# Film Vs Lens resolution.

Discussion in 'Film and Processing' started by acm, Feb 11, 2012.

1. ### acm

Nikon introduces D800. 36 Megapixels. Photographers start questioning are there lenses there to do justice to 36 MP resolution?
My question is, did film ever surpass the best resolution available in some best of lenses?

2. ### richardsperry

Film grains are microscopic.

I think your question is hypothetical because the human eye cannot resolve to film resolution, regardless of quality of the
lens. Or is that theoretical? Don't know.

3. ### alan_marcus|2

As you know, we focus our camera changing the lens to focal plane distance. Light rays from each point on the subject transverse the lens. The camera lens is a converging lens. The light rays are guided to form a cone of light. Sharp focus is achieved when the cone of light just kisses off on the film or digital chip. In theory, this causes a point of light to play on the film or chip. Now a point is so tiny that it is immeasurable. What happens in reality a tiny circle of light plays on the film or chip. This circle of light is often called the circle of confusion because when examined under the microscope it has poorly defined margins. It is also called the Airy Disk after George Biddell Airy (English 1801 - 1892 Astronomer Royal).
Airy's disk was studied at length by John William Strutt, 3rd Baron Rayleigh (English 1842 - 1919 Nobel Prize 1904).
The Rayleigh Criterion of theoretical resolving power:
Resolving Power = 1392 ÷ f/number
Table of R.P for wavelength 589 millimicrons as to resolved lines per millimeter

f/1 = 1392 f/2 = 696 f/2.8 = 497 f/4 348 f/5.6 = 249 f/8 = 174 f/16 = 87 f/22 = 63
The resolving power of any lens decreases as the lens is stopped down due to diffraction. This is due to the light rays passing in close proximity to the blades of the Iris diaphragm aperture. These close passing rays bleed into the
shadow regions and into the path of light that clears the aperture. The result is the boundaries of the Airy Disk become less clearly defined. Note: a good lens operating at apertures f/8 and larger exceed the resolving power of pictorial film. This creates reserve resolving power that will cover inaccuracies of focus, poor film flatness, residual aberrations
within the useful circle of good definition (central portion of the image).

Under favorable conditions, the human eye can resolve a disk held 1/3000 of its diameter way from the observer. Because the resolving power of the eye is less under most photographic condtions charts and dept-of-field tables use 1/1000 of the viewing distance. This works out to a circle 1/100 inch in diameter at 10 inches viewing distance.
I like to call this stuff gobbledygook. What do you call it?

4. ### randrew1

As a practical matter, some films have exceeded the resolution of high quality lenses. Some years ago, Kodak put out a chart with magnified images comparing all of the then available color neg films. There was a note beside the Ektar 25 image explaining that the image sharpness on that film was limited by the lens.
IMHO, resolution doesn't count for much in most images. Sure, the human eye has some ability to detect patterns out to about 50 cycles per degree, but the peak of the eye MTF curve occurs at about 1 cycle per degree. The contrast of a lens at 1 cycle per degree has much more to do with the perceived sharpness of an image than anything near the resolution limit.

5. ### franklin_polk

There are a few slow films that have no problem resolving higher than many lenses. Ektar 25, as pointed out, and rebadged aerial films such as Tech Pan, Adox CMS, and Rollei ATP should handily out resolve most lenses out there. Kodachrome 25 and Panatomic-X were probably higher resolution than many of the lenses they were used with as well. IIRC, Zeiss tested 400 lp/mm with the new Biogon 25mm on film.

6. ### alan_marcus|2

Kodachrome 25 resolved 157 lines per mm
Velvia 160 lines per mm
Ektar 25 200 lines per mm

8. ### acm

Thanks everyone for the inputs. Now my next question is this: D800 at 36 MP and Ektar or any other sharp film: which resolves better?

9. ### randrew1

I believe a more useful question would be: Which produces sharper pictures?

10. ### charles_stobbs|3

Alan, I think resolving a 3000' diameter disk held a foot away is no great task for the eye. The arms maybe.

11. ### mjferron

"Now my next question is this: D800 at 36 MP and Ektar or any other sharp film: which resolves better?"
As one who just hung a roll of Tmax 400 to dry I have to say I personally cannot get any 35mm film including Tmax 100 to match the detail I get from my my old Canon 5d. But that's just me and I still love to shoot film so I can wet print if i want.

12. ### alan_marcus|2

Mr. Stobbs,
That's a 1 inch disk viewed from 3000 inches (10mm disk viewed at 30 meters). This is the limit for a young person with 20 20 vision in bright sunlight. For older people the resolving
power of the eye is much reduced. It is common practice when preparing depth-of-field
tables to use a lower standard. This works out to be 3.4 minutes of arc. This translates to a circle 1/50 inch in diameter viewed from 20 inches. That's 0.50mm. That value is considered a practical limit for a contact print.
Now a 35mm negative must be enlarged to make a reasonable size print. Commonly 10X magnification is planned. Thus, the size of the circle when calculating depth-of-filed tables is commonly 0.05mm on the film.
Lens makers simplify this math by stating the circle size as a fraction of the focal length. The most common figure tossed about is 1/1000 of the focal length. Using this value for a 50mm
lens, the maximum size of the circle of confusion is 50 ÷ 1000 = 0.05mm. For precision
optics Leica uses 1/1500 = 0.33mm and Kodak 1/1750 = 0.0286mm.
Again, the size of the circle on the finished print is 0.5mm based on viewing at normal reading distance. For prints in a gallery or on a mantel, the standard is relaxed.

