Computer Graphics with Photoshop - Opening raw files in CS

Author

Opening raw files in CS

Jytzel

2004-05-22, 11:28 pm

I was wondering if there is any benifit of using CS RAW dialogue for
correcting raw images (contrast, exposure, etc.) Would it jus be
better, or at least the same, not to use any of those corrections and
applying the ordinary corrections (curves, levels, etc.) in Photoshop?

thanks

Lionel

2004-05-23, 12:08 pm

Kibo informs me that jytzel@netscape.net (Jytzel) stated that:

>I was wondering if there is any benifit of using CS RAW dialogue for
>correcting raw images (contrast, exposure, etc.) Would it jus be
>better, or at least the same, not to use any of those corrections and
>applying the ordinary corrections (curves, levels, etc.) in Photoshop?

In general, it's better to do as many of your adjustments as possible in
the RAW dialog. The advantage is that the RAW converter has more data
available to it at that point, & it get stripped out when the image is
transferred to the main program.

--
W
. | ,. w , "Some people are alive only because
\|/ \|/ it is illegal to kill them." Perna condita delenda est
---^----^---------------------------------------------------------------

Alfred Molon

2004-05-23, 12:08 pm

Jytzel <jytzel@netscape.net> wrote:
>I was wondering if there is any benifit of using CS RAW dialogue for
>correcting raw images (contrast, exposure, etc.) Would it jus be
>better, or at least the same, not to use any of those corrections and
>applying the ordinary corrections (curves, levels, etc.) in Photoshop?

In my opinion it's best to do all optimisations before JPEG compression.
Besides if you optimise the RAW file all optimisation parameters (WB,
exposusre, shadows, contrast, etc.) are stored in the Exif data and can
be retrieved later - something you cannot do if you optimise with
ordinary corrections.
--

Alfred Molon
------------------------------
http://groups.yahoo.com/group/Olympus_405080/
Olympus 5050 resource - http://www.molon.de/5050.html
Olympus 5060 resource - http://www.molon.de/5060.html
Olympus 8080 resource - http://www.molon.de/8080.html

Jytzel

2004-05-23, 12:08 pm

Lionel <nop@alt.net> wrote in message news:<ql40b0pofht029n2muv69r298rf9fflhdc@4ax.com>...
> Kibo informs me that jytzel@netscape.net (Jytzel) stated that:
>
>
> In general, it's better to do as many of your adjustments as possible in
> the RAW dialog. The advantage is that the RAW converter has more data
> available to it at that point, & it get stripped out when the image is
> transferred to the main program.

But the image is still in RAW format even after it is transfered to
the main program, rigth?

David Kilpatrick

2004-05-23, 12:08 pm

Jytzel wrote:

> Lionel <nop@alt.net> wrote in message news:<ql40b0pofht029n2muv69r298rf9fflhdc@4ax.com>...
>
>
>
> But the image is still in RAW format even after it is transfered to
> the main program, rigth?

no. the moment it opens in the CS main window, it's Photoshop data,
8-bit usually, and all the information you had access to in the raw file
is no longer there. You can't make further exposure adjustments, white
balance changes etc, without losing data.

Do all this in the raw conversion stage.

David

Johan W. Elzenga

2004-05-23, 12:08 pm

David Kilpatrick <iconmags3@btconnect.com> wrote:

>
>
> no. the moment it opens in the CS main window, it's Photoshop data,
> 8-bit usually, and all the information you had access to in the raw file
> is no longer there. You can't make further exposure adjustments, white
> balance changes etc, without losing data.

Not neccessarily. You can also output the data in 16 bits format, in
which case you haven't lost anything and you could make your corrections
in Photoshop if you like.

--
Johan W. Elzenga johan<<at>>johanfoto.nl
Editor / Photographer http://www.johanfoto.nl/

Alfred Molon

2004-05-23, 12:08 pm

Johan W. Elzenga <nomail@please.invalid> wrote:

>Not neccessarily. You can also output the data in 16 bits format, in
>which case you haven't lost anything and you could make your corrections
>in Photoshop if you like.

Well, if you choose your black and white points in such a way to clip
away some of the histogram (you might want to do that to have stronger
colours and/or a more vibrant image) you cannot go back without losses.
The best idea is to do all image processing at the RAW stage.

