debate 2008: digital vs. film quality
*Let me preface this post with a reminder that Epic Edits Weblog is having their very cool $50 Film Camera Project, running through mid-September. Shooting film can be a great way to break your usual photo routine.
There seems to be some curiosity about shooting film among younger photographers and it’s reviving the medium (just when you thought the film vs. digital debate was dead, the bones get dug back up!). I enjoy shooting film but never, ever made a head-to-head comparison of film vs. digital on the same day / subject. I’ve been a bit curious so this weekend was a good occasion to have a little imaging shootout.
The truth is, a client hasn’t paid me to shoot film in over four years. It’s been all digital, all the time at the studio. Film is considered expensive to produce, purchase and process. There’s also an environmental cost (though we’re not necessarily ahead of the game if all of these digital cameras are destined for landfills). Black and white film and paper materials are on the wane and concessions in formulas are continually being made. But there’s something about film and those solid film cameras that I love.
This test is intended to compare sharpness and rendering detail of a full-frame Canon 5D, a 12.8 megapixel camera, against an aging Nikon 35mm film camera loaded with color neg film. A big question has to be dynamic range, an area where color neg film is considered superior. And what about color? I’m curious about those things, too.
To be honest, I don’t know how to test those things outside of a lab, where all conditions can be measured equally. But the real world is not a lab, the real world is software pushing and pulling on dynamic ranges and various ways to enhance all of these settings. The dynamic range of digital images can be expanded and the range of film images can be compressed. I don’t know how to draw a conclusion other than to suggest trying this test for yourself under the conditions you normally shoot and process under.
On the left side of the ring is a Nikon FE, circa 1981, on the right side a Canon 5D, circa 2007. Each are sporting standard 50mm lenses, a Nikkor 50mm f2.0 on the Nikon and a 50mm f1.8 EF on the Canon. The Nikon is noticeably smaller and lighter, despite the entirely metal camera body.
Nikon settings: Kodak Gold 200 color neg film, straight from the corner drug store shelf. I would have liked to use ISO 100 film for maximum detail but found that it’s hardly available. Instead, let’s go with something more “real world” like the 200.
Just like digital images, film doesn’t tolerate underexposure as much as overexposure, so this was shot at ISO 160. As a matter of fact, overexposing C-41 film a little tightens the grain up a bit, just like choosing a lower ISO with a digital camera. The image was scanned with a Nikon Coolscan V ED scanner using Digital Ice to mask any dust or scratches. The cool thing is the neg required no dust spotting at all (the 5D did a little). 1/500 sec @ f8.
Canon settings: ISO 160 to equal the film camera, 1/500 sec @ f8. The Canon image came out of Adobe Camera Raw at 4,368 x 2,912 pixels at 240 ppi, so that’s what the neg was scanned to, also.
An interesting note: the viewfinder on the 5D was maybe the slightest bit brighter but it was also considerably smaller. Manual focus with the manual focus camera was easier. Autofocus was, well, not an option on the FE, of course.
The results? Well, the Nikon Coolscan scanner is SHARP and ruthless. A cheaper scanner wouldn’t have shown this grain but also wouldn’t have shown the detail. You can find the full sized images available on Flickr if you would like to check out the fine details.
 |
 |
| Kodak Gold 200 35mm film | Canon 5D Digital |
 |
 |
| Kodak Gold 200 35mm film | Canon 5D Digital |
To my eye, the film renders fine detail but in a hail storm of grain at 100%. The 5D image looks very…digital. After years of shooting push-processed Tri-X I’m perfectly happy printing images at 6″ x 9″ at 300 dpi, so the grain is not nearly as noticeable at that size.
Film has grain and I’m used to it. I like film and enjoy the experience of using my old Nikon FE. It forces me to slow down and that’s not a bad thing, sometimes. But it’s not for everyone. Hey, that’s not a bad thing, either.
Hi,
Nice comparison. However I don’t think its a fair one. Kodak Gold is not exactly a match for highly developed imaging chips combined with smart image processors.
Its fact that digital imaging is the present and future of photography. They reduce the costs, are highly flexible and even in terms of quality and resolution they will soon match even the best film material. No matter if we want to shoot still images or video, they’ll do the job.
