[Leica] T Stops

[henningw at archiphoto.com (Henning Wulff)]

One of the main issues is with dof. If the cameras compensate for 
loss of steeper angle rays with gain, it's not that much of a deal 
for me either. If you have a noticeable gain in dof (or 'loss of 
shallowness') then you're not really getting what you paid for when 
you buy the fastest lenses. You should be able to check this by 
comparing shots at various apertures taken with CMOS sensors, 
especially the worst offenders like the Nikon D40 and D60 and shots 
taken on film. If there is a noticeable difference in the dof at 
large apertures then I think the manufacturer's better come clean.

CCD sensors apparently don't have quite as deep 'wells'. The article 
mentions though that MF cameras use CCD's so that camera movements 
can be used more readily. The Canon TSE lenses I use can be shifted 
and tilted quite extensively with excellent quality and no obvious 
problems due to oblique rays, but they are of quite retrofocus 
design. Still, with the amount of shift I would think that there 
would be significant shading if the lens were shifted/tilted to its 
extreme. The rays are about as oblique as any coming from the exit 
pupil of the 85/1.2. Also, when there is this extreme tilt/shift 
there is no auto gain-up, as this would be noticeable from the part 
of the image that still passes through the other side of the lens and 
remains perpendicular to the sensor.

In view of the above I remain quite skeptical about this article as a 
whole. I don't think they have a true handle on what's happening, 
just as DXO doesn't seem to have a handle on some of their 'quality' 
criteria or measurements.

On the M9 I get approximately the same difference in exposure in the 
readout and in the final results with the Noctilux as I did on film, 
which seems to me to indicate that the M9 (and M8) don't do this 
gain-up.




>I'm not sure what the problem is.  There is some light loss due to an
>optical effect.  The sensor's gain is increased to compensate for this.
>This lets the imaging chain be exposed in a predictable way.  As long as
>the noise level is within the acceptable range then it should be ok.
>Obviously the cameras at the top of the chart perform better and those seem
>to be the more expensive full-frame models.  The way sensors are improving
>noise levels overall seem to be getting lower all the time, at least for
>the decently sized sensors.
>
>Mike Durling
>
>Original Message:
>-----------------
>From: Jayanand Govindaraj jayanand at gmail.com
>Date: Wed, 27 Oct 2010 19:33:05 +0530
>To: lug at leica-users.org
>Subject: [Leica] T Stops
>
>
>Interesting study:
>
>http://www.luminous-landscape.com/essays/an_open_letter_to_the_major_camera_
>manufacturers.shtml
>
>Cheers
>Jayanand
>
-- 

    *            Henning J. Wulff
   /|\      Wulff Photography & Design
  /###\   mailto:henningw at archiphoto.com
  |[ ]|     http://www.archiphoto.com