DPR Forum

Welcome to the Friendly Aisles!
DPRF is a spin-off of dpreview. We are a photography forum with people from all over the world freely sharing their knowledge and love of photography. Everybody is welcome, from beginners to the experienced professional. From smartphone to Medium Format.

DPRF is a community for everybody, every brand and every sensor format. Digital and film.
Enjoy this modern, easy to use software. Look also at our Reviews & Gallery!

Film vs digital camera pros and cons

Lynn,

"... am very fussy about quality slide processing. I have had slides are scratched, dusty, or have droplet marks from chemicals in some cases. I'm still playing the "test the lab" game of seeing who does the most consistently good job.."

I made that road already. I ened up buying a Nikon Supercoolscan 4000 Ed filmscanner and an Epson Stylus Photo 1290 printer. I get better resuklts as from the so called prolabs in Frankfurt Germany. They tried to charge me up to 35 Euro for one A4 print of a slide. My results are still better with my scanner. I found that shocking.

Nikon brought out 3 new filmscanner a couple of days ago:

http://www.nikoninfo.com/discus/messages/4/8133.html

you might want to look at them closer...

My dream is a Contax/ Zeiss filmscanner for 135 and Medium Format. I know there have been plans for it already years ago, but obviously Kyocera did not dare to do it. Seems to get a habit at Kyocera...
 
Wow! This thread has heated up!

Marc,

I think that there are several factors that play into my being "less thoughtful" as a photographer lately - I have two young sons, ages 4 and 7. So, I now get 7 minutes per day to myself, and it's non-contiguous.. That, factored in with zero marginal cost of taking a bad digital shot (I delete bad shots directly from the camera), has affected my behavior and lowered my yield of good shots.

Mike,

It is, of course, the marginal cost of having to process my bad shots on film that affects my behavior. I'm a rational person (just as economists assume). So I disagree with your point about the costs of digital because of the context of my decision. Those are costs that you must incur to play. The marginal cost of a bad shot, as I've described, is zero. Hence, the cost of cards, etc., is irrelevent to the decision to snap away.
 
Well I wrote that there have been too many manipulations going on with Photoshop.
I didn't mean it is wrong to do that. Of course the use of lens filters don't make you better photographer, but the use of Photoshop gimmicky effects doesn't make a better photgrapher either.
People just overkill images with manipulation.
I use PS myself with curves/histogram for retouching scanned slides and negatives ; adjusting contrast, saturation and dust removal.

The point I tried to make was that I don't really like the idea of creating spectacular sunset artificially using computer generated fake clouds and special effect filters, simulating blurred motion, out-of-focus background,using PS. All the photo magazines are now flooded with articles to tell you how to do these kinds of tricks. While I have nothing against it, I think it is a bit of a cheat.

Film photography is more pure and organic: its about capturing the right light, at the right moment...etc Now you can cut and paste images to modify reality. No doubt Photoshop image manipulation itself is sophisticated craft by itself requiring lots of skills and time, but it is not my cup of tea. This my personal opinion and I respect other poeople's tastes.
 
[snip]
> However, some of the points you make are hypothetical, like the loss
> of data from a CF card from some magnetic field

Hypothetical? That's a joke, right? As I said, I'm computer engineer. There is a thousand things that can go wrong with any electronic storage devices. Hell, even EMC systems that cost millions of dollars can have a multiple point failure resulting in irrecoverable loss of lots of important information. You can even lose that data due to computer virus! Not to mention that in about 10 years you will have to worry about converting all that data from DVDs or CDs or whatever you use into some new format. Not so with film.

. In three years of
> shooting digital, I've not experienced that problem nor have any of my
> fellow digital shooters. I have not read of that happening either.

Good. Because in over 10 years of film photography I had only two-three badly developed films. I'm very picky, I don't just go to the lab everyone is raving about. I "test" them personally first. If I'm not happy - I bring my business elsewhere. Technically I don't care - I can mail my film for development to East Siberia, if they will happen to have the lab staffed with best people and best equipment. And that's the way this business is going I believe. Thousands of BAD labs are going out of business (which is a good thing) and better, more successfull labs emerge as a result.
With current global nature of things, I can really send my film over to Germany, Switzerland, you name it and get it done at high quality and decent price.