If you want to make some changes to the image later, you can again start
with the RAW image, because the JPEG file contains all information about
the paramaters you used for the RAW conversion (in File Info).
--

Alfred Molon
------------------------------
http://groups.yahoo.com/group/Olympus_405080/
Olympus 5050 resource - http://www.molon.de/5050.html
Olympus 5060 resource - http://www.molon.de/5060.html
Olympus 8080 resource - http://www.molon.de/8080.html

Johan W. Elzenga

2004-05-23, 12:08 pm

Alfred Molon <alfred_molon@REMOVEyahoo.com> wrote:

> Johan W. Elzenga <nomail@please.invalid> wrote:
>
>
> Well, if you choose your black and white points in such a way to clip
> away some of the histogram (you might want to do that to have stronger
> colours and/or a more vibrant image) you cannot go back without losses.

That's obvious, but not very realistic. If you plan to do your
corrections in Photoshop, it's obvious that you will *NOT* clip anything
while opening the RAW file. Choosing black and white points will *ALSO*
be done in Photoshop in that case.

--
Johan W. Elzenga johan<<at>>johanfoto.nl
Editor / Photographer http://www.johanfoto.nl/

Bill Hilton

2004-05-23, 12:08 pm

>> In general, it's better to do as many of your adjustments as possible in

That's the way I prefer to work too.
[color=darkred]
>From: jytzel@netscape.net (Jytzel)
>
>But the image is still in RAW format even after it is transfered to
>the main program, rigth?

No, typically it's converted from camera specific RAW format to TIFF or
something similar. Maybe you're thinking of the Photoshop RAW format but
that's different from the camera RAW (or NEF or whatever).

Bill

Alfred Molon

2004-05-23, 7:14 pm

Johan W. Elzenga <nomail@please.invalid> wrote:

>
>That's obvious, but not very realistic. If you plan to do your
>corrections in Photoshop, it's obvious that you will *NOT* clip anything
>while opening the RAW file. Choosing black and white points will *ALSO*
>be done in Photoshop in that case.

It is possible to do all optimisations at the RAW stage - white balance,
white and black points, saturation, sharpness, contrast, noise and CA
reduction etc. That's the beauty of RAW - further processing after the
RAW converter stage is not necessary.
--

Alfred Molon
------------------------------
http://groups.yahoo.com/group/Olympus_405080/
Olympus 5050 resource - http://www.molon.de/5050.html
Olympus 5060 resource - http://www.molon.de/5060.html
Olympus 8080 resource - http://www.molon.de/8080.html

Johan W. Elzenga

2004-05-23, 7:14 pm

Alfred Molon <alfred_molon@REMOVEyahoo.com> wrote:

> Johan W. Elzenga <nomail@please.invalid> wrote:
>
>
> It is possible to do all optimisations at the RAW stage - white balance,
> white and black points, saturation, sharpness, contrast, noise and CA
> reduction etc. That's the beauty of RAW - further processing after the
> RAW converter stage is not necessary.

I know that, and that's the way I work myself. But that's not what the
discussion is about, though. The discussion is about that is not BETTER
to do all the optimisations at the RAW stage, it's a matter of choice.
If you open the file in 16 bits in Photoshop, you *CAN* do it there if
you want. After all, Photoshop has more uptions than the RAW opener has.
It's up to you to use what you like best.

--
Johan W. Elzenga johan<<at>>johanfoto.nl
Editor / Photographer http://www.johanfoto.nl/

Lionel

2004-05-23, 11:14 pm

Kibo informs me that jytzel@netscape.net (Jytzel) stated that:

>Lionel <nop@alt.net> wrote in message news:<ql40b0pofht029n2muv69r298rf9fflhdc@4ax.com>...
>
>But the image is still in RAW format even after it is transfered to
>the main program, rigth?

Nope. When you hit that "Ok" button, the image is converted into
Photoshops internal format.

--
W
. | ,. w , "Some people are alive only because
\|/ \|/ it is illegal to kill them." Perna condita delenda est
---^----^---------------------------------------------------------------

Lars Ekdahl

2004-05-24, 4:14 am

Alfred Molon <alfred_molon@REMOVEyahoo.com> wrote in message news:<MPG.1b1a56d96d3f99a298a65b@news.supernews.com>...
> Jytzel <jytzel@netscape.net> wrote:
>
> In my opinion it's best to do all optimisations before JPEG compression.
> Besides if you optimise the RAW file all optimisation parameters (WB,
> exposusre, shadows, contrast, etc.) are stored in the Exif data and can
> be retrieved later - something you cannot do if you optimise with
> ordinary corrections.

