Subjective Observations on Print Size
The last thing I want is to start another useless film vs. digital debate. I did, however, want to share the results of a recent experiment that may be of interest to many of you who struggle with the issue of “how large can I print?” This seems of special concern to those coming into digital capture from large format film (as I have).
I originally intended to experiment with new mounting and presentation techniques for mural-size prints. To that end, I made a series of 30×40″ prints for my own personal use, with the intent of sending them to a couple of service bureaus for mounting using different methods.
Since I planned to hang the prints on my own walls, I picked a few personal favorites I didn’t yet have on display. It was a mix of 4×5 film scans and digitally-captured images.
In the past, I subscribed to the common wisdom that up to a given print size the difference between digital files and film scans will be a wash but beyond a certain threshold, the large format scans will consistently have the upper hand. I was surprised to find (by my own subjective judgment) that this was not the case and in fact at this size some of the digital files actually produced sharper-looking prints.
This seems to contradict not only common perception but mathematical calculations (which I generally try not to argue with.) The explanation came to me as I was working on the files, preparing them for print. The raw 4×5 scans were indeed significantly larger but as anyone who worked with large scans knows, straight out of the scanner the files looks gritty; and lines, detail, and transitions appear a bit “fuzzy”. In comparison, a good digital capture, albeit containing less pixels, is very clean with clear detail and well-defined lines.
There is a difference between resolution and sharpness, the former being actual measurable detail and the latter being perceived/subjective appearance of detail. Until recently I thought they were well correlated but I’m no longer convinced.
At large print sizes, the digital files lend themselves well to interpolation while retaining the appearance of smooth lines and detail. The film scans, on the other hand, becomes fuzzier and harder to sharpen.
I now believe that rather than a clear threshold, there is an interim area where large format film still has sufficient detail and require little interpolation and can appear sharper than the digital capture but beyond that, where both formats require significant interpolation, the digital capture gives a sharper appearance despite having somewhat less detail.
In the coming few weeks I will ask friends who share my experience with various formats to evaluate these prints and see if they share my subjective impressions.
For now, though, I wouldn’t be too quick to assume that a large format scan will necessarily produce a better print. I will also think twice about the need for higher resolution digital capture (medium format or future camera upgrades).
This is one of the prints in the set. It is cropped from a file captured using a first-generation Canon 5D. To my eye (and surprise) it turned out to be the sharpest-looking print of the batch.
Category: Images, Printing, Technology and Technique
Guy,
I wonder if the reason for this is that with the digital camera files you’re starting with pixels from the outset; whereas with film you’re starting with grain and converting that to pixels. Sensor pixels are entirely consistent. Film grain is much less so and converting a less consistent starting point may exacerbate the ‘errors’ inherent in interpolation.
I recently had an experience similar to yours. I took an image from a 5D and found a small section of it that I felt would look good as a crop. This was a section that I rezzed up about 2 1/2 times to generate a 14×22 print at 180 ppi. I was surprised at how well detail was retained.
Great thoughts Guy. I recently printed several 40 x 60′s for a client from 20D files and even those look surprisingly detailed at that size. From two feet back they looked amazing but more surprising was that when I stuck my face up to the print, it didn’t look like crap either.
Guy, great article – and thanks for taking on some dearly held beliefs about format and print size. Like you, I think things are a bit more complex that “in format A you can print at size X, but in format B you can print at size Y.”
To some extent I think that film and digital images “break down” in different ways as they get larger, perhaps creating different upper size boundaries from a given original size and perhaps also how rapidly the image “fails” above the “boundary” size. (Roughly, I’m thinking that digital images – from good captures – can go larger than you might expect but that once you get just beyond whatever the upper limit might be that they fail in more obvious and unpleasant ways… if that makes any sense.)
Dan
This is nice to hear. I’m planning on making a 19×24 print from one of my D300 captures to see how it stand up.
BTW, I love that photo that you posted. It will be an awesome large pring.
- Dan.
Is there a site where I can see the comparisons? I shot 4×5 for 20 years before reluctantly switching to digital. So far my digital files can’t even come close to my 4×5 scans. I’m desperately trying to upgrade from my D200 ($ issues you know) and maybe your results can help me decide which camera to go with. Thanks.
I don’t have much experience with film scans, but I can say that I have printed as large as 42×60″ from a little Canon Rebel XT (can’t remember what the MP of that camera were). I thought the large print was equal to another copy printed at 16×24″. When comparing the two side by side I couldn’t see any difference.
So far in my experience, the large format scans look a lot more detailed in larger prints than the digital images I can get with my D300. I’ve seen digital images prints large that look sharp from afar, but they get fuzzy when you approach to read the caption… the large format scans I have printed don’t.
I’m looking forward to getting some of those 4×5′s drum scanned…
I know you don’t want to start a film vs. digital debate, but as is the nature of those debates…this can feed into the debate for either side. I, of course, am of the belief that either could be made to be just as sharp if the interim process (from film/file to print) – a process riddled with human interaction – is handled with enough care, wisdom and patience. There is no exact science to doing scans or sizing up digital photos, but there is certainly a level of finesse that must be applied. I’m not saying your scanning process is wrong…I certainly don’t believe that’s the case…but as you experiment more, you may find some things help to improve the quality of the scans. Be it different scan settings, some post-processing filters or so on.
I guess the point comes down to process…one can make beauty and/or crap out of anything.
Thank you for these great responses!
I did want to emphasize that this is a subjective evaluation. I don’t believe there are universal metrics for better/worse when it comes to print quality. Unfortunately there is no practical means for me to offer a way of judging the results for yourself short of seeing the actual prints, and I do plan to solicit some input from friends.
I also agree wholeheartedly that the processes, tools, and operator skill play an immense role in the quality of the outcome.
Guy
“one can make beauty and/or crap out of anything.”
I think that’s probably the best answer to the film vs digital debate in the end, but for some reason it’s also the one answer that gets the least positive response, because too many photographers focus so much on their gear that they forget about the art.
Wow… outstanding image! Truly out of this world.