Sony FF mirrorless lens size

In article , Eric Stevens
wrote:


I don't think so. With any lens, except at the very center, light rays
approach the sensor at an angle to the surface. The closer that the
lens is to the surface, the greater that the angle will be. There is a
limit to the angularity that a sensor will accept and hence a limit on
how close to the sensor that the lens can be.

what matters is the exit pupil of the lens, not the back focus distance.

Yes, but keeping the exit pupil forward requires a longer lens, hence
defeating the attempt to obtain a shorter lens.

you're ignoring that the closer the exit pupil is, the smaller the
coverage circle is.

you're *not* going to get a 4mm lens covering full frame, for example,
but that length is common on cellphone cameras.

But so what? We are talking about full frame, not cellphone cameras.

doesn't matter. it scales.

the point is that a 4mm lens *can't* cover a full frame without major
optic tricks (otherwise the rays would be nearly parallel to the
sensor, it's that close), however a lens that short can easily cover a
cellphone-sized sensor.

In article , Alfred Molon wrote:

If you check this article:
http://www.fujix-forum.com/threads/w...-professional-
mirrorless-was-a-fatal-mistake.54299/

the author is claiming that because the body is small, the lenses
must be bigger to maintain the distance to the sensor.

But is it perhaps because Sony is using an old lens design (suitable
for DSLRs where the lens is further away from the sensor)? And if
Sony used a different lens design, the lenses could be smaller (with
the rear element closer to the sensor)?

Only semi-valid point is that a short flange distance and narrow mount diameter
makes IBIS harder for wide angle lenses. Only, wide-angle lenses rarely ever
need IS to begin with, it's more useful for longer lenses.

And comparing the size of the FE 55/1.8 to the Canon 50/1.8 lens is stupid as
hell, the FE 55/1.8 is such a massively superior lens that it's just stupid.
Look at the sheer size of the Nikon Nikon 58/1,4G compared to the Nikon
50/1.4G.

Also, quite stupid to compare the length of the package instead of the weight
(penis pun intended), the A7 with any given lens will be smaller in volume and
weight than a comparable DSLR setup. The first image of native 24-70/2.8 lenses
is quite misleading:

Length Weight Flange
Sony 24-70/2.8 136 mm 886 g 18 mm
Canon 24-70/2.8 113 mm 805 g 44 mm
Nikon 24-70/2.8 133 mm 900 g 46.5 mm

It's clear that size and weight is not inherently due to the flange distance of
the mount, but by the design decisions made by the camera manufacturer.

Why aren't we comparing the tiny manual focus Nikon 50mm/1.2 to the huge Canon
50mm/1.2 using the same logic?

--
Sandman

On Wed, 06 Apr 2016 00:23:01 -0400, nospam
wrote:

In article , Eric Stevens
wrote:


I don't think so. With any lens, except at the very center, light rays
approach the sensor at an angle to the surface. The closer that the
lens is to the surface, the greater that the angle will be. There is a
limit to the angularity that a sensor will accept and hence a limit on
how close to the sensor that the lens can be.

what matters is the exit pupil of the lens, not the back focus distance.

Yes, but keeping the exit pupil forward requires a longer lens, hence
defeating the attempt to obtain a shorter lens.

you're ignoring that the closer the exit pupil is, the smaller the
coverage circle is.

you're *not* going to get a 4mm lens covering full frame, for example,
but that length is common on cellphone cameras.

But so what? We are talking about full frame, not cellphone cameras.

doesn't matter. it scales.

the point is that a 4mm lens *can't* cover a full frame without major
optic tricks (otherwise the rays would be nearly parallel to the
sensor, it's that close), however a lens that short can easily cover a
cellphone-sized sensor.

Duh!
--

Regards,

Eric Stevens

On 2016-04-06, nospam wrote:
In article , Eric Stevens
wrote:

If you can keep the rays of light coming from the lens more or less at
right angles to the sensor regardless of the effective focal length and
right across the picture then all your lenses will inherently be well
matched to the sensor. Keeping a good distance between the sensor and
the back of the lens achieves that. Lenses designed to make room for an
SLR mirror will also meet that requirement, if fitted to a lens mount or
adaptor that makes up for the slimmer body of a mirrorless camera. That
isn't a bad thing.

longstanding myth.

No.

yes.

lenses have been retrofocus well before digital so the rays are already
fairly parallel and the refractive index of silicon makes it impossible
for the angle to matter all that much.

Retrofocus lenses were first designed for use on 35mm SLRs as there was
no other way to get wide-angle lenses to work with them. Even standard
lenses were sometimes affected - the alternative being to use a
longer focal length, as was done for the Soviet Zenit cameras (56mm
standard prime lens) and for the Leica Visoflex (60mm).

