White balance

Is white balance information used to determine the way a (raw) image is
displayed, or is it used to cause the sensor to record the information
differently in the first place?

Daniele

Is white balance information used to determine the way a (raw) image is
displayed, or is it used to cause the sensor to record the information
differently in the first place?

Daniele

If you shot in RAW you can forget the White Balance - You can set it later
in your RAW converter
DAVE
Bristol UK
http://djmp.co.uk/slr/

D.M. Procida wrote:

Is white balance information used to determine the way a (raw) image is
displayed, or is it used to cause the sensor to record the information
differently in the first place?

I believe it actually changes the data in the analog to digital
conversion but it's not usually all that big of a deal to adjust with
the raw converter. In other words, it rarely matters.

I could be wrong.

People talk about using filters for slight improvements, and there was a
fuss over being able to read the as-shot WB in the Nikon D2x when it
came out in non-nikon raw converters, presumably that all had to do with
getting the optimum image quality by knowing the actual WB used for the
file.

--
Paul Furman Photography
http://edgehill.net
Bay Natives Nursery
http://www.baynatives.com

Paul Furman wrote:

D.M. Procida wrote:

Is white balance information used to determine the way a (raw) image is
displayed, or is it used to cause the sensor to record the information
differently in the first place?


I believe it actually changes the data in the analog to digital
conversion but it's not usually all that big of a deal to adjust with
the raw converter. In other words, it rarely matters.

I could be wrong.

People talk about using filters for slight improvements, and there was a
fuss over being able to read the as-shot WB in the Nikon D2x when it
came out in non-nikon raw converters, presumably that all had to do with
getting the optimum image quality by knowing the actual WB used for the
file.

I just did a test and there is only a very very slight change in the
histograms when I shoot and adjust to the far extremes of kelvin temp in
camera and the raw converter. It's not identical but close enough for
practical purposes. That's on a tripod in manual mode.

--
Paul Furman Photography
http://edgehill.net
Bay Natives Nursery
http://www.baynatives.com

Dave wrote:

Is white balance information used to determine the way a (raw) image is
displayed, or is it used to cause the sensor to record the information
differently in the first place?

If you shot in RAW you can forget the White Balance - You can set it later
in your RAW converter

But what as a reference? It might be easy to do it if your image has a
handy white wall or something to take a measurement from, but not all
images do?

And just to be clear: are you saying that a raw image shot with wildly
incorrect white balance, and one shot with perfect white balance,
actually contain the same data apart from the white balance information?

Daniele

D.M. Procida wrote:
Dave wrote:

Is white balance information used to determine the way a (raw) image is
displayed, or is it used to cause the sensor to record the information
differently in the first place?

If you shot in RAW you can forget the White Balance - You can set it later
in your RAW converter

But what as a reference? It might be easy to do it if your image has a
handy white wall or something to take a measurement from, but not all
images do?

That can be a problem. E.g. a red sunset. In such cases,
a calibrated monitor and remembering the colors is about
your best hope.

And just to be clear: are you saying that a raw image shot with wildly
incorrect white balance, and one shot with perfect white balance,
actually contain the same data apart from the white balance information?

Raw, at least in some cameras is a straight linear
analog to digital conversion with no changes to the data
regardless of settings on the camera. Some cameras
then do post processing before writing the raw
to the memory card (Nikons, for example do some minimal
processing including lossy compression via a look-up table,
although in practice this processing is negligible
except in rare cases). Canon's raw appears to be a true
raw, for example.

Roger

Roger N. Clark (change username to rnclark) wrote:

And just to be clear: are you saying that a raw image shot with wildly
incorrect white balance, and one shot with perfect white balance,
actually contain the same data apart from the white balance information?

Raw, at least in some cameras is a straight linear
analog to digital conversion with no changes to the data
regardless of settings on the camera.

OK, but it is at least theoretically possible - is it not? - that a
camera could use its white balance settings to affect the way the sensor
records the light (as I understand it does with sensitivity settings in
order to get it to behave differently).

