Sony A100 to A700

In article , Alfred
Molon wrote:

In article 1212478815.530753@ftpsrv1, frederick says...

And I stand behind my comments above - it's dishonest / unscientific to
put the colour-resolution chart test in place of the normal chart tests.

Since the world is not black and white, it makes sense to use colour
resolution test targets instead of black and white ones.

the world is indeed colour, but it's not saturated stripes with
constant luminance. bayer sensors are designed to work the way human
eyes do, and that's maximize luminance resolution. foveon is nice in
theory, but it has a lot of shortcomings, such as metamerism, noise,
very critical white balance and significantly more data to move.

put a colour luminance chart on the wall next to a b/w chart and see
how well *you* can resolve it (assuming you don't get a headache first
from the colour chart). you'll find that you won't see anywhere near
as much detail on the colour chart as the b/w chart. why bother
capturing what you can't see?

In article , Alfred
Molon wrote:

In article , nospam says...

there is no truth in the 'foveon pixel count.' pixels are pixels,
regardless what sigma/foveon claim (and they even use the term
inconsistently, further proof that their method is bogus).

the reason for the difference is that sigma omits the anti-alias filter
and thereby the sensor can resolve closer to nyquist, along with a lot
of aliasing and heavy sharpening, which some people mistake for real
resolution. it also does better on colour resolution charts but that
isn't anything that matters to human vision. it just makes bayer look
worse in an edge case that never occurs in nature.

This has been discussed to death and there is no point to reopen the
discussion.

yes, it has been hashed to death on numerous forums, however, the myths
continue.

Do a Google search on the archives to find out why with
Bayer sensors the effective resolution is lower than the nominal pixel
count.

it's the lack of an anti-alias filter in the sigma cameras along with
sharpening in the raw processing (which bayer cameras generally are set
to much less), and not anything inherent to foveon.

In article , Ilya Zakharevich
wrote:

You can count angels at the point of a needle as long as you want.
The fact remains the fact.

what fact is that? a pixel is a spatial element of an image, and not a
layer within. the sd14/dp1 sensor has 4.7 megapixels. period.

DPreview (which is often suspected for prejudice in Canikon direction)
compares shots from a 5MP x 3 sensor, and from Bayer sensors with
(several) larger MP count. The conclusion was quite surprising (for me).

nothing surprising about an edge case designed to exploit a weakness in
bayer, and one which doesn't matter to humans. there are edge cases
where foveon falls apart too, and even more so than bayer does.

Hey, a lot of good info'
But still undecided

Ta
Ron


"Ron and Gail Smith" wrote in message
...
Hi All
Thinking of upgrading from Sony A100 to Sony A700 and wondering if anyone
has done
that and what your thoughts are?
I'd been interested to know

Regards
Ron

Alfred Molon wrote:
In article , Ilya Zakharevich says...

Even kit lenses *at their best settings* outperforms 10-12MP
half-frame sensors. (What one needs is to have MTF about 50% on the
most area of the sensor, at the frequency about 70% of Nyquist [the
latter number assumes Bayer sensors]. Just look at dpreview graphs.)

And well-below-$1000 wide-range QUALITY zooms start to appear, which
are getting "close" to such performance even when wide-open...

A 10MP 3:2 sensor has about 2600 lines.

The Sony 18-70mm f/3.5-5.6 DT kit lens has been reviewed by photozone:
http://tinyurl.com/3rqwtf

The MTF50 value reaches at best 2200 lines (centre, F5.6, 18mm).
Otherwise it's between 1400 and 2000 lines. In other words, this kit
lens is not good enough for a 10MP sensor. By the way, these MTF50
values are for MTF at just 50%.

I'm puzzled by this arithmetic. A 10MP 3:2 sensor doesn't have about
2600 lines, it has about 2500 pixels down. To reproduce a series of
closely spaced lines you need twice as many pixels, so that one line
can be black and the next white.

So aren't these lenses about twice as good as you're suggesting? Or
is my arithmetic confused?

--
Chris Malcolm DoD #205
IPAB, Informatics, JCMB, King's Buildings, Edinburgh, EH9 3JZ, UK
[http://www.dai.ed.ac.uk/homes/cam/]

In article , Chris Malcolm says...
Alfred Molon wrote:
In article , Ilya Zakharevich says...

Even kit lenses *at their best settings* outperforms 10-12MP
half-frame sensors. (What one needs is to have MTF about 50% on the
most area of the sensor, at the frequency about 70% of Nyquist [the
latter number assumes Bayer sensors]. Just look at dpreview graphs.)

And well-below-$1000 wide-range QUALITY zooms start to appear, which
are getting "close" to such performance even when wide-open...