Yet I think film material will ever be teplaced completely. Especially for artistic reasons, in B/W photography or simply because of the low equipment prices, it will remain, even if its just for fun
cu
Roman
Great post, thanks for that.
Most of the time I use my Nikon DSLR but for 2 month now I try to use again my Leica R3 (film camera form the late 70’s) and I love it.
The viewfinder is much better (brighter, larger …) and the 50mm F2 lens is sharp as a lense can be. The only drawback would be the weight. I swear this camera is made of lead but it feel unbreakable.
I think going back to the root of photography with a film camera (no zoom, no auto focus, patience) make you a better photographer at the end.
btw, congratulation for the videos, really appreciated.
I’d be much more interested how these two pictures would compare in print, since I find print to be much more interesting than pixel-peeping.
You don’t happen to have printed and compared those? Best would be to show both printed pictures to others and let them choose which they like better…
Thanks for that. I was *pushed* away from film about five years ago when my processing lab ruined 8 rolls of Velvia that I had shot on vacation in England. I bought a D70 a few months after its release, and have not put film in the F5 since. I still have my Coolscan 8000 and I still remember the long, long days of waiting for the scans to complete. Can’t say that I miss that part. I do miss the feel and smell of film, though.
Regarding environmental impact, as I *out grow* my digital cameras I pass them along to my children or relatives. My daughter is now the proud owner of my D70 and an in-law actually had a need for a little 2 mpxl Nikon P&S that I won from Nikon a few years ago. He records the condition of motion picture rental equipment as it leaves and is then returned to his facility. I think an idea (especially for P&S cams) would be to donate to youth organizations to foster an interest in the creative arts within the community. We may outgrow the cameras, but someone else may be able to grow into them.
Delos
To be fair in the comparison though, Kodak Gold is hardly a worthy film. I mean if you’re going to compare it to a 5D a pro level film might have been a better comparison. For example Kodak Portra is rumoured to have around 8 stops of exposure latitude! The grain is quite light too. Though you did say that even 100ISO film was difficult for you to source.
Interesting how the two shots came out similar in colour. Did you process the film scan much? I always find I don’t need to do anything from scans on the Coolscan as it seems to get accurate colour renditions.
Thanks, all, I’m glad you found the comparison interesting. I see now that I chose Gold 200 for personal reasons: if I’m going to shoot film, this is a film I’m very likely to shoot. It’s really (really) inexpensive and readily available, unlike the pro neg and slide films.
Ahhhh, yes, prints. I haven’t printed them yet and that’s a great point. There’s another project!
The film scan was hardly processed. I’m not an expert with the Coolscan but have found two things: (1) I don’t know how to get complete control over the scanned images yet and (2) it doesn’t seem to be necessary because it does an incredible job on it’s own.
Thanks Jim for the comparison. I’ve mentioned before that I was interested in shooting film and hope to in the near future. It won’t be for a while, though, as our film SLR is being used by my niece for a photography class. It’s an inexpensive Canon Rebel that my wife bought for college. I think it has a built in light meter and it definitely has an auto focus system.
It would be fun to try out one of the classic manual film SLRs at some point, but I still have a lot to learn about photography and exposure. That will keep me real busy for quite a while.
Ah..that reminds me, I still have a batch of unscanned Velvia slide film from 2 years ago. Should probably scan that before the film goes bad.
I gotta second the “unfair comparison” comment re Kodak Gold. The 5D is hardly a “run down to the drugstore and get what you can find” type of camera. So your film shouldn’t be either. You’re using a pro level camera, and comparing it to a consumer level film. Perhaps it would have been fairer to get a point-n-shoot digital camera to compare instead?
If you were for example doing a studio shoot, you’d probably be shooting chrome rather than neg film (never mind ‘Gold’
). Perhaps another try, using some mailorder Provia or something?
My problem is always my scanner. I use an Epson 4990, and the resolution isn’t so hot. It’s either that or have the lab scan it ($$$$).
