« Quote O' the Day | Main | Photo L.A. »

Monday, 12 January 2009

TrackBack

TrackBack URL for this entry:
http://www.typepad.com/services/trackback/6a00df351e888f8834010536bec172970b

Listed below are links to weblogs that reference Film vs. 12MP Digital in Large Prints.:

Comments

Umm, if you want to be a pixel peeper, they are technically comparing straight digital capture to a 35mm film scan. That being said, I think many of us have known this for quite some. I am still awed by the images my trusty Fuji S2 produces.
PS I would love a D700 if anyone forgot to get me a Christmas present...

(Warning: this comment contains a plot spoiler.) OK, the digital print won; that doesn't surprise me. But, because the printer works from a digital file, the 35mm film negative had to be digitized, and that's an important point that everyone seems to ignore. This scanning process, like any conversion, introduces entropy into the process.

For a valid comparison between film and digital, film should be processed and printed the traditional way - in a wet darkroom, with chemicals.

I'm not defending film - I shoot 100% digital. I'm just saying that it's not a valid comparison.

I'm sure you'll receive numerous comments from people arguing whether or not the comparison was valid or not. ("They should have used ISO 100 film!" etc.) What I find amazing is that either camera can produce a decent billboard-sized print from an image that measures roughly 24x36mm. And don't forget, the Nikon D700 they used has "only" 12MP. Imagine the results they could get with the Sony A900.

I've had so many debates with my pal about how bad the tests are on the gadget show, especially the blow up test. Fair play, he's not in the least bit interested in photography. I'm just waiting for the day when he makes a 50ft print from his old Ixus

this is typical of most media today, leaving out the most important details. this is simply infotanement, not information of any real value.

since they leave out how they got the film to a digital file (drum scan....or...?) and if they used the same uprezing approach, etc. too many unknowns to be of any real value.

Regards,

Robert

Well, I guess this settles the debate whether 12 MP is actually enough for large prints :-)

Love this one, holy cow! 17m high images! To me that kind of says, if I want no bigger than say half a meter high, than the much maligned 10MP E-520 for sure will do rather nicely.

And at the same time, over at his site, Ken U-know-hoo has yet another of his "film looks better" posts up ... Well, it's a free world, innit.

"...blown away by their budget..."

I'm cynical enough to think that the budget was provided by Nikon. I'll bet that the "quality" of an image that large is 99%+ due to the process used to enlarge it and maybe 1% due to how it was initially recorded. This bit of fluff was only meant to sell cameras.

(PS - I shoot Nikon.)

only one quibble with this. If I was going to print out a picture measured in yards instead of inches, I would not use ISO 400. I understand this shows the strength of digital (low noise at higher ISO) but it pretty much was a foregone conclusion as soon as they announced it. Would also have been interesting to know which film they used.

Hello Mike,

Re: Film vs. 12MP Digital in Large Prints.

As with most test involving digital cameras there is usually an element about which we may not agree with.

The Nikon D700 may not show much of a difference in term of image quality if the image was originally shot at ISO 100 or 400. Not so with film.

The test is valid for measuring ISO 400 in both film and digital.

To measure film vs digital capture, a film having an ISO 100 or less should be used, image quality being tested, not convenience.



Don't tell Ken Rockwell!

Choices are nice to have.

The problem with film now is price. If I shoot a wedding in film, my lab costs add at least $1500 to the bill of the wedding, same thing with any commercial shoot. I am all for digital, but how long can the working stiff really afford film? It's almost impossible to be competitive with film at the mid/low end, the squeeze is on.

Pretty soon the only people shooting film will be Sultan's or big name fine art photographers who charge $$$, are they really going to keep the film factories open just for them?

100+ years of film technology got flushed faster than a blink of the eye.

After close and lengthy scrutiny of the lovely woman in the catsuit (OK, OK; but can you blame me?), I must say that the green tint in the man's shoes made me think the digital photo had the off-color shoes. Maybe a 100-speed film would do better than the 400-speed, but then the digital camera would probably do better at an ISO 100 setting too.

Quite amazing to see that such quality is possible from a full-frame 12MP sensor. I won't be waiting for a full-frame digital camera from Pentax before I take the plunge. Less than full-frame should be plenty good enough for me!

P.S. That video reminded me of the segments shown on "Top Gear". Must be something in the English genes. (But why do they pronounce it "Nikkon"?)

Quite a few comments about the "test" side of things already... so does that mean I was the only one to lose focus when Suzi Perry turned up on screen in a cat suit?!

Ho hum... I must just be weak!

