35mm Scans VS FF digital

[Вormental_old]

higher dynamic range overall? No, film has more dynamic range if you take info account highlights.

I like film as much as you do, but you are wrong here. No need to argue over something manufacturers are publishing, just count stops on your film datasheets, then look up BSI CMOS specs adjusted for noise. https://www.photonstophotos.net/ has a nice database which is actually quite conservative as he subtracts noise quite aggressively, most would agree to add 1-2EV to his charts, especially if you're comparing to a grainy ISO400 film. It's not even hard to prove it using your own tools. Try making a C41 negative (or B&W, doesn't matter) which will cover 100% of the histogram if you camera-scan it. It won't be even close.

Moreover, high DR on film (even if existed) wouldn't even be usable in practice, as DR of photo paper is abysmal and sensors used in film scanners are junk from 20 years ago.

 

[George Mann]

Digital specs are nothing but smoke and mirrors. There isn't a single digital device that has truly lived up to its claims.

 

[138S]

I’ve stopped engaging with 138S in these matters.
He clearly has a hoppy horse he has to go ride till it drops and then beat with a stick some more.
Even if said horse is a figment of his own imagination.
Let me just suffice to say buyer beware.
That is, don’t buy his BS and FUD.

Helge, I am not a newbie that you can get nervous with a mild personal attack, if we go to the technical discussion I make you look ridiculous.

This is what film does, if any doubt:

http://www.tmax100.com/photo/pdf/film.pdf
https://www.onlandscape.co.uk/2011/12/big-camera-comparison/

One of the things you don't catch is that film resolving power depends on if you test the ultra fine crystals that are sensitive at +6 overexposure (1000:1 contrast test) or the coarser crystals that are sensitive at +/-0 or -2.

You have been missleading yourself by not catching that a lab test and real photography are different things.

The other thing you don't catch is that most of us don't shot film because of resolving power, at 8x10" format I can tell you that I've no flaw in Image Quality. We shot film because we master a crafting delivering the aesthetics we love.

 

[George Mann]

Anyway, in staying on topic I will state that I have seen consistantly better results from dedicated professional scanners using fine grain, professional quality films than I have from digital cameras.

It's more about how well the image presents the reality of what it captured than the far less meaningful technical arguments that are exhaustively bandied about.

 

[Pieter12]

Anyway, in staying on topic I will state that I have seen consistantly better results from dedicated professional scanners using fine grain, professional quality films than I have from digital cameras.

It's more about how well the image presents the reality of what it captured than the far less meaningful technical arguments that are exhaustively bandied about.

But one cannot always shoot fine-grain film. I have obtained excellent results with a Nikon D800, making 20x20" piezography prints that are better than what I could achieve making optical prints from 35mm film with the same equivalent ISO.

 

[Pioneer]

I shoot FF digital, and I'm learning about printing digital negatives etc. A question that I have though is, what are the differences between a scanned 35mm negative, and a full frame digital file? Why would someone prefer to start with a scanned digital file, instead of a straight digital file?

I know that's a very generic question, because preferences will differ from person to person, and the end justifies the means, but from a quality perspective what are the differences?

In short,

• I shoot film because I enjoy the entire process not because there may be some quality difference.
  
• I use digital cameras to produce images that will ONLY be used for on-line purposes. I typically do this to sell things or to illustrate a procedure for accomplishing a task. These are not typically printed.
  
• I scan film as a digital shortcut to producing a contact sheet, one that allows me to see in general terms how a given negative will look if I decide to print it.
• I also scan film to allow me to share images on-line, usually in forums like this one.
• I don't (usually) scan film so that I can produce a print although I do occasionally print these scans, usually in a color process. There are things that I can do with Photoshop that I cannot easily recreate using an analog process. This is almost certainly due to a skill deficiency of my own in the analog printing process, particularly on the color side.
  
• I much prefer the black and white images I create using film and enlarging. Again this could certainly be due to a deficiency of my own in creating acceptable digital black and white files and prints.
• I do not feel that one process is better than the other. I have personally seen (and purchased) amazingly creative work done using both digital processes, film processes and a hybrid of the two.
• Finally, I do not personally believe that the technical capabilities or deficiencies of the digital sensor or of film have any bearing on what I consider to be beautiful prints. Rather it is the composition and the skill of the photographer in the use of their chosen tools.

Well...maybe that wasn't so short after all.

 

[Helge]

I like film as much as you do, but you are wrong here. No need to argue over something manufacturers are publishing, just count stops on your film datasheets, then look up BSI CMOS specs adjusted for noise. https://www.photonstophotos.net/ has a nice database which is actually quite conservative as he subtracts noise quite aggressively, most would agree to add 1-2EV to his charts, especially if you're comparing to a grainy ISO400 film. It's not even hard to prove it using your own tools. Try making a C41 negative (or B&W, doesn't matter) which will cover 100% of the histogram if you camera-scan it. It won't be even close.

Moreover, high DR on film (even if existed) wouldn't even be usable in practice, as DR of photo paper is abysmal and sensors used in film scanners are junk from 20 years ago.

