Canon FD Film Scans Softer than Digital Camera with Canon FD

[ifaynshteyn]

Hi everyone,

Longtime mirrorless shooter with legacy glass. I recently began to shoot film, 35mm and MF. Prior to shooting with my Canon A1 and FD lenses, I used my FD Lenses on my Sony A7 Mirrorless camera.

Is it normal to have my old film glass appear sharper on my mirrorless digital than from my lab's professional film scans? Is it due to film's inherent grain structure that there is a perceived difference? Normally, I shoot zone-focused, so I know very well when I have outright missed focus. Even when I adjust focus on a fixed object, I don't get the same sharpness.

I have attached a couple sample images for review/discussion. I'm at a bit of a loss because I love the tones and process of film, but see myself pining for the critical sharpness from digital.

The shot from the Mermaid Parade is digital with Canon FD. The other three are 35mm, Canon A1 with FD glass.

 

[xayraa33]

All four shots look good and plenty sharp with FD glass.

 

[ifaynshteyn]

All four shots look good and plenty sharp with FD glass.

Thanks for the review! Perhaps it is just me. Perhaps it’s ‘resolution’. Perhaps the result will be different printing from negative versus scanning.

 

[ddutchison2]

As I understand it, digital sensors do not react to focused images in quite the same way as film does (non-retrofocus wide angles being a well known example) so digital image quality can change from what you'd expect with film.

 

[Darthfeeble]

I'll say it. Digital nowadays is sharper than 35mm film. Particularly if the film is scanned and printed from the scans. If you want to play hybrid, go medium format at least.

 

[ifaynshteyn]

I'll say it. Digital nowadays is sharper than 35mm film. Particularly if the film is scanned and printed from the scans. If you want to play hybrid, go medium format at least.

Okay, so I am NOT crazy/losing my eyesight/looking for a unicorn. Yes, I am also shooting 6x6 now and see a difference even in just the scan alone, from 35mm.

I think my solution is simple enough. Develop negatives, use scans to determine which ones I want to wet print from negative, and voila...I get to learn the darkroom while continuing to use my both digital and film workflow with vintage glass.

 

[peterm1]

I have frequently heard of people talking about the inherent sharpness of digital sensors over film. I have, however, never inquired further in any systematic manner as I pretty much no longer shoot film so I am not all that concerned about it. A factor which does come to mind though is that of course film sharpness will be affected by the type of film used - fast film being an example of a medium which is less sharp due to grain. Which goes to your point about film grain being a culprit. I recall that I used to use Ilford XP2 a lot for black and white photography and that it was known for the smoothness of its rendering and fineness of grain. Both of which could affect apparent sharpness.

I can say that I recently scanned a large number of old color slides from back in the late 1990's using a bellows outfit, a Sony camera with 24 megapixel sensor and a high quality 50mm lens at f8. I cannot recall being impressed with the sharpness of any of the slide shots even though I am sure the set up I used to digitize them was both well calibrated and used high grade equipment. Some of this was due to my crappy technique back in the 1990's when I was still a novice but even slide images that seemed sharp at first never impressed me as actually being sharp once I digitized them and examined them onscreen. I think we are used to a different standard of sharpness today even when using old lenses on digital.

 

[Coldkennels]

This is the key:

A factor which does come to mind though is that of course film sharpness will be affected by the type of film used - fast film being an example of a medium which is less sharp due to grain.

Film can be plenty sharp, even compared to modern mirrorless cameras... but it's going to depend on what you're using and how it's developed. There is a huge difference between Delta 100 in Perceptol, HP5+ in LC29, and Pan F+ in Rodinal, for instance.

There's usually a trade-off involved, though. Film speed, sharpness, and grain are all interlinked; you can usually have fine grain and sharp images with slow film, but if you want fast film, you'll have to decide which of the two other aspects to lose.

