Interesting article: "36 Megapixels vs 6×7 Velvia" and 4 x 5

[timparkin]

With so many factors it would be tough to do.. I could do a 'cheapest film camera that still trumps digital' which could possibly be the Bronica? but then it loses by weight? So how do you score things? I'll write articles descibing pros and cons but I can't do much more than that I imagine..

[Ari]

Thank you very much for this, Tim; greatly appreciated.
Though had you come to a wholly different conclusion, I'd still be shooting my 8x10 no matter what!

[David A. Goldfarb]

My only reservation/observation is that the Mamiya 7 used in the test is a rangefinder camera: precise focus is dependent on calibration of the rangefinder mechanism.

It's not as if SLRs are without calibration issues. Looking through the prism gives the illusion that the image is in focus, but presumes that the mirror is aligned and the focusing screen is in register. Usually it is, and a rangefinder can be a delicate mechanism, but I wouldn't take SLR focus for granted.

[Ken Lee]

I've never had an issue with SLRs but had to send one of my Mamiya 7 cameras back for calibration several times.

Mamiya USA required several tries before I eventually gave up on them and sent it to Precision Camera Works.

Short wide angle lenses with abundant depth of field are less troublesome but longer lenses like the 150mm require precise calibration: otherwise those world-class optics are... for naught. Getting my camera calibrated to use both an 80mm and 150mm is even more challenging.

According to Precision Camera Works "The most common repair request we receive for the 6, 7 and 7II models is for the proper calibration of a camera's rangefinder. We also receive occasional complaints about a discrepancy in the rangefinder accuracy between one or more lenses in a set..."

[paulr]

The real world comparison for me takes into consideration the cost of high quality drum scans. I can only afford them under rare circumstances.

If you do a comparison like this with desktop scans or with darkroom prints, things skew 100% the other way. 60 inch prints from my 36 megapixel files look better than 50" darkroom prints from 4x5 (don't argue ... I have examples. It's no contest).

Prints from high end drum scans of those same negatives would look considerably better than the 36mpx prints, but that's not going to happen.

[Christopher Barrett]

I was playing with tests of this nature last summer (posted the results at LuLa) and preferred the look of film over digital, though many of their qualities were very similar. I'll do another shootout shortly and post the tests for you guys. Since I happen to have an M-Line 2, a roll film back, an IQ 260 back and an A7r that will mount on the view camera, I can shoot everything through the same lens and then drum scan the film (Portra 160) to do some highly critical tests. Let me know if there is anything specific you want to see.

I'll likely compose a still life set and light it with extreme contrast in part of the scene.

CB

[Bernice Loui]

Digital imaging as a technology must have filtering and varying degrees of numeric processing which can and does alter the results. How the results are altered can be widely adjusted to deliver a specific perception of the resulting image, i.e. the illusion of extreme sharpness-resolution when it is not really there in the original data acquired.

Some time ago during a LF get together, one of the participants show me a digital print that appeared really, really sharp with extreme resolution. Within moments of looking at this digital print, I noted that it has been digitally sharpened. He was very surprised that this was visible in the print, with that an explanation of how and what visual clues in this print pointed to the fact is has been digitally sharpened. As that discussion went on, it turns out that image was made at f90, then digitally sharpened and worked over to produce a sharp, high resolution, contrasty and snappy look.

Point being, digital is a completely different technology to film and their results are inherently difference is baked into how they work to produce images. IMO, comparing the two is more academic than realizing they should be treated as different imaging technologies producing different results.



Bernice

Not sure what you mean - there is definitely moire in the test images, exacerbated because it was a D800E.

Tim

[timparkin]

The real world comparison for me takes into consideration the cost of high quality drum scans. I can only afford them under rare circumstances.

If you do a comparison like this with desktop scans or with darkroom prints, things skew 100% the other way. 60 inch prints from my 36 megapixel files look better than 50" darkroom prints from 4x5 (don't argue ... I have examples. It's no contest).

Prints from high end drum scans of those same negatives would look considerably better than the 36mpx prints, but that's not going to happen.

Not going to argue I'm just going to say $23 for a drum scan and my 4x5s blow away my A7R files even with an Epson flatbed..



In fact my Mamiya 7 gets pretty close to matching my A7R on an opticfilm 120



I'll let you extrapolate that out to 5x4 if you like..

My only argument would be that it's easy to get the best out of digital but it takes some skill to get the best out of scanners. I'm not going to argue with the results you get though...

Tim

[timparkin]

Digital imaging as a technology must have filtering and varying degrees of numeric processing which can and does alter the results. How the results are altered can be widely adjusted to deliver a specific perception of the resulting image, i.e. the illusion of extreme sharpness-resolution when it is not really there in the original data acquired.

Some time ago during a LF get together, one of the participants show me a digital print that appeared really, really sharp with extreme resolution. Within moments of looking at this digital print, I noted that it has been digitally sharpened. He was very surprised that this was visible in the print, with that an explanation of how and what visual clues in this print pointed to the fact is has been digitally sharpened. As that discussion went on, it turns out that image was made at f90, then digitally sharpened and worked over to produce a sharp, high resolution, contrasty and snappy look.

Point being, digital is a completely different technology to film and their results are inherently difference is baked into how they work to produce images. IMO, comparing the two is more academic than realizing they should be treated as different imaging technologies producing different results.



Bernice

Agreed but barring gross stupidity if you take a photographer who uses both film and digital seriously and specialises in scanning film then I would hope that you can get the best out of each medium. And yes, if you read the articles it mentions that digital has a peak contrast at close to it's nyquist and then drops to zero whereas film's contrast drops slowly as detail gets finer. The grain in film limits applying sharpening too.

Which is why the comparisons weren't really 'interpreted' and no gross conclusion drawn. The results should stand for themselves...

Tim

[Greg Miller]

This seems to come down to priorities. The CMS20 seems to resolve line pairs very well, nut has noise (grain) issues. Mamiya 7 Provia has less noise, but doesn't resolve lines pairs as well. The IQ180 seems to have better resolution of the numbers (6, 8, 100) than the Mamiya 7 Provia and much less noise, but quite a bit of digital artifacting in the line pairs. All of which is seen at a magnification factor that is not declared (unless I missed it) so had to know if meaningful in the real world. One thing I know for sure is that every time I look at my old 35mm Velvia slides I am disgusted and the amount of grain or dye blobs in many skies.

So much agonizing over technical details. I wish people spent as much time working on the artistic side of photography.