Quite cringe making as we say over here, so embarrassing for those of us who feel that British TV is usually better than American TV.

Perhaps they used a Nikon film scanner?

Dave I said "But why do they pronounce it "Nikkon"?"
Dave, we would say "But why do they pronounce it "Nike on"?"

Cheers, Robin

Dawson - it's all about priorities. I'm sorry, I was looking at Suzi - what was the question?

Ach! Painful presentation. But I only shoot colour film now if I want it to look like colour film. And I shoot black and white film if I want the picture to look like that, but I do still shoot both as well as digital.

I'm currently working on a prject to scan (the worthwhile part of) 1,000 35mm Kodachromes I took on an African trip in 1995, and eventually to make a short book from some. They are clearly film and technically nothing like what a 5D etc can do, but they hold some very precious memories for me and I find the look of prints from them deeply evocative.

Mike

'But why do they pronounce it "Nikkon"?'

That's more or less the intended pronunciation, AFAIK. Saying it with a long "i" seems to be an American thing (though that includes Nikon US).

This whole scenario is easy to debunk. Please tell me which fashion photographer shot with an F5? Ok I'm sure a few did but really that camera was for sports, reportage and nature. Any photographer worth his/her pay was shooting these scenes with a Hassy or Mamiya RZ. Not 35mm. Digital for a price has topped 35mm film but you better be one rich SOB to afford the best equipment.

I do agree with the comments regarding ISO 400. So let me add a newbie question.

If I want to have digital pictures of the best possible quality from my 35mm film camera, how do I get it? Slide film or negative? How do I choose the film? What do I have to watch out for when having the film developped and scanned? Is the Fujifilm SP-3000 at my local shop good enough to give my F80 a fair chance against digital?

Thanks for any help,

C.

It was a bogus test for many reasons.

First, the mostly monochromatic shot favored the Bayer sensor-equipped dSLR. What if the two had worn '60s-era flower suits? Color resolution, especially with lots of color detail, would have favored film.

Second, ISO 400 color film is not the standard by which 35mm film resolution should be judged. Why not ISO 50 Velvia? It's like pitting a new Porsche against an old Plymouth with a 6-banger.

Third, as others have commented, what happened during post-processing? Were the same up-rezzing techniques used on both digital files? What film scanner was used, and at what resolution? After all, a 4000 dpi scan would result in a 24MP image from film. And, was noise reduction used on the film scan?

In short, this 'test' really didn't tell us what we already know: a 12 MP dSLR image (full-frame or otherwise) is better in terms of resolution and noise than the best 35mm film image. But it will sell a lot of D700s to the moneyed dilettanti.

"That video reminded me of the segments shown on "Top Gear"."

Definitely. Not as funny as Top Gear, but similar in approach. And in budget as well.

"(But why do they pronounce it "Nikkon"?)"

Maybe because it's a Japanese name and Japanese doesn't have diphthongs?

http://en.wikipedia.org/wiki/Nikon

Dave I - I have always pronounced it "Nikkon", and I assume that most UK photographers do. May not be the authentic Japanese pronunciation, but then we (British) feel that sometimes our good friends in the USA do not always pronounce English words quite correctly! I find that reading TOP each day I often envy you all in the USA, but today I can gloat over that fact that here we can see Suzi Perry on TV very often.

"The problem with film now is price. If I shoot a wedding in film, my lab costs add at least $1500 to the bill of the wedding, same thing with any commercial shoot. I am all for digital, but how long can the working stiff really afford film? It's almost impossible to be competitive with film at the mid/low end, the squeeze is on."

yunfat, I hope you're charging your clients for all your time at post processing your shoot. If not, you're ripping yourself off.
That $1500 lab bill will save you hours of your own time which you can put toward towards other aspects of your business or time with your family.
I don't shoot weddings full time but when I do it's on film. My clients get guaranteed archival images that are also digitized so they can share them/reproduce them, I don't have to keep upgrading my cameras every two years, the guests don't bother me to see the preview of what I just shot, and I get far more time for myself!

I told Ken Rockwell.

Aside from the issues of the validity of the comparison, who shoots 35mm ISO 400 film in the studio with strobes unless they particularly want a grainy effect (which they might, but that's another issue)?

"The problem with film now is price. If I shoot a wedding in film, my lab costs add at least $1500 to the bill of the wedding..."

You either need to find a different lab or you should disable the burst mode on your camera.

Ohmigawd!

"The Attack of the 50-Ft Avengers!"

Seriously silly. I love it.