 

[ifaynshteyn]

I have frequently heard of people talking about the inherent sharpness of digital sensors over film. I have, however, never inquired further in any systematic manner as I pretty much no longer shoot film so I am not all that concerned about it. A factor which does come to mind though is that of course film sharpness will be affected by the type of film used - fast film being an example of a medium which is less sharp due to grain. Which goes to your point about film grain being a culprit. I recall that I used to use Ilford XP2 a lot for black and white photography and that it was known for the smoothness of its rendering and fineness of grain. Both of which could affect apparent sharpness.

I can say that I recently scanned a large number of old color slides from back in the late 1990's using a bellows outfit, a Sony camera with 24 megapixel sensor and a high quality 50mm lens at f8. I cannot recall being impressed with the sharpness of any of the slide shots even though I am sure the set up I used to digitize them was both well calibrated and used high grade equipment. Some of this was due to my crappy technique back in the 1990's when I was still a novice but even slide images that seemed sharp at first never impressed me as actually being sharp once I digitized them and examined them onscreen. I think we are used to a different standard of sharpness today even when using old lenses on digital.

This is the key:


Film can be plenty sharp, even compared to modern mirrorless cameras... but it's going to depend on what you're using and how it's developed. There is a huge difference between Delta 100 in Perceptol, HP5+ in LC29, and Pan F+ in Rodinal, for instance.

There's usually a trade-off involved, though. Film speed, sharpness, and grain are all interlinked; you can usually have fine grain and sharp images with slow film, but if you want fast film, you'll have to decide which of the two other aspects to lose.

Thank you both for your informative responses. I'll have to play with different film speeds and development to see what I like best. I've liked the speed of 400 films, and find the grain pleasant, but find myself seeking that clearly defined sharpness so I may try a roll or two of 100 and see how that looks!

 

[ifaynshteyn]

I have frequently heard of people talking about the inherent sharpness of digital sensors over film. I have, however, never inquired further in any systematic manner as I pretty much no longer shoot film so I am not all that concerned about it. A factor which does come to mind though is that of course film sharpness will be affected by the type of film used - fast film being an example of a medium which is less sharp due to grain. Which goes to your point about film grain being a culprit. I recall that I used to use Ilford XP2 a lot for black and white photography and that it was known for the smoothness of its rendering and fineness of grain. Both of which could affect apparent sharpness.

I can say that I recently scanned a large number of old color slides from back in the late 1990's using a bellows outfit, a Sony camera with 24 megapixel sensor and a high quality 50mm lens at f8. I cannot recall being impressed with the sharpness of any of the slide shots even though I am sure the set up I used to digitize them was both well calibrated and used high grade equipment. Some of this was due to my crappy technique back in the 1990's when I was still a novice but even slide images that seemed sharp at first never impressed me as actually being sharp once I digitized them and examined them onscreen. I think we are used to a different standard of sharpness today even when using old lenses on digital.

I think you hit the nail on the head with your last comment regarding different standard. I have been spoiled by my mirrorless camera, so the difference definitely threw me off. I will try a roll of 100 speed film to see how that looks.

 

[Darinwc]

A very fine grain film and a sharp lens matches about 36mp.
Also consider that:
1. Digital cameras have sharpening and grain smoothing algorithms baked into their processing chips. This makes a huge difference. Anso the algorithms used to generate the JPG's make a difference.
2. Film scanners are at least 10 years old. Algorithms have improved dramatically since then. Scanners and scanning software have not. A good TIFF or RAW scan processed in a current software package and moderate sharpening added would be closer to an average digital sensor.

So its not necessarily film vs sensor. Total workflow vs workflow.

 

[ifaynshteyn]

A very fine grain film and a sharp lens matches about 36mp.
Also consider that:
1. Digital cameras have sharpening and grain smoothing algorithms baked into their processing chips. This makes a huge difference. Anso the algorithms used to generate the JPG's make a difference.
2. Film scanners are at least 10 years old. Algorithms have improved dramatically since then. Scanners and scanning software have not. A good TIFF or RAW scan processed in a current software package and moderate sharpening added would be closer to an average digital sensor.