As for the environmental and monetary costs: you have to factor in all the computers and peripherals one uses (and disposes of) when calculating the true cost of digital. Film never had that: you had chemicals, and you or your lab had a sunk cost of enlarger and other gear. Which never needed upgrading! Not to mention the film cameras, which don’t need upgrading and take much longer to grow obsolete. How much have you spent on cameras, computers and software of the past five years? Film doesn’t seem quite so expensive after that.
Mind you, I shoot digital because it’s convenient, and film because I love it. If there were some way for me to quickly and cheaply get scans of my film, I’d shoot it even more.
Disappointing article.
Kodak Gold is not on par with a 5D. You should be comparing Kodak Gold 200 w/ a 200 ISO image from a Rebel.
Or, Fuji Velvia II vs. 5D @ 100 ISO.
Others have touched on the unfair comparison, but I have a different take.
With film one of the first steps is choosing which film to use. It is the start of the creative process. You also don’t hear people say very often that they will fix something in photoshop. Film requires discipline.
With digital everything is done post exposure. They don’t call it RAW for nothing
The dynamic range with digital is getting better with each generation of sensors. Blocky shadows and blown highlghts are getting to be less of a problem. Unfortunately optical scans of a negative have the same problem. A real comparison is just not possible using this method.
Matt, I agree totally on the cost of digital photography. Factor in cameras, computers and software and it certainly costs more to operate a commercial studio now than in the film days. Film was an easy profit center, the cost was passed on to the client with a markup. Yes, we charge digital fees but we’re also investing a whole lot of time after the job that must be recovered. We’re now the photographer, photo lab and color separator, all rolled into one.
Conversely, wedding photographers that shoot a lot do consider digital a savings, since film and processing costs are customarily rolled into package prices. I was told by one photographer that they shoot 4000 images a wedding and consider digital a huge savings.
It looks like my film vs. digital comparison wasn’t deemed fair and I certainly understand - the 5D is a professional camera and the Gold 200 is an amateur film. I wanted to use a full-frame camera for comparison (50mm lens vs. 50mm lens) and wanted to use a film I was likely to pay for and shoot. Here in my home office are stacks and stacks of binders containing Fuji Velvia slides shot over the years. No doubt comparing Velvia 50 would have yielded a very different result.
Incase you are looking to compare dynamic range, from my experience with black and white negatives, the Nikon Coolscan V software clips shadows and highlights to a huge extent. Switching to Vuescan software gave me a lot more detail (and since I don’t do anything with my negatives aside from scanning, that basically translates to an extra stop of film speed
). I’ve also read several suggestions that the positive mode in Nikon Scan doesn’t do that clipping either, just use Photoshop to invert, but haven’t tried this myself.
And as to the Kodak gold 200, just had a look at Calumet’s website, they’ve got a lot better stuff! Even got Kodachrome still! Although its nice to see the cell phone camera of the film world stand up against a 5D.
Thanks, Marcus, I’ll be sure to explore Vuescan, that sounds great. I’ve never fully figured out the Coolscan software for controlling highlights / shadows.
Wow, Kodachrome brings back memories (there’s still some in the ‘fridge). It seems like yesterday there were Kodak labs all across the country. Now we’re down to one Kodak certified Kodachrome processor in the U.S.: good ol’ Dwayne’s Photo in Parsons, KS. Hmmm, I better shoot that stuff fast…
Interesting article. I’d like to try film again but there is only one Kodak lab left near me and the last two rolls I gave him years ago he ruined — I think he processes film so rarely he has forgotten how to do it properly.
All he does these days is digital prints — and he sells handbags, belts and clocks as well in his shop. No other way to stay in business it seems.
Film shall live! Digital must die! Just kidding, but Kodak Gold is not the greatest film out there, so it doesn’t show the best result. Another thing I found, for emerging photographers striving for top quality, film medium format might be a good solution, it’s way better quality than digital SLRs, looks cool and just kicks ass:) Plus it doesn’t cost 30+ thousand dollars. Cheers
P.S. Jim, please take my apologies for not calling you back last week about lunch, I got slammed with work and still can’t catch up. It’s 12-51 and I made a little break from PS.