> Besides, when shooting important work, I download the CF cards as I go
> to a portable HD. Which means I have 2 exact copies of the work while
> I'm still there. Something not possible with film.

yeah, two copies are good. And both can be lost. trust me on that one
happy.gif


So, although it's a good point, it's of no use to me. Quality is irreplaceable. And there is no way I can get quality I'm used to getting on film by using digital. So, having two copies of bad photograph doesn't make me feel any better.

As I mentioned in one of previous posts, just check out new CLN from Zeiss, film still beats the hell out of digital and it will only keep improving while getting faster. Again, if Kodak abandons it - it will only bring more business to Fuji and Ilford that are ALREADY doing a better job than Kodak (besides, Kodak announced they are not going to spend more money for R&D of consumer film. Don't remember seeing anything about professional film). How many serious hobbyists and pros you know that are using exclusively Kodak? Because I know plenty that are using exclusively Fuji!

Just take Fuji NPZ 800. 110lp/mm. That translates into (24*2*110) x (36*2*110) ==> 5280 x 7920 "pixels" image. That's 41.8 MILLION. True color "pixels". No bayer pattern, post-processing, anti-aliasing, layer of micron-lenses, cross-cell noise, issues related to extremely small sensor sizes, degradation of quality of applied "filter" on photo cells and other crap of this sort.

For me digital just doesn't cut it. And probably will never do. I enjoy watching movies shot on film and can't stand movies shot on video for the same reason.

It works for you, you get paid for you work, then be happy!

> I have not lost any images on CD-ROMs or DVDs due to scratches, but
> acknowledge that it could happen. That is why, for less than a dollar
> more, I burn a second one (one for the client's records and one for
> mine). The wonder of it is, duplicates of RAW or Tiff images are exact
> copies. I also have a 200 gig drive with often used images stored on
> it. But a HD should not be trusted as a sole form of storage.

You still keep forgetting one thing. TIME. Time is money. And even more so in my case, since I don't do photography for a living. It's like having a car and changing oil, tires, lubricating chassis, changing cooling liquid, brake fluid, transmission fluid, replacing shocks, etc and all that by yourself. Most of people would rather take their car to the shop. And that's what I do with my film.

> On the other hand, my recent loss of images due to poor lab work is
> not hypothetical. It actually happened. The fact is that good labs are
> under stress due to a majority of commercial work going digital. For
> ex&le, mail house labs that once thrived on wedding orders are
> getting less and less from that source as that industry goes digital.
> And we have to remember that everything that ends up in printed form,
> which constitutes millions of printed images, is basically now all
> digitally processed (either scanned or digital in the first place). In
> addition, financial pressures from the economy are fueling the switch
> to digital. For ex&le, more and more automotive print is being done
> digitally, which I thought would be the last to switch. The situation
> was personally brought home to me recently when my favorite Pro Lab
> (in my area) stopped developing Tri-X. : -( And during a recent trip
> to Las Vegas I requested info from pros there as to where to get my
> film processed. The reply was that the main labs had shut down. Only
> one remained, and it's quality had slipped. Again, that info was from
> working pros from the area.

You know, funny you should say it. A&I can't keep up with demand for film processing and printing. Over the last 2 years, they doubled number of film-processing machines and printers and STILL demand exceeds their abilities. See my note above - I'm glad BAD labs will go out of business and good ones will remain. It will narrow the field of competitors for all of us and offer us a better choice - you won't be choosing between 100 bad labs and 1 good lab anymore. You will be choosing between 3-4 good labs and that's it!

And I disagree about all prints made digitally. As I said, even lightjet prints don't look like well-made optical ones. And that's why A&I still offers this service even for basic 4x6 proof prints. As someone here noted, true resolving power of prints made on Frontier is 180dpi (actually I'm not sure if that figure is entirely accurate. I have seen doc somewhere that was stating that Frontier's maximum resolution is 200dpi. It's not much better though) and Lightjets are 305dpi. Difference is visible between the two AND optical prints.

> As to the price tag for a Mac G5 and attendant gear. As a business
> expense it is figured into my overhead and reflected in the rates I
> charge. In the end, it pays for itself...so in essence it is free.
> Again, it's the difference between professional services and a hobby.
> Like any business, you have to spend money to make money.

