NB. This article was written in 2010 - things have moved on a lot since then and a mid-range smartphone in 2020 produces sharper results than either of these. I enjoyed writing this article however, so I decided to keep it when I moved my site over to here.
I'm very happy with digital photography, and wouldn't ever want to go back to 35mm.
However, I've found a few sites recently that have comparisons between film and digital photos and been inspired to test this out myself.
For this experiment I shot several subjects on a manual SLR and a digital camera within a few minutes of each other. The film was processed and lab-scanned so that it could be compared with those from the digital camera. The negative scan size was 7200 x 4800 pixels, which is the best I could find online at a price I was prepared to pay. This turned out to be overkill, but I didn't want to leave the possibility of wishing I'd paid a bit more, and it gave the film the best possible chance of performing well.
The equipment and consumables used were as follows:
Canon Powershot SX20 IS (12mp / 20X optical zoom) - all photos taken with ISO100 setting, except as noted below.
Fujica st705w fitted with max F1.8, 55mm 'Standard' lens
Hoya HMC max F2.8, 135mm 'Telephoto' lens
100ASA Fujicolor 35mm film
As I only had fixed focal length lenses available for the SLR in 55mm and 135mm sizes, I tried to adjust the digital camera's optical zoom so that the overall area of the subject was roughly similar.
(Obviously a Digital SLR with similar lenses would be a closer comparison, but as I rejected that route for myself in favour of a 'prosumer' model, this is the best I could do)
So why do some more comparisons when there are several out there ? Well, the best ones I have seen aren't quite what I'm looking for. These are my favourites:-
- Ken Rockwell's "Film vs. Digital Cameras" shows how 4"x5" film can completely blow 35mm film and DSLRs out of the water.
Definitely eye-opening, but most of us have never considered using a camera the size of a bucket that costs £3 per negative before it's developed.
- Michael Reichmann's balanced essay "Like It Is"
- Roger Clark's fairly technical "Film versus Digital Information"
As good as these articles are, they don't really feature enough comparison images to show the pros and cons of digital and conventional photography. I wanted to see the full-sized images to gauge their overall appearance as well as close-ups of selected details. I was also keen to see how the two systems coped with different lighting conditions, such as long exposures, heavily back-lit subjects, artificial and natural light, and sunny/cloudy days.
NB. The film was developed and scanned by Fuji Digital Imaging using an Agfa d-lab 2 minilab, and I was really pleased with the speed of service and degree of personal attention given.
Here is the final list of 28 photographs and 23 subjects / lighting conditions that I compared.
Comments and Conclusions
Here are my impressions. Note that the 7200x4800 film scan image will take you to the relevant image on Flickr and you can download it or view it there by clicking on the 'All Sizes' option.
The links in the left column take you to a framed page that lets you compare the images side-by-side.
| 01 Standard|| ||The digital image is a clear winner in terms of sharpness, though I think the lighter exposure on film creates a more balanced photo. The different exposure results in certain details being clearer on film and others being better on digital. For example, the windows in the house appear to have net curtains on the film image but this may just be the effect of grain - I'm not certain even looking at the original scan. However, the sign on the children's play area is over-exposed on film, whereas you can discern some detail on the digital image.|
| 01 Telephoto|| ||The digital image shows much more detail than the film scan.|
Worthwhile details: the fence at far left, the notice board on the play-park, the tops of the trees, and the connecting barrel in the middle of the climbing frame.
| 02 Standard|| ||Detail, colour, depth of field are all better on the digital image.|
Worthwhile details : anywhere!
| 02 Telephoto|| ||Detail, colour, depth of field are all better on the digital image. (Try reading the label on the teddy at far right)|
| 03 Standard|| ||Detail, colour, depth of field are all better on the digital image.|
| 03 Telephoto|| ||Detail, colour, depth of field are all better on the digital image. (Look at the markings on the penknife blade)|
| 04 Standard|| ||Not exactly sure what happened here but the 35mm image is a disaster, probably focus if you look at the white rail on the right. Digital Win|
| 04 Telephoto|| ||Colour, sharpness, depth of field are all better on digital.|
| 05 Standard|| ||This one is fairly close. I'm surprised the sky detail was not better on the film scan. Overall, I can see more detail and better colours on digital.|
| 06 Telephoto|| ||Close again but digital is clearly sharper and more detailed.|
| 07 Telephoto||Not much to choose between them in sharpness, but I think the colour details are better on film. Have a look at the light reflections on the red cars, the brick work on the quay near the water, and the two wooden benches to the left of the grey car. 35mm Film Win|
| 08 Telephoto||Sharpness and detail are better on digital, but the colour tones look better on the film scan. In particular, where have the clouds gone on the digital photo ?|
It does appear that the sun is edging behind a cloud on the digital shot; otherwise I'd have to give this one to the film.
| 09 Standard|| ||The difference in exposure makes it hard to evaluate these two. However, the film is clearly better at coping with the backlit tree than the digital which show complete white-out.|
If you look close to the base of the tree on the digital image where it backlit by the reflection off the water, it appears to narrow. There's also a fair bit of purple fringing (which is mainly a lens effect).