Hmm… Film Vs. Digital Comparison…
I think this was a good idea… but using Kodak Gold 200 Vs the Canon 5D is not a good idea. Comparing a low-end pro camera to a cheap common film is not something that should be done…
Why don’t you write a follow up with a high end film that’s stored and processed correctly.
Looking forward to that.
BTW, i’ve not looked at the comments till now… just looked up and noticed that people are talking about your choice of film. Do consider the follow up
Cheers!
The people have spoken! Expect a rematch next week: same format, same subject, 2 DSLRS (full-frame and APS-C), professional C-41 film, amateur C-41 film and professional slide film. See you here next Tuesday with the results (to be continued)…
Don’t forget to make and compare prints!
Oh, no, I knew I was leaving something out! I can’t promise gallery quality prints but I’ll print equal sections from each and scan them in for comparison. This may be spread over a couple posts!
Great! And you will need to think if you want to print the film directly from the neg or from the scan (my guess is there should be a difference).
Well, but printing sections directly from negative is pain, so probably not…
Great, a rematch sounds interesting, but can you fit in some Ilford Pan F developed in DD-X (or ID-11, but I doubt you’ll want to use powder)? Or at the least some slow black and white film? I’m pretty sure B&W has an order of magnitude greater resolution capability than print or negative (see http://photo.net/film-and-processing-forum/00A3er ), although comparing black and white film to a sensor with a Bayer matrix does feel a bit unfair.
Marcus, I’d be interested in those results, too, but I think we’ve hit the limit of what I can accomplish in a week’s time. Sorry if the hand developed black and white doesn’t make it in.
Nice brief comparison. As an old shooter, I always like this kind of stuff, but please try to use a high-end film like Fuji Velvia/Provia or Kodak 100 slides to make the most out of the analogic set. I would advise to use also a grain reducer software like Noise Ninjia or Nest Image to make the comparison as fair as possible.
I shoot both digital and film. I love digital, and hate film. But i also love my trusty old Chinon CM-4 (works flawlessly even without batteries) and all my manual lenses, and i really hate that big clumsy digital camera, i hate all its buttons, i hate its absolutely useless keyhole viewfinder, i hate its menus, i hate autofocus and all those silly big lenses, i hate anything automatic. Complicated? Yes. Unfortunately i cannot buy the camera i want for money, no matter how much. And believe me, i would be prepared to pay a lot. Leica M8? Well.. yes, but no, it is a rangefinder.
A tip all camera makers, there is a lunch to be eaten here.. I do not think i am alone either, i even guess this niche is bigger moneywise than the whole market share of Pentax.
Hi. Interesting article and i think all that can be said about the choise of film for the test have been said =) But i would like to add another thought that people tend to forget. Your film scan is before any noise-reduction have been made (if its made). the 5D is always WITH noise reduction as this is an integrated part in capturing the image. Even the RAW format does contain post-processing in how the signal is handled and filtered and tone-curved to become a usable RAW image.
Me, i shoot digital and film and actually do get requests to shoot film for a particular look sometimes. I mean, why spend a few hours at $200 / hour in Photoshop to create a image similar to a Holga camera when you can just shoot the damn thing on a $50 camera right away
And i must say the skills you train when shooting with film do help your digital photography. Especially the bad habit of not trusting your viewfinder and your skill when taking the image, but always checking the display if the image is any good =)
Might add that i mostly shoot medium format film where there is a larger difference and a MUCH larger price threshold to go digital.
Interesting test!
I have not read all the comments but here are a few thoughts:
• both media will excel at some elements of photography - it’s your job to choose the right tool.
• film offers the ability to produce better large sized prints with appropriate tools - I have been asked to submit work for very large projects BECAUSE I have a good collection of film stock. I use an Imacon (Hasselblad) scanner.
• with software like Lightroom, digital editing and processing is getting easier.
• Digital work is still more expensive in many business models because of the massive depreciation costs associated with quickly depreciating camera bodies and DX lenses.
