Focus accuracy on film vs. digital

[kennylovrin]

Hi

I read somewhere a while back, can't recall where now, that focus accuracy is more important on digital than on film as film has a thickness and sensors doesn't. Apparently then it isn't so critical on film as there is a depth where the light will hit the grain and within that depth it will result in equal sharpness.

My understanding of this is that on a sensor the plane of focus must hit the sensor dead on it's surface, where on a film it just has to be within the thickness of the film (emulsion?).

So I have a few questions regarding this:

1. Is this in fact true? Does anyone actually know?
2. How does this relate to focus shift? Is focus shift less apparent on film than on digital, or it makes no difference at all on focus shift?

Thanks,
Kenny

 

[jsrockit]

2. How does this relate to focus shift? Is focus shift less apparent on film than on digital, or it makes no difference at all on focus shift?

It is more apparent on digital.

 

[zuiko85]

Not sure but I do know I'm 'talented' enough to get out of focus images with either.


Seriously, although this could be true technically it probably does not matter in the real world. Here is what I think is happening;

1. Auto-focus is now ubiquitous so DSLR cameras have screens optimized for brightness and not for easy accurate manual focus. (The great unwashed masses think auto-focus actually works optimally but it can miss just as we can with manual focus)

2. Because of the potential for higher resolution and the relative ease of making BIG enlargments with digital files compared with traditional darkroom prints then bigger prints are made demanding very precise focus. (In my own case I live in a small apartment, set up my small 35mm enlarger in a small bathroom and so.....make small prints, 6X9 inch on 8X10 paper for 35mm.)

That is my theory anyway and I'm sure there are a few out there in RFF forum land ready to blow some cannonball holes in it.

 

[kennylovrin]

Not sure but I do know I'm 'talented' enough to get out of focus images with either.


Seriously, although this could be true technically it probably does not matter in the real world. Here is what I think is happening;

1. Auto-focus is now ubiquitous so DSLR cameras have screens optimized for brightness and not for easy accurate manual focus. (The great unwashed masses think auto-focus actually works optimally but it can miss just as we can with manual focus)

2. Because of the potential for higher resolution and the relative ease of making BIG enlargments with digital files compared with traditional darkroom prints then bigger prints are made demanding very precise focus. (In my own case I live in a small apartment, set up my small 35mm enlarger in a small bathroom and so.....make small prints, 6X9 inch on 8X10 paper for 35mm.)

That is my theory anyway and I'm sure there are a few out there in RFF forum land ready to blow some cannonball holes in it.

You are probably right, but in this specific case I am not talking about perceived sharpness or whatever it may be called. Of course focus errors could be caused by loads of stuff.

I may have phrased my question wrong. What I'm thinking about here specifically is that if we say it's a given that the same lens is focused exactly correct wide open on both a digital and a film camera. Let's ignore potential user errors or any calibration needed, and just say that the focus is spot on.

Let's then say that this lens focus shifts when you stop it down. So you stop the lens down and take the photo.

Will it then appear sharper on the film camera than it will on the digital, because of the film thickness leading to film being more forgiving?

(I guess this is what jsrockit is saying above with his post).

If this is the case, would it be a negligible difference, or would it be still enough so that a problematic lens on digital could appear without focus shift problems on film?

This question stems from the fact that I'm browsing different potential 50mm lens purchases. And both the C Sonnar and Nokton 1.1 are said to exhibit focus shift, which is why I am wondering if this is less so on film or not.

And also because I find it interesting. 😉

 

[x-ray]

Out of focus is a little more apparent on digital but the big problem is pixel peeping. When we examine a negative we use a 4x - 8x magnifier. When we pixel peep we look at the image many time that a foot away from the monitor. My Hasselblad digital files are roughly 100 inches on the long dimension at 100% on the screen. I've seen some people even examine their digital files at 200-400%. The hard core digital guys feel everything must be pixel sharp, every pixel. I never enlarge anywhere close to that. My 6x6 film negs are rarely enlarged more than 12x if that. If you examined your film negatives at the same enlargement you would dee the same sharpness problems plus a lot of grain.

I think this pixel sharp mentality plus over saturation is what makes digital look digital. It doesn't have to look that way.

 

[kennylovrin]

But I'm thinking that if you miss focus by 2cm due to the lens exhibiting focus shift, you wouldn't have to view that photo at 100% to see that.

I'm not talking about how sharp a photo is, I'm asking how much film neutralizes focus shift. It may be a completely stupid question, but I still want to ask it as I read about that sharpness thing a while back. It just made it sound like when your focusing the image on a sensor there is only one absolute point, but when you're focusing on a film there is effectively a range within which sharpness will be equal.

 

[zuiko85]

I think see what your question is about now, thanks for the clarification. I would think that the amount of focus shift would of be the same if the lens were on a film Leica or digital Leica. So the smallest focused spot would still be the same diameter on film or silicon. Therefor I would think the best resolution you could pull off of either system would be similar. That is assuming the film is actually as flat as is technically possible.

 

[DtheG]

Well, suppose the depth of the film emulsion was significant —and the depth, if that is the right term, of microlenses and AA filters on a digital sensor were not— and that the film were perfectly positioned and flat. Then, if the focus was exactly right at the emulsion surface the bottom of the emulsion layer would be out of focus. So it could be argued that if the depth of emulsion matteres then some part of the film is always out of focus.

