The GetDPI Photography Forum

Great to see you here. Join our insightful photographic forum today and start tapping into a huge wealth of photographic knowledge. Completing our simple registration process will allow you to gain access to exclusive content, add your own topics and posts, share your work and connect with other members through your own private inbox! And don’t forget to say hi!

MFD ... Emperor's New Clothes?

fotografz

Well-known member
i'm with bob, everything is digitized at some level.

as far as film, all of my work, B&W 6x6 and 4x5, consisted of deliberate, slow shooting and careful darkroom printing. form many of the images taken many years ago, I can still remember the details of where I set the tripod. my enlarger used a cold light to de-emphasize the crystalline nature of the emulsion and minimize surface artifacts. all of this is different from digital shooting today; no wonder the film has a different look
Everything digitized ... that may be true regarding the internet, but not necessarily everything ... like I said, a majority of the prints on my wall have no digital involved at all.

However, my comparisons were of digitized film images, and to my eye they still look better over-all. What especially looks better are B&W silverprints that were scanned on a high-end flatbed.

Also, 99% of my film work wasn't done in a slow, plodding manner at all ... instead in a more candid, intuitive and spontanious way.
 

fotografz

Well-known member
Marc I have experienced the same regarding personal favorites and after much thought I had to put my mathematicians head back on to explain the phenomenon. I think we are experiencing a simple statistical error in concluding film produces more keepers versus digital. The error is a sampling error involving a number of potential biases - most notably the so called 'survivor bias'.

Simply put - if a person has been shooting for many years - by definition the film years ( still?) outnumber the digital years. Added to this is the fact that in using personal work to define 'favourites' and personal work often having a bias towards film shooting ( I know that is the case of me) a person is actually biasing towards film again.

Only time will tell...

Regarding film 'profiles' versus digital processing - I still prefer to use B&W, although recently I have become impressed with SilverFX.

One thing for sure as Stuart alludes to - the whole process of making images with film is different and I agree in many ways more involved and contemplative. Again a different process will deliver different personal value beliefs again biasing the outcomes! -:)

I wont be selling my Leica MPs, my XPan or my MF film backs any time soon.:ROTFL:
Peter, I totally grasp what you are saying about statistics and biases.

And that could well be the answer except for the fact that the "subjective review" I did was of my wedding work for which I have been shooting digital far longer than film.

I'd say the math would be 30 to 1 in favor of digital in terms of output and selection of " potential bests" chosen primarily on the basis of content. I don't pick annual keepers based on the media used, but instead all the other creative criteria. Of those in the final selection edit, the film stuff stands out as visually more 3D and all the other stuff I mentioned above.
 

fotografz

Well-known member
Interesting. I really prefer digital for that reason -- because I am more involved in the process. Heck, I control the whole process. At the time of capture I can check the histogram to make sure I am capturing all the information available in a scene (or at least not losing anything I need), and then in post I can craft the picture exactly how I want it.

With film, I feel like I am just along for the ride. I can control the general tonal curve of the image by selecting a film with that response and deciding whether to process straight or push, and trying to guess at an exposure that records my vision of the scene faithfully. BUT, I don't really know what it looks like until I get it back from the lab, if there is a mechanical problem with the camera I won't know until I get the film back, if my lab screws it up, if the film scans poorly, if if if if if if if. My 10D was an eye opener because for the first time I was in total control of the entire image making process and it was like having a polaroid for every image!

The other thing that really shaped my view of the photographic process was when I started selling all my Canon gear and bought Leica, but had not yet found a DMR. At the time I was doing digital work for a well know travel photographer who would come back from a trip with 5,000 - 10,000 images from his digital Nikons. Sorry to toot my own horn, but I worked digital magic and made the images really look like film. Every. Single. One. But as fast as I was, it was still a lot of time, but there is no way anyone could turn around that volume of film on the deadlines we had. Since I was sans digital I was back to shooting film. It seemed like every frame that came back from the lab was perfect -- no processing necessary.

I had neither time nor energy for my own work (even dealing with clients) so I loaded up the R8. The film that came back blew me away. Staring back at me was the look I had spent years trying to create with Canon and Nikon digital. Even with the DMR, I have created "recipes" for NPH, Provia and Velvia.
Interesting, in all the time I shot film I guess I was lucky ... I never had a camera failure that I didn't sense immediately (except a shutter bounce that clipped 1/8" off the negs), and only lost 3 rolls of film to errors ... one I gave away to a tourist in the mountains who had run out of film and I graciously gave an exposed roll to because I didn't wind the leader all the way in ... LOL. The other 2 were lab errors.

In contrast, I've had a digital camera that displayed the jpgs on the LCD but didn't write the file, an undetected CF card failure, and various user errors ... one of which was the catastrophic failure of a Western Digital hard drive while I was backing up the contents to DVDs .... 2500 images irretrievably lost because I didn't back-up as I went and waited to long ... my bad.
 

fotografz

Well-known member
Yeah, that would be me (amongst others, I'm sure). Hey, what happened to those incredible "fat pixels" of the CFV?

