been a major influence on what the average Sunday shooter lusts after. With the demise of the large format news magazines (such as Life and Look)a few decades ago, and with the much smaller Time an Newsweek being more text orientated,

the average photo is seldom used much larger than a 4"x6" (10x15 cm)from the local mini-lab.

 

Since the 1960's resolution performance of the best lenses at their optumum apertures hasn't changed much. Performance wide open has improved somewhat, and zooms have gone from being newspaper sports photographers' bright light play toys to being more all-around useful. The biggest improvement has been in films, whether Tri-X or color. ISO 160 High Speed was really too grainy for an 8x10 from 35mm, as was the wedding shooters' standby, first ISO 80 CPS followed by ISO 100 VPS. 120 film ruled that market. The answer to your question is no.

There was an article along these lines in the March/April 2004 issue of Photo

Techniques. The tester set out to compare the resolution with Kodak Gold

100 and digital resolution using a Canon 10D (also I assume at ISO 100).

Although lens test charts were used, it seems to have been a realistic

practical test. The results were that the film (and the lens) could resolve 90

line pairs per mm in the centre and 55 lp/m in the corners, and the 10D (with

same lens) could resolve 32 both centre and corners. The surprise was that

11 x 14 prints from both were virtually indistinguishable. Before people rush

in to point out that this is impossible, not real world, etc, please read the

article. You might not agree with its conclusions, but it is very interesting and

at least someone has given it a try. Not sure if there is a link to their web site

for this article. It is on page 27 and is by Neil Lipson.

I would agree that the 35mm 8-perf frame (let alone a 6x6 frame from a MF camera) yields a heck of a lot of detail with good lenses.

 

However I would take a different approach. For all practical purposes, a 4Mpx camera (e.g. D2H) and up can replace 35mm film provided the sensor is a modern one. Not because of resolution but because of overall image clarity.

 

Having said that I have a personal bias toward 35mm film for several reasons. One is to do with long exposures.

"I want to know if the resolution of digital SLR's has exceeded film"

 

Here's the point where, in my view, film's resolution finally exceeds digital.

 

It's a full-frame 11MP DSLR versus 6x9 ISO 100 film (with a Schneider lens giving virtually identical coverage). The film was scanned at 4000dpi with a glass negative carrier and a good scanner. It produced a file size that brought my computer to its knees, something that should be remembered in these kinds of discussion.

 

In any print up to about 11"x14"/A3 I can't spot the slightest difference in resolution. Around about 20"x30" film slowly starts to have a small advantage, at least in terms of raw resolution.

 

For me the most significant learning that comes out of all this is a pyscholigical or perceptual one. Academic assesments of resolution forget that all images are not created equally, IMO there's not that many shots that contain sufficient recognisable detail for resolution to be a practical or meaningful issue. IMO Zeiss got it right when they claimed it's the acutance of the larger structures that contribute most to our judgements on sharpness, and that's why more and more photographers are preferring the digital look.

And here's a tiny section on film at 100%<div></div>

And here's digital ressed up to match.<div></div>

There are two aspects to it: first the theoretical side of the debate, second - by far more important, the practical side.

 

In theory, you go after the optimal performance of film, for what reason ever, and go to a lot of efforts to find the limits, you have to create a perfect workflow and perfect variables (like the scanner). If you go full analog, especially the color part will require a master at work. As an experiment, this is worthwile. But for taking pics and printing them - I'm afraid the results are useless. In real world, this is not very practical to get your frames printed. The digital comparison just needs a shutter press and raw file processing, and you are done.

 

So while film might have more resolution in theory, it is not practical to extract it all from the negative.

 

So for a real world decision, the 4mp and higher digicams will be undistinguishable from film in terms of resolution, and it comes to your personal preference if you like to have a little computer taking your pics, or prefer the characterisitcs of film cameras and images (DOF comes to mind, dynamic range, grain, responsiveness of the camera..).