A 10MP 3:2 sensor has about 2600 lines.

The Sony 18-70mm f/3.5-5.6 DT kit lens has been reviewed by photozone:
http://tinyurl.com/3rqwtf

The MTF50 value reaches at best 2200 lines (centre, F5.6, 18mm).
Otherwise it's between 1400 and 2000 lines. In other words, this kit
lens is not good enough for a 10MP sensor. By the way, these MTF50
values are for MTF at just 50%.

I'm puzzled by this arithmetic. A 10MP 3:2 sensor doesn't have about
2600 lines, it has about 2500 pixels down. To reproduce a series of
closely spaced lines you need twice as many pixels, so that one line
can be black and the next white.

So aren't these lenses about twice as good as you're suggesting? Or
is my arithmetic confused?

In the above figures lines = rows of pixels. Otherwise they would be
called libe pairs.
--

Alfred Molon
------------------------------
Olympus 50X0, 8080, E3X0, E4X0, E5X0 and E3 forum at
http://tech.groups.yahoo.com/group/MyOlympus/
http://myolympus.org/ photo sharing site

[A complimentary Cc of this posting was sent to
frederick
], who wrote in article 1212482281.344227@ftpsrv1:
Now we know that it is not a myph. Visual comparison of Foveon images
and Bayer images gives an advantage of about 2x in pixel count.
(Myself, I was expecting something about 1.5x...)

I don't know what you are arguing about.
4.69 x 2 = 10mp

I don't know what you are arguing about. As dpreview images show,
there are some situations where 4.69x3 sensor gives better "visual
impression of resolution" (sorry, can't find a better term) than 10mp
Bayer sensor. *This* is what surprised me.

IIRC DPreview already said the ("3.4mp x 3") SD10 was similar to 6mp
dslrs in resolution - so nothing new here.

Can't agree here - I do not remember anything about sd10 which would
suggest getting close a factor of to 2x. 6/3.4 = 1.76 which is
similar to what one would expect from inspection of images shot with
Bayer sensors (see threads about "dead pixels" in relation to MTF).

"So what we'd really like to see is a sensor that combines the Foveon's
per pixel resolution with the output size of a modern DSLR.

Do not think so. What "we'd really like to see" is a
separation-in-silicon sensor with pixel count about 40MP per square
inch, and capacitance/area and read noise similar to current Canon
sensors. *This* would start to utilize the potential of the best
lenses...

Hope this helps,
Ilya

In article , Ilya Zakharevich
wrote:

Now we know that it is not a myph. Visual comparison of Foveon images
and Bayer images gives an advantage of about 2x in pixel count.
(Myself, I was expecting something about 1.5x...)

I don't know what you are arguing about.
4.69 x 2 = 10mp

I don't know what you are arguing about. As dpreview images show,
there are some situations where 4.69x3 sensor gives better "visual
impression of resolution" (sorry, can't find a better term) than 10mp
Bayer sensor. *This* is what surprised me.

the 'visual impression' is nothing more than alias artifacts and a lot
of sharpening in raw processing. some people like the look and others
don't. in any event, it's not true resolution.

[A complimentary Cc of this posting was sent to
nospam
], who wrote in article :
I don't know what you are arguing about.
4.69 x 2 = 10mp

I don't know what you are arguing about. As dpreview images show,
there are some situations where 4.69x3 sensor gives better "visual
impression of resolution" (sorry, can't find a better term) than 10mp
Bayer sensor. *This* is what surprised me.

the 'visual impression' is nothing more than alias artifacts and a lot
of sharpening in raw processing.

??? There is no sharpening involved. Moreover, as resolution chart
shots show, there is no significant aliasing either.

some people like the look and others don't. in any event, it's not
true resolution.

Can't agree less. "True" resolution is what I can or can't see.

Any other "objective" metrics are just, in some sense, a window
dressing: trying to capture this - elusive - "visual impression of
resolution" via some "scientific metric". People who look at images of
resolution chart with Bayer sensors understand quickly how these data
correlate with "visual sharpness" of images.

It turns out that the (B&W) resolution charts do not fully correlate
with "visual impression of resolution" if one compares them between x3
and Bayer sensors. (AFAIU, the "color resolution charts" of dpreview
is an attempt to introduce an objective metric which reflects better
the way our visual system works.)

Yours,
Ilya

[A complimentary Cc of this posting was sent to
Ron and Gail Smith
], who wrote in article :
Hey, a lot of good info'
But still undecided

I forgot another point where people report that a100 is "just a toy",
and a700 flies: flash exposure. There is something very fishy with
the algorithm of a100 (*very* unreliable). a700 is reported to
perform very reasonably.

Hope this helps,
Ilya