 

[kennylovrin]

Well, suppose the depth of the film emulsion was significant —and the depth, if that is the right term, of microlenses and AA filters on a digital sensor were not— and that the film were perfectly positioned and flat. Then, if the focus was exactly right at the emulsion surface the bottom of the emulsion layer would be out of focus. So it could be argued that if the depth of emulsion matteres then some part of the film is always out of focus.

That is also a very interesting perspective I must say.

 

[jschrader]

Well, suppose the depth of the film emulsion was significant —and the depth, if that is the right term, of microlenses and AA filters on a digital sensor were not— and that the film were perfectly positioned and flat. Then, if the focus was exactly right at the emulsion surface the bottom of the emulsion layer would be out of focus. So it could be argued that if the depth of emulsion matteres then some part of the film is always out of focus.

but if this is so then film can never be a sharp as digital; because point sharp at one surface of the emulsion wouls mean a certain OOF circle at the other surface. When enlarging, You would always see this circle.
So, then focus shift might be less relevant because the film would produce less sharp pictures than the sensor.
That would mean You should not worry about focus shift but simply be less picky for resolution when buying for a film camera.

Is that so?

 

[Calzone]

My understanding of "Sharpening" in effect is like creating a physical space/offset between the image on the negative and say the paper in making a wet darkroom contact print.

Compare that first analog normal contact print to another analog contact print made with the emulsion side up. The offset/space caused by the physical thickness of the film will display a perceived sharpening on the resulting positive. I am no expert in digital, but it is my understanding that this is what is manipulated digitally in post processing in what is called "sharpening." Of course I already learned that sharpening should be the final and last manipulation of any digital image.

I would agree with the above post that suggests that some of the perceived higher resolution of digital that resembles MF is due to use/overuse of this sharpening effect.

Now the inverse: to understand why the physical thickness of the actual film emulsion (less than the physical thickness of the film base) has a similar effect of increased perceived sharpness, but on a diminished/lesser scale due to the emulsion being thinner than the film base.

Anyways, hope I didn't confuse everyone.

I would like to add to this discussion that I look upon digital as a separate medium, and that the clean look of digital that some people object to or dismiss makes this medium (digital) look to me to be more in step with looking like MF in 135, except perhaps in tonality, because of it's perceived higher resolution perhaps because of manipulation of this "sharpening" effect.

I wonder if its real higher resolution in digital that gives the perception of more critical focus over film or is it the manipulation or overuse of sharpening? Please do not interpete or spin this thread into a film verses digital thread, and please look upon analog and digital as separate mediums.

Cal

 

[kennylovrin]

but if this is so then film can never be a sharp as digital; because point sharp at one surface of the emulsion wouls mean a certain OOF circle at the other surface. When enlarging, You would always see this circle.
So, then focus shift might be less relevant because the film would produce less sharp pictures than the sensor.
That would mean You should not worry about focus shift but simply be less picky for resolution when buying for a film camera.

Is that so?

Personally I would say this is the case, but I didn't dare to say it for risk of starting a flame war. 😉

 

[Calzone]

Personally I would say this is the case, but I didn't dare to say it for risk of starting a flame war. 😉

Surely we can safely say that digital offers a level of precision and control that is different than analog. Analog is analog and digital is digital. It's when we go beyond that its when things break down.:bang:

Cal

 

[jschrader]

Surely we can safely say that digital offers a level of precision and control that is different than analog. Analog is analog and digital is digital. It's when we go beyond that its when things break down.:bang:

Cal

nice to hear an unemotinal opinion.

What I like about digital: I can produce an acceptable b&w image in minutes and give it to someone to print in excellent quality on baryt paper. That is impossible with film.

What I like about film: I can buy a camera 30 years old and not worry how much value it looses in 2 years.

All these digital vs film tonality and unpleasant look discussions: that is lightyears away from my decision making.

J"rgen

 

[jsrockit]

But I'm thinking that if you miss focus by 2cm due to the lens exhibiting focus shift, you wouldn't have to view that photo at 100% to see that.

Exactly. I've used the c-sonnar and the nokton on a M9, and your point of focus between f/2 and f/4 (if the former is optimized for 1.5) will not be what you intended if you do not compensate.

 

[DNG]

If you are using the C-Sonnar on a non-M Digital.. Then the topic is moot, because you focus with LIVE-VIEW, no a RF Arm..

The C-Sonnar is made to have a focus shift from f/1.5 to f/2.8. As Zeiss has stated on their web site about this special lens.

I would think, the DOF is also a consideration at the Focus Plane. More will be in focus, yes, but, does this allow for different thickness's in sensors?

 

[kennylovrin]

Exactly. I've used the c-sonnar and the nokton on a M9, and your point of focus between f/2 and f/4 (if the former is optimized for 1.5) will not be what you intended if you do not compensate.

Yes, and I find that annoying (the Ultron I owned for a while did it a bit), which is why I was wondering if it is less of a problem on film (based on what I had read earlier online).

I suspect the difference, if any, would be negligible, and I might go for the Planar instead.

How much does the C Sonnar focus shift in reality? Are we talking millimeters or centimeters, or even a decimeter? I know if depends on focus distance as well, but maybe there is some approximation to be made anyway?

 

[jsrockit]

How much does the C Sonnar focus shift in reality? Are we talking millimeters or centimeters, or even a decimeter? I know if depends on focus distance as well, but maybe there is some approximation to be made anyway?

Unfortunately, centimeters. It's significant.

http://www.rogerandfrances.com/photoschool/ps%20firstlook%20sonnar%2050.html