Tend to use slide film with a Nikon 5000ED scanner. Usually ISO100 slide film ( velvia 100F or Kodak 100GX). The resolution I get seems to equate to something north of a 10MP DSLR (meaning the Canon 1ds or Kodak 14n) - the others (10D, D70, D2H) or 6-8MP DSLRs don't seem to be quite there ( though all will print fine at A3 size). The scanned image from silm though do seem a bit "noiser" in the shadows but have not compared it using Nikon's multi pass scans which are supposed to reduce noise ( anyway it takes much longer for one of these)....... What we really need is for someone with an Imacon Flextight to weigh in as I think the Nikon scanner is not the final word in low noise.

 

 

Notice this doesn't answer the question exactly. As this was a test of a DSLR sensor vs. Film scanner sensor.

I'm a film man. However, there is an interesting comparison done on Luminous Landscape when he tested the 1Ds. I hope Michael doesn't mind me quoting him: "I'm afraid that film has definitively lost the battle. The 1Ds's full-frame 11MP CMOS sensor produces a 32MB file ? as big as a typical scan. But this file is sharper and more noise free than any scan I have ever seen, including drum scans. There simply isn't a contest any longer."

 

That said, I have still not rushed out to buy a 1Ds!

But look at his evidence and make your own decision...

funny how MR's opninion differs from Ken Rockwell & Co's in <a href="http://www.kenrockwell.com/tech/filmdig.htm">THIS ARTICLE</a>. </p>

 

I haven't extensive scientific testing myself, however, in my case I find my prints based on film looks more pleasing, i.e. better to me, than what I got out of a Canon 10D I used use/own. More important than lp/mm for me is the latitude of film vs. a prosumer DSLR. If I had something like a Canon 1DMkII or 1Ds, I'm sure I would shoot digital a lot more.

 

Funny enough, it is interesting to see what you can achieve with a simple Leica M, a 35/2 ASPH and Fuji Velvia 50 vs. e.g. a Canon 1DMkII. If you were strolling around Paris for 5 hrs in a suit, which one would you bring?

The fallacy in all the film vs digital "tests" that find digital equal or better, is that in all cases the comparison is made to film that has been scanned--in other words, digitized, and in most instances not on high-end drum scanners. The fact is, there can't really be a truely fair comparison between digital capture and film, because there is no way to do so without introducing a host of variables. We can't look at a slide and a digital shot side-by-side on a lightbox under a loupe. We need to view the digital shot on a monitor or make a print, and to place a film image on the computer or make a print means digitizing it, or making a wet-print...in which case the quality of the enlarger and lens and the skill of the person doing the printing are variables. So perhaps the only thing we have is theory, such as Martin's calculations.

 

It seems to me that it is the convenience and to a certain extent the exciting novelty of digital, in a medium that has changed little in 75 years, that is its biggest draw; however feeling that they are not giving up any image quality is essential to the digital user, hence the number of biased "tests" that "prove" the math is wrong and they are right. Add to that the billions that the various manufacturers and suppliers are pouring into slick marketing, and it is almost at the point of irrelevency whether digital is worse, equal or better than film. In the sales arena, the battle is over without a shot being fired, digital has won. I'm just sitting back, shooting film for as long as I can get it. I don't believe digital is anywhere near film in resolution, it is sufficient enough proof for me just the fact that every digital image--from camera or scanner--must be "sharpened" artificially by software. Moreover, since I am almost universally dismissed for making that statement, I don't forsee digital *ever* attaining the native resolution of film, since the manufacturers can see no reason to spend the money developing it when people are satisfied with what's already out there.

 

Look at it this way: 60 years ago the quality of lenses and film were nowhere what they are today, and people were happy and a lot of spectacular, historic photographs were made. Though digital's resolution limits take us a step or three backward, there is no reason to believe that great photographs will not be made with them. in fact, believing themselves relieved of the cost of film and processing there may be many people who will do photography that would not have, had they needed to buy film and pay for processing. Eventually they may wake up and realize that they have spent much more on the eqipment, memory, and ancillary doodads and software, not to mention their time spent learning and using image processing software, but by then there will probably be some new emerging technology that is even less sharp than digital but which offers some even more seductive allure to the weaknesses of human nature.