So its not necessarily film vs sensor. Total workflow vs workflow.

Well received. So many factors involved. I'll continue to play with my two workflows and learn from it.

 

[peterm1]

This is the key:


Film can be plenty sharp, even compared to modern mirrorless cameras... but it's going to depend on what you're using and how it's developed. There is a huge difference between Delta 100 in Perceptol, HP5+ in LC29, and Pan F+ in Rodinal, for instance.

There's usually a trade-off involved, though. Film speed, sharpness, and grain are all interlinked; you can usually have fine grain and sharp images with slow film, but if you want fast film, you'll have to decide which of the two other aspects to lose.

I previously mentioned how good I thought Ilford XP2 was in terms of relative sharpness - for a film stock. Just to reinforce how good some film can be in terms of sharpness I show this link (see below). It definitely looks like the images have been sharpened after scanning but, unless they were basically pretty sharp to begin with, they would not show as they do here. And for a film stock it is pretty damn impressive. Never the less, most prints and slides I have made with other film stocks back in the day do not compare with this nor with digital images at their best.

Ilford XP2 - My Go-To Black and White Film — Peter Jeffrey

I have finally found “the one”. The black and white film to compliment my color work on Portra 400, with it’s silvery highlights, shadow detail and deep blacks: Ilford XP2 Super 400.

www.peterjeffrey.photography

 

[Hilmersen]

Try adox cms20 with the recommended chemicals. I shoot it at 12 iso - and I am blown away by the resolution. Even when using a macro lens on 1/4 of the negative using a 24mp camera, it is obvious that there are more details in the negative. I am not using all that resolution, but I do love the tonality and how it handles the highlights: surpasses what I have achieved with mf and other films...except that dust are just as large.
A long time ago I scannned ilford delta 100 and compared it to a darkroom copy I made. The darkroom copy was sharper, and with less grain than the digital scan (using Nikon coolscan V) . I have also been experimenting with developers, finding that developers that are famous for their large grain actually make better scans than what I would have preferred if wet-printing.
That said I worry less and less about sharpness - to me it just isn't that important, and I only shoot handheld. ADOX CMS20II is still my default film...as I just love the tonality and freedom to use open aperatures without nd-filters.

 

[BernardL]

Apples and oranges.

To reach any conclusion you should shoot the same subject (framing, light, sensor size, aperture, tripod) in succession with a digital sensor and film. Do not reframe after focusing (some geometry involved here); choose your test scene such that the focus point is near center.

In the first of you four test pics, the main subject, in focus, takes up a major area, enhancing the subjective impression of general sharpness. In the second one (film, like the remaining pics) the DOF must be quite shallow, since the rear foot is visibly out of focus. If you compare the third pic to the first one, the main subject is not as close (which would other things equal, throw the background less out of focus) yet the background is more blurred than in the first pic. And in the fourth pic, what is in focus? maybe the top of the head of the reclined character?

Just a guess. Because digital allows it, when using auto ISO, you (or your camera's software) can use a smaller aperture and higher speed without a grain penalty. Both conducive to higher sharpness. This has to do with he noise performance of the sensor, not its resolution.

 

[dab]

One more thing to look into - how do you process your digital files? In Lightroom it set up Sharpness much higher then 0 by default. So you need to increase the Sharpness for your film scans as well to match digi- look.

 

[Coldkennels]

1. Digital cameras have sharpening and grain smoothing algorithms baked into their processing chips. This makes a huge difference. Anso the algorithms used to generate the JPG's make a difference.

This is something I hadn't even thought about.

Yep, if you're using a hybrid workflow (even if it's a lab doing the scanning for you), the digital files will need sharpening added in post.

