Any technical explanations as to how or why film sharpness looks different from digital sharpness ??

[runswithsizzers]

With film, the sharpness is quite subtle and soft even on medium format meanwhile with digital the sharpness is much more glaring and just pops out. Basically, the two have different character when it comes to sharpness. Is this just as a result of the grain sort of masking everything ?

What are you looking at when are comparing film to digital?`

That is, are you comparing an analog print to a paper print made from a digital camera? Or something else?

 

[Pioneer]

How much of this is confusing lossy "sharpness" for the real deal? Digital edge enhancements do so by subtracting, not adding, actual detail. But film to film unsharp masking preserves the detail, while potentially enhancing edge acutance, provided one understands how to do that. Yet it can be overdone and look annoying either way. Basically, a lot of this line of debate is based on superficial generic stereotypes of things actually involving all kinds of variables. As usual, the Devil is in the details, either way.

And this is why the people who are really skilled at these techniques are paid the big bucks whether in is digital or analogue printing.

 

[Sirius Glass]

How much of this is confusing lossy "sharpness" for the real deal? Digital edge enhancements do so by subtracting, not adding, actual detail. But film to film unsharp masking preserves the detail, while potentially enhancing edge acutance, provided one understands how to do that. Yet it can be overdone and look annoying either way. Basically, a lot of this line of debate is based on superficial generic stereotypes of things actually involving all kinds of variables. As usual, the Devil is in the details, either way.

The software over sharpening of film and digital images has led to unobtainable sharpness that some now think is normal yet beyond what is optically obtainable.

 

[Nodda Duma]

I read through the responses, but failed to see anyone point out the fundamental technical difference:

The detectors in a digital sensor is the pixel, all of uniform shape and size.

The detectors in photographic film vary in size, with sizes ranging within a distribution determined by the design of the emulsion.

The contribution to the MTF (google it) of the imaging system due to the differences in the detectors are inherently different.

In addition, the scanning of the film contributes a factor, but the effect is negligible if you apply scanning rules of thumb now known for 25+ years.

Microcontrast is just layman slang for what’s going at higher spatial frequencies at or near the sampling frequency (including aliasing) of the system.

Anyone who says film isn’t capable of as high/sharp resolution as digital hasn’t looked at high-altitude or LEO reconnaissance imagery from the 1990s.

 

[Steven Lee]

high-altitude or LEO reconnaissance imagery from the 1990s.

What's the negative size in those LEO cameras? Standard 35mm?

 

[chuckroast]

Of late, I've been doing a lot of fiddling with developers to optimize acutance - which typically is expressed as images appearing more "sharp". A lot of this has to do with edge transitions (see the examples below) rather than absolute sharpness.

It is also my experience that maintaining good local contrast in the midtones, especially, improves the perception of sharpness.

There is no magic here. Digital can do unsharp masking to improve edge transitions, and there is plenty of post processing possible to improve local contrast. Moreover, tools like Topaz can do all manner of automated image repair.

The perception the digital is magically sharper stems from two things I think.

First, analog is a lot clumsier to post process in the darkroom than something from a digisnapper going onto a Mac for quick fixes. I have worked in the darkroom for 50 years and I am well aware how much effort it takes to really get a silver print right. I see tons of promising analog images that have not been sufficiently "tuned" in the darkroom after capture. The photographers just seem to give up after a certain point or - perhaps - lack the darkroom technique. IOW, it's more likely - IMHO - that someone will take the extra time with digital post production, because it's just easier.

Secondly, the overwhelming majority of digital images I see are dreadfully over produced with post processing adjustment, Excessive HDR "fixes", over saturation, and - yes - too much unsharp masking gives so many promising digital images a kind of garish, bordering on cartoon like look. And that look absolutely can appear astonishingly sharp, if not terribly aesthetically interesting.

Here are a couple images shot handheld with a Nikon F3 and a 35mm Ai-S f/1.4. I was experimenting with what Double-X would do in a small format with respect to edge transitions and apparent grain. I have no doubt that a mirror lockup and tripod would have made these even sharper. Even so, note the nice edge transitions and how the midtone local contrast is held by means of highly dilute, low agitation (EMA) in Pyrocat-HD.

These are scans of my "workbook" silver prints. The only post scanning adjustments I make is to try to match the physical print as best as possible to overcome scanning artifacts. They're not final and could still bear futher refinement. But these are 8x10s I make as a record of work I may later want to make into wall hangings. These are printed pretty much edge-to-edge from the negative onto an 8x10 sheet:

 

[MattKing]

Of late, I've been doing a lot of fiddling with developers to optimize acutance - which typically is expressed as images appearing more "sharp". A lot of this has to do with edge transitions (see the examples below) rather than absolute sharpness.

