Olympus and others. Greater mega pixel capacity expected soon? Skinnyon E-3

On Sat, 12 Apr 2008 15:35:50 -0700 (PDT), "."
wrote:

I use an E-510 and like it a lot for my needs but am thinking of a
second body to have two different zooms availably quickly without the
need to change lens. The E-3 seems to have some real advantages but is
pricey.

2 questions:

Do any of you here have extensive experiance with the E-3 and your
opinion of it.

Are there expected increases in mega pixel capacity any time soon for
Olympus or others?

Thanks for any info.

Charlie

Noise is not really an issue on E3, read the reviews. However noise
and the removal of it regardless of camera brand should be part of
your workflow. There is enough software out there cheap enough or even
free that deals with noise. So noise is simply a non issue for you.

If you already have Zuiko glass why don`t you consider either hanging
on until 520 comes out ( very soon ) and grabbing a 510 at fireside
prices or grabbing a second hand E1.

Be warned though about the E1. Many 510 users have got one for a back
up body and it became their camera of choice!

Do not listen to the crap on here about 4/3, noise and other so called
limitations, a camera is just a tool and it is the glass that matters
really. Forget megapixels as well, it is only a small part of the
equation.

Email replies remove REMOVE
Powered by Agent 4.2 Mail/News
http://www.forteinc.com/main/homepage.php

"JG" wrote in message
...

snip

Do not listen to the crap on here about 4/3, noise and other so called
limitations, a camera is just a tool and it is the glass that matters
really. Forget megapixels as well, it is only a small part of the
equation.

Isn't that a little like saying, "Don't worry about the film you're using,
it's the glass that matters"? Yes, the glass does matter, but so does
whatever the image actually falls on.

Robert Brace wrote:
"Steve Sherman" wrote in message
...
Tony Polson wrote:
"flaming-o" wrote:

The phenomenal rate of development of digital sensors has made a
laughingstock of many of the older posts on this forum about how smaller
sensors will never have low noise at high ISOs.

Someone should tell Olympus that. The Four Thirds sensor in the 10MP
Olympus E-3 is noisy.

Given that the larger APS-C sensor in the 12 MP Nikon D300 performs so
much better than the sensor in the E-3, it would appear that smaller
sensors still suffer critical limitations to their ability to deliver
low noise images.

What is or defines this "critical limitation" of the E-3.
I guess I'm looking for, just where or how does this occur in E-3, but a
Nikon
has no noise. Is it just the sensor in the E-3 or is it all sensors used
in 4/3
cameras?
Steve

Steve:
The "Reader's Digest" version is that the prevalent common wisdom is
that the 4/3 sensors just don't have the real estate available to create
pixels of sufficient size, in sufficient numbers to perform at such a low
noise level to give sufficient signal without visible noise becoming
obvious. In other words, and all things being equal and forgetting all
in-camera processing, given the same number of pixels, sensors of larger
size (with larger pixels) will generate images with a lower noise level than
smaller sensors.
By the way, the D300 is far from having "no noise" in real world use
(and don't believe anyone who tells you differently). However, whether or
not the noise is visible (or bothersome in your use) is the crux of the
matter. I've seen noisy images from the D300 and amazingly clean
enlargements from the E-3.
Now be prepared for the influx of the "experts" on this site with any
and all opinions (some very valid) leaving you where you started i.e.- you
need to make up your own mind, based upon your own use!!
Bob




Bob,
Thanks. That is a very nice description.
I have a Canon 30D and my son has a E-500, both 8meg devices. I can not
see any
real difference in picture quality. The 4/3 image does need more enlargement
to match the physical image size of the 30D. But, I'm not a color or
resolution wizard.
The one thing I do know is that I will never buy another Nikon. I have a
pile of AIS lens
and they don't work for crap on a Nikon, except for a D200 and up. But,
they work and
meter fine on both the Olympus and the Canon. There is no way I will
ever let Nikon stick
it to me again.
By the way, I really like the Canon 30D. It's not the pictures, it the
physical part and
the speed of operation.

Regards,
Steve

Wolfgang Weisselberg wrote:
flaming-o wrote:

Regardless of brand their will always be something newer and better on the
horizon so you must decide if you really need that second camera body right
now.

True.

The phenomenal rate of development of digital sensors has made a
laughingstock of many of the older posts on this forum about how smaller
sensors will never have low noise at high ISOs.

There are only so many photons to go around, and they are emitted
randomly. That gives an upper limit to noise-freeness in regard
to pixel size. That's an unalterable physical fact.

