Welcome to our community

Be a part of something great, join today!

Log in
  • Hey all, just changed over the backend after 15 years I figured time to give it a bit of an update, its probably gonna be a bit weird for most of you and i am sure there is a few bugs to work out but it should kinda work the same as before... hopefully :)

Please help me interpret the MTF chart for this Schneider-Kreuznach lens.

Id like to know this too. The better Kowa 10 Mega pixel lenses hold 200 line pairs per mm.
 
Erich Ocean said:
Please help me interpret the MTF chart [...] I'm trying to figure out how many line pairs per mm this lens can resolve.
Click to expand...

Unfortunately, it's not possible. The chart doesn't show MTF extinction. It only shows 10, 20, and 30 l/mm. If you plot those lines and can fit a curve, you could extrapolate the MTF at any frequency (e.g. 200 lp/mm) -- but I don't think that it would be accurate at all. ("30 lp/mm performance is not a guarantee of 200 lp/mm lp/mm").
 
MTF charts plotted for "in AIR" are not valid for camera's with OLPF+IR?UV cut filters (or ones with a sensor cover glass) since the thickness of the "plate" will introduce negative spherical aberration and some curvature of field and chromatic aberration. More so when you are talking about 200lp/mm.

There is also more to know, since the chromatic aberration is not clearly reflected in the MTF curves. Many lenses are rated for very small MTF values, often unstated, perhaps 2% MTF, so the numbers show that with extreme anti-OLPF sharpen you may see some signal off a test chart, but the shape of the CoC will also be amplified by the high sharpen applied, so the result does not look like a image shot on a film camera and printed film-to-film, the sharpen needed adds artifacts and increases the effective DoF when a OLPF filter is used.

You cannot omit the OLPF filter in Bayer sensor cameras since it is needed to suppress the chroma moire and aliasing.

Since the lens maker does not know what exact OLPF+IR/UV cut filter will be in the camera he cannot publish absolutely accurate MTF curves for the lens when used on some unknown digital camera using a Bayer filter with OLPF+IR/UV and sensor cover glass.

Curves at 200lp/mm would read close to zero MTF at smaller f/ stops anyway (f/6.3 to 32), since you cannot stop down much past f/6.3 and get anything near 200lp/mm at 30% MTF because diffraction starts to enlarge the CoC. Likewise at f/ stops wider than f/3.5 (f'3.5 to f/0.7) the extra aberrations from the OLPF thickness become large enough to begin to have an impact on the MTF at 200lp/mm.

So for shooting at 30% MTF at 200lp/mm you are more or less limited to about shooting at f/4.5, larger or smaller stops will probably decrease the MTF at 200lp/mm.

Images people think "look fine" have better than 512 lines per image height at about 30% MTF, so try that as the measure of needed resolution as less should be considered too soft.

I don't think there are any C mount lenses with more than a few percent MTF at 200lp/mm wide open, if you know of one let me know what the link to its data sheets that state so are. The CA may be ten or more times as large, so the resolution at 80% MTF may only be 20lp/mm full spectrum for a lens rated at 200lp/mm monochromatic computer simulation at a few % MTF.

A better measure of any lens it a photograph of its CoC, as some photographic magazines had done in the past, you can see then what the size of the CoC is in total, which may be 20 to 50x as large as some fractional percent MTF "limiting" resolution.
 
Thanks guys.

I think the 5MP 1/2" sensor I'm looking at for the lens is just too small for the image quality I need. Instead, I'm going to move up to an 8MP 4/3" sensor from Kodak and see about pairing that with a lightweight Panasonic Lumix G 14mm f/2.5 ASPH lens. That lens seems to meet the resolution requirements, and is tiny as well.
 
OLPF

OLPF

Erich Ocean said:
Thanks guys.

I think the 5MP 1/2" sensor I'm looking at for the lens is just too small for the image quality I need. Instead, I'm going to move up to an 8MP 4/3" sensor from Kodak and see about pairing that with a lightweight Panasonic Lumix G 14mm f/2.5 ASPH lens. That lens seems to meet the resolution requirements, and is tiny as well.
Click to expand...

If you are thinking about the Kodak sensors for movie use you should look at the FPN in the sample Acam dII DNG files by making shadow lift and increasing the contrast in the night photos and the Steenbeck (that may have removed those shots, but you may know someone that has them).

http://www.ikonoskop.com/dii/footage

Kodak may tell you that you don't need a OLPF but for a movie camera you do need an OLPF, I went around that subject with Acam until they seem to have stopped answering my emails, you can look at the early "Nick" footage to see what happens without an OLPF. One problem with finding an OLPF with a good fit is that the Kodak pixels are a bit large so the OLPF you find may not have a large enough blur radius.

If you are talking about the Aptina sensor 'P401 you might try using Super8 lenses rather than lenses made for 2/3" sensors. It can produce nice results at maybe EI/ISO 400. (The de-Bayer software I have been working on was developed to get "usable" images from it at 2592x1104x12bit@24fps etc.)

http://www.aptina.com/assets/downloadDocument.do?id=865

I like its "look" when processed right, the colors are a bit "dull" but natural looking under good lighting, the Kodak chip makes somewhat brighter colors and less noise in the bright areas but the splits are an issue, and you may have some problems getting the Kodak chip to work at all near the needed fps if you are making a digital movie camera because its not as much a "camera on a chip" as the Aptina are.

What are you building?

The resolution is not dependent on the lens specs in a digital camera as the OLPF cuts that MTF at high frequencies, the way you sharpen the images after the OLPF is used have a huge impact on the SYSTEM transfer function.
 
Erich Ocean said:
I'm trying to figure out how many line pairs per mm this lens can resolve. (I'm looking for a compact C-mount lens at this focal length with 200 lp/mm in the center, and hopefully not too much lower at the edges.)
Click to expand...
Why do you specifically need a C-mount lens?

Not to be flippant, but there's a lot of factors involved with apparent sharpness, including MTF, contrast, lighting, and filtration. My advice would be to shoot some tests and then look at them on a calibrated 20' screen and see what they look like.

All the major LA lens companies, like Clairmont, Otto Nemenz, Panavision, and so on can assist with providing focus charts. Realistically, sharp lenses are not cheap, and you can't skimp on them. I often tell my clients, you're better off just renting the lenses for a specific production, and buying just the lenses you can afford. C-mount lenses wouldn't be my first choice for the RED.
 
You must log in or register to reply here.
Back
Top