As for direct comparisons, I did something similar to check dynamic range and exposure some years back. I used a 50/3.5 Elmar on a Leica IIIf and some (old, pre-Kenmere) Agfa APX 100 in Rodinal for the film, and an X-Pro 2 with Fuji's 35/2 XF lens for the digital exposure (I didn't want to use the Elmar on both cameras due to the crop factor of the X-Pro).

Here's the full frame of both:



It should be pretty obvious, even at this size, which is film and which is digital. But here's a screenshot of the two side-by-side at 100% in Lightroom at the point of focus:



The Fuji shot is an out-of-camera JPG. Both shots have the same sharpening settings. I was honestly a little surprised by how well the Elmar and APX 100 did in this comparison; it's very non-scientific as I was mostly testing exposure, but it does illustrate that film is still more than capable of producing good results in the modern age.

 

[ifaynshteyn]

Try adox cms20 with the recommended chemicals. I shoot it at 12 iso - and I am blown away by the resolution. Even when using a macro lens on 1/4 of the negative using a 24mp camera, it is obvious that there are more details in the negative. I am not using all that resolution, but I do love the tonality and how it handles the highlights: surpasses what I have achieved with mf and other films...except that dust are just as large.
A long time ago I scannned ilford delta 100 and compared it to a darkroom copy I made. The darkroom copy was sharper, and with less grain than the digital scan (using Nikon coolscan V) . I have also been experimenting with developers, finding that developers that are famous for their large grain actually make better scans than what I would have preferred if wet-printing.
That said I worry less and less about sharpness - to me it just isn't that important, and I only shoot handheld. ADOX CMS20II is still my default film...as I just love the tonality and freedom to use open aperatures without nd-filters.

Don't get me wrong, I'm learning to like the look of Tmax400 and HP5+. I love the tonality and the soft grain. I was just making sure I wasn't doing something wrong during my shooting workflow causing the stark difference. That being said, I do need to just take my negatives and darkroom print them rather than just relying on the TIFF scans I am getting back from my lab. Part of the journey!

 

[ifaynshteyn]

One more thing to look into - how do you process your digital files? In Lightroom it set up Sharpness much higher then 0 by default. So you need to increase the Sharpness for your film scans as well to match digi- look.

For digital, my process is actually very simple. I use Photoshop CC for SilverEffex plugin. I rarely if ever sharpen. As for the film scans, the more I think about it, the more Im considering a dedicated film scanning setup, whether it be through a digital scan or a flatbed scan. I would love to get the most out of my negatives.

 

[ifaynshteyn]

Apples and oranges.

To reach any conclusion you should shoot the same subject (framing, light, sensor size, aperture, tripod) in succession with a digital sensor and film. Do not reframe after focusing (some geometry involved here); choose your test scene such that the focus point is near center.

In the first of you four test pics, the main subject, in focus, takes up a major area, enhancing the subjective impression of general sharpness. In the second one (film, like the remaining pics) the DOF must be quite shallow, since the rear foot is visibly out of focus. If you compare the third pic to the first one, the main subject is not as close (which would other things equal, throw the background less out of focus) yet the background is more blurred than in the first pic. And in the fourth pic, what is in focus? maybe the top of the head of the reclined character?

Just a guess. Because digital allows it, when using auto ISO, you (or your camera's software) can use a smaller aperture and higher speed without a grain penalty. Both conducive to higher sharpness. This has to do with he noise performance of the sensor, not its resolution.

Bernard,

All valid points. The film shots were all shot between f8 and f11 with different distances and situations, so hard to replicate identically. That being said, I have done general tests like you said, same subject, tripod, same settings both digital and film, same lens. The film scan still ended up looking less "sharp" to my eye than the same scene rendered on my mirrorless.

On the street, I tend to shoot zone, usually at f8 unless light dictates more stopped down, usually focused at hyperfocal distance or at least 6ft (for the most latitude front and rear).

Signed:
FrustratedByStreetPhotography