H
I like the article a lot, and would really like to see the rematch results. I suspect a much more balanced outcome, and not so much for the better film to be used, but mainly because of the workflow. The first round workflow is reasonable if what you want is see a photo on screen, but it doesn’t compare just film with digital. To see this, consider that the comparison is of:
Digital -> Jpeg -> Screen
compared with:
Film -> Print -> Scan -> Jpeg -> Screen
The extra steps (Print and Scan) introduce both loss and noise to a much greater degree (IMHO) than that of different films. To compare film with digital (e.g. for printing purposes) and share it with us you may need to use a workflow like:
Digital -> Print -> Scan -> Jpeg -> Screen
Looking forward to see the results.
Mita.
p.s. Just to see how bad it can go, I printed the full resolution 5D image (on a crappy inkjet printer) and scanned it (on an even crappier scanner) and believe me, you don’t want to see the results. (in case you don’t believe me: http://picasaweb.google.com/mitalapo/FilmVsDigital# )
Thanks! Very illuminating!
I’ve blogged it:
http://eolake.blogspot.com/2008/09/film-vs-digital.html
To Mita: I think the Nikon Coolscan is a high-quality film scanner, not a print scanner. Right, Tim?
Which means it’s pretty much the best you can get if you need your pictures from film to be on the computer.
Ooops, sorry for calling you “Tim”, Jim.
Mita,
Why the print stage on film? My workflow is
Develop -> Scan -> Jpeg
And if you really wish to compare digital to film (which a moot point in the beginning really. Different tools for different jobs.) then you need to compare printed output. So the workflow to compare is really
Digital (Shoot -> process -> print)
Film (Shoot -> develop -> scan -> process -> print)
Since if you compare on screen jpegs you will not be taking the advantage of any higher resolution or dynamic range simply because your screen wont be able to handle 48bit scanned data on film. And if its web output then well a cell-phone image might even suffice =)
But as i said before, different tools for different jobs. And i can take one slightly lo-fi example of this that i recently was faced with. I had a client that wanted me to shoot a series of fashion images for her which she wanted to have a old, 60’s/70’s retro feel to, and i was faced with either shoot digital and spend several hours getting it to look like say a holga camera or just go out and buy a $30 camera and shoot it with film.
I obviously opted for the last solution and the pictures turned out so well directly after scanning that the client did not want any additional post processing or retouching done to them.
Now obviously lo-fi holga photography is not the only time you use film and the point im trying to make is that if i want to get that velvia feel in a shoot or make it look like a polaroid, i will go and shoot velvia film or polaroid or whatever. I wont spend hours trying to get the same results in photoshop.
And dont get me started on B&W
Eolake said: “I think the Nikon Coolscan is a high-quality film scanner… Which means it’s pretty much the best you can get if you need your pictures from film to be on the computer.”
The Nikon Coolscan is certainly not the the “best you can get”! It is a reasonable scanner but far from excellent! I have had two Nikon scanners and after disappointing results starting asking for advice from one of my stock agencies and another stock shooter, Darwin Wiggett. Both urged me to investigate the Imacon (now Hasselblad) line of virtual drum scanners. I ended up buying one. It produces far better results than any Nikon scanner I used.
Nikon tries to scan a piece of film while it is flat. The film will not lay perfectly flat - causing focus challenges. Imacon curves the film while scanning - as does a drum scanner - so that on one axis the film plane is perfectly flat - better success focusing. I like Nikon’s cameras but there are many reasons the Nikon scanner is not the best you can get!
Harry
Thank you, I stand corrected. I don’t know much about scanning.
I notice though that the grain in the picture is dead sharp, so focus does not seem to be an issue.
@Eolake: I’ve found, time and again that one can be astounded. I’ve seen lenses that I thought were the epitome of sharpness and colour rendition only to have that thought shattered by an even better lens.
It never hurts to keep an open mind, and I’m sure that you’ll discover some scans that are truly mind-blowing some day. Seek, and ye shall find.
The focus should be ‘ok’ on the Nikon but in side by side comparisons the Nikon will not be on par with high quality scanners. Many people get acceptable results with Nikon scanners.
My point is I would not bill the Nikon scanner as ‘high quality’ or ‘the best you can get’!
