With or without 6-bit lens coding and M8/9 firmware correction

F

Frankie

Speaking Frankly
Local time
4:54 PM
Joined
Sep 13, 2007
Messages
752
There are much talks in these forums that RF is too difficult for digital; that the oblique angles of illumination especially by super-wide angle lenses are too difficult to tame...

I have seen postings elsewhere that plain uncoded and uncorrected CV 15mm images used in M8's look fine...nice and even, even in shots containing open sky's.

More examples from members would help prove or disprove this conundrum.

Please state details of lens used, coded or encoded, and whatever processing firmware/software used or not.
 
Bless:angel: CV 15 shots without coding and IR filter have horrible cyan drift in the corners, with filter and without coding horrible red drift in the corners, B&W have strong vignetting, which might be fobbed off as "artistic" . Solutions? For the colour vignetting use Cornerfix, for the exposure vignetting Photoshop has a correction tool. Much simpler to code the lens and use an IR cut filter....
 
Don't know what your point is.?

I have the 15mm it works beautifully, however, for me it seems to work best coded and with a IR filter on the M8, on the M9 it works perfectly well coded as a WATE with no IR filter. I do not understand the need to prove or dis-prove my reality.

Your mileage may vary.


Bo

www.bophoto.typepad.com
 
bo_lorentzen said:
Don't know what your point is.?

I have the 15mm it works beautifully, however, for me it seems to work best coded and with a IR filter on the M8, on the M9 it works perfectly well coded as a WATE with no IR filter. I do not understand the need to prove or dis-prove my reality.

Your mileage may vary.


Bo

www.bophoto.typepad.com
Click to expand...

The objective:

To learn how bad wide angle lens could be...in pixel-vignetting, colour shifting, or other ailments...as used in a RF or an M8/9 without the benefits of 6-bit and/or firmware correction.

An image example would be nice.
 
Ahh, now I understand better. the lens is not particularly bad, its a great little lens, and I use it naked on my M6 and CL being very happy with it. There is really no excuse for not coding it though, all the new M mount 15mm also have a recessed grove for sharpie coding, I got mine from cameraquest.com as always.

2009-11-14-hollywood-blvd-B.jpg


2009-11-14-hollywood-blvd-A.jpg




Bo

www.bophoto.typepad.com
 
Does the new cv15 lense bring up correct framelines (28/90) in the M8 for it to be sharpie coded as WATE or is mount modification necessary?

//Juha
 
Frankie said:
There are much talks in these forums that RF is too difficult for digital; that the oblique angles of illumination especially by super-wide angle lenses are too difficult to tame...

I have seen postings elsewhere that plain uncoded and uncorrected CV 15mm images used in M8's look fine...nice and even, even in shots containing open sky's.

More examples from members would help prove or disprove this conundrum.

Please state details of lens used, coded or encoded, and whatever processing firmware/software used or not.
Click to expand...

What you have heard is basically true.

Sure, the CV 15 produces some nice photos with the M8. But then this is a camera with a 1,33 crop factor. You don't even see the extreme corners. It seems to be like this so far in digital sensor development that RF cameras isn't that suited as SLRs. The M9 do show Full Frame, but only through hefty software manipulation of the corners.
 
People always refer to the CV15 when talking about wideangle vignetting on sensors because they have this "short focal length = oblique angles" thing in the back of their head. However, the CV15 isn't actually that much of an extreme lens - it's a retrofocus construction whose rear element sits quite far from the film plane. From a technical point of view it's a rather non-challenging example.

It would be much more interesting to try this with more extreme constructions - such as a Hologon, or an old (1950s) Zeiss Biogon 21/f4,5, or pre-"R" (non-retrofocus) Canon 19/f3.5, or a Nikon 21/4, or a Super Angulon 21/f3.4. With the Hologon the 15mm version won't even fit the M8, maybe the 16mm version converted from the Contax G would. You could even try this with a Jupiter-12 as a more affordable example of a challenging lens.
 