Just come across this article which might be of interest for the
readers of this thread
http://www.creativepro.com/story/feature/21351.html

Lars Ekdahl
webmaster http://www.ekdahl.org/digital.htm

Bill Hilton

2004-05-24, 11:14 pm

>From: lars@ekdahl.org (Lars Ekdahl)

>Just come across this article which might be of interest for the
>readers of this thread
>http://www.creativepro.com/story/feature/21351.html

Thanks Lars, that was a really good article. Since I was just shooting my
Macbeth card at different ISOs to check for noise I think I'll use one of the
files to do this calibration.

I wonder if others have compared CS RAW to Capture One's RAW converter? I get
excellent results from C1 with a Canon 10D and Canon 1Ds ...

Bill

Stephen H. Westin

2004-05-25, 12:14 pm

nomail@please.invalid (Johan W. Elzenga) writes:

> David Kilpatrick <iconmags3@btconnect.com> wrote:
>
>
> Not neccessarily. You can also output the data in 16 bits format, in
> which case you haven't lost anything and you could make your corrections
> in Photoshop if you like.

Unless, of course, there are problems like overexposure or out-of-gamut
colors that get clamped.

--
-Stephen H. Westin
Any information or opinions in this message are mine: they do not
represent the position of Cornell University or any of its sponsors.

Darryl Okahata

2004-05-25, 7:14 pm

westin*nospam@graphics.cornell.edu (Stephen H. Westin) writes:

>
> Unless, of course, there are problems like overexposure or out-of-gamut
> colors that get clamped.

True. However, with Canon 300D/10D RAW data, "slightly blown
highlights" can often be corrected.

[ That's because the "overexposure" exists only in the thumbnail/LCD
jpeg preview image, and may not actually exist in the 300D/10D RAW
data, because RAW data has almost another full f-stop of headroom
above the processed Canon jpeg image. ]

--
Darryl Okahata
darrylo@soco.agilent.com

DISCLAIMER: this message is the author's personal opinion and does not
constitute the support, opinion, or policy of Agilent Technologies, or
of the little green men that have been following him all day.

Johan W. Elzenga

2004-05-25, 7:14 pm

Stephen H. Westin <westin*nospam@graphics.cornell.edu> wrote:

>
> Unless, of course, there are problems like overexposure or out-of-gamut
> colors that get clamped.

I don't see the difference. Overexposure (meaning washed out highlights)
is beyond repair in the RAW opener too. You cannot repair what isn't
there anymore. And out of gamut clamping will occur when the file is
read into Photoshop in a (too) small color space. This will happen no
matter where you do the corrections.

--
Johan W. Elzenga johan<<at>>johanfoto.nl
Editor / Photographer http://www.johanfoto.nl/

Stephen H. Westin

2004-05-25, 7:14 pm

nomail@please.invalid (Johan W. Elzenga) writes:

> Stephen H. Westin <westin*nospam@graphics.cornell.edu> wrote:
>
>
> I don't see the difference. Overexposure (meaning washed out highlights)
> is beyond repair in the RAW opener too.

As someone else pointed out, that's not true. Most raw files contain a
broader brightness range than is normally represented in the converted
image. With Kodak DSLR's, you can pull back two full stops!

> You cannot repair what isn't
> there anymore. And out of gamut clamping will occur when the file is
> read into Photoshop in a (too) small color space. This will happen no
> matter where you do the corrections.

Not if you change color balance so as to bring the image within gamut.
Imagine an image that was shot with an electronic flash with white
balance set by mistake to tungsten. Everything will be shifted towards
blue; objects in the image that started as a saturated blue may be
beyond gamut. Switching to "flash" white balance during the raw
conversion will allow the converted image to represent these colors
better.

--
-Stephen H. Westin
Any information or opinions in this message are mine: they do not
represent the position of Cornell University or any of its sponsors.