Strictly speaking, "sharpness" is a subjective characteristic, and is almost entirely a function of acutance - edge contrast. Overall contrast has a smaller role in the perception that we refer to as "sharpness". Resolution of fine detail has an even smaller role. It is for that reason that extremely fine grain films like T-Max 100 often appear less "sharp" - they resolve very fine detail, but they don't (artificially) accentuate edges by adding grain.
"Absolute sharpness" isn't really a thing, because there is no objective way of expressing that which is mostly perceptual and subjective.

 

[Nodda Duma]

Well said, Matt. Loose terms to describe the relationship between contrast and spatial frequency as well as the system’s sampling frequency.

What's the negative size in those LEO cameras? Standard 35mm?

IIRC, either 5” or 9” wide. Fun fact: for wide field high altitude reconnaissance, use of film was only discontinued about 2-3 years ago. It took decades of continuously evolving digital imaging technology to make the replacement of reconnaissance film feasible.

But here’s the thing: an entire *system* is designed to optimize high resolution (or whatever the requirement is), of which film or a digital sensor is only one “subsystem”. That is important to keep in mind. From an imaging system perspective, either imaging media can be used. A simple example of this is converting to sheet film to out-resolve a digital camera.

 

[Helge]

This is the data the demosaicing and all subsequent processing mechanisms has to work with. An ordered mess, low contrast with low colour information.

Especially interesting is the naive, super simple demosaicing in the upper right.

The demosaiced image is to big degree an invention.
You get a big delta if you compare an uncompromised image scanned with a synthetic simulated Bayer photo.

Edges and contrasty textured areas, are treated much differently than the rest of the image.

The very apparent moire is also present as harshness in a less obvious way in random textures. It’s baked into the data.

Monochrome sensors, rare as they are in consumer cameras, are not much better. The temptation and necessity of sharpening to simulate more information is just too big.
Good fine grain B&W always beats the resolution and lack of aliasing of monochrome sensors.
And the tonal content of almost all B&W film is better.

The “cleanness” of the signal and substrate structure that is predictable is what saves digital.
But it’s only through decades of research.
Research of a magnitude no one has ever extended to high resolution film scans.

Transformer networks will make it possible, even trivial, to raise the low contrast detail of a scanned negative (or any film or plate really) in the near future (if it hasn’t already happened), and even to remove the grain seamlessly if you should so desire.

 

[DREW WILEY]

When it comes to big expensive surveillance cameras, and both ground-based and space telescopes, and even very high end optical microscopes (in distinction from electron microscopes), there are engineering, design, and budget options that go well beyond anything likely to be taken into account here among us in either a film or digital capture frame of mind.

But more down to my own humble scale of gear and budget, the limitation of TMX100 mentioned by Matt, having high detail capacity but relatively poor edge acutance or perceived detail, can be cured by special development tweaks allowing a little more grain growth.

 

[faberryman]

When it comes to big expensive surveillance cameras, and both ground-based and space telescopes, and even very high end optical microscopes, there are engineering, design, and budget options that go well beyond anything likely to be taken into account here among us in either a film or digital capture frame of mind.

Unsurprisingly, lenses for national security and scientific applications are better than those available to photographers.

But more down to my own humble scale of gear and budget, the limitation of TMX100 mentioned by Matt, having high detail capacity but relatively poor edge acutance or perceived detail, can be cured by special development tweaks allowing a little more grain growth.

Unsurprisingly, some developers provide more acutance than others, so pick the one most appropriate for the look you are trying to achieve. None of this is rocket science.

 

[chuckroast]

But more down to my own humble scale of gear and budget, the limitation of TMX100 mentioned by Matt, having high detail capacity but relatively poor edge acutance or perceived detail, can be cured by special development tweaks allowing a little more grain growth.

Funny you should mention that. I just pulled a couple of 120 rolls of TMX out of the freezer with the intention of giving it 60min EMA development in Pcat to see how it cleans up the edges. We'll see.

 

[DREW WILEY]

Pyros work fine with TMX100, but not in relation to improving edge effect. For that purpose, I use Perceptol 1:3 @ 16 min, which at that greater dilution favors just enough grain growth.

 

[chuckroast]

Pyros work fine with TMX100, but not in relation to improving edge effect. For that purpose, I use Perceptol 1:3 @ 16 min, which at that greater dilution favors just enough grain growth.

I plan to use highly dilute, low agitation over a 60 min standing time. This tends to promote edge effects in most films. We'll see.

 

[DREW WILEY]

Well, keep us updated per results. I shoot a lot of Tmax in 120 formats. It's the best all-round film out there in terms of scene contrast range, spectral sensitivity, development gamma potential, and fine grain. The only thing that needs curing is the edge acutance. TMax 400 solves even the latter issue by itself; but I generally use that in sheets instead.

