35mm film vs. Digital - where is the border?

<small><i><blockquote>

 

My FM10 cost $200. At $4 a roll for film, $4 a roll for processing, 36 slides per roll that 1Ds/II is equivalent to 30,914 images!

 

</blockquote> </i> </small><p>

 

Bernard, what are you going to do with your Wal-Mart-processed slides -- stare at them? Have people over to look at them projected on a (free?) projector? Email them via (free?) scanning? Print them for free? <p>

 

Your FM10 is a cheap, Cosina-made camera with a Nikon namplate on it. Do you really think the kit lens that came with it will result in images that are better than those from a DSLR with a modern kit lens?

This comparison between <a href="http://www.luminous-landscape.com/reviews/shootout.shtml">State-of-the-Art 35mm Digital Vs. Medium Format Film</a> may be of interest.

Comrad,

 

I faced this exact same dilemma earlier this year having spent 35 years shooting film. I hate grain. I considered 5x4 and bought a 5D. I love it. The border is the 5D. Dont even bother testing anything less. Go test the 5D against the 1Ds/11. But I think the only difference is features not quality and enlargement size. However I dont enlarge the 5D images beyond the 10x15 inches that comes straight out of the camera. Most of my work is even shot at 1600 ISO.

 

Regards.

That is Canons paper written by them. And while there is no doubt that 35mm film especialy slow slides films and slow B&W film out resolves the half frame 6-8mp DSLRs your own tests prove this, you can't escape that fact that grain is often intrusive and much of the low contrast detail is lost in a sea of grain. If you take your Canon 20D sample VS the Fuji Reala Sample. In the fuji sample the blue grid lines are barely visible and the blue rivers what happened to them as for the gradient shadings you can't even see them but they are all clear on the Canon 20D samples. So while I can read smaller text on the fuji sample I would still prefer the Canon 20D. There is also no doubt that fuji Reala will have more dynamic range so in a situation where contrast would be a problem Reala could be a better choice. There are situations where one may be better than the other so why not choose based on the situation and experience.

Les, nevertheless, The Luminous Landscape has done such a comparison, and on February of 2003 wrote:

 

<p><i>In the above artcile I make the claim that the 1Ds produces images superior to medium format drum scans. I took a lot of flack for this when the piece first appeared, even though many pros who I have spoken with and who have started working with the camera are also coming to the same conclusion. It therefore gives me a chuckle (and no small satisfaction) to report that the review of the 1Ds in this month's Shutterbug magazine by Jay Abend comes to the same conclusion as I did.</i></p>

 

I would think that if the 1Ds is better than medium format drum scans, it's obviously better than 35mm drum scans.

Les, I'm not sure if we're reading the same thing. I believe he wrote that the Canon EOS D60 (6 Megapixels) surpassed 35mm film quality. If I'm mistaken, would you please quote the part of his review that indicates otherwise?

True Les that the UC100 is better than the Reala but still the rivers map grid lines and gradient shadings are not so visible and clear like the D20, UC100 has better resolution in terms of the black text on the light background there is also still the grain to be dealt with. I would still pick the 20D against the UC100 unless I needed the better dynamic range of UC100. We are looking at the what would be huge enlargements much bigger than most would expect to use either 35mm or a half frame 8mp camera for. Remember that the E10 Olympus has a very small sensor much smaller than a half frame DSLR so I won't really measure up that well at all but small prints from it would probably still be OK.

Additionally, I'd like to suggest that the reason your Olympus E-10 can't outresolve film in the way the D60 can, is because the E-10 has a sensor that is 8.80 x 6.60 mm as compared to 22.7 x 15.1 mm for the D60. That's a bit like comparing the image quality of <a href="http://en.wikipedia.org/wiki/Disc_film">Disc camera</a> to that of an <a href="http://en.wikipedia.org/wiki/Advanced_Photo_System">APS camera</a>. Though, the sensor size of the E-10 is actually smaller than the film size of the Disc camera.

I stand corrected! You have a great setup there. I only wish that I had too :) I have tried the XTi and I think it is fantastic, I'm still using my 20D as I got one of the very first ones. Still, I am not at all saying that digital is not up to the standards of film -- it can be far better in fact. It's just that to get that quality I've seen many film people struggle quite a bit, especially long-time photographers who are not so computer savvy. In fact I wrote virtualPhotographer so that film photographers could get artistic effects in one click. I am just trying to give a heads up to the OP so he can decide if this is a journey he wants to take, I hope it is.

 

I don't dis-agree with you at all -- it's just that there's a lot of surprises in digital, like having to sharpen the image, that are alien to film photographers. I think if people understand that, and have the right expectations going in, then the chances of success are high.

 

Thank you for keeping at it and setting the record straight.

 

Kind regards,

--Peter

John McPhereson, read my posts again....you were qouting Lucas Bennet, not me, Pavel. In my post I too was quoting Lucas, my bestest buddy here on PN.

