[Leica] [LRflex] Erwin Puts

Don Dory don.dory at gmail.com
Wed Mar 24 12:07:53 PDT 2021


I agree with your comments,  we are seeing multiple processors in almost
all prosummer and above bodies.

Two complications arise.   Video is becoming the driving force: 8K video is
a huge resource hog at capture,  processing,  and viewing.   But 8k isn't
even comparable to a 24mp still sensor.   The volume desire will drive the
market.

The second driver is the desire for larger sensors for two major reasons.
One is dof control driven by sensitive size and aperture of taking lens.
 The second driver is color depth/ISO performance.

On Wed, Mar 24, 2021, 1:49 PM Frank Filippone via LUG <lug at leica-users.org>
wrote:

> I am not an Olympus person, so I will leave that specific example to
> you.  However, it fits the bill.....
>
> Let me comment on the 24x36 and or hi res sensors....
>
> The issue is not the body size to handle the increased firmware, it is
> the amount of processing power and speed of that processor, and the
> associated electronics that you require for the resolution of that sensor.
>
> In effect, if you want enough pixels and only have so much physical
> processing power per processor, it requires 2 processors to get the job
> done.  Same with the bandwidth of that processor.... you need more I/O
> capability at a higher speed, maybe necessitating an additional
> processor.  These are all trade offs. ( note that this is really true
> when it comes to video.... high frame rates, high resolutions = more
> processing power required)
>
> The one truth is that by adopting computational photography, you CAN
> make a 0.95 Canon 50mm Lens look like it is a 50 APO ASPH Summicron.
> Weight and size could be like a 50mm Elmar from the 30's.  Cost could be
> in the $100's.
>
> And THAT is the Holy Grail.
>
> I think that Leica is in board, except for my $100 comment........
>
> BTW, there is a problem for Leica in this area... how to get someone to
> pony up $9K to buy a $100's lens with attached Firmware..... So far they
> have been successful.   But not with the CP ( computational photography)
> at any significant level.
>
> This is a much better strategy for the N and C guys.
>
> IOW, in my estimation, Erwin is right ( for Leica).
>
> Frank Filippone
> BMWRed735i at gmail.com
>
> On 3/24/2021 11:22 AM, Don Dory via LUG wrote:
> > I believe that Olympus was the major example of this approach: fix the
> > optical flaws in the camera software.  Olympus has relatively inexpensive
> > high performance lenses when used on their bodies.  One advantage is the
> > micro 4/3 coverage circle.  I believe that Olympus's sin was not moving
> > aggressively enough in the pixel wars(Leica seems to be coming around to
> > high pixel count sensors even though many of their lenses aren't designed
> > around that concept: the APO Summicrons being the exception at a steep
> > price).
> >
> > I hope that another major manufacturer will come around to 24x36 sensor
> > coverage lenses where the camera corrects flaws in the lens.  I believe
> > this is best done by the body manufacturer as they have the room to put
> the
> > ROM that would correct during internal processing for lens flaws.  In a
> > mature state the camera and lens would talk to each other about focus
> > distance, aperture, and zoom setting if applicable.
> >
> > Imagine where you could go from 0.95 Canon massive spherical aberration
> to
> > L mount 35 APO Summicron correction just by a menu selection.
> >
> > My bias is I really like extremely well corrected lenses but cringe at
> the
> > size and weight to get there.  Look at the MTF charts for the L mount APO
> > 35 Summicron compared to the MTF of the M mount version: over 90%
> transfer
> > at F2 across the field at even high frequencies compared to dipping to
> only
> > 60% transfer for the M mount on the edges.  But then compare the size and
> > weight of the two; I believe most people would make the compromise for
> the
> > lower size and weight.
> >
> > On Wed, Mar 24, 2021 at 1:08 PM Frank Filippone via LUG <
> lug at leica-users.org>
> > wrote:
> >
> >> If anything, Erwin was a believer in optical excellence.  Np tricks, no
> >> excuses.  Not GOOD engineering and production.  Rather, GREAT
> >> engineering and production.
> >>
> >> he whole theory of COMPUTATIONAL PHOTOGRAPHY would be  (?) anathema to
> >> him.  It is NOT about optical excellence, it is about learning to
> >> correct that lack of excellence with digital manipulation.  IOW, let the
> >> costly optical components go down in COST, measure what you got, then
> >> figure how to digitally correct them.
> >>
> >> It makes sense.... in a away..... The repetitive costs of expensive
> >> optical and mechanical components goes down, and the one time only costs
> >> of figuring out a digital correction works "as well" as if you did it
> >> right in the first place.
> >>
> >> Costs Down = More Profit at the same selling point.  ( I think we can
> >> all agree that Leica keeps the same selling point: High)
> >>
> >> The Huawei and Panasonic connection, etc. is all about computational
> >> photography.  It was getting popular in the interviews with Leica
> >> personnel in the 2018-2019 time period.
> >>
> >> It is still popular in the interviews.  The anti-keystoning software in
> >> the M10 recent firmware release is a perfect example and product of this
> >> type of thinking....
> >>
> >> My guess is that this contributed to his frustration with Leica....
> >>
> >> Frank Filippone
> >> BMWRed735i at gmail.com
> >>
> >> On 3/24/2021 10:22 AM, Douglas Sharp wrote:
> >>> http://joerivanderkloet.com/erwin-puts-says-bye-bye-leica/
> >>>
> >>> With a statement from EP saying what annoyed him about
> >>> Leica.
> >>>
> >>> Understandable, too.
> >>> Cheers
> >>> Douglas
> >> _______________________________________________
> >> Leica Users Group.
> >> See http://leica-users.org/mailman/listinfo/lug for more information
> >>
> >
>
> _______________________________________________
> Leica Users Group.
> See http://leica-users.org/mailman/listinfo/lug for more information


More information about the LUG mailing list