 

[chuckroast]

Well, keep us updated per results. I shoot a lot of Tmax in 120 formats. It's the best all-round film out there in terms of scene contrast range, spectral sensitivity, development gamma potential, and fine grain. The only thing that needs curing is the edge acutance. TMax 400 solves even the latter issue by itself; but I generally use that in sheets instead.

I will report back what I find ...

 

[Nikon 2]

Is it true film can never achieve the sharpness of good digital photography…?

 

[MattKing]

Threads merged - with a caution. Photrio rules prohibit digital vs. analog - it is only the qualitative comparison inherent in tis thread's premise that allowed for its continuance.

 

[Helge]

When it comes to big expensive surveillance cameras, and both ground-based and space telescopes, and even very high end optical microscopes (in distinction from electron microscopes), there are engineering, design, and budget options that go well beyond anything likely to be taken into account here among us in either a film or digital capture frame of mind.

But more down to my own humble scale of gear and budget, the limitation of TMX100 mentioned by Matt, having high detail capacity but relatively poor edge acutance or perceived detail, can be cured by special development tweaks allowing a little more grain growth.

I love exactly that characteristic of TMX.
I usually describe it as velvety.
The “cure” is to use Delta 100, it has exactly that extra little “bite”.

 

[Helge]

Light travels, just like sound, in waves.
Till a certain extend both, light and sound, can be approached equally, so you can convey a theory about sound to light.
I think that the illustration in appendix can explain something...

How so? It’s just the usual fallacy of a sampled digitized signal as a bar chart.

 

[Rudeofus]

How so? It’s just the usual fallacy of a sampled digitized signal as a bar chart.

More than that, it completely ignores the factor noise, and it compares a wave, which is observed over time, with a wave, which is observed as a whole packet in one. Whatever the differences between analog vs. digital may be, they are not explained in this chart.

 

[Helge]

More than that, it completely ignores the factor noise, and it compares a wave, which is observed over time, with a wave, which is observed as a whole packet in one. Whatever the differences between analog vs. digital may be, they are not explained in this chart.

People are focusing on the wrong part with the “digital”.
What gives new fangled cameras their characteristics is to a large extent the very analog sampling sensor.
The integration, interpolation of the data from the sensor could in theory be done via analog means too. Like it was done on analog comb filters for video.

 

[Nikon 2]

Of late, I've been doing a lot of fiddling with developers to optimize acutance - which typically is expressed as images appearing more "sharp". A lot of this has to do with edge transitions (see the examples below) rather than absolute sharpness.

It is also my experience that maintaining good local contrast in the midtones, especially, improves the perception of sharpness.

There is no magic here. Digital can do unsharp masking to improve edge transitions, and there is plenty of post processing possible to improve local contrast. Moreover, tools like Topaz can do all manner of automated image repair.

The perception the digital is magically sharper stems from two things I think.

First, analog is a lot clumsier to post process in the darkroom than something from a digisnapper going onto a Mac for quick fixes. I have worked in the darkroom for 50 years and I am well aware how much effort it takes to really get a silver print right. I see tons of promising analog images that have not been sufficiently "tuned" in the darkroom after capture. The photographers just seem to give up after a certain point or - perhaps - lack the darkroom technique. IOW, it's more likely - IMHO - that someone will take the extra time with digital post production, because it's just easier.

Secondly, the overwhelming majority of digital images I see are dreadfully over produced with post processing adjustment, Excessive HDR "fixes", over saturation, and - yes - too much unsharp masking gives so many promising digital images a kind of garish, bordering on cartoon like look. And that look absolutely can appear astonishingly sharp, if not terribly aesthetically interesting.

Here are a couple images shot handheld with a Nikon F3 and a 35mm Ai-S f/1.4. I was experimenting with what Double-X would do in a small format with respect to edge transitions and apparent grain. I have no doubt that a mirror lockup and tripod would have made these even sharper. Even so, note the nice edge transitions and how the midtone local contrast is held by means of highly dilute, low agitation (EMA) in Pyrocat-HD.

These are scans of my "workbook" silver prints. The only post scanning adjustments I make is to try to match the physical print as best as possible to overcome scanning artifacts. They're not final and could still bear futher refinement. But these are 8x10s I make as a record of work I may later want to make into wall hangings. These are printed pretty much edge-to-edge from the negative onto an 8x10 sheet:

Now, that’s sharp…!

 

[Nikon 2]

I love exactly that characteristic of TMX.
I usually describe it as velvety.
The “cure” is to use Delta 100, it has exactly that extra little “bite”.

Interesting…!

 

[Nikon 2]

Threads merged - with a caution. Photrio rules prohibit digital vs. analog - it is only the qualitative comparison inherent in tis thread's premise that allowed for its continuance.

Fine line for film, sharp line for digital…!