 

And yes, digital has surprassed 35mm film...I agree with you.

 

And as for carrots, they are better then pickles...all true...I read it on the internet someplace so it's true!

Les, yes I've read the links. I'm quite well versed with the concept of links, so it's not necessary to go through the extra step of making them bold. ;-) After carefully reading the Canon white paper, I've come to the conclusion that the sentence you quoted means to say that full frame sensors are better than partial frame sensors, and not that partial frame sensors do not exceed the image quality of film.

 

As for the review of the D30, I too find it surprising. However, not having any experience with the D30, all I can do is refer to the samples provided. Nevertheless, regardless of how surprised I am by the D30 results, I can subjectively confirm that the Digital Rebel (which also has a 6 Megapixel sensor, similar to the D60) has better image quality than my own film shots taken with an Elan IIe, and subsequently scanned (the negatives) by a local pro shop. This holds true at ISO 100-400, and even more so at 800-1600. Furthermore, shots I took at ISO 800 on my Elan IIe are not as sharp and detailed as shots I took at ISO 1600 on my 20D (replaced the Digital Rebel). Regarding the ease of reproducing the LL comparison, I don't understand why that's difficult to reproduce. Is the difficulty in obtaining the equipment?

 

I have also reviewed your tests, but have had a difficult time making comparisons, because it appears the focal length is different from one example to another. Maybe it's the cropping, and not the focal length, but still...

 

Also, in your chart listing the varying combinations and their respective resolutions, you've got mixed units. The column that shows 4000dpi should actually contain the Megapixel equivalent of the film, which I believe you can get by multiplying the pixel width by pixel height, for each of the sources (and divide by one million).

Z-

 

I buy Fuji mailers from BH. Walmart does not process slide film. If I need faster turnaround I bring them to a local lab for overnight service which costs $11.

 

I took the kit lens off the FM10 and replaced it with a $300 Nikon 45/f2.8 pancake. Very sharp lens, sharper than my Nikon 50/f1.8 with better color and contrast. The FM10 with that lens makes a very nice package, small and light, has a nice velvety covering that is easy to grip when wet.

 

As far as being made by Cosina, so what? I couldn't care less who makes it or where it's made. All I care about is how well it performs.

 

I rarely look at slides on a projector. I use a viewer and a loupe. The good ones I scan (16 mega pixel file) and print.

Digital vs film? pfff I don't know, best thing for me would be to buy a DSLR and see for my self whether I like it better then my film camera I use now. Point is, at the moment I'm happy I can pay for a couple of slide rolls each month and have it developed. For now a DSLR is budgetaraly out of the question.

However since I have been following this site for the last two years I lately came to a strange conclusion:

Way back 2 years ago (?) the pro- and con-Velvia discusion was THE debate on this site. Some loved Velvia, others found it too much Disney colors. A minority had the clever insight: all film & slide have there usage (sounds familiar?). However the pro-Velvia crowd was so fanatical that Velvia was the total summit of film/slide and it was unreplacable. Those who suggested that Velvia was easily and more cost effective imitated by using photoshop, were burned down. No way that Velvia was to be simulated, it was so unique in appearance. Combined with the fact that for landscape only Velvia was to be used, I'm now totally confused that in this digital vs film rantings no one ever brings up this element in the discusion. Lot of Digital shooters are landscape photographers but apparently their DSLR's made them forget their precious Velvia.

Point I'm making, if any, is that memory is short, first only Velvia was good enough for landscapes, now it seems that a Canon 5D delivers better results? I'm not want to make a judgement which is true but it's funny how THE debate changed so fast in a such a short time.

for the record: all I said above is a general statement.

 

cheers

 

Marc

When I read the question, out of sheer curiosity, I went to get two prints that I have to make a visual comparison.

 

The first print is an 8x10 B&W fine print made by the late Fred Picker from a 4x5 Tri-X negative. Fred did some training under Ansel Adams. Fred always used to say that people didn't know how to print because they had never seen a good fine B&W print. Fred learned to print while having an Ansel Adams' fineprint at his side which he used as a reference point. Hence, through his Zone VI catalogue, Fred sold selected fine prints, made in bulk, so that beginners could have a yardstick as to what a good fine print looked like. Over the years, I've seen lots of B&W prints made by others, that had unrealistic contrast, and harsh blacks and burntout highlights. I've come to see Fred's wisdom in having a fine-print, done by an experienced photographer, as a yardstick.

 

The second color 8x10 print is one which I made from my HP 8450 printer from an uncropped Nikon D70 DSLR 6MP raw file. In fact, it was the first print I made on that printer. Hardly any manipulation. Took me a few minutes to bang it out.

 

As I compared both of these 8x10 prints, under a strong light, the 6MP print looks pretty good, side by side with the fine print made from a 4x5 negative.

