proper exposure metering

"leo" wrote in message
ink.net...
With 300D, I read that for proper exposure, the trick is set the exposure
to
+2, point the camera to the white area or highlight, lock the exposure,
and
recompose. This works but a a bit combersome. What do you do?

What I do? Just meter using common sense, shoot, and check the histogram. If
it's very off (clearly blown highlights or all too dark), I compensate and
shoot again. If it's not very off, I keep the shot and trust that any minor
over- or underexposure is fixable when converting from RAW -- and it usually
is.

However, if you shoot JPEG (which has inferior dynamic range), it's
naturally much more important to get the shot right from the start. It just
sucks to get home and notice that you lost some important detail in the
shadows or that you blew some highlights. Which is why I couldn't be
bothered to shoot JPEG (but I don't want _that_ war to get started again

"mch42" wrote in message
...
"leo" wrote in message
ink.net...

snip

However, if you shoot JPEG (which has inferior dynamic range), it's
naturally much more important to get the shot right from the start.

JPEG has essentially the same dynamic range as RAW, as the JPEG of a photo
consisting of two adjacent pure black and pure white squares will
demonstrate. JPEG has a deliberately reduced spatial-frequency range, which
is not the same thing as reduced dynamic range.

"mch42" wrote in message
...
"leo" wrote in message
ink.net...

snip

However, if you shoot JPEG (which has inferior dynamic range), it's
naturally much more important to get the shot right from the start.

JPEG has essentially the same dynamic range as RAW, as the JPEG of a photo
consisting of two adjacent pure black and pure white squares will
demonstrate. JPEG has a deliberately reduced spatial-frequency range, which
is not the same thing as reduced dynamic range.

In article ,
wrote:

You can now plot JPEG vs. raw values, or even (if your illumination is
particularly constant (e.g. LED's fed by a constant current source))
discern the linearity of the sensor (via raw), and look at the
linear-non-linear function ('gamma'). In any event, you'll see that
the dynamic range for a JPEG is basically the same as that of a raw
image.

Not on Canon DSLRs it's not - the inbuilt JPEG conversion clips the
highlights at up to 2 stops the actual point at which the raw file maxes
out. You can't use this extra headroom in Canon's software, but Photoshop CS
lets you get at it.

I think they do it to presrve colours as blowout approaches, and to give
them room for white-balance adjustments.

"Paul H." wrote in message
...
[]
JPEG has essentially the same dynamic range as RAW, as the JPEG of a
photo
consisting of two adjacent pure black and pure white squares will
demonstrate. JPEG has a deliberately reduced spatial-frequency range,
which
is not the same thing as reduced dynamic range.

No, that's not quite right. JPEG has an 8-bit display range which,
coupled with the gamma correction typically used by the camera conversion
firmware, results in a 10-bit dynamic range. RAW has an 11-12 bit dynamic
range.

JPEG does not have a reduced frequency response, but provides a
compression which can under certain circumstances reduce both edge
sharpness and low-contrast detail.

Cheers,
David