35mm Film Win
| 10 Standard||I'm surprised how different these images are. Surprisingly, the digital camera has coped much better with the range of light conditions through shade to sunlit areas. The film image looks harsh and the sunlit areas are washed out.|
| 11 Telephoto|| ||Fluffed both these shots. Exposure was not fast enough to avoid camera-shake, and the focus isn't perfect on either one.|
| 12 Standard|| ||Not much to choose between these two. The digital image is a bit more detailed, but I prefer the colours from film.|
| 13 Standard||Exposures are quite different again, though the digital image will take more alteration in xnview to match the look of the film version than vice-versa. The digital image has more detail and generally looks sharper so it's another..|
| 14 Standard|| ||This was a shot I expected the film to do well at but it seemed very over-exposed and appears quite grainy.|
| 14 Telephoto|| ||As above.|
| 15 Telephoto|| ||I was really pleased with the photo taken on digital but they couldn't get a sensible scan of the film one unfortunately. If you read some comparisons, long exposures are supposed to be something that film does really well, but this isn't my experience.|
| 16 Standard||At last a night shot that worked OK on film, but the digital image has much nicer colours and appears sharper.|
| 17 Telephoto|| ||I expected the film to do well here, and it did preserve the lace curtains which are white-outs on the digital version. However, the film image was under-exposed and had to be brightened considerably, so this will have contributed to the retention of the lace curtains. Digital is a clear winner when you zoom to 100%.|
| 18 Standard||Very little to choose between these images - similar levels of detail.|
| 19 Standard||Detail levels are very similar but the digital image looks very harsh and over-exposed.|
35mm Film Win
| 20 Standard|| ||Slightly unlucky here because the light changed due to cloud movement. Overall the digital image is much sharper.|
| 21 Standard||I prefer the film colours but the detail and sky is better on the digital image.|
| 22 Standard||A close run but the digital image edges it due to the sharpness|
| 23 Telephoto||The writing is on the wall (literally !) - the digital image can resolve the 'LADIES' sign between the two red doors. However the film looks pretty good as well - the bricks and slates have nice textures (though this could just be a grain effect)|
So there you have it; six ties, three better captures on film, and nineteen photos that I think are better on digital. That said, I am a bit rusty at using film, and no doubt it would be possible to get better scans of some of the film negatives individually. On the other hand, I can get similar images on digital with a faster ISO setting, and I would expect the film to suffer increased grain with a faster film. A comparison with 400ASA film(*) would probably be a lot more one-sided in favour of digital, whereas 50ASA Velvia might have performed better than the digital camera's minimum 80ISO. I don't think this is that realistic however - some of my images suffered in the comparison due to lack of depth of field on the 35mm SLR. Ideally, I would have liked to have been using a higher f-stop to increase the depth of field but this would be even harder with a slower film.
* I realise I'm being inconsistent here since 100ASA = 100ISO, but before digital we always referred to 'ASA' for film speeds, and nowadays we only speak about ISO settings for digital cameras !
A couple of the images highlighted the digital camera's wide depth of field. This is due to the sensor size being significantly smaller than a 35mm film negative. If you repeated the exercise with a digital SLR, the sensor size would be a closer match to the film and the depth of field similarly reduced at large apertures. It's also likely that a dSLR would show higher definition and less noise since the larger sensor would improve both and a shorter zoom or fixed lens would be expected to perform better than the SX20 IS's 20X optical zoom.
One person has commented that the film scans would be better if done individually on a Nikon film scanner, or by professional drum scanning. I'm sure this is true, though I still doubt it would be that significant in terms of detail levels. The suggested scanner costs about £1500 in the UK, and drum scanning seems to come in at around £15 per negative - a bit beyond the scope of this article! Nevertheless, if I get the chance to have a few of the negatives rescanned on better equipment, I'll add these examples too.
This exercise has demonstrated to me that film really isn't dead. Who knows, I may even run another film though my trusty old st705w some time in the future!