But s/n ratio should be looked at from the POV of an entire image, not a
fixed number of pixels.
Of course, you can "denoise" any image, turn it into an oil
painting ...

What is spectacular in the Nikon D300 today will be baseline specs in
two years.

The D300 is not, exactly, a "smaller sensor", and the pixel pitch
is 5,5µm. The E-3 has a pixel pitch of less than 5µm, less than
80% of the possible pixel size of the D300 --- and the D300 has
22% *more* pixels. Larger ones!


I've seen E3 and other Olympus sensor analysis that showed the main
problem isn't sensor/sensel size so much as the fact that the olympus
sensors clearly underperform on a square measure basis compared to the
better sensors of Canon, Sony/Nikon. If they could get their act
together on that, then 4/3 might have a future.

On Mon, 14 Apr 2008 16:49:11 -0500, "Paul"
wrote:


"JG" wrote in message
.. .

snip

Do not listen to the crap on here about 4/3, noise and other so called
limitations, a camera is just a tool and it is the glass that matters
really. Forget megapixels as well, it is only a small part of the
equation.

Isn't that a little like saying, "Don't worry about the film you're using,

Nope
it's the glass that matters"? Yes, the glass does matter, but so does
whatever the image actually falls on.

To a certain degree, but nowhere near critical as to what is spouted
on here.

Email replies remove REMOVE
Powered by Agent 4.2 Mail/News
http://www.forteinc.com/main/homepage.php

"Steve Sherman" wrote in message
news
Robert Brace wrote:
"Steve Sherman" wrote in message
...
Tony Polson wrote:
"flaming-o" wrote:

The phenomenal rate of development of digital sensors has made a
laughingstock of many of the older posts on this forum about how
smaller sensors will never have low noise at high ISOs.

Someone should tell Olympus that. The Four Thirds sensor in the 10MP
Olympus E-3 is noisy.

Given that the larger APS-C sensor in the 12 MP Nikon D300 performs so
much better than the sensor in the E-3, it would appear that smaller
sensors still suffer critical limitations to their ability to deliver
low noise images.

What is or defines this "critical limitation" of the E-3.
I guess I'm looking for, just where or how does this occur in E-3, but a
Nikon
has no noise. Is it just the sensor in the E-3 or is it all sensors used
in 4/3
cameras?
Steve

Steve:
The "Reader's Digest" version is that the prevalent common wisdom is
that the 4/3 sensors just don't have the real estate available to create
pixels of sufficient size, in sufficient numbers to perform at such a low
noise level to give sufficient signal without visible noise becoming
obvious. In other words, and all things being equal and forgetting all
in-camera processing, given the same number of pixels, sensors of larger
size (with larger pixels) will generate images with a lower noise level
than smaller sensors.
By the way, the D300 is far from having "no noise" in real world use
(and don't believe anyone who tells you differently). However, whether
or not the noise is visible (or bothersome in your use) is the crux of
the matter. I've seen noisy images from the D300 and amazingly clean
enlargements from the E-3.
Now be prepared for the influx of the "experts" on this site with any
and all opinions (some very valid) leaving you where you started i.e.-
you need to make up your own mind, based upon your own use!!
Bob




Bob,
Thanks. That is a very nice description.
I have a Canon 30D and my son has a E-500, both 8meg devices. I can not
see any
real difference in picture quality. The 4/3 image does need more
enlargement
to match the physical image size of the 30D. But, I'm not a color or
resolution wizard.
The one thing I do know is that I will never buy another Nikon. I have a
pile of AIS lens
and they don't work for crap on a Nikon, except for a D200 and up. But,
they work and
meter fine on both the Olympus and the Canon. There is no way I will ever
let Nikon stick
it to me again.
By the way, I really like the Canon 30D. It's not the pictures, it the
physical part and
the speed of operation.

Regards,
Steve

Steve:
OK, now you've got me totally confused!
If I understand what you wrote above, "But, they work and meter fine
on both the Olympus and the Canon", you have Nikkor AIS lenses which, when
mounted on the Canon or the Olympus, "work and meter fine".
I don't think so.
As much as I can identify with your dissatisfaction with Nikon's
improvements to their lenses when progressing from AIS to the current, I
can't agree they have "stuck it" to you.
I've been shooting Nikon for over 45 years (among others) and have
always stuck to their Pro bodies. Because of that I am still using the 50mm
lens I bought with my first F Nikon on my current D2 series. I think I had
it AI converted somewhere around 1979 or so.
So I have always looked upon the Nikon Pro series bodies as an
investment and not an expense!
My son shoots a 20D and loves it but also has a FM2n I bought him in
about 1988, as I recall.
So YMMV but keep shooting!!
Bob

.. wrote:
On Apr 14, 3:58 am, Wolfgang Weisselberg

The D300 is not, exactly, a "smaller sensor", and the pixel pitch
is 5,5µm. The E-3 has a pixel pitch of less than 5µm, less than
80% of the possible pixel size of the D300 --- and the D300 has
22% *more* pixels. Larger ones!