Not that I regret it for a minute....film or fat pixels, either way works for me. :thumbs:

Gary Benson
Eagle River, Alaska
Proud owner of a CFV-II
Yeah, that love for the CFV is predicated on one main thing isn't it? ... It looks more like film than most other digital solutions ... at least the color files do :ROTFL:
 

harmsr

Workshop Member
I only shot 35 mm film and never anything larger, so I can't say that I have the same experience as some of you other guys. I also got so soured on poor film processing by the "pro" labs in the area, that I gave it up completely except for B&W which I process myself and shoot from an M7. However, that mainly is just for fun.

I switched to MFD, to get the tonality / DR / color / & resolution which I was not getting from 35 MM digital or film.

I do have to say that I personally like my digital files better, except when I'm going for a more classic B&W look.

The funny thing is that, I'm not even sure that you can truly call my film shots as film shots. Once developed, they get scanned and then printed digitally. Nothing goes through a full wet darkroom anymore for me.

So, hello digital and loving it.

Ray
 

Jack

Sr. Administrator
Staff member
Everything digitized ... that may be true regarding the internet, but not necessarily everything ... like I said, a majority of the prints on my wall have no digital involved at all.
Marc, I'm going to take you to task here :).

If you look at the science of silver-halide exposure based emulsions, you will find it takes a fixed and measurable number of photons to 'convert' a single crystal of silver halide. Once that happens, it remains in the activated state, and as such, that grain has a binary response to light -- it is either exposed or it isn't, on or off, a zero or a 1.

The grains are either of differing sizes or stored at different depths in the emulsion, or both, and this is how some get converted and others don't, but it is still a binary process -- a single grain either receives enough photons to expose or not. So a group of the crystals together form a sort of matrix that allow for a range of tones within any given "volume" of emulsion. That volume of emulsion then re-acts to varying quantities of light in a fashion very similar to a ........

single digital sensor pixel. :D

Now the above *may be* poitning us in a direction to help explain why certain DB's tend to resemble film emulsions -- at least our favorite ones -- more than others; the photometric response of their single pixel (or more accurately the Bayer set of 4 pixels) is for whatever reason very similar to the photometric response of a similar volume of film emulsion.

Cheers,
 

carstenw

Active member
True enough, but film still has a large random factor, in the size, shape, depth and number of layers in the grains, and this isn't matched by digital. So, although each crystal acts in a binary fashion, the distribution still has a random, yet statistically distributed look, whereas digital is deterministic.

On the other hand, and perhaps some of the more expensive film-look filters do this, it ought to be possible to "play back" the photons recorded by the sensor over a randomly generated grain soup, and get something closer to film.
 

Jack

Sr. Administrator
Staff member
Agreed Carsten -- so just you wait until the next MF digital sensor has a random array of 6, 6.8. 7.4 and 9 u pixels... Or perhaps more likely, the 6 u sensors will get slightly randomized conversion algorithms for the deBayering.

:D,
 

carstenw

Active member
It would already help if they would build in a slight, random offset in each microlens.

In fact, and this is a current area of research in graphics chip design, if they could even pick a specific, repeating, *apparently random*, tile-able pattern, that would help.
 

Jack

Sr. Administrator
Staff member
Except not all the backs have microlenses...

As for 'randomizing' the pixel patterns for images, there are already ways to implement this after the fact digitally -- and at least one method I'm working with actually works quite well; I have used the technique for a few *very* large prints to amazing effect. I am still in the research stages of my idea so not ready to share it, and add it is relatively time-consuming in it's present form, but at least actionable in CS so maybe has some promise.
 

bensonga

Well-known member
Yeah, that love for the CFV is predicated on one main thing isn't it? ... It looks more like film than most other digital solutions ... at least the color files do :ROTFL:
Hi Marc,

Ahhh....so the CFV is the best of both worlds. :)

The convenience of digital and that "almost like film" look. I'll take it! Now if Hasselblad would only make a full frame square sensor CFV-III back, I might be able to retire my A12 film backs. I know, not likely. :(

To tell you the truth, I still don't see that "film is inherently better". Either in BW or color. I guess my eyes are just not that good. In which case, I'll continue shooting both, because it works for me.

I will say this....in my own work, I've yet to surpass the look of my scanned 4x5 film with anything I have in digital capture. The CFV-II comes so close I can almost taste it. If I had the money for a better Leaf, Sinar or Phase One back then I really think digital would equal the best that I have personally ever gotten from scanned film....with the caveat....I'm not using an Imacon scanner like Marc, which I'm sure must up the ante considerably.

Gary Benson
Eagle River, Alaska
 

charlesphoto

New member
Marc,

I 100% agree with you. That's the reason I shot my new book on breakdancers with medium format film (Neopan 400), a seemingly crazy thing to do considering the cost and difficulty of shooting live action with MF (and it was all out of pocket). But when you look at the images they really look larger than life, and the tonality is esp nice - and forgiving as I was using a flash. And it sets it apart from the rest of the pack. All scanned for the book on my Imacon with 8.5 x 11 match prints on my 4800.