The question was hand held digital, against hand held film. I have wondered that myself. I see no economical, conclusive way to run tests. You would have to have some way of measuring camera shake with a statically valid cross section of the population under differing circumstances. Excitement, cold, tired? Then you would get into the weight of different cameras, methods of holding them?. One of the reasons that is cited for digital not being as sharp as it could be, is the fact that some people hold the cameras out and use the screen for a view finder.

 

As resolution of an optical system will always be less than the weakest element of the system, I suspect that hand held would almost completely mask any differences in resolution between 35mm film and say a 10 meg digital.

 

If you look as tests showing system resolution vs. individual resolution of components, you see results where tripling the resolution of an individual component only increases system resolution by a small percentage.

 

However, having said that entire why can?t people just enjoy their digital cameras instead of trying to prove the impossible.

 

Resolution is a measurement, not an opinion. In careful testing using laboratory testing and scientifically defendable methods, 200 asa film has been shown to have the resolution equivalent to a 16meg camera. (F.B.I. crime lab testing.) You are free to prefer the Empire State Building to the Sears Tower but the Sear?s tower is taller.

 

People who want to prove that digital resolves more detail than film always take photographs of subjects where there is no reference to measure. I.E. faces, workbenches, burglar alarm boxes anything but something with diminishing detail. I.E. test charts, parking lots with license plates at various distances, newspaper type, and city scapes with signs, and other detail at various distances.

 

The nature of film is to attempt to resolve finer and finer detail less and less clearly. Digital on the other hand destroys any detail that it can?t present clearly. The choice is a personal one. What do you want?

 

I chose film because, if I capture the information, I can always manipulate the image to look like digital. I.E. if I scan a negative at the same resolution as a 5 meg camera, the results will look like an image shot with a 5meg camera.

 

However, I can?t go the other way. If I can?t read a license plate in a digital image, no amount of Photoshop manipulation will create the information.

The problem with these discussions is that it doesn't matter.

 

What does matter is the appearance of the final print in the conditions it is meant to be viewed.

 

Everything else, including pretty much everything above, is irrelevant and a waste of words and time.

Speaking of the FBI crime lab, does anyone know where one could get that software they use on CSI where they zoom in on a surveillance photo and get a sharp image of some guy's wristwatch a block and a half away?

<<Everything else, including pretty much everything above, is irrelevant and a waste of words and time>>

 

I assume by "above" you refer to the first two lines of your post, in which case I agree 100%.

If people don't care about photography, but just about technical analysis of output, sure, it's irrelevant. But for a <i>photographer</i>, the point is prints that do what they are supposed to, maybe that's difficult for you to understand, but that's what <i>photographers</I> do. For equipment geeks, sure, everything above my post means something. Just don't confuse that with <i>photography.</i>

http://www.imx.nl/photosite/technical/highres.html

 

Here is the link that discusses 35mm film resolution in detail. It might answer you question.

 

Jay, don't ever use the word "ever" :) To use that word makes your statement false because nobody know what will transpire in the near/distant future.

 

Regards

 

Steve

Jeff you are hilarious in a pathetic sort of way. The assertion that real photographers don't give a crap about sharpness and resolution is simply your way of bolstering your own sagging self-esteem by demeaning everyone else. Fortunately your antics are as transparent as your point is groundless.

"People who want to prove that digital resolves more detail than film always take photographs of subjects where there is no reference to measure. I.E. ... burglar alarm boxes"

 

Neal, you're wrong in assuming I've got an agenda one way or another. I don't, I shoot both digital and film. Furthermmore, what I posted was actually an example where scanned film outresolves digital.

real photographers dont sweat over it and yap about it over and over and over. they learn it, transcend it, and do it.....

Tell it to your grandchildren Grant, they might be naive enough to believe you. Anyone who's been to any convention where pro photographers gather as a group knows that gear-related p'ssing contests are frequent and heated. Maybe the mincing artsy-types aren't, I wouldn't know as I don't hang out with them except here.