Exactly. But previously, you spent that money on paying the lab and saving your time. Now, you're spending your time and money on bunch of electronic equipment that loses value faster that the Earth rotates around its axis.

> The rest, like liking or not liking Ink-Jets is personal preference
> and not something that can be empirically touted or dismissed except
> from a personal point of view. That my clients accept them, and many
> prefer them, is the only opinion that I need be concerned with.


Precisely. They like it, they pay for it, then why bother? But the point is - a lot of people out there don't really know what they want. Once they see two very good prints made using two different methods - they might change their minds. A lot of wedding photographers here in California are basically pushing digital down the throat of their clients, who don't know better to make their own decisions. And these "photographers" do it only because it's easier for them to make money. Not because of quality.

> Making great Ink-Jets or Dye-Sub prints is something of an art. Some
> are better at it than others...just like analog

Actually, results seem to be so inconsistent among the labs that do that, that it's absolutely impossible to predict final results.

> printing. Personally, like you, I think a well made silver print, by a
> master printer, is still unmatched in the digital domain. But the gap
> is rapidly closing.

So far I don't see that happening.

And still, my point stands - to each hiw own.


Dirk:

Yeah, I'd love to see Zeiss engineered film scanner myself. And Zeiss certified film-processing and printing lab as well. A lot of optical printers out there will do well by using higher quality enlargement lenses.


Mike.
 
Mike,

Based on the tone of your letter, I'm beginning to think that "Clive" is back. Your statement: "So, although it's a good point, it's of no use to me. Quality is irreplaceable. And there is no way I can get quality I'm used to getting on film by using digital. So, having two copies of bad photograph doesn't make me feel any better." is very strange.

You say you are a software engineer. Hmmm. Your lack of faith in technological solutions is remarkable since you work in the industry. Millions of us fly every year, and just think of the faith system there ... and the electronics. I, recently had to have developed/scanned 17 rolls of film. My sweat glands were on overactive waiting for the results. You seem to think that developing accidents are rare compared to the loss of digital files. In my book, digital files are by far a safer backup then slides or film. The are secure, they are organized, they are highly portable. And Marc's comment on being able to make two copies is exactly right. And each one an exact copy of the other. For your observations to even begin to make sense, you need to quantify the word quality. For me, if the image has impact, has interest, has form, shape and makes me later "think" about that image, then the image is a quality image. From a purely technical point, are you saying the image must hold detail as a 24 x 36? Or has a 5 x 7? Where is the line? What is your definition of quality?

Also, most digital workflows are far superior and safer their anolog counterparts. However; today's reality is that alot of photographers merge film into a digital workflow, hoping to merge the strengths of film capture with digital processing. But, it is obvoius that the pure digital photographer has the upper hand in the workflow environment.

You also made the statement "For me digital just doesn't cut it. And probably will never.". Technological advances are remarkable, and even if you could argue that quality digital has not arrived (which I believe an eroneous statement ... digital quality is here and it is here now), then even suggesting that in the future, digital quality will still not be there really makes any of your other observations suspect.

Michael.
 
Mike - really an excellent assessment of film vs digital - and why I, who have been building computers for a scarily long time, don't trust any of them for archival purposes. Would any of you who isn't an ostrich regard any software by Microsoft as able to make anything that will be around in ten years, or hardware by Intel? Let me tell you, most of it is base consumer level lowest bidder CRAP.

I just got a digicam. I use it to shoot instant pix of my cat. I have not found much more that it is good for yet. It doesn't inspire me to take great pix, it just sits around and looks cute.
 
> Based on the tone of your letter, I'm beginning to think that "Clive"
> is back. Your statement: "So, although it's a good point, it's of no
> use to me. Quality is irreplaceable. And there is no way I can get
> quality I'm used to getting on film by using digital. So, having two
> copies of bad photograph doesn't make me feel any better." is very
> strange.