I don't know why, but most MFD I see leaves me cold. But then I'm not shooting product, etc and have no need for it. A few rolls of film MF here and there and my trusty Imacon come in a lot cheaper than investing in a MFD system. I'll admit I am hooked on the M8 and D3 and my M7 sits and gently weeps.
 

BradleyGibson

New member
Apropos?

Not exactly medium format digital, but certainly a very large print...

Parallels some opinions on this thread...

http://fwd.five.tv/videos/challenge-blow-up-part-3

Regardless of the outcome of this amusing little test, I still say beauty is in the eye of the beholder, and feel that is far more of the answer than any of the objective differences between the media.

Enjoy,
-Brad
 

Stuart Richardson

Active member
Not that I will necessarily disagree with the results of the test, but who would choose 400 iso 35mm film to make a 100 foot tall print? Anyway, I agree with Brad here.
 

carstenw

Active member
A quality per dollar per kilogram test would be interesting. You could get a really nice, compact, light, cheap, high-quality MF film system for the same price as a 35mm DSLR, and that comparison would hardly be as lopsided. 35mm film was always a compromise.
 

beamon

New member
Thanks, all, for a good and fun read. Though it will be discussed for as long as there is basically different and competing methodology, this exploration has been particularly well done. The fact that no animosity :angel: has crept into the thread has made it even more enjoyable.
 

Jeremy

New member
I have been contemplating a D700 for a bit lately, but it makes zero economic sense for me.

I buy my 35mm film in big lots off of eBay for about $1/roll and then I process them at home (if B&W) or more likely at Target for $1.08/roll. That's about $2/roll for 36 exposures, but we'll even hedge and say $3/roll.

I can scan these with a Canoscan 4000 at 4000dpi and get a 12" x 18" file @360dpi which is about the largest I print outside of my 4x5 shots.

I would have to shoot 30,000* images with the D700 before I broke even and that's not including the CF cards, extra batteries, and more hard drive space (I don't scan every negative shot) I would have to purchase while my Nikon F3hp is already bought and paid for and runs on rechargeable batteries I already own. I am also not a shoot everything and let the editor sort them out type of shooter, but much more contemplative. I have no need for instant review as I turn off the LCD anyway on the digital cameras I use and I don't need the extra speed as I'm already shooting at f/1.2. Instead I'm going to funnel this money into saving up for a Mac Pro.

I have a feeling that I could probably print larger images with the D700 than my 35mm scans, though it would be close, but if I want to go large I can wetmount 4x5 or 8x10 and make some ridiculous non-interpolated files.

Of course, this is just for me where I don't have any turn-around windows as I'm firmly in the fine art camp and not commercial. Everyone's needs differ, but lots of times for those without the need for quick turnaround the math, IMO, doesn't work out in favor of digital at the higher quality threshold.

* D700 priced at $2500/$3 per roll of film = 833.33 rolls of film * 36 exposures/roll = 30,000 exposures.
 

Dolce Moda

New member
I have been contemplating a D700 for a bit lately, but it makes zero economic sense for me.

I buy my 35mm film in big lots off of eBay for about $1/roll and then I process them at home (if B&W) or more likely at Target for $1.08/roll. That's about $2/roll for 36 exposures, but we'll even hedge and say $3/roll.

I can scan these with a Canoscan 4000 at 4000dpi and get a 12" x 18" file @360dpi which is about the largest I print outside of my 4x5 shots.

I would have to shoot 30,000* images with the D700 before I broke even and that's not including the CF cards, extra batteries, and more hard drive space (I don't scan every negative shot)

* D700 priced at $2500/$3 per roll of film = 833.33 rolls of film * 36 exposures/roll = 30,000 exposures.

Wow ... that's perspective. I have been thinking about getting a Phase One back for my Hasselblad H2. I did your little math thing and worked out that I would have to shoot and process somewhere in the neighborhood of 700 rolls of film (11,000 frames) to break even. As I don't yet shoot commercially yet, this is a no brainer.
 

Stuart Richardson

Active member
While I don't mean to deny your conclusion, it is not necessarily as clear cut as it sounds. If I am going to be a bit of a devil's advocate, I would say that that kind of calculation assumes equal results and equal effort expended. For example, the D700 will dramatically outperform film from a high ISO perspective. It also has many different features which may or may not mean anything to you (Auto ISO, live view, instant review). Film may look better in black and white. Also, how easy is it to get the files you need? Scanning is a time consuming process, while modern programs like Lightroom, Aperture and Capture One can allow you to process many files at once.

They are really different things and it is hard to choose one over the other for anyone other than yourself.
I find that I generally prefer working with film for my own photography, but when I am doing something professionally or under the gun, the digital really adds a sense of security.
 
Top