«

SMS wrote:

> They also dropped the tilt, twist and swivel LCD, which was a very
> useful feature on the G series.

And it's hard for me to envision spending that kind of money on a
camera whose widest focal length is 35 mm equivalent.

Bill

On Thu, 14 Sep 2006 09:55:43 -0700, SMS <> wrote:
> Daniel Silevitch wrote:
>> There was some speculation here recently about whether Canon would
>> continue their G series. According to a press release on DP Review, the
>> answer is "yes".
>>
>> http://www.dpreview.com/news/0609/06091405_canon_g7.asp
>
><snip>
>
> They also dropped the tilt, twist and swivel LCD, which was a very
> useful feature on the G series.
>
> Still, the G7 is probably going to end up being the best
> image-stabilized compact camera in terms of lens quality, noise, and
> features (other than the lack of RAW).

From all reports, the Panasonic LX series has a very good lens, though
of course it has the notoriously noisy sensor. It does have RAW support,
which helps somewhat. That's one vaguely equivalent model; there are
probably a few others across different makers' lines.

-dms

On 14 Sep 2006 12:43:24 -0700, wrote:

>> They also dropped the tilt, twist and swivel LCD, which was a very
>> useful feature on the G series.
>
> And it's hard for me to envision spending that kind of money on a
> camera whose widest focal length is 35 mm equivalent.

The lens IIRC is smaller than before and the distinctive G series
shape is gone. Assuming that its lens is still of high quality, the
camera is stealthily trying to hide its heritage, resembling just
another humdrum P&S with a traditional camera silhouette.

"Daniel Silevitch" <> wrote in message
news:...
[ . . . ]
> "Canon has finally announced the long-awaited new G series compact to
> replace the G6 (launched August 2004). The $550 PowerShot G7 features a
> new Digic III processor, 10 megapixel (1/1.8-inch) CCD

Heh. Pixel cramming.


> and a 6x
> (35-200mm equiv.) image stabilized zoom lens.

Which (inevitably, I suppose) loses the earlier G series' nice f/2.0 at the
short end.

Neil

On 2006-09-14, Kinon O'Cann <> wrote:
> Canon dropped RAW support on this one? What an amazingly stupid move!

Looks like it =( I really liked my G5 and really weird if Canon has dropped
RAW on G7... RAW would make G7 pretty nice camera.

--
Petri Lopia :: lid
StormChasing/Myrskynbongaus etc. Lightnings/Salamoita yms.
http://www.petrilopia.net/
http://www.pbase.com/whig/

"Neil Harrington" <> wrote in message
news:ZcOdnQIW_-...
>
> "Daniel Silevitch" <> wrote in message
> news:...
> [ . . . ]
>> "Canon has finally announced the long-awaited new G series compact
>> to replace the G6 (launched August 2004). The $550 PowerShot G7
>> features a new Digic III processor, 10 megapixel (1/1.8-inch) CCD.
>
> Heh. Pixel cramming.

Yes, that equates to slightly less than 2 micron pixel pitch.
It will be interesting to see what the noise reduction quality of the
Digic III will be able to do, as it would indicate what can be
expected from newer Digic III based dSLRs.

Too bad it will also inevitably hurt Dynamic Range, unless the user
has an option to apply binning/averaging.

--
Bart

In article <450aea49$0$4519$>, Bart van der Wolf
says...

> Too bad it will also inevitably hurt Dynamic Range, unless the user
> has an option to apply binning/averaging.

If they offered double exposure (a very short exposure followed by a
longer exposure) they could solve the dynamic range issue.
--

Alfred Molon
------------------------------
Olympus 50X0, 7070, 8080, E300, E330 and E500 forum at
http://tech.groups.yahoo.com/group/MyOlympus/
Olympus E330 resource - http://myolympus.org/E330/

On Sat, 16 Sep 2006 10:01:48 +0200, Alfred Molon <> wrote:
> In article <450aea49$0$4519$>, Bart van der Wolf
> says...
>
>> Too bad it will also inevitably hurt Dynamic Range, unless the user
>> has an option to apply binning/averaging.
>
> If they offered double exposure (a very short exposure followed by a
> longer exposure) they could solve the dynamic range issue.

I'm sure that it will support exposure bracketing, which would amount to
the same thing. Better use a tripod if you are planning on merging
multiple shots, though.

-dms

In article <>, Daniel Silevitch
says...

> >> Too bad it will also inevitably hurt Dynamic Range, unless the user
> >> has an option to apply binning/averaging.
> >
> > If they offered double exposure (a very short exposure followed by a
> > longer exposure) they could solve the dynamic range issue.
>
> I'm sure that it will support exposure bracketing, which would amount to
> the same thing. Better use a tripod if you are planning on merging
> multiple shots, though.

Yes, but it's messy and time consuming assembling two bracketed images
after the fact. If you took two measurements for each pixel, one under-
and one overexposed and integrated these in the camera, you would
effectively increase the dynamic range.

For instance, if one pixel was overexposed with the long exposure time,
it would most likely not be so with the short exposure time. So you
would not use the overexposed measurement and instead use the
underexposed measurement, which would then be multiplied by a constant
factor to account for the shorter exposure time.
--

Alfred Molon
------------------------------
Olympus 50X0, 7070, 8080, E300, E330 and E500 forum at
http://tech.groups.yahoo.com/group/MyOlympus/
Olympus E330 resource - http://myolympus.org/E330/

"Alfred Molon" <> wrote in message
news: ...
> In article <450aea49$0$4519$>, Bart van der
> Wolf
> says...
>
>> Too bad it will also inevitably hurt Dynamic Range, unless the user
>> has an option to apply binning/averaging.
>
> If they offered double exposure (a very short exposure followed by a
> longer exposure) they could solve the dynamic range issue.

Not really, because the image will be swamped by photon noise relative
to the well depth. Even if we were to very liberally assume a 2x2
micron area per sensel, and say 1000 electrons per square micron, that
would result in a well depth of 4000 electrons maximum. The dynamic
range would then be something like 4000:63 (assuming a read noise
significantly below photon noise of 63), or almost 6 stops, at best
(!). Binning a 4x4 array of sensels would add another stop of DR.

That is even the case if you tonemap the image by a non-linear
operation on silicon. Combining multiple seperately recorded exposures
like in HDR imaging is always possible, but not too useful for moving
subjects like e.g. children, or cars passing by, etc., without a lot
of postprocessing not available in current cameras.

--
Bart