Luminous Landscape Forum
Equipment & Techniques => Medium Format / Film / Digital Backs – and Large Sensor Photography => Topic started by: Dick Roadnight on October 23, 2010, 12:29:48 pm
-
Does any body use the Schneider Digital centre filter?
Does it work?
Does it do a better job than PS?
How do you get it if you do not speak German?
Can you get an analog (glass) center filter for the Apo-Digitar 47 XL?
-
There was a post on that matter here: http://www.luminous-landscape.com/forum/index.php?topic=45314.0
Does that helps?
-
Yes there is a CF for the 47 XL. Ask your dealer. I use it only on extreme vertical shifts, otherwise its not really necessary.
-
There was a post on that matter here: http://www.luminous-landscape.com/forum/index.php?topic=45314.0
Does that helps?
Thank you, Fred... I should have done a search before asking the question.
-
Yes there is a CF for the 47 XL. Ask your dealer. I use it only on extreme vertical shifts, otherwise its not really necessary.
Yes, thanks, I think I will get the 16190 glass one - the less the software messes about with the file the better.
-
Does any body use the Schneider Digital centre filter?
I have one but I don't use it.
Does it work?
not yet.*
Does it do a better job than PS?
by now: no.
How do you get it if you do not speak German?
http://digitalcenterfilter.com/
You have to have a license. It's included when you buy a new lens... but AFAIK you won't get one for older lenses (i.e. those you have bought before the release of the digital centerfilter).
Can you get an analog (glass) center filter for the Apo-Digitar 47 XL?
yes, and WRT to noise at the edges when shooting with large movements it's a way better solution than any software based light falloff correction. I also think the degradation of IQ due to the glass filter is absolutely negligible (at least on the 47XL).
* The previous version produced blown out highlights. The just updated version works much better but the final file is undercorrected (i.e. too dark at the edges). The falloff correction refers to the lens data, not to the real capture.
Another issue is the correction depends on the Gamma resp. TRC of the TIF (i.e. you get different results depending on the color profile the source-TIF contains).
I have been in contact with the software designer of the DCF and showed him where the DCF is incorrcet. He is aware of the issues and is very concerned to solve the issues as soon as possible... which is very positive.
Basically I don't think the DCF will work better than the LCC based light falloff correction in Capture One and/or the respective tools in other MFD raw softwares… as these corrections are based on a white reference shot and therefore refer to the real lens and the real sensor… so to the real image, not just to lens data.
By now the DCF is also quite slow…
It won't replace a glass filter anyway when you have to deal with a falloff of about 2 stops or so...
-
Does any body use the Schneider Digital centre filter?
Does it work?
Does it do a better job than PS?
How do you get it if you do not speak German?
Can you get an analog (glass) center filter for the Apo-Digitar 47 XL?
A glass centre filter will in theory produce a better approximation, and in practice probably as well. The reason is that one attenuates the photons at the center of the image circle, which requires an increase of exposure which in turn will get more photons at the edges/corners. More photons = higher signal to noise (S/N) ratio. However, glass filters introduce an optical element that's foreign to the lens design, so it's not going to be a totally free lunch.
A digital filter has it's share of issues. From a S/N ratio point of view it will be a sub-optimal. What's more, depending on the lens it's going to require per aperture value settings, and it has to be applied to Raw not-gamma-corrected data for the best results (it's at least computationally more efficient).
Lenses exhibit 2 types of darkening towards the corners, as seen by the image sensor array / film. First is the vignetting potentially caused by the tubular design of a compounded (multiple element) lens. When light strikes the 'tube' at an angle other than the optical axis, there may come an angle where the light will no longer be a circular cone of light, but it'll become somewhat elliptical in diameter and thus restrict light. This will disappear when stopping down the aperture far enough, hence the dependency on aperture value.
The other component is light fall-off, which is due to the incident angle of the light (and travel distance or magnification factor) towards the corners. This is only minorly influenced by the aperture.
Since the light striking the sensor does so while in linear gamma space, it's most effectively dealt with in linear gamma space (RAW). A lens filter will automatically do (most) of that, although it cannot take the aperture value into account. An LLC taken at the aperture used for the actual shot will be able to correct the uneven lighting (including sensor inefficiencies), but it can't compensate for the difference in photon count (= S/N ratio). It also requires to have an input about lens shift if applicable, because that will cause a decentered vignetting/light fall-off.
Cheers,
Bart
-
A lens filter will automatically do (most) of that, although it cannot take the aperture value into account. An LLC taken at the aperture used for the actual shot will be able to correct the uneven lighting (including sensor inefficiencies), but it can't compensate for the difference in photon count (= S/N ratio).
this is why it is the best workflow to use a glass filter and still use LCC based falloff correction. The glass filter decreases the light falloff from e.g. 2 stops to e.g. 1 stop (at the edges of the image cricle).
The result is a perfectly corrected image (re light falloff) without too much noise at the edges (as the LCC correction does not have to push the edges that much).
It also requires to have an input about lens shift if applicable, because that will cause a decentered vignetting/light fall-off.
Schneider's DCF provides settings for lateral resp. rise/fall movements... i.e. it takes the entire image circle into account; therefore it also provides settings for the sensor dimensions. (It also provides settings for different apertures.)
With LCC based corrections you don't have to log the movements... if you shoot the LCC files in the respective scene.
However it is still required to log the movements if you want to apply for instance distortion correction (e.g. in Alpa's Lens Corrector).
-
Oh, I have an idea.... how about designing wide lenses with the correct center filters built into them?!!! An evenly illuminated field across the image circle? Radical, I know.
;)
-
I second CBarrett's movement and raise him an adapted "f stop" notation to adapt for stops lost for easier metering. ;)
-
Oh, I have an idea.... how about designing wide lenses with the correct center filters built into them?!!! An evenly illuminated field across the image circle? Radical, I know.
;)
This is because, without the center-filter you get a brighter image to focus with, and, if you are not using the edges of the image circle, the lens performs better without it.
...They could, of course supply the (glass) center-filter with the lens... but they do now, of course, supply the Digital Center-Filter with the lens.
-
is there any Center rectangular filter ?
-
is there any Center rectangular filter ?
That could be useful for rectangular pictures, but image circles are round, so all photo images are round before being cropped by the sensor/film.
...You could have an off-center focal plane graduated filter for extreme shifts?
-
to place the dark area where is the sun it will be useful I think
(http://www.urpix.fr/files/gu079sqdiqeizv6614t.jpg)
-
is there any Center rectangular filter ?
a centerfilter with a sliding mechnism would be great...
you could remove the filter to get a brighter image for composition/focusing and simply could slide it over the lens without the need to screw/unscrew the filter all the time...
-
a centerfilter with a sliding mechnism would be great...
you could remove the filter to get a brighter image for composition/focusing and simply could slide it over the lens without the need to screw/unscrew the filter all the time...
Doesn't The Sinar filter system work like that?