I think anyone who tries to seriously nitpick this "test" needs to send their senses of humor and absurdity in to the shop for refurbishing.

pax / vastly-amused Ctein

While we're on the subject of film vs. digital, I chuckle at the idea of spending $250 for software that makes digital images look like they were captured on film. Perhaps there also would be a market for third-party firmware that permanently disables autofocus and autoexposure on a modern DSLR, mandates the use of a maximum size 64 Mb memory card, and fries the memory card if you insert it in the camera a second time.

The video brings up several very interesting issues:

a) Processing, both in digital and film, are key to a good final image. Did anyone notice how crappy and yellow the digital photos looked in the previews? There was quite a bit of post processing done to those digital files, even if it was to get the color balance right. Was the same care taken with the film files? Was it even possible without treating the film files as digital and negating the whole test? After all, 'uprezzing' of film is not really possible without it becoming 'digital, yet 'uprezzing' of digital is treated as an exotic art, and expected from any proficient photographer.

b) The size of the sensor is relevant to the size of the final print and the distance at which it will be viewed. As a pro photographer, I am so tired of clients expecting a 30 MP MF back when they will be printing a poster print on a commercial printer. I have seen great poster prints from an 8MP camera and horrible similar sized prints from a MF back. 'Darkroom' skill still counts.

c) The subject matters. Remember the 'softening' filter that traditionally was used in portraiture? If they has photographed a highly detailed and contrasty scene of 'Angkor Wat', then the results may have been different, with both prints being judged as crappy. Photograph people, and softness and lack of grain and 'accutance' are a often a benefit of digital when you pass the 'acceptable' limit.

d) Skin tones. I still have high end clients that refuse to have their portraits done with a digital camera, because film gives their skin a better 'feel', less digital is the way they explain it. And I agree with them. There is a different 'feel' to the images, and one may be better than the other regardless of the technical measurements.

All in all a good test IMHO, more real-world than most, but subject to a large number of factors that could easily skew the test either way.

And lets be real... it was a side by side test... in the real world both prints would be just fine for the way ion which they were presented.

I think they decided what result they wanted before the test. I'm sure Jon Bentley was told what to say and look for. He commented on the shadow detail in the blacks of the shoes - something we know digital is good at and ignored the highlight detail - something it doesn't do too well.
It was really just a test of Nikon D700 vs. whatever scanner and scan method was used(which was not shown).
If you want to do a digital vs. film test (and I don't) you would need to use the best (or near best) of both systems e.g. a 5x4 film camera with a good ISO 100 film and possibly the same camera with a digital back. Not a near top of the range DSLR against a 35mm SLR with ISO 400 film.

>>If I want to have digital pictures of the best possible quality from my 35mm film camera, how do I get it? Slide film or negative? How do I choose the film? What do I have to watch out for when having the film developped and scanned?

I know this is not a forum or tutoring site, but I share Carsten's curiosity and would appreciate any pointers to relevant sites/ books.

Their pronunciation of Nikon is closer to the Japanese (something like "neekon" but don't drag out the ee) than the American "Naikon".

Hey Serious Printer Guys:

When I passed the link for this amusement to Mike, I mentioned that one thing the test might suggest (to me, anyway) is that the benefits of >12MP digital are possibly very limited. Mike doesn't agree; he says the difference between 12MP and 24MP will be easily seen in a print made to be studied closely in a gallery. I wonder...

As a practical matter, I have no interest in full-frame, high resolution digital cameras until they make one as small and light as a D40/D60. BUT...it would be really cool if a seriously talented printer would make 20x30 gallery prints of the same scene shot with four different cameras (with equivalent lenses, ISOs...you know, the way YOU would do it):

- D700 or equivalent 12MP digital
- A900 or equivalent 24MP digital
- 35mm film camera (best quality commercial film scans)
- 120/220 medium format camera (ditto)

A blind comparison judged by other talented printers or other known experts would be the best entertainment reading of 2009.

Not exactly easy or cheap to accomplish, but a bargain compared to the Gadget Show Nikon commercial and a heck of a lot more relevant to boot.

Ken N: "Ordinary" wedding photographers today are promising to shoot around 4000 exposures (and present 600 proofs), at least in my local market. That's over 100 rolls of film, so the $1500 lab cost mentioned is completely reasonable (particularly since "lab + film" is the relevant difference).

The burst mode on the camera is often very useful for getting the best shot.

If you find it more convenient to shoot less, and you're good enough to get the shots anyway, market only the number of proofs guaranteed and the quality of your samples; that might be a viable strategy. Although you'll be missing a number advertised by a lot of wedding photographers, so you'll have to explain why. I'm not marketing myself in wedding photography, but I'd think you'd get decent results saying "I don't think it's fair to count multiples I take to get the one first-rate shot I present to you". With some clients, at least.