Frankie said:
70+ visitors and no real contributor...maybe few M8 owners tried real world lenses on the M8/9; subscribing only to the 6-bit myth...

Here are a couple of real examples:
http://www.rangefinderforum.com/forums/showpost.php?p=1214864&postcount=30

http://www.rangefinderforum.com/forums/showpost.php?p=1214874&postcount=31
Click to expand...

That is an interesting concept - seeing that I have been using the lens on my M8 bodies since 2006, and on my M9 for three months now....:rolleyes:

I can assure you 6-bits coding is no myth. It may be some users have exceptionally high tolerances for colour shifts.
 
Best thing is to just experiment on your own by coding as a WATE or whatever you think the CV 15/4.5 is closest to (Leica doesn't have a designated code for CV lenses ..., and CV lenses are not the same as a Leica lens).

Also try with filter, and without UV/IR filters and in different lighting conditions.

See what you like best. There is probably no single "best" all the time solution. I've definitely seen over-correction with the coding, so has Sean Reid and others.

You can tell by my real-life photos that great un-photoshopped images can be taken without any coding on the CV lens.
 
Olsen said:
What you have heard is basically true.

Sure, the CV 15 produces some nice photos with the M8. But then this is a camera with a 1,33 crop factor. You don't even see the extreme corners. It seems to be like this so far in digital sensor development that RF cameras isn't that suited as SLRs. The M9 do show Full Frame, but only through hefty software manipulation of the corners.
Click to expand...

I understand that...in an M8 at 1.33 crop factor, a 21mm is now a 28mm, thus with a FoV of only 75-degrees.

However, Leica seems to be expecting trouble with the M8, thus offset micro-lens over sensor and 6-bit lens encoding...

If a lens had no encoding, the firmware would not know what to correct, thus micro-lens only. If there was lens encoding, then the firmware helps a little/lot (?) more...

But if the lens used is way beyond 21mm, such as the CV 15mm or 12mm, then what?

With the WATE at 16/18/21mm, taming the CV 15 is doable, what if WATE encoding is not available or not used?

What about the CV 12mm?

I am not a super-wide lens user and only care about bread-and-butter lenses form 24/28~75mm. But I am curious now because the M9 allows FF FoV. Will the M9 kill the 24/5mm with no crop factor [at 90-degrees]?
 
Frankie said:
But if the lens used is way beyond 21mm, such as the CV 15mm or 12mm, then what?
Click to expand...

The problem wouldn't usually be the focal length. It's irrelevant at what angle the light falls into the lens - the question is at what angle it exits. The problem is the optical design, in particular the position of the rear element. The closer the rear element is to the film plane, the more difficult it gets. The CV 15 and 12 Heliars are both retrofocus constructions whose rear element sits quite far from the film plane (for their focal length anyway), which is why people don't have too many problems with them.

If you want a more challenging lens, look at pictures taken with a Super Angulon, which has a more challenging, symmetrical optical design and delivers quite a bit of vignetting and corner effects on the M8, in spite of being "only" a 21 on a crop camera. Here's some random pictures pulled off Flickr:

3992189965_1acd1f060e.jpg


2898212925_39958b2f76.jpg


2911799674_e74e965444_d.jpg
 
rxmd said:
The problem wouldn't usually be the focal length. It's irrelevant at what angle the light falls into the lens - the question is at what angle it exits. The problem is the optical design, in particular the position of the rear element. The closer the rear element is to the film plane, the more difficult it gets. The CV 15 and 12 Heliars are both retrofocus constructions whose rear element sits quite far from the film plane (for their focal length anyway), which is why people don't have too many problems with them.

If you want a more challenging lens, look at pictures taken with a Super Angulon, which has a more challenging, symmetrical optical design and delivers quite a bit of vignetting and corner effects on the M8, in spite of being "only" a 21 on a crop camera. Here's some random pictures pulled off Flickr:
Click to expand...

I understand how a retro-focus lens works...

In other words, don't use the older glass not designed for dRF.
 
Last edited:
You must log in or register to reply here.
Back
Top Bottom