Imacon Flextight resolution

[Peter_S]

Hi!

I am new to the forum, and hope you can answer me a question. I have used both analog and digital in the past, but have gone back to analog (and Foveon when digital).
Every once in a while I have a small collection of my best photos scanned at a scan service with a Imacon Flextight scanner. The output files have up to 6300dpi and 8732 x 5928 px in size. The size would correspond to over 51MP resolution, but following the "Is my Bronica SQ-A an 81-MegaPixel camera??" thread I am wondering - what digital camera resolution does this correspond to roughly in real world, when a sharp film and sharp lenses are used?

I I shot a Contax G2 with the 28/45/90mm lense and use mainly Ilford XP2, some other b/w films on occassion, and will try Kodak Ektar 100 for color.

Not that is matters much (I shot film for other reasons than quantifiable IQ) but I am still curious.

Thanks,
Peter

 

[Rudeofus]

There's a number of issues to consider. Neither system will give you nearly as much resolution as the numbers claim, and for many reasons:

• Film grain. 'nough said. In addition to the grain you'd already get with regular B&W film, you picked an ISO400 film based on C41, so you have dye clouds instead of smaller grain.
• The digital cameras either have a bayer sensor, which counts all the pixels despite the fact that they register just one color channel each. With foveon the joke is even worse, the have below 5 megapixels and then upscale to 14+ MP just to look competitive.
• Lens resolution. You use high quality fixed focal length lenses, possibly stopped down, yet don't forget that 6300 dpi is 248 dots per mm, so it's well over 100 line pairs per mm. Few lenses can do this, especially off center. Also, don't forget the diffraction limit if you stop down.
• Antialiasing filter for digitals. Either you have them and lower your resolution, or leave them out and have artefacts. Or have something in between.
• DOF. The typical equations for DOF and hyperfocal distance are stated for a CoC radius of 25m, which is well above sensor pixel size or this scanner's resolution. And DOF gets smaller nearly proportional to CoC limit.
• Camera shake, mirror slap, motion blur, shake from the shutter going, ....
• Air between lens and subject. If you read the lens tests on photozone, the super teles came out worse than the kit lens. Why? Because the air between lens and test chart blurred the view.
• Postprocessing artefacts. People say "you have to sharpen" and often don't realize what this means to actual reproduction of details.

There used to be an article online with the title "Beating the 50 lpmm barrier" or similar, unfortunately it got lost in bit heaven. It essentially boiled down to all the problems one has to overcome just to reach 50 lpmm consistently.

 

[pellicle]

Hi Pete

Hi!

I am new to the forum, and hope you can answer me a question. I have used both analog and digital in the past, but have gone back to analog (and Foveon when digital).

...
I have a small collection of my best photos scanned at a scan service with a Imacon Flextight scanner. The output files have up to 6300dpi and 8732 x 5928 px in size. The size would correspond to over 51MP resolution, but
...

its a tough question. I recently had some stuff done on an X-5 and was stunned by the output. However despite knowing that I still take my G1 along with me instead of my 35mm rig because its lighter and 'good enough'

they're pretty close though. I've found that slide gives cleaner results than negs but that negs give better range in harsh light.

swings and roundabouts ... wish there was an easy answer, but I use digital for the convenience and 4x5 when I feel like film.



not sure if any of the following help you, but if nothing else you'll know others are thinking in similar circles

http://cjeastwd.blogspot.com/2009/09/g1-and-wide-angles.html the purpose of which wasn't intended to be a film vs digital comparison, but it led me there.

http://cjeastwd.blogspot.com/2010/01/noise-about-film.html

http://cjeastwd.blogspot.com/2009/10/two-films-and-digital.html

http://cjeastwd.blogspot.com/2009/10/film-cost-effective-still.html

 

[Peter_S]

Hello Rudeofus and Pellicle,

thanks for your answers! They make sense.
And Pellicle - the links are awesome, I looked for something like that for a while, thanks. I was contemplating m4/3, but went the film route again, for various reasons.
I am starting to wonder if the full res. on X1/X5 is overkill and only enlarges grain in the end? It will depend on the grain I guess. For a Pan F 50 it makes probably more sense than for a XP2 or Tri-Max.

Anyhow, I was impressed with the results of my scans, too, and for the little (but selected) occassions I use film this hybrid works just fine.

Thanks,
Cheers,
Peter

 

[Rudeofus]

I am starting to wonder if the full res. on X1/X5 is overkill and only enlarges grain in the end? It will depend on the grain I guess. For a Pan F 50 it makes probably more sense than for a XP2 or Tri-Max.

From a signal processing stand point: yes, 6300 dpi is overkill IMHO. But these nice theories sometimes fail because our eyes and brains are not computers. Even if all you do is enlarge grain, the result may still give you the impression of higher sharpness and resolution. Look at pointilist images. Although these images consist of large visible dots, they appear tack sharp. Try to blur such an image until the dots blend, and it will look extremely soft despite containing essentially the same information.

 

[Peter_S]

Thanks again, Rudeofus!

 

[toyotadesigner]

If you want resolution there is no way around a drum scan. The Imacon is a nice machine, but a Nikon LS 9000 delivers almost identical results for slides up to 6x9 cm. For 35mm the Nikon LS 5000 is just perfect, unless you need prints at 3x2 meters...