As someone who already own an Olympus what, in your opinion, is going
to be noticed as a difference in my images and the images produced
with the D300?

Dear Fullstop,
If you don't do low light photography of any kind (birding
being a not-so-obvious example) nor do (quite) shallow DOF
photography (some types of portraits, for example), you'll not
see much differences, since it'll still be you behind the camera.
Otherwise, you will see differences as you reach the limits of
your technology:


You can crop a bit less --- or have a bit less resolution on
large posters. Quite irrelevant in most cases.

Your lenses need to be 10% better in absolute spatial resolution
(due to the smaller pixels) for a same pixel-by-pixel resolution,
or a bit more than 30% more resolution for full-picture size.
That means against a high-end lens on the D300 you need to pit
an even higher quality lens.
According to photozone.de, most lenses reach the sensor resolution
only in the center and only a stop or maybe 2 around f/5.6 (for
APS, that'd be f/8 for full frame), so your best resolution
against comparable equipment will probably suffer somewhat.
This will only be a problem where you have the resolution to see
the difference (i.e. 100% crops or larger posters seen close up).

If you keep your ISO setting below 400 or (maybe) 800, you'll
probably not see much of a difference --- but the D300 can
utilize higher ISO settings with the same or less noise. This
will affect every situation where light is in short enough
supply, be it 1/800s to freeze a bird lifting off in the shadow
or an evening in the pub (or even normal indoor photography)
without flash, especially if your objects move ... I happen to
run into such situations regularly, others do not.

The D300 has about a stop shallower DOF. That means:
- full frame camera: f/2.0
- D300 and other 1.5/1.6x: f/1.4
- 4/3rds: f/1.0

Now, there are 85mm f/1.2 and f/1.8 for Canon, 50mm (80mmKB) f/1.4
(and f/1.2 and even the old f/1.0), also for Canon, but what
is in the ~40mm(80mmKB) class for Olympus? 14-35mm f/2.0 and
35-100mm f/2.0 are the fastest lenses I could find on
Olympus' website for any focal length! So:

FF ~ APS ~ 4/3rds
85mm f/1.8 ~ 50mm f/1.2 ~ 40mm f/0.9
85mm f/4 ~ 50mm f/2.8 ~ 40mm f/2

As you can see, playing with shallow DOF is, at least in the
'portrait' length of ~80mm, doable with APS, but the small
sensor and the dearth of really fast lenses cost you lots of
shallow DOF potential. That is, of course, irrelevant, if
you plan on only shooting at f/8, oops, f/4.

You'll also find that the amount and breath of lenses offered
is quite low: there are very few fixed focal length lenses,
especially there seem to be no ultra-fast lenses (faster than f/2).
How that hurts or influences you only you can know.
(Note that I would worry if these f/x-5.6 lenses are already
beyond the best f-stop for resolution wide open (and may or may not
need stopping down) --- just as I would worry about an f/11-f/16
(cheapo-)tele for full frame.)

-Wolfgang

frederick wrote:
Wolfgang Weisselberg wrote:

There are only so many photons to go around, and they are emitted
randomly. That gives an upper limit to noise-freeness in regard
to pixel size. That's an unalterable physical fact.

But s/n ratio should be looked at from the POV of an entire image, not a
fixed number of pixels.

Most certainly not.

Just today I have seen an image (printed!), that was fine in
the sunny parts but very disturbingly and visibly noisy in the
shadows --- even though it wasn't a large poster or something
like that. If you averaged that to the whole image, you't have an
acceptable noise level --- which would be similar to describing
a car behaviour 'satisfactory' that runs well above 5 MPH, but
bucked and screeched and randomly jumped about between 1-4 MPH!

Or describing nitroglycerine as "mostly" unexplosive (it is not
explosive, unless you disturb it badly enough --- when it becomes
extremely explosive).