13. ### chris_antidote

A 35mp camera will most more than likely surpass most films and their resolution. That may be a different story with larger format films. However, that question doesn't really interest me. I'm more interested to see the color tonality and dynamic range of this new camera. Digital always seem to lose it's mind when it hits highlights and deep shadows. Trying to render them both always turns digital photos into a cartoonish look. I imagine with their new processors you'll be able to really tweak blending options and have that little camera shooting ugly HDR all day.
Either way, the progression is astounding.

14. ### randrew1

For older people the resolving power of the eye is much reduced.​
The primary vision problem with "older" people is presbyopia which is a hardening of the crystalline lens. The effect is to turn the eyes into fixed focus image sensors. With proper vision correction, the resolution is close to that of much younger people. With my trifocals, I still have 20/20 vision.
For elderly people (elderly is my age + 20) the iris starts to get a bit smaller which will reduce resolution.
The other frequent vision problem of older people is cataracts. In this case the lens becomes translucent. It is a fairly simple outpatient procedure to remove the existing lens and replace it with a manufactured intraocular lens. With an accommodating intraocular lens, some focusing adjustment ability can be returned to the eye.

15. ### a._t._burke

Gentlepersons:
Hmmm, my math is not what it was, so you’ll have to help me out. I mistakenly thought that the Nikon D800 produced 4912 rows of receptors (lines) in about 24mm vertically. Kindly humoring an old fool that learned “old math” please don’t laugh all that hard when I come up with about 102 lp/mm of resolution. Again, doddering that I am, I think that an information unit in a Bayer Array comes from using two green and one each of red and blue pixels. This gives a true resolution of about 51 lp/mm, which is then interpolated back up and reported to the file at 102 lp/mm.
Prior posters have claimed particular films resolve at 800 lp/mm, 200 lp/mm, 160 lp/mm, and 157 lp/mm. I’ll throw out Techpan at 300+ lp/mm and Bluefire (probably Adox) at 500+ lp/mm. Mr. Polk posted that Zeiss tested 400 lp/mm with their new Biogon 25mm on film. That means that both the Biogon lens and the film have more than 400 lp/mm capability as the results are lowered when used together.
I attended San Pedro, CA’s Dana Jr High in 1930. There I took some math classes. I’ll bet that my math teachers would have thought that any of the lp/mm figures in the second paragraph exceeded any of the lp/mm figures in the first paragraph.
Math sure has changed.
A. T. Burke

16. ### derek_hofmann

This gives a true resolution of about 51 lp/mm...​
That's the chroma resolution. The luma resolution is the 102 lp/mm figure. Both are "true" resolutions, depending on the context.

17. ### a._t._burke

Mr. Hofmann...
I’m puzzled by your answer and see it as a non sequitur. Chroma and Luma are ways that the eye sees. Am I not reading you right?
I was posting about needing four pixels in a Bayer Array to establish a point of reference. Then the camera’s computer chip interpolates or resses up the file by making a guess as to what each pixel under each of the four filters should be. In other words a 36mp chip does have 36mp but because of the Bayer Array is really a ressed up 9mp result. Because a two dimension chip has two axis the lp/mm is divided or multiplied by two rather than four. That is where my 51 or 102 numbers come from.
The math I’m using is as follows:
4912/24 (mm) is 204 2/3. Divide by two to get pairs is 102 1/3, say 102 lp/mm for pixel count. Then divide by two to get the lp/mm for reference points or 51 lp/mm.
All of this happens before the eye sees anything. Hence neither Croma nor Luma are introduced yet.
A. T. Burke

18. ### derek_hofmann

But there are twice as many green pixels as red or blue. Therefore, if the red and blue channels each have 51 lp/mm, then green has 102 lp/mm.

19. ### john_stockdale|2

"My question is, did film ever surpass the best resolution available in some best of lenses?"
I think I read in some old Kodak literature that for improvement in lens resolution to make no difference to the final result it had to be better than three times better resolution than the film. Likewise, for improvement in film resolution to make no difference in final result, it has to be three times better than the lens resolution.
Given that improvements are expensive in either dollar cost or inconvenience (low speed etc) the "sweet spot" might be where the two are about equal.

20. ### Karim Ghantous

a 36mp chip does have 36mp but because of the Bayer Array is really a ressed up 9mp result.​
Not so. As Derek pointed out, luminance data is still 102 lp/mm. Only colour data is interpolated (but this means that sharpening is recommended to undo the softness induced by this process). This might mean less accurate colour but the resolving power of the sensor is always 102 lp/mm.
As one who just hung a roll of Tmax 400 to dry I have to say I personally cannot get any 35mm film including Tmax 100 to match the detail I get from my my old Canon 5d​
35mm film easily matches the 5D's 12Mpx. In fact it's well over 24Mpx last time I checked (i.e. Ektar 100). Scanning methods are important.

21. ### Dave Luttmann

Karim, I'd love to see your scans of 35mm Ektar 100 that matched or exceeded a 24mp sensor, in a side by side test.

22. ### Robin Smith

So would I. "Scanning methods are important" is the understatement of the year!

23. ### bebu_lamar

Uhm! I thought I am the only one still using 35mm film. I thought most people went on to digital or larger format.

24. ### mauro_franic

To the OP's question, many/most 35mm lenses can exceed 36MP detail.
"As one who just hung a roll of Tmax 400 to dry I have to say I personally cannot get any 35mm film including Tmax 100 to match the detail I get from my my old Canon 5d."
That is not a limitation of the film or the lens. (not even close)
Tmax 100 35mm will have higher resolution than the D800-36MP in high contrast areas. I believe even a scan of the film using a Coolscan will exceed it. I will test it and post the results in due time.

25. ### mauro_franic

Bottom line though, the D800 will be a fantastic improvement in technology for landscape photographers (who actually print and sell prints).