I’d be interested where you think you see those stops you count in the data sheet?
Output media always has far worse DR than the recording media (that’s why slide is considered an end medium. Though a good projection probably is the output medium method extant with the highest real DR) you are meant to pick and choose and “correct tonal relationships” to fit in the medium.
What is called tone mapping in digital.

 

[Helge]

Helge, I am not a newbie that you can get nervous with a mild personal attack, if we go to the technical discussion I make you look ridiculous.

This is what film does, if any doubt:

http://www.tmax100.com/photo/pdf/film.pdf
https://www.onlandscape.co.uk/2011/12/big-camera-comparison/

One of the things you don't catch is that film resolving power depends on if you test the ultra fine crystals that are sensitive at +6 overexposure (1000:1 contrast test) or the coarser crystals that are sensitive at +/-0 or -2.

You have been missleading yourself by not catching that a lab test and real photography are different things.

The other thing you don't catch is that most of us don't shot film because of resolving power, at 8x10" format I can tell you that I've no flaw in Image Quality. We shot film because we master a crafting delivering the aesthetics we love.

You are your own “mild personal attack”.

 

[Вormental_old]

I’d be interested where you think you see those stops you count in the data sheet?

Where I think I see? This passive-aggressive tone is not how you ask a stranger to teach you something, Helge.

 

[138S]

You are your own “mild personal attack”.

Scanners are a heated debate, the thing comes from there, isn't it? Specially if commercial interests are in the middle...

This is a mild personal attack to me because it's you that get qualified, insults qualify the offender most of the times.

That is, don’t buy his BS and FUD.

 

[Helge]

Where I think I see? This passive-aggressive tone is not how you ask a stranger to teach you something, Helge.

“Passive aggressive”‽
It often becomes a question of tone when argument ability lapses.
I ask you where you think you see it, because I don’t think I see it on the data sheets.

 

[138S]

Moreover, high DR on film (even if existed) wouldn't even be usable in practice

Вormental, negative film is designed to record an insane amount of Dynamic Range, well beyond monitors or paper may show.

This excessive DR it is usable in practice in two ways.

A) Deciding in the post processing what share of the captured dynamic range you print or display, selecting the bottom-top levels to decide what you clip or not. You keep your options open.

B) Compressing shadows and highlights to take a certain amount of the extremes of the DR in the end medium, usually sparing a way larger proportion of thr central DR for the scene mids. In that way you depict a way larger dynamic range of the scene in the more limited end medium, Historically this has taken a notable creative effort. Desing of the film/paper toe-shoulder, regular print manipulation (dodging-burning), advanced print manipulation (SCIM, HLM, CRM...)... and of course curve edition (etc) in Photoshop and the like.

For darkroom printing, if you shot a dull scene in then shadow then all will mostly fit in the paper range, no problem. But as you have a contrasty scene then you have two main choices, one is taking some advantage from film toe/sholder to compress the extended DR, or you may take a more linear capture and later deciding what you do. The later delivers a more flexible linear negative, but it can be more difficult to print. With hybrid curve edition is easy. Instead in the darkroom control can be quite challenging, and controlling what you do requires mastering several tools, but you get a genuine traditional craft, this can be fairly appreciated or not.

 

[George Mann]

Вormental, negative film is designed to record an insane amount of Dynamic Range, well beyond monitors or paper may show.

At mid and high frequencies film is superior. But the sensors used in digital cameras show greater sensitivity at low frequencies.

The dynamic range of the original capture medium is of greater importance than that of a scan or print.

 

[138S]

At mid and high frequencies film is superior. But the sensors used in digital cameras show greater sensitivity at low frequencies.

The dynamic range of the original capture medium is of greater importance than that of a scan or print.

Of course, film shines in the highlights, digital shines in low light conditions.

An interesting exercise is exploring high budget movie production, those reaching $200 million, and understanding when the cinematographer uses Vision 3 stock film or when he uses an Alexa digital camera.

Those cinematographers are not rookies, they are a kind of true artisits... they know exactly what they do and why they do it, and they master their tools. I'm fortunate to have a friend usually working as First Assistant Camera (Focus Puller) and I've got explained a bit how they make those decisions.

Today, being this one a digital world, what is clear is that if stock film is still used in some selected movies this is because several factors, and a principal one is highlight depiction. In cinematography (and sometimes in photography) shadows are way less important than highlights, we may want shadow detail, but many times we use shadows to clip distracting stuff, but highlights and glares (in cinematography) are always important to depict volumes, specially in the faces of the actors, creating a sense of reality and immersion that it would be challenging with digital capture. Still every day digital cameras for cinematography are better and the gap is narrower.

That friend working as 2nd assistant explained me many times that with digital captures illumination is way more critical than with film, once adjusted it can happen that the actor moves the face a bit in another direction than the exactly intended and the capture is bad, they may have many captures and the one with the good expression is not good enough. With film when the actors nail the role they are done, this saves money because the shooting time is shorter, they can focus in the important things, and all the captures in what the actors performed best are good. Also actresses feel themselves very beautiful in the scanned captures they may preview, so they are more self-confident. Film is film...