H
Henrik: Thanks for pointing out that the scan is from film. I didn’t notice this on first read. I also agree that comparing film to digital for printing purposes requires the following workflow:
Digital (Shoot -> process -> print)
Film (Shoot -> develop -> scan -> process -> print)
My point is that in order to share the comparison with us on the web, the prints will need to be scanned. It is true that the final scan from print involves excessive noise and loss of details, but it will be the same process for both. The problem might be that if the final scan introduces too much noise and blur, it will mask the difference.
Mita.
Thanks for the article. My wife and I were thinking about going back to film. When the brightness or colors were incorrect, we had the developer reprint it. With digital, now we do the corrections. Even with all the fancy software that can batch process the digital photos, we spend more time adjusting individual photos for brightness, sharpness, and colors than we did with film. But for my casual photos where a perfect print is not necessary, I can’t beat the digital for the instant response.
I am thinking about adding a Coolscan 5000 to the 4000 I already have before starting a oceanographic research project that requires 45,000 35mm film images be digitized. However, I am finding that the 5000 is out of stock at B&H and well as Amazon.
Is something else coming out soon or is Nikon getting out of the scanner business?
Thanks for any answers.
… I’m waiting eagerly for the follow-up article.
Follow up coming soon, I shot 2 DSLRs and 4 film stocks for comparison just yesterday! Tomorrow they’re off to get processed…
Gosh, Bob, I surely don’t know the future of Nikon film scanners. It all depends on demand, I’m sure, but I really can’t imagine it’s all that great since film cameras continue to dwindle.
Anyone?
One thing that I will never understand is why is a raw film scan being compared to an ultra-processed digital image.
Try putting a quality 100 ISO film - Astia is a good example - through a quality scanner, post-process it with quality software - hey, the 5D firmware post-processes the raw file like it or not, why are you denying that to film?
This is the Astia result (greatly downressed, of course, after grain treatment and sharpening with Focus Magic:
http://fc75.deviantart.com/fs37/f/2008/264/c/2/c24b608b197caf3677f3b6a3cc798893.jpg
This is a very small 300X200 crop of the post-processed full size scanned file (in excess of 20MP):
http://members.iinet.net.au/~nsouto/photos/latest/colourful%20rescue%20-%20crop.jpg
This is an equivalent 300X200 size crop of an image produced with the latest whitest and brightest dslr, the 5DMarkII - about TWICE the resolution of the 5D:
http://members.iinet.net.au/~nsouto/photos/latest/5d2%20cap.jpg
I can tell you straight away which has the best colour saturation, detail and sharpness. And it most definitely is NOT the 5DM2…
But then again, you oughta be comparing apples with apples.
Not a cooked-up “comparison” designed from the start to show how “bad” and “dirty” film is (what’s with this “dirty” thing anyway? Sounds like one of those cheap subliminal low quality detergent ads of 30 years ago…)
It would be appreciated if FOR ONCE, in the interest of truth, a REAL comparison was published.
Instead of the usual nonsense.
Digital has MANY advantages over film, but ultimate resolution and sharpness is NOT it.
A D3 at 25600 ISO versus the equivalent in film is what I have in mind. Plus of course the speed and convenience of the workflow. Resolution, saturation and sharpness? No way, Jose!
The Coolscan VED has been recently discontinued although of course there are still plenty of places with stock. And it’s a darn good deal, at the price being asked: one of the best 35mm scanners one can get.
The Coolscan 5000ED and the 9000ED will not be discontinued while Nikon makes film cameras.
The F6 is still in production. The online places will be out of stock due to low demand, just ask them to order a scanner in, they MUST supply it or they will be in breach of their agreement to carry Nikon gear.
Watch the prices being paid in ebay for second hand Nikon scanners and then tell me there is no demand for them!
Leave your response!
Recent Articles
Pages
Categories
Archives
Top Commentators this month
Recent Comments
Further Reading
Related Posts
Most Popular Posts
Most Commented
Tags
approach around the net assisting bulbs business creativity customer service education equipment exposure film fine art gear history insurance interview lenses lighting logo design marketing nature photographer photographers pinhole portrait preview professional organizations RAW review stick-in-a-can studio taxes technical video workflow