Those same lens designs are useful for getting all the lenses on your
mirrorless digital camera to work well with the sensor built into the
camera body. Or would you prefer that each lens comes with its own
sensor?

Look at some rangefinder wideangle lenses and see just how close to the
film the back of the lens can come. Get yourself a Jupiter-12 35mm
rangefinder lens (a surprisingly good budget lens) and try to imagine
what happens to the light coming out of the back of it and hitting a
digital sensor designed to work with retrofocus lenses.
http://cameras.alfredklomp.com/jupiter12/

--
-- ^^^^^^^^^^
-- Whiskers
-- ~~~~~~~~~~

In article , Eric Stevens
says...
I don't think so. With any lens, except at the very center, light rays
approach the sensor at an angle to the surface. The closer that the
lens is to the surface, the greater that the angle will be. There is a
limit to the angularity that a sensor will accept and hence a limit on
how close to the sensor that the lens can be.

Ok, but m4/3 lenses don't have this size problem. Is it perhaps because
they are telecentric (while the Sony lenses are not)?
--
Alfred Molon

Olympus E-series DSLRs and micro 4/3 forum at
http://tech.groups.yahoo.com/group/MyOlympus/
http://myolympus.org/ photo sharing site

In article , Alfred Molon wrote:

Eric Stevens:
I don't think so. With any lens, except at the very center, light
rays approach the sensor at an angle to the surface. The closer
that the lens is to the surface, the greater that the angle will
be. There is a limit to the angularity that a sensor will accept
and hence a limit on how close to the sensor that the lens can be.

Ok, but m4/3 lenses don't have this size problem. Is it perhaps
because they are telecentric (while the Sony lenses are not)?

No it's because their sensors are a quarter the size of full frame.

A full frame lens needs to project an image circle that covers 35mm, which
means that for the same equivalent focal length, the lens needs to be bigger.

But "bigger" is relative. Using modern glass and autofocus parts, the sizes
seem to converge on about 50mm/1.4, meaning that for full frame, a 50mm/1.4
lens can be made fairly small, while going short or longer makes it bigger.

For APS, that convergence is at 35mm/1.4, meaning that a APS 35mm/1.4 is
roughly the same size as a FF 50mm/1.4. This convergence is no coincidence,
since a 35mm lens on a APS body is equivalent to a 52mm on a FF body.

If you move to a FF 35mm/1.4 they grow in size, as do a FF 85mm/1.4, going
further in either direction increases size.

There are exceptions. Some cheap 85mm and 35mm lenses are smaller, but also
optically inferior to the current top of the line selection from Nikon and
Canon.

--
Sandman

On Wed, 6 Apr 2016 19:43:00 +0200, Alfred Molon
wrote:

In article , Eric Stevens
says...
I don't think so. With any lens, except at the very center, light rays
approach the sensor at an angle to the surface. The closer that the
lens is to the surface, the greater that the angle will be. There is a
limit to the angularity that a sensor will accept and hence a limit on
how close to the sensor that the lens can be.

Ok, but m4/3 lenses don't have this size problem. Is it perhaps because
they are telecentric (while the Sony lenses are not)?

Almost certainly. I wasn't trying to say that it can't be done but
that it is not a way to get a compact lens.
--

Regards,

Eric Stevens

On 07/04/2016 08:33, Sandman wrote:
In article , Alfred Molon wrote:

Eric Stevens:
I don't think so. With any lens, except at the very center, light
rays approach the sensor at an angle to the surface. The closer
that the lens is to the surface, the greater that the angle will
be. There is a limit to the angularity that a sensor will accept
and hence a limit on how close to the sensor that the lens can be.

Ok, but m4/3 lenses don't have this size problem. Is it perhaps
because they are telecentric (while the Sony lenses are not)?

No it's because their sensors are a quarter the size of full frame.

A full frame lens needs to project an image circle that covers 35mm, which
means that for the same equivalent focal length, the lens needs to be bigger.

But "bigger" is relative. Using modern glass and autofocus parts, the sizes
seem to converge on about 50mm/1.4, meaning that for full frame, a 50mm/1.4
lens can be made fairly small, while going short or longer makes it bigger.

For APS, that convergence is at 35mm/1.4, meaning that a APS 35mm/1.4 is
roughly the same size as a FF 50mm/1.4. This convergence is no coincidence,
since a 35mm lens on a APS body is equivalent to a 52mm on a FF body.

If you move to a FF 35mm/1.4 they grow in size, as do a FF 85mm/1.4, going
further in either direction increases size.