Stephen H. Westin

2004-05-25, 7:14 pm

Darryl Okahata <darrylo@soco.agilent.com> writes:

> westin*nospam@graphics.cornell.edu (Stephen H. Westin) writes:
>
>
> True. However, with Canon 300D/10D RAW data, "slightly blown
> highlights" can often be corrected.

But only in the raw conversion: if you convert with default
parameters, the information is gone.
j
<snip>

--
-Stephen H. Westin
Any information or opinions in this message are mine: they do not
represent the position of Cornell University or any of its sponsors.

Johan W. Elzenga

2004-05-25, 7:14 pm

Stephen H. Westin <westin*nospam@graphics.cornell.edu> wrote:

>
> As someone else pointed out, that's not true. Most raw files contain a
> broader brightness range than is normally represented in the converted
> image. With Kodak DSLR's, you can pull back two full stops!

It all depends what a "converted image" is. You seem to be talking about
an 8 bits per color image, I'm not. I'm talking about a 16 bits image,
as I have stated from the start. Those two stops will be in the 16 bits
converted image too, providing that you didn't unneccessarily clip them
when you opened the file.

Clipping may happen using the RAW opener for the following reason:
Photoshop's RAW opener has default settings that are NOT zero. Use the
default settings and you may indeed get clipping. Set all settings to
zero (except of course for color temperature, you cannot set that to
zero) and you will have no washed out highlights in the resulting 16
bits image. But if you still do, the RAW opener won't be able to fix
that either, because the information simply isn't there.

>
>
> Not if you change color balance so as to bring the image within gamut.
> Imagine an image that was shot with an electronic flash with white
> balance set by mistake to tungsten. Everything will be shifted towards
> blue; objects in the image that started as a saturated blue may be
> beyond gamut. Switching to "flash" white balance during the raw
> conversion will allow the converted image to represent these colors
> better.

I still disagree. If you open the file in Photoshop in a color space
that is as wide or wider than your camera's color space, I see no reason
why correcting the white balance in the RAW converter should be better
than doing so in 16 bits in Photoshop. Besides, the ONE thing you HAVE
to do in the RAW opener anyway is setting a white balance. After all,
you cannot open an image without any white balance at all. So yes,
setting/correcting the white balance would be something one would
normally do in the RAW opener, even if all other corrections were -for
whatever reasons- made in Photoshop.

--
Johan W. Elzenga johan<<at>>johanfoto.nl
Editor / Photographer http://www.johanfoto.nl/

Darryl Okahata

2004-05-25, 7:14 pm

westin*nospam@graphics.cornell.edu (Stephen H. Westin) writes:

>
> But only in the raw conversion: if you convert with default
> parameters, the information is gone.

How is the RAW conversion being done? While I do not have PS CS,
I've been told that converting from RAW to 48-bit Adobe PSD will not
result in loss of information. I don't see how converting RAW (12-bits
per channel) to 48-bit PSD (16-bits per channel) could result in loss of
information.

[ I'm talking about using dcraw to do the conversion to 48-bit PSD,
where the default PSD conversion settings do not do any brightness or
gamma adjustments. ]

--
Darryl Okahata
darrylo@soco.agilent.com

DISCLAIMER: this message is the author's personal opinion and does not
constitute the support, opinion, or policy of Agilent Technologies, or
of the little green men that have been following him all day.

Stephen H. Westin

2004-05-25, 7:14 pm

nomail@please.invalid (Johan W. Elzenga) writes:

> Stephen H. Westin <westin*nospam@graphics.cornell.edu> wrote:
>
>
> It all depends what a "converted image" is. You seem to be talking
> about an 8 bits per color image, I'm not. I'm talking about a 16
> bits image, as I have stated from the start. Those two stops will be
> in the 16 bits converted image too, providing that you didn't
> unneccessarily clip them when you opened the file.

You've never actually tried this, have you? Notice that an image
converted to 16 bits is just as bright as if you convert it to 8
bits. That means the same camera value is being mapped to the
maximum. You can, however, move the exposure slider to -1.0 and
recover highlight information. I just tried this with Photoshop CS on
a Canon CRW image with overesposed highlights. Converting to 16 bits
doesn't change the histogram; the same pixels are blown out. I can't
move the "Levels" slider beyond 255 to get beck the pixels that were
clamped to maximum. And yes, I understand that 255 on the Levels
dialog really means 32768 on a 16-bit image. If those bright pixels
are mapped to, say, 40,000, the information is gone. If, however, I
convert to 8-bit mode and move the exposure slider to -2.0, some of
the blown pixels can be saved.

