JakTrax From United Kingdom, joined Jun 2005, 4936 posts, RR: 6 Posted (4 years 2 months 2 weeks 7 hours ago) and read 5127 times:
Digital has often been fiercely compared to slide, but I've not seen too many debates comparing it directly to 35mm film. I've heard quoted that slide is around 21mp in digital terms, and I was curious as to what you guys think is a fair mp assessment of film (Fuji Superia 200 and Kodak Gold 200 tended to be my preferred choices) by a similar comparison.
Also, do you think the equivalent mp count would be much greater if the negative is scanned, rather than a paper print?
Here are a couple of examples of mine, scanned from prints...
How do you think these particular images square up to those taken a few years back with, say, 6mp DSLRs? Also, do you think, with the right editing, it is possible to tart an old scan up enough to compete with today's standards? I know some scanners are way better than others, but for reference the above was scanned using a Canoscan 4400F.
cliffak From Sweden, joined Aug 2011, 77 posts, RR: 0
Reply 1, posted (4 years 2 months 2 weeks 6 hours ago) and read 5109 times:
A good scan can look very good even with cheap consumer films. Scanning the prints is nowhere near as good as scanning the negs/slides.
A properly exposed film shot taken with a good lens will blow away a 6MP DSLR shot if scanned correctly. After all, with digital you're stuck with whatever resolution came out of your camera while it's entirely possible to scan the film on better equipment and get better quality out of it (up to a limit, of course). When most people think of film shots as low quality, they're usually thinking about all those snaps shot with cheap, poor (optical and mechanical) quality cameras and lenses.
I see film and digital as different mediums with their own distinct characteristics and advantages, it's simply a canvas on which to produce an image. Just as I love digital for its convenience and clean images, I also love film for its much more controlled handling of highlights (especially reds) and subtly different rendering style. For an aviation-related example, if I were to shoot classic planes I would not hesitate to grab my F5 and a few rolls of B/W. If pretty much everyone else at such an event shoots color digital, it's fun to walk away after a day's shooting with something different that will stand out from what the next guy shot.
JakTrax From United Kingdom, joined Jun 2005, 4936 posts, RR: 6
Reply 2, posted (4 years 2 months 2 weeks 2 hours ago) and read 5081 times:
I was really happy with the quality of the two examples above. Took a lot of work in Photoshop (way more than a standard digi shot) and Neatimage and it's surprising how different the end product looks compared to the original. Unedited scans look pretty awful to be honest!
What sort of quality increase over my two examples do you think scanning the negatives would offer? What would be a comparable mp count in digi terms?
vikkyvik From United States of America, joined Jul 2003, 11659 posts, RR: 25
Reply 3, posted (4 years 2 months 2 weeks ago) and read 5070 times:
I don't have much experience with film cameras (I had a cheap one when I was a kid), but I did do some scanning of family photos at my parents' house earlier this year. They had bought a scanner (can't remember what model) for scanning old slides and photos that they have.
I was pretty impressed by the scan quality I got, considering that everything I scanned was a print, and most were in the 3x5 or 4x6 range. Of course, I scanned them at high quality, and did some Photoshop touchups when I got back to my apartment (dust removal, a bit of NR, and some sharpening).
One thing that I noticed that surprised me is that the black and white scans are much better looking (in my opinion) than the color ones. Don't really know why, but it's made me think about doing some shooting in black and white.
Here's an example that I think is the best out of that batch of scans:
ckw From UK - England, joined Aug 2010, 813 posts, RR: 14
Reply 4, posted (4 years 2 months 2 weeks ago) and read 5063 times:
Difficult to say, but as a general rule, photography is a "lossy" process - each stage (whether digital or film) from exposure to final print involves the loss of some data. Hence, all else being equal, a scan of a negative should contain more data than a scan of a print produced from that negative.
But things are seldom equal - are the film and print scanners of equal capability (and scanners designed for both operate differently for each format)? Can the negative be held absolutely flat in the scanner? Do you know how to get the most out of the negative? - in general its harder to get optimum results from a neg than a print.
What is the equivalent mp to a scanned neg? Impossible to say - it depends on the scanner, and of course the film. There also seem to be many ways of counting 'pixels' on film (remembering of course that these don't exist - it is a theoretical arguement based on how many pixels can be extracted). I've seen figures ranging from 20mp to over 80mp. And of course are the pixels 'meaningful' - how much of that data is noise?
Perhaps a fairer way of measuring is to use line pairs per millimeter (the ability to resolve distinct line pairs) to compare resolution. In theory a top quality film like Velvia is capable of resolving around 160lpmm. The Canon 5D's 12mp sensor can record 61lpmm. Film wins right? But then consider that a top quality consumer lens can probably manage 60lpmm at best.
Ultimately, while you can measure a specific aspect of film or digital - and depending which aspect you measure, either may come out as 'better' - the final image is a composite of all aspects. I think we have got to the stage where such comparisons are pointless. Film and digital are have a different look ... its like comparing oil paintings and water colours. However good a scan is, it will never look like a quality digital print, and nor can a digital image look exactly like a quality film print. Perhaps the biggest mistake is to try an make a film scan look digital!
There is also a matter of what you, as an individual, find acceptable. Both film and digital are abstractions of the 'real world'. Some feel that digital looks less natural, though I suspect this is changing as more people have had more exposure to digital than film, and now film is becoming perceived as more 'arty'
You see exactly the same sort of arguements over digital and analogue sound recordings.
My bottom line - a well scanned negative can look wonderful. Personally I much prefer film for B&W ... ultra clean digital B&W just doesn't work for me. In time people will stop worrying about comparing the two media altogther and just choose the media that looks right for their work.
cliffak From Sweden, joined Aug 2011, 77 posts, RR: 0
Reply 5, posted (4 years 2 months 1 week 6 days 5 hours ago) and read 5003 times:
Quoting ckw (Reply 4): Perhaps a fairer way of measuring is to use line pairs per millimeter (the ability to resolve distinct line pairs) to compare resolution. In theory a top quality film like Velvia is capable of resolving around 160lpmm. The Canon 5D's 12mp sensor can record 61lpmm. Film wins right? But then consider that a top quality consumer lens can probably manage 60lpmm at best.
To complicate matters more, resolution also depends on aperture. I've seen very high diffraction limited resolution figures quoted for some pro lenses, but then again you need an extremely good sensor (or film) to really see it.
Quoting ckw (Reply 4): Personally I much prefer film for B&W ... ultra clean digital B&W just doesn't work for me. In time people will stop worrying about comparing the two media altogther and just choose the media that looks right for their work.
Oh yes.. I really like processing my own B/W, pulling a strip of freshly developed negs out of the reel provides a completely different feel of accomplishment compared to loading a bunch of RAW files into Lightroom.
Both great all round films that I tended to use myself. BW I opted for Tmax 400. No matter what I do, however, I just can't seem to nail my print scans. They need a heck of a lot of NR and the overall final result is a photo that's way too smooth and inherently colour casted, which I try to remedy but don't seem to be able to 100%.