There are exceptions. Some cheap 85mm and 35mm lenses are smaller, but also
optically inferior to the current top of the line selection from Nikon and
Canon.

Are u4/3 lenses telecentric?
IIRC Olympus made a song and dance about this years ago - with 4/3 but
before u4/3.

Leica used offset microlens design in older M digital cameras to correct
vignetting which may have been an issue with their lenses, but that was
with old CCD sensors and quite some time ago. The maker of their new
sensors (CMOSIS) goes in to quite some detail about sensor, thin sensor
stack, and microlens design, but don't mention offset microlens as a
feature. If offset microlenses aren't needed any more in an FX Leica M,
then u4/3 users shouldn't be too worried about "needing" telecentic lenses.

I suspect that microlens design has made it a non-issue.

In article , Me wrote:

Eric Stevens:
I don't think so. With any lens, except at the
very center, light rays approach the sensor at an angle to the
surface. The closer that the lens is to the surface, the
greater that the angle will be. There is a limit to the
angularity that a sensor will accept and hence a limit on how
close to the sensor that the lens can be.

Alfred Molon:
Ok, but m4/3 lenses don't have this size problem. Is it perhaps
because they are telecentric (while the Sony lenses are not)?

Sandman:
No it's because their sensors are a quarter the size of full
frame.

A full frame lens needs to project an image circle that covers
35mm, which means that for the same equivalent focal length, the
lens needs to be bigger.

But "bigger" is relative. Using modern glass and autofocus parts,
the sizes seem to converge on about 50mm/1.4, meaning that for
full frame, a 50mm/1.4 lens can be made fairly small, while going
short or longer makes it bigger.

For APS, that convergence is at 35mm/1.4, meaning that a APS
35mm/1.4 is roughly the same size as a FF 50mm/1.4. This
convergence is no coincidence, since a 35mm lens on a APS body is
equivalent to a 52mm on a FF body.

If you move to a FF 35mm/1.4 they grow in size, as do a FF
85mm/1.4, going further in either direction increases size.

There are exceptions. Some cheap 85mm and 35mm lenses are smaller,
but also optically inferior to the current top of the line
selection from Nikon and Canon.

Are u4/3 lenses telecentric?

Any lens can be telecentric, really, and it has nothing to do with flange
distance. The 4/3 standard heavily leaned upon telecentricity since it was a
standard developed solely for digital sensors.

It's *harder* to make telecentric lenses the closer you are to the sensor,
but as long as the rear element is larger than the sensor, it's quite
possible. For for wider lenses, the process of creating a telecentric lens on
a short flange distance would be quite a feat.

Leica used offset microlens design in older M digital cameras to
correct vignetting which may have been an issue with their lenses,
but that was with old CCD sensors and quite some time ago. The maker
of their new sensors (CMOSIS) goes in to quite some detail about
sensor, thin sensor stack, and microlens design, but don't mention
offset microlens as a feature. If offset microlenses aren't needed
any more in an FX Leica M, then u4/3 users shouldn't be too worried
about "needing" telecentic lenses.

The more telecentric a lens is, the better. But it's no magic bullet, and it
says nothing about chromatic aberrations or distortions, which is due to lens
design, not whether or not it is telecentric.

I suspect that microlens design has made it a non-issue.

I'd assume so.

--
Sandman

On 4/5/2016 11:57 PM, Eric Stevens wrote:
On Tue, 05 Apr 2016 22:06:05 -0400, nospam
wrote:

In article , Eric Stevens
wrote:

If you check this article:
http://www.fujix-forum.com/threads/w...-professional-
mirrorless-was-a-fatal-mistake.54299/

the author is claiming that because the body is small, the lenses must
be bigger to maintain the distance to the sensor.

But is it perhaps because Sony is using an old lens design (suitable for
DSLRs where the lens is further away from the sensor)? And if Sony used
a different lens design, the lenses could be smaller (with the rear
element closer to the sensor)?

I don't think so. With any lens, except at the very center, light rays
approach the sensor at an angle to the surface. The closer that the
lens is to the surface, the greater that the angle will be. There is a
limit to the angularity that a sensor will accept and hence a limit on
how close to the sensor that the lens can be.

what matters is the exit pupil of the lens, not the back focus distance.

Yes, but keeping the exit pupil forward requires a longer lens, hence
defeating the attempt to obtain a shorter lens.

you're ignoring that the closer the exit pupil is, the smaller the
coverage circle is.

you're *not* going to get a 4mm lens covering full frame, for example,
but that length is common on cellphone cameras.

But so what? We are talking about full frame, not cellphone cameras.


You seem to have an insatiable urge to feed trolls.



--
PeterN