I've seen E3 and other Olympus sensor analysis that showed the main
problem isn't sensor/sensel size so much as the fact that the olympus
sensors clearly underperform on a square measure basis compared to the
better sensors of Canon, Sony/Nikon. If they could get their act
together on that, then 4/3 might have a future.

They can only win by either offering similar noise (and thus
reducing resolution), being way ahead technologically (unless Canon
and Nikon and Co. catch up, and you can bet your life on that
they will!) or being so much better in another quality (size,
price, weight, ...) that the lower image quality or resolution
is not as important[1].

At the moment I don't see how 4/3rds will find a large niche
between the convenience point&shoot and the DSLR rig.

-Wolfgang

[1] Compare 35mm still cameras to medium and large format,
even back in the film days, where the same sensor could be
used by all: 35mm "won" because of the size and mobility,
not because of image quality!

Comare also your point&shoot or mobile phone "camera" you
always carry with you versus the vastly more capable DSLR rig.

Wolfgang Weisselberg wrote:
frederick wrote:
Wolfgang Weisselberg wrote:

There are only so many photons to go around, and they are emitted
randomly. That gives an upper limit to noise-freeness in regard
to pixel size. That's an unalterable physical fact.

But s/n ratio should be looked at from the POV of an entire image, not a
fixed number of pixels.

Most certainly not.

Just today I have seen an image (printed!), that was fine in
the sunny parts but very disturbingly and visibly noisy in the
shadows --- even though it wasn't a large poster or something
like that. If you averaged that to the whole image, you't have an
acceptable noise level --- which would be similar to describing
a car behaviour 'satisfactory' that runs well above 5 MPH, but
bucked and screeched and randomly jumped about between 1-4 MPH!

Or describing nitroglycerine as "mostly" unexplosive (it is not
explosive, unless you disturb it badly enough --- when it becomes
extremely explosive).


Lol - that wasn't what I meant - and I suspect that you probably knew that.
I've seen E3 and other Olympus sensor analysis that showed the main
problem isn't sensor/sensel size so much as the fact that the olympus
sensors clearly underperform on a square measure basis compared to the
better sensors of Canon, Sony/Nikon. If they could get their act
together on that, then 4/3 might have a future.

They can only win by either offering similar noise (and thus
reducing resolution), being way ahead technologically (unless Canon
and Nikon and Co. catch up, and you can bet your life on that
they will!) or being so much better in another quality (size,
price, weight, ...) that the lower image quality or resolution
is not as important[1].

At the moment I don't see how 4/3rds will find a large niche
between the convenience point&shoot and the DSLR rig.

-Wolfgang

[1] Compare 35mm still cameras to medium and large format,
even back in the film days, where the same sensor could be
used by all: 35mm "won" because of the size and mobility,
not because of image quality!

Comare also your point&shoot or mobile phone "camera" you
always carry with you versus the vastly more capable DSLR rig.
Small film camera are "hobbled" by having the same resolution across
formats.

frederick wrote:
Wolfgang Weisselberg wrote:
frederick wrote:
Wolfgang Weisselberg wrote:

There are only so many photons to go around, and they are emitted
randomly. That gives an upper limit to noise-freeness in regard
to pixel size. That's an unalterable physical fact.

But s/n ratio should be looked at from the POV of an entire image,
not a fixed number of pixels.

Most certainly not.
Just today I have seen an image (printed!), that was fine in
the sunny parts but very disturbingly and visibly noisy in the
shadows --- even though it wasn't a large poster or something
like that. If you averaged that to the whole image, you't have an
acceptable noise level --- which would be similar to describing
a car behaviour 'satisfactory' that runs well above 5 MPH, but
bucked and screeched and randomly jumped about between 1-4 MPH!

Or describing nitroglycerine as "mostly" unexplosive (it is not
explosive, unless you disturb it badly enough --- when it becomes
extremely explosive).


Lol - that wasn't what I meant - and I suspect that you probably knew that.

Sorry - perhaps that deserved more explanation.
Downsample a 10mp image to 6mp, and if the s/n ratio "comparing pixels"
is no worse than a "native" 6mp image, then there's arguably nothing
"lost" by having the extra pixels, but something is potentially gained
through having the extra pixels.
For Olympus sensors, they underperform - ie if there was a canon sensor
with similar pixel density - it would probably perform much better.
I guess that could be seen comparing the new 450d vs a 10mp Olympus 4/3
camera. At the pixel level, I'll bet that the canon sensor outperforms
the Olympus sensor by a considerable margin. If true, then 4/3
performance could be improved.