Daniele

In article , change username to rnclark
wrote:

Some cameras
then do post processing before writing the raw
to the memory card (Nikons, for example do some minimal
processing including lossy compression via a look-up table,
although in practice this processing is negligible
except in rare cases). Canon's raw appears to be a true
raw, for example.

nikon has both an uncompressed and compressed raw. lower end nikon
cameras only offer the compressed variety, while higher end ones offer
both.

(D.M. Procida) wrote:
Roger N. Clark (change username to rnclark) wrote:

And just to be clear: are you saying that a raw image shot with wildly
incorrect white balance, and one shot with perfect white balance,
actually contain the same data apart from the white balance information?

Raw, at least in some cameras is a straight linear
analog to digital conversion with no changes to the data
regardless of settings on the camera.

OK, but it is at least theoretically possible - is it not? - that a
camera could use its white balance settings to affect the way the sensor
records the light (as I understand it does with sensitivity settings in
order to get it to behave differently).

In fact that does happen.

Exposure calculations, histograms, and
blink-on-over-exposure LCD displays will all change
depending on the WB, because WB is applied to the JPEG
image produced and it is _that_ image, not the raw
sensor data, which is used to generate those particular
data sets. (That is true even if you shoot RAW only.)

Also it appears that some cameras might (the
manufacturers are tight lipped about sensors and digital
to analog converters, so we don't really know) actually
have some analog gain adjustment done, based on WB
sensing, in the analog channel between the sensor and
the conversion to digital.

Regardless of that, if you shoot RAW the cameras
determination of what the WB should be (whether that is
manual or automatically generated from a WB light
sensor) is embedded into the RAW data file, and can be
used by an external program that converts to JPEG or
another format for viewing. Most such programs have a
built in algorithm that also attempts to automatically
determine WB from the image data. And of course most of
them will allow you to manually select some part of the
image and use that as a reference to set WB.

Hence it is often interesting to compare between the
camera, the program, and using a white reference (and of
course your memory of the scene or you idea of what you
want to to look like).

--
Floyd L. Davidson http://www.apaflo.com/floyd_davidson
Ukpeagvik (Barrow, Alaska)

On Sep 3, 12:47 am, (Floyd L. Davidson) wrote:
(D.M. Procida) wrote:
Roger N. Clark (change username to rnclark) wrote:

And just to be clear: are you saying that a raw image shot with wildly
incorrect white balance, and one shot with perfect white balance,
actually contain the same data apart from the white balance information?

Raw, at least in some cameras is a straight linear
analog to digital conversion with no changes to the data
regardless of settings on the camera.

OK, but it is at least theoretically possible - is it not? - that a
camera could use its white balance settings to affect the way the sensor
records the light (as I understand it does with sensitivity settings in
order to get it to behave differently).

In fact that does happen.

Exposure calculations, histograms, and
blink-on-over-exposure LCD displays will all change
depending on the WB, because WB is applied to the JPEG
image produced and it is _that_ image, not the raw
sensor data, which is used to generate those particular
data sets. (That is true even if you shoot RAW only.)

Also it appears that some cameras might (the
manufacturers are tight lipped about sensors and digital
to analog converters, so we don't really know) actually
have some analog gain adjustment done, based on WB
sensing, in the analog channel between the sensor and
the conversion to digital.

In fact, the external WB sensor of the D2x makes me suspect that they
planned to make the camera WB the data before digitisation (since then
you'd need to know WB before you did anything to the data!). Also
Nikon seems to imply that something is done to the data in the
analogue domain (for both the D2x and d200), but they don't say what.
And the D200, at least, certainly doesn't WB the data.

However, if you use the multiple exposure function of the D200, WB is
applied to the raw data. It's a neat function, by the way, you can set
the camera to take a multiple exposure of 10 frames at ISO 100 and the
result is silky smooth. You cannot see noise no matter how much you
lift the shadows and sharpen. It's like having very low ISOs
available. But you have to set WB accurately (I suspect they apply
other adjustments too, such as contrast etc, but I'm not sure).