 

[Helge]

At mid and high frequencies film is superior. But the sensors used in digital cameras show greater sensitivity at low frequencies.

How exactly would that work? Low frequencies are the easiest.
Might you be thinking of the misleading MTF plots, with a steady fall from film and a pretty flat-ish, only slightly slanting line with a fall off a cliff at the sensors site size limit (with some light chattering after due to aliasing)?

 

[George Mann]

Of course, film shines in the highlights, digital shines in low light conditions.

Actually, low speed, fine grain film outperforms digital for low light captures.

 

[George Mann]

How exactly would that work?

The way that sensors capture light make them unusually sensitive to low frequencies.

 

[chris77]

“Passive aggressive”‽
It often becomes a question of tone when argument ability lapses.
I ask you where you think you see it, because I don’t think I see it on the data sheets.

Sorry, but your lack of self-reflection is... rare. May i ask you what is your profession? How old are you? I am just curious.. excuse the off topic.

 

[138S]

Actually, low speed, fine grain film outperforms digital for low light captures.

Yes... but.. how fine ? what speed ?

Image Clarity book, by John Williams, Figure 11.1 compares different crystal sizes to the Red-Blue wavelength.

Lippmann emulsion, 0.01µm to 0.05µm
Photopaper, 0.1µm to 0.2µm
Process films, 0.3µm to 0.5µm
Popular fine grain film: 0.6µm to 1µm
Coarse grains in fast film: 2µm to 4µm

Compared, a Nikon D850 has a pixel pitch of 4.34 µm.

On the paper film should resolve more, but in practice the halide effectiveness is way lower.

There are sensors with very small grains, the one in the Sony XZ3 smartphone has a 1.22 μm pixel pitch, sensor is only 1/3"... but this would allow a D850 with 500MPix in the future if wanted.

What is clear, if you "design" a crystal and a pixels of the same size, the pixel will be faster. Still film has a particular advantage: you mix crystrals of different sizes, while sensors have a fixed size, beyond some rarities.

 

[Вormental_old]

Вormental, negative film is designed to record an insane amount of Dynamic Range, well beyond monitors or paper may show.

Kodak and Fuji both disagree with you, as they publish different information in the data sheets for the films they sell. The only advice I can offer is "go and study". And as you'll be learning about DR, stick to just film and sensors, leave monitors for a later time (you're not up to date on monitors as well).

 

[138S]

Kodak and Fuji both disagree with you, as they publish different information in the data sheets for the films they sell. The only advice I can offer is "go and study". And as you'll be learning about DR, stick to just film and sensors, leave monitors for a later time (you're not up to date on monitors as well).

I'm a good student...

This is an exposure test with Portra 400:


https://petapixel.com/2018/02/05/test-reveals-exposure-limits-kodak-portra-400-film/

It shows a 10 stops dynamic range without exhausting the film capability, and in the +6 shot you have sun bathed walls that would be additional +2...

TMX with POTA developer may reach 18 stops latitude, or more...

This is the Vision 3 (Cinestill) similar to Portra, each H unit is 3+1/3 stops:



So it has 10 linear stops and 4 additional stops in the shoulder, without pulling development. If you pull you add several additional captured stops to 16 or 17 total.

 

[George Mann]

Yes... but.. how fine ? what speed ?

Ektar 100 is an upper end example of such a film.

Both it and Pan-X (50 ASA) have given me the best nighttime pictures I have ever taken.

 

[138S]

Ektar 100 is an upper end example of such a film.

Both it and Pan-X (50 ASA) have given me the best nighttime pictures I have ever taken.

OK, but Ektar is less sharp than a D850, for the same format, even with at 30:1 "micro-contrast". Once you go to low contrast (the most common situation in textures) film lowers the yield a lot, while the sensor holds a similar performance. If going to extreme contrasts say (300:1) not usually found in scenes then Ektar perhaps may beat that sensor.

The ektar advantage is that we can shot 6x7cm or 8x10" if we want, blowing any digital bug around.

 

[Lachlan Young]

Kodak and Fuji both disagree with you, as they publish different information in the data sheets for the films they sell. The only advice I can offer is "go and study". And as you'll be learning about DR, stick to just film and sensors, leave monitors for a later time (you're not up to date on monitors as well).

And the further complexity is that while the films when scanned may have theoretically larger gamuts (and some degree of greater latitude) than the chromogenic papers can deliver, they tend not to deliver their design colour balance when handled in the wrong colour space (the limits of the paper gamut effectively act to stop the colours going 'wrong' - if everything is working correctly), and (as you say) their curve behaviour is such that while information may be retrieved from the extremes of the scale, it may not be qualitatively 'good'.

 

[George Mann]

OK, but Ektar is less sharp than a D850, for the same format, even with at 30:1 "micro-contrast". Once you go to low contrast (the most common situation in textures) film lowers the yield a lot, while the sensor holds a similar performance. If going to extreme contrasts say (300:1) not usually found in scenes then Ektar perhaps may beat that sensor.

This is why I stated that digital numbers and figures are not comparable to that of film.

Ektar beats the 850 in all but shadow detail.