> Clipping may happen using the RAW opener for the following reason:
> Photoshop's RAW opener has default settings that are NOT zero. Use
> the default settings and you may indeed get clipping. Set all
> settings to zero (except of course for color temperature, you cannot
> set that to zero) and you will have no washed out highlights in the
> resulting 16 bits image.

No, that's not correct. Setting exposure to -2.0 will preserve the
full brightness range. Leaving it at zero will, in general, clip some
bright values.

> But if you still do, the RAW opener won't be able to fix
> that either, because the information simply isn't there.

Really, there is information in the raw file that doesn't make it into
Photoshop with the default settings, or with setting everything to
zero. Would you like me to dump a raw file to show you?

>
> I still disagree. If you open the file in Photoshop in a color space
> that is as wide or wider than your camera's color space, I see no
> reason why correcting the white balance in the RAW converter should
> be better than doing so in 16 bits in Photoshop.

And which color space would that be? Why do you think that any color
space exists that includes the entire raw gamut of your camera? How do
you know what the camera raw gamut is?

> Besides, the ONE thing you HAVE to do in the RAW opener anyway is
> setting a white balance.

Or let it default to the setting of the camera at capture time.

> After all, you cannot open an image without any white balance at
> all.

Trust me; you wouldn't want to.

> So yes,
> setting/correcting the white balance would be something one would
> normally do in the RAW opener, even if all other corrections were -for
> whatever reasons- made in Photoshop.

--
-Stephen H. Westin
Any information or opinions in this message are mine: they do not
represent the position of Cornell University or any of its sponsors.

Stephen H. Westin

2004-05-25, 7:14 pm

Darryl Okahata <darrylo@soco.agilent.com> writes:

> westin*nospam@graphics.cornell.edu (Stephen H. Westin) writes:
>
>
> How is the RAW conversion being done?

1. Read raw data values from the file. These may be 10 or 12 bits.

2. Convert raw data to linear. A CCD is inherently linear, but
many cameras apply a nonlinear transfer function to save bits.
And the Canon D60, for example, seems to be slightly nonlinear.

3. Transform color. This includes not only white balance correction,
but also color space conversion.

4. Demosaic: fill in the missing color values, since each sensel
picks up only a single color (except for Foveon-based cameras).

5. Correct for display gamma. All common color spaces include a
nonlinear transfer function that compensates, at least approximately,
for the display nonlinearity.

> While I do not have PS CS,
> I've been told that converting from RAW to 48-bit Adobe PSD will not
> result in loss of information. I don't see how converting RAW (12-bits
> per channel) to 48-bit PSD (16-bits per channel) could result in loss of
> information.

It's actually intentional. If you take some value less than the sensor
maximum and scale that up to the maximum pixel value, you retain the
ability to compensate for some degree of overexposure. Different
cameras do this differently; the Kodak DSLR's probably provide the
most headroom.

Oh, by the way, you don't get a full 16 bits per channel in Photoshop;
the value range is 0:32768, which is 15 bits and one extra value.

<snip>

--
-Stephen H. Westin
Any information or opinions in this message are mine: they do not
represent the position of Cornell University or any of its sponsors.

Alfred Molon

2004-05-25, 7:14 pm

Stephen H. Westin <westin*nospam@graphics.cornell.edu> wrote:

>2. Convert raw data to linear. A CCD is inherently linear, but
> many cameras apply a nonlinear transfer function to save bits.
> And the Canon D60, for example, seems to be slightly nonlinear.
>
>3. Transform color. This includes not only white balance correction,
> but also color space conversion.
>
>4. Demosaic: fill in the missing color values, since each sensel
> picks up only a single color (except for Foveon-based cameras).

Wouldn't step 4 have to happen before step 3 ?
--

Alfred Molon
------------------------------
http://groups.yahoo.com/group/Olympus_405080/
Olympus 5050 resource - http://www.molon.de/5050.html
Olympus 5060 resource - http://www.molon.de/5060.html
Olympus 8080 resource - http://www.molon.de/8080.html

Darryl Okahata

2004-05-25, 9:06 pm

westin*nospam@graphics.cornell.edu (Stephen H. Westin) writes:

> nomail@please.invalid (Johan W. Elzenga) writes:
>
>
> You've never actually tried this, have you? Notice that an image
> converted to 16 bits is just as bright as if you convert it to 8
> bits. That means the same camera value is being mapped to the
> maximum.