Well, first of all, I don't know who is Clive. I can assure you that my name is Mike
happy.gif

Also, I don't know about the tone of my letter. These were just my thoughts, it's not like I'm trying to offend anyone. Everyone has their opinion, which is the one and only one reason why Marc and I have been talking about it. He sees things from his perspective, I see them from mine. And we both can enjoy watching a good game of basketball probably
happy.gif
No offense intended, and no offense taken. I'm not trying to prove him wrong, I'm trying to make a point that he isn't always right. Neither am I
happy.gif



> You say you are a software engineer. Hmmm. Your lack of faith in
> technological solutions is remarkable since you work in the industry.

I said I'm a computer engineer. At no point in time I ever said I'm a software engineer.

> Millions of us fly every year, and just think of the faith system
> there ... and the electronics.

Based on time-proven techniques, technologies and manual-overrides. Boeing 737 is so reliable because it's designed to be first and foremost reliable and safe. I don't think Canon G2 is designed as well as Boeing.

> I, recently had to have
> developed/scanned 17 rolls of film. My sweat glands were on overactive
> waiting for the results. You seem to think that developing accidents
> are rare compared to the loss of digital files.

That's my personal experience.

> In my book, digital
> files are by far a safer backup then slides or film. The are secure,
> they are organized, they are highly portable. And Marc's comment on
> being able to make two copies is exactly right. And each one an exact
> copy of the other. For your observations to even begin to make sense,
> you need to quantify the word quality. For me, if the image has
> impact, has interest, has form, shape and makes me later "think" about
> that image, then the image is a quality image. From a purely technical
> point, are you saying the image must hold detail as a 24 x 36? Or has
> a 5 x 7? Where is the line? What is your definition of quality?


My definition of quality is the same as yours - whether it makes impact or not. Except one thing - lack of detail almost always reduces quality. And if I can't make an impeccably looking 8x12 or 11x13 enlargement - I'd like to get my money back.

And that's what I can't have on digital. Detail, tonality range, contrast, depth of field - all of this doesn't satisfy me as it's currently done in digital cameras. If I want to have two exact copies of an image - I would order a drum scan of my slide/neg, burn a copy on DVDs, flash disks, CDs, whatever - I can do it then.

> Also, most digital workflows are far superior and safer their anolog
> counterparts. However; today's reality is that alot of photographers
> merge film into a digital workflow, hoping to merge the strengths of
> film capture with digital processing. But, it is obvoius that the pure
> digital photographer has the upper hand in the workflow environment.


Safer? In what sense? Digital photographers having upper hand? I don't think so...
By the way, film is not "analog". It's not an LPs vs. CDs situation. It's totally different.

> You also made the statement "For me digital just doesn't cut it. And
> probably will never.". Technological advances are remarkable, and even
> if you could argue that quality digital has not arrived (which I
> believe an eroneous statement ... digital quality is here and it is
> here now), then even suggesting that in the future, digital quality
> will still not be there really makes any of your other observations
> suspect.


Digital can replace the film. When it becomes as good as film that is. But I don't see how that can happen considering the direction in which this technology is heading. The more "pixels" you get on that small sensor - the more problems you have and getting images of higher quality becomes much more difficult. Bigger sensors with bigger photocells (and often smaller count of "pixels") produce better results than smaller sensors with higher count of pixels of smaller size. The industry is however chasing the Megahertz here. UltraSPARC III running on lower frequency beats any Intel-based systems in floating-point performance, processing of huge amounts of data, etc. Hell, the bus width on Enterprise servers is 512 bits! It's 4 (FOUR!) times what Intel has.
Yet the industry is still chasing the Megahertz... Same thing with digitals - the race is on who has more pixels stuffed on that nickel-sized sensor...

When and if they will finally concentrate on getting better overall image quality and giving us, the users, flexibility that film offers - maybe then things will look better from my perspective. But not sooner than that.


Dana,

I know exactly what you mean.


Mike.

P.S. I think it's about time we wrap this thing up. It's quite obvious that we will keep our opinions.
 
Scott,

I will take a look at it later. But name Bob Atkins sounds very familiar. I think I talked to him a while ago....

Mike.
 
Scott,

I read the article. It accomplishes its goal of making you think, but generally speaking it would be a stretch to say that it's very accurate or very informative. First of all, it's unclear to me how MTF of the lenses for all three systems is calculated, because it seems to me that they're based on a purely theoretical assumptions and calculations.