 

I make this comparison only in terms of detail, grain and sharpness. I do not comment on dynamic range or smoothness of tones, although I mention that the 6MP photo is of the Taj Mahal close to sunset when the blue sky just starts to get its first pink tinge, before the full-blown red-sunset time. The digital 6MP print seems to have captured the subtlety of the just-pinky color.

 

I'm sure there'll be some on this forum who'll retort that I can't recognise a good photo when I see it. But, hey, I'm giving a comment here, not from theory, but from an actual side by side comparison between a good B&W 4x5 fine print, and an actual 6MP 8x10 uncropped print.

How is your result when you look at much larger prints?

I don't have any larger prints to compare. My HP 8450 only does up to 8x11, and the only fine print I have is 8x10.

Dark, when making your comparisons, keep in mind that if making a print in the size you describe, it's possible that the original would need to be upscaled. This is true of film or digital, but since with film you are doing so optically, the act of doing so is forgotten. Divide the pixels in one dimension by the ppi/dpi you expect to be printing. If this doesn't give you the dimensions you want in inches, you'll need to upscale. However, since you're likely starting with a cleaner original, the enlargement should look much cleaner.

...and you might try <a href="http://www.ononesoftware.com/detail.php?prodLine_id=2">Genuine Fractals</a> for the enlarging. Here's a <a href="http://www.imaging-resource.com/SOFT/GF/GF.HTM">review</a>. They also have a <a href="http://www.ononesoftware.com/download.php">free 30-day demo</a> available for download.

True, but it should mean that it can better interpolate the results, without also magnifying the imperections (that do not exist in the digital image, like dust on the film, etc.). I'd be interested in your opinion of the results of using Genuine Fractals with a <a href="http://www.canon.com/camera-museum/camera/digital/data/2004_eos-1ds_markii_pix.html">sample from a 1Ds Mark II</a>.

...or from the <a href="http://www.canon.com/camera-museum/camera/digital/data/2002_eos-1ds_pix.html">EOS 1Ds</a>.

I think at the end of the day Dark you are right on the border line as to whether film or digital could be better for you. The only way for you to be really sure would be to try and get files from the cameras you are looking at and make some prints to see if you would be happy with them. For me a 6mp DSLR replaces 35mm color neg film most of the time. If I wanted slides I would shoot slide film but they are not much use to me for most of what I do. I don't doubt velvia would be better for me if I shot landscapes but I don't that often. For B&W I still like Trix, HP5 and other B&W films but since getting the DSLR I can do nice B&W prints using the HP photo grey catridge. I still do shoot some film I like the workflow of dropping film at the lab and getting prints back without having to mess around with digital files. In that way I prefer film for some things.

<p><i>Also, since you provided a link to a free trial of Genuine Fractals, I am sure you would prefer to form your own opinion using your own photos</i></p>Except, I don't have any MF format photos, and I was under the impression that you did. :-)

i noticed something comparing les #1 6x7 scan to his d2x image. if you scale the d2x image up to the same dimension as the 6x7 crop in photoshop, it doesn't look as good due to pixelization. fair enough. i wouldn't expect an aps sensor to capture as much data as a piece of film over 10x larger in surface area.

 

but if you scale the 6x7 crop down to the dimensions of the d2x crop, the d2x crop is actually the one that looks better. it has MUCH better contrast and color. (see attached image.)

 

a d2x crop viewed on screen like this is like looking at a huge print, around 50 or 60 inches depending on your monitor settings.

 

so while saying a 6x7 piece of film out resolves a 12mp aps sensor is technically true, it doesn't look like it matters for anyone printing smaller than around 60 inches. for most subjects that people actually shoot, it may not matter for even larger sizes.

 

i can see how reichmann came to his conclusions. he was shooting a skyline with a full frame body. the skyline didn't have lots of really tiny curves like the letter shapes in this map test. most subjects don't. a little bit more resolution from full frame, a test with a little less stress on really tiny curves, and you have two images with the same detail but better contrast and color from the digital. just like reichmann's crops show.

 

i don't know about you guys, but i wouldn't bother with the trouble and expense of processing and scanning 6x7 film over a small advantage that only shows in 60 inch prints, and then only with some subjects. 10mp aps sensor cameras are now normal. if an aps sensor camera is this close to film with over 10x the surface area, then digital is truly higher resolution (imagine a 6x7 sensor!) and this debate is pretty much over.<div></div>

How about <a href="http://www.dpreview.com/news/0609/06093007seitzd3.asp">6x17</a>? ;-)

you would have to back far away from the map to give that 6x17, 160mp camera a challenge! :-)

 

it may seem impossible right now, but with the way electronic equipment improves and drops in price...something like that may very well be affordable to the masses in a decade or so.

 

btw, les - responding to a test between canon ff and mf by claiming a marketing piece says you cannot compare them, when it says no such thing AND the images shown clearly show the opposite...STUPID on your part.