For me the take-away here is that a properly-exposed, properly-focused image recorded with a 12MP camera can be enlarged to just about any size you can afford to have printed, practically speaking.

Yep, you gotta stand back if you want to oggle a billboard-size image or it won't be pretty. But neither do you stand scant inches away from a 20 x 30" print targeted for your living room. (My living room, alas, is too small for a print of that size.)

As for those puny 13 x 19" (or smaller--usually smaller) prints, don't even worry about it. You're good.

You'd think also that for a test like this they'd tether the camera to a somewhat rare device known as a "tripod" to remove a few more variables from the equation.

Ctein rules, as usual!

Dear Steve,

The test you speak of has been done, definitively and to death. Digital wins, period. Has for some time. By every objective metric one can think of. There is no reason to ever do it again.

"Mine is bigger than yours" tests are fundamentally boring and miss the point. Which is to make PHOTOGRAPHS. And make them good enough to make you happy, in the medium that you like the best, for whatever personal reasons. And once you're convinced your tools are good enough to make you happy, you should move on.

This reminds me way too much of people arguing about whether slides were/are better than negatives. Most of them didn't understand what they were talking about.

And, really, folks! Do you think the folks on this TV show are utterly stupid? Do you think that for one moment they seriously considered that doing an avengers spoof, making 50-foot high prints from it, and hanging them on the side of a large office building was anything remotely like what their typical viewer did? Honestly, how seriously do you think they were taking this?!

People, you just have to lighten up!


~ pax \ Ctein
[ Please excuse any word-salad. MacSpeech in training! ]
======================================
-- Ctein's Online Gallery http://ctein.com 
-- Digital Restorations http://photo-repair.com 
======================================

"You'd think also that for a test like this they'd tether the camera to a somewhat rare device known as a "tripod" to remove a few more variables from the equation."

Robin,
You don't need a tripod when you're using studio flash. The flash "freezes" the subject at an effectively high "shutter speed." In fact, here's a little secret--if you really want to "test" (i.e., find out) how sharp a lens is, shoot with high-powered flash. Won't get any sharper than that.

Mike J.

"- D700 or equivalent 12MP digital
- A900 or equivalent 24MP digital
- 35mm film camera (best quality commercial film scans)
- 120/220 medium format camera (ditto)"

Paul,
Here are your results, where check equals "good enough" and no check equals "not good enough":

- D700 or equivalent 12MP digital--check
- A900 or equivalent 24MP digital--check
- 35mm film camera (best quality commercial film scans)--check
- 120/220 medium format camera (ditto)--check

Done and done.

Mike J.

Apart from all the other comments made about the validity of the tests my question is why, this topic is 5 years dead. My 10D has been turning out better prints then medium format print film could ever do. As soon as you scan film you loose so much including color fidelity hence the greens present in the black pants and shoes, thats not from the enlarging proccess but from the scanning. If you dont scan film and print from an enlarger you get a better image then you would from the scan but you are left without all the tools provided from Photoshop and raw proccesing. Without that its an end to film and has been for a long time.

>Done and done.

Oh, you're just a big tease.

QUOTE:As soon as you scan film you loose so much including color fidelity hence the greens present in the black pants and shoes, thats not from the enlarging proccess but from the scanning.

First of all it's lose not loose.
But more importantly, film was not meant to be scanned, it should be optically printed.
A test at about 16" x 20" should be plenty for either medium. Nobody needs building sized prints.

Dear Steve and Richard,

Enough! You're both wrong.

If your film scans don't look good, it's your technique that's at fault. If you're having a "professional lab" do your film scans for you, it's their technique.

Poor film scans are the consequence of poor technique (and that includes not having appropriate equipment), not any inherent unscannability of film nor what film was "meant" for. (That isn't even technically correct-- manufacturers have been designing their films for scanning for quite a few years).

I state this as someone who gets excellent film scans from any of my films, no matter how old, and who is widely considered to be one of the very best traditional color printers alive today. So you may be assured my standards are higher than most.

Take it to the bank. If you're not happy with your scans, look to your methods. Otherwise you're in the same position as the neo who complains that their new, expensive camera makes lousy pictures.

pax / Ctein

Great find. I am thoroughly impressed with the size of those prints and how a 12mp camera could blow up an image that large. Same goes for the F5.

Jeff

Who cares? Film vs Digital is old hat. Both have advantages over the other.

Get out more and go take some pictures people!! Film or Digital.

The comments to this entry are closed.