For instance, this part "So for the D30, we need 40.4 lp/mm on the sensor, and at 40.4 lp/mm a lens operating at f8 will have an MTF of 0.75" left me totally puzzled. It's true that for most 35mm based systems, MTF determined at spatial frequency of 40 lp/mm is the defining characteristic of the lens performance when you need large enlargements (in other words, it defines the resolving power of lens and its ability to provide good contrast). But how was that value of 0.75 determined in this case? And why at faster apertures MTF is going higher? Usually it's the opposite - MTF is better at stopped down apertures, not wide open.
Also, it seems that the assumption taken is that lenses in all three systems are of equal quality - "assuming we use a lens of the same optical quality on each camera."

That's just plain unscientific I believe. Microscope can resolve up to 1000 lp/mm. But only over surface area of 2mm and at very steep price. In contrast, good 35mm camera lens is designed to resolve up to 200 lp/mm (center), covering entire 24x36mm frame.

So, it's totally unclear to me how all this plays together, because we know that Coolpix and Minolta Xi are definitely NOT using same quality lens as whatever could have been mounted on D30 in real life. Perhaps I missed something in the article, but then it would be nice to hear some explanation of that.


Also, sensor resolution limit listed is basically Nyquist limit and its *purely* theoretical.

Now, there is even more interesting thing there. Here it is: "Now lets take an ex&le of an 8x10 print and let's say we want a fairly sharp print, so we are going to need a resolution of at least 3 lp/mm in the print"

This definition of "fairly sharp" print is quite strange. There is a pretty much established standard in the publishing and imaging industry about acceptable level of resolution in prints, publications and any other "visual" media. It's based on scientific findings and assumptions of 70s and even earlier. It's based on resolving power of human eye. It is widely believed that human eyes can resolve up to 6-7 lp/mm. In reality though, there will be people that can "see" 8 lp/mm and there will be ones that can see even up to 10 lp/mm (and just a personal observation - I think nearsighted people and kids have eyes with higher resolving power. Don't know if there is any scientific data on it though, but I'm really curious to find out).

Anyway, back to imaging and publishing. For the most part, the standard is 300-305dpi for what they consider to be high resolution prints. One inch is 25.4mm. Each dot can represent any color. Taking human eye resolution of 6 line pairs per mm it means => 6 * 2 * 25.4 = 304.8 dots per inch. Prints at this resolving power are believed to be smooth and sharp as seen by human eyes.

Also, one poster there included a link to Norman Koren's website. There is something that you need to realize what you read on Norman Koren's website. In most of his articles, a lot of findings are based on theory and some number-crunching. Very few are really real-life results. What's even more important is that just like luminous-landscape.com they're doing the same mistake by evaluating film vs. digital. That mistake is in... surprise... judging the resolution of film and digital cameras. A lot of people that read that for first time (including myself) got bait basically. You see, in both places, resolution of film is determined by... SCANNING IT on film scanner! Now, what's wrong with that picture? Everything!

If you scan the film to determine resolution, you're simply introducing another step in a chain of "source image - final print". And in this case, the quality of that final print will depend more on the quality of scanner than your film + lens combination. You see, if you're using Canon FS4000, advertised scanning resolution is 4000dpi. The problem though is that it doesn't mean that it can actually truly resolve 2000 lp/inch (78.7 lp/mm). Because it depends not only on CCD inside scanner, but also on the quality of that lens and how well focused it is on the negative/slide.

So, although it will create you a 20 Megapixel picture, it doesn't mean it captured all the details from film, nor does it mean that it actually captured them at 2000 lp/inch. The only proper way of determining final resolving power of a film is using microscope or something close enough in terms of true optical resolving power.

Now, to make it easier to understand what I'm talking about, here is a quote from Norman Koren's article ( http://www.normankoren.com/Tutorials/MTF7.html ):

"We compare digital camera resolution with Provia film scanned at 4000 dpi and sharpened. This provides a fair comparison because it contains nearly all the information on the film-- a sharpened 4000 dpi scan results in a sharper image than you can get through an enlarging lens. "

This assumption is absolutely, totally flawed! The comparison is not only unfair, but it also misleads you to believe that film's performance somehow has to be judged by the quality of an enlarging lens!

Well, I think that pretty much sums it up.


Mike.
 
Back
Top