I think he's talking about converting 12-bit (per channel) RAW data
directly into 16-bit (per channel) data and not 8-bit data.

[ And, I think a key, unstated assumption here is that the CAMERA'S
HISTOGRAM (not CS PS) is saying that the highlights are blown, which
means that they might not really be blown -- see below. ]

> You can, however, move the exposure slider to -1.0 and
> recover highlight information. I just tried this with Photoshop CS on
> a Canon CRW image with overesposed highlights. Converting to 16 bits
> doesn't change the histogram; the same pixels are blown out. I can't
> move the "Levels" slider beyond 255 to get beck the pixels that were
> clamped to maximum. And yes, I understand that 255 on the Levels
> dialog really means 32768 on a 16-bit image. If those bright pixels
> are mapped to, say, 40,000, the information is gone. If, however, I
> convert to 8-bit mode and move the exposure slider to -2.0, some of
> the blown pixels can be saved.

Going off on a tangent: while this may be true in PS CS, the jpeg
(8-bit) conversion in a Canon 300D/10D is quite different (note that the
camera shows a jpeg thumbnail image in the LCD display, even when
shooting RAW, and that the camera's histogram appears to be that of the
jpeg, and *NOT* RAW, data). The Canon in-camera jpeg conversion will
blow jpeg highlights long before they are really blown in the RAW data.
The red channel is the worst; someone estimated that the red channel
blows (clipped to 255) around 1350 (out of 4K values). The blue and
green channels blow around 2100.

So, if the camera's histogram says that the highlights are blown,
they are blown in the jpeg, and might not really be blown in the RAW
data.

--
Darryl Okahata
darrylo@soco.agilent.com

DISCLAIMER: this message is the author's personal opinion and does not
constitute the support, opinion, or policy of Agilent Technologies, or
of the little green men that have been following him all day.

Bill Hilton

2004-05-25, 11:14 pm

>From: Darryl Okahata darrylo@soco.agilent.com

>How is the RAW conversion being done? While I do not have PS CS,
>I've been told that converting from RAW to 48-bit Adobe PSD will not
>result in loss of information. I don't see how converting RAW (12-bits
>per channel) to 48-bit PSD (16-bits per channel) could result in loss of
>information.

There is no conversion from 12 bits/channel to 16 bits/channel ... it's just 12
bits of actual information with the upper 4 bits set to zeros.

All 10 bit/channel, 12 b/c, 14 b/c and 16 b/c files default to 16 bit mode in
Photoshop but the upper bits are just zeroed out (except of course for true 16
bits/channel input files). This is because it's much simpler to store the data
in two eight bit bytes.

Bart van der Wolf

2004-05-26, 12:14 pm

"Bill Hilton" <bhilton665@aol.comedy> wrote in message
news:20040525214759.17812.00001986@mb-m16.aol.com...
SNIP
> There is no conversion from 12 bits/channel to 16 bits/channel ... it's
just 12
> bits of actual information with the upper 4 bits set to zeros.

The zero padding is usually in the least significant bits.

Bart

Stephen H. Westin

2004-05-26, 12:14 pm

Alfred Molon <alfred_molon@REMOVEyahoo.com> writes:

> Stephen H. Westin <westin*nospam@graphics.cornell.edu> wrote:
>
>
> Wouldn't step 4 have to happen before step 3 ?

Yup. My mistake.

--
-Stephen H. Westin
Any information or opinions in this message are mine: they do not
represent the position of Cornell University or any of its sponsors.

Stephen H. Westin

2004-05-26, 12:14 pm

bhilton665@aol.comedy (Bill Hilton) writes:

>
>
> There is no conversion from 12 bits/channel to 16 bits/channel ... it's just 12
> bits of actual information with the upper 4 bits set to zeros.
>
> All 10 bit/channel, 12 b/c, 14 b/c and 16 b/c files default to 16 bit mode in
> Photoshop but the upper bits are just zeroed out (except of course for true 16
> bits/channel input files). This is because it's much simpler to store the data
> in two eight bit bytes.

That's obviously not true; if it were, a 12-bit camera would show no values
above 31 or 32 in the "Levels" histogram, as that maps the full 0:32768 range
to 0:255, and an image with the top 4 bits zero would have a maximum value of
4095, about 1/8 of the 32,768 maximum for Photoshop 16-bit data.

--
-Stephen H. Westin
Any information or opinions in this message are mine: they do not
represent the position of Cornell University or any of its sponsors.

Stephen H. Westin

2004-05-26, 12:14 pm

Darryl Okahata <darrylo@soco.agilent.com> writes:

> westin*nospam@graphics.cornell.edu (Stephen H. Westin) writes:
>
>
> I think he's talking about converting 12-bit (per channel) RAW data
> directly into 16-bit (per channel) data and not 8-bit data.

So was I, hence the phrase "an image converted to 16 bits". I should have
said "a raw image converted to 16 bits". By point is this: if you take
the same raw file and conver twice in CS, once in 16-bit mode and once
in 8-bit mode, the "Levels" histograms will be essentially identical on
the screen. That is to say, the same raw value that is converted to 255
in the 8-bit image will be converted to 32,768 in the 16-bit image;
any raw value greater than that will be clamped to 32,768 (or 255).
You don't gain any headroom simply by using 16 bits.

> [ And, I think a key, unstated assumption here is that the CAMERA'S
> HISTOGRAM (not CS PS) is saying that the highlights are blown, which
> means that they might not really be blown -- see below. ]

That's what I mean: the normal raw conversion for any camera will
actually lose the highest raw values from the sensor. The "exposure"
slider lets you recover these values, and there's no other way to do this.

>
> Going off on a tangent: while this may be true in PS CS, the jpeg
> (8-bit) conversion in a Canon 300D/10D is quite different (note that the
> camera shows a jpeg thumbnail image in the LCD display, even when
> shooting RAW, and that the camera's histogram appears to be that of the
> jpeg, and *NOT* RAW, data). The Canon in-camera jpeg conversion will
> blow jpeg highlights long before they are really blown in the RAW data.
> The red channel is the worst; someone estimated that the red channel
> blows (clipped to 255) around 1350 (out of 4K values). The blue and
> green channels blow around 2100.
>
> So, if the camera's histogram says that the highlights are blown,
> they are blown in the jpeg, and might not really be blown in the RAW
> data.

Exactly what I'm trying (unsuccessfully) to say: the default
conversion (and Photoshop's default conversion is, I'm sure designed
to mimic what Canon does in this respect) will lose some values. The
point of those sliders is to let you intervene in the process and
recover this information, which can't be recovered after the
conversion.

--
-Stephen H. Westin
Any information or opinions in this message are mine: they do not
represent the position of Cornell University or any of its sponsors.

Darryl Okahata

2004-05-26, 7:14 pm

westin*nospam@graphics.cornell.edu (Stephen H. Westin) writes:

>
> So was I, hence the phrase "an image converted to 16 bits". I should have
> said "a raw image converted to 16 bits". By point is this: if you take
> the same raw file and conver twice in CS, once in 16-bit mode and once
> in 8-bit mode, the "Levels" histograms will be essentially identical on
> the screen. That is to say, the same raw value that is converted to 255
> in the 8-bit image will be converted to 32,768 in the 16-bit image;
> any raw value greater than that will be clamped to 32,768 (or 255).
> You don't gain any headroom simply by using 16 bits.

True, but the "headroom" is in the camera, not PS CS. By shooting
RAW, you gain a little headroom in the camera, to (hopefully) avoid
blown highlights. If you can avoid blowing the highlights in the RAW
data (the camera's LCD histogram may still say that the highlights are
blown, though), then none of the values will be converted to 255/32768
in PS CS.

> Exactly what I'm trying (unsuccessfully) to say: the default
> conversion (and Photoshop's default conversion is, I'm sure designed
> to mimic what Canon does in this respect) will lose some values. The
> point of those sliders is to let you intervene in the process and
> recover this information, which can't be recovered after the
> conversion.

[ I don't have PS CS, so take the following with the appropriate chunk
of salt. ]

While this may be true of the PS CS converter, others seem to be
using a different approach to RAW & PS CS: they do not use the RAW
converter in PS CS. Instead, they use dcraw (which is free, third-party
software -- not Canon's) to do a plain linear conversion (no brightness
or gamma correction) of the RAW data to 48-bit PSD data, and then read
the PSD file into PS CS.

Note that, even though the *camera* may say that the highlights are
blown, they may only be blown in the camera's LCD jpeg preview, and may
not be blown in the RAW data. Because of this, there is no clipping (by
dcraw -- I have no idea what PS CS does) or loss of information during
the dcraw conversion process. This is because, assuming that nothing in
the RAW data is blown, all of the RAW data are below the maximum value
(and because no gamma or brightness correction is done).

--
Darryl Okahata
darrylo@soco.agilent.com

DISCLAIMER: this message is the author's personal opinion and does not
constitute the support, opinion, or policy of Agilent Technologies, or
of the little green men that have been following him all day.

Johan W. Elzenga

2004-05-26, 7:14 pm

Stephen H. Westin <westin*nospam@graphics.cornell.edu> wrote:

>
> So was I, hence the phrase "an image converted to 16 bits". I should have
> said "a raw image converted to 16 bits". By point is this: if you take
> the same raw file and conver twice in CS, once in 16-bit mode and once
> in 8-bit mode, the "Levels" histograms will be essentially identical on
> the screen. That is to say, the same raw value that is converted to 255
> in the 8-bit image will be converted to 32,768 in the 16-bit image;
> any raw value greater than that will be clamped to 32,768 (or 255).
> You don't gain any headroom simply by using 16 bits.

No, indeed you don't and that is not what I meant either. I realize that
I haven't been clear enough. What I meant is this:

If you open a RAW image in 16 bits rather than 8 bits, you can do all
the *FURTHER* corrections in Photoshop, without losing data by doing so.
Why would you want to do that? The RAW opener does not have anything
like 'Shadow/Highlight' or Adjustments Layers with layer masks, for
example, so there can be many reasons why someone may prefer to use
Photoshop for fine tuning the image. Personally, I make as many
corrections as I can in the RAW opener, but I still use 16 bits output,
because I may indeed want to do some extra work involving Adjustment
layers and/or the Shadow/Highlight correction. Only when I'm really
satisfied with the results, I convert to 8 bits. I believe that gives
the best possible results. It's not either/or, it's and/and.

> Exactly what I'm trying (unsuccessfully) to say: the default
> conversion (and Photoshop's default conversion is, I'm sure designed
> to mimic what Canon does in this respect) will lose some values. The
> point of those sliders is to let you intervene in the process and
> recover this information, which can't be recovered after the
> conversion.

I think it's a matter of semantics. What is 'the default conversion'
anyway? I don't think there is anything like that. There is only the
conversion using default values, but we agree that those aren't always
the best settings for every image. You should obviously convert the RAW
file without causing any unnecessary clipping in the first place.
Usually, if your image was correctly exposed, that simply means you set
the correct white balance and you can leave everything else at zero.
However, if your image was overexposed, you may indeed have to set the
exposure at -1 or -2 in order not to cause any EXTRA clipping. So be it.

--
Johan W. Elzenga johan<<at>>johanfoto.nl
Editor / Photographer http://www.johanfoto.nl/

Bill Hilton

2004-05-26, 7:14 pm

>> There is no conversion from 12 bits/channel to 16 bits/channel ...

[color=darkred]
>From: "Bart van der Wolf" bvdwolf@no.spam
>
>The zero padding is usually in the least significant bits.

You're right of course, I was sloppy ... what I meant was the extra four bits
are set to zeros instead of containing any actual useful data :) It's still 12
bits of data even though you open the file in 16 bit mode in Photoshop.

Bill

Bill Hilton

2004-05-26, 7:14 pm

>bhilton665@aol.comedy (Bill Hilton) writes:

[color=darkred]
>From: westin*nospam@graphics.cornell.edu (Stephen H. Westin)
>
>That's obviously not true

You and Bart are right, the upper bits aren't zeroed out ... what I meant to
say was "the unused bits are just zeroed out" instead of "the upper bits ...".

My point was that a 12 bit/channel file still only contains 12 bits of data
even though it's "converted" to 16 bit mode. Sorry if that point got lost in
the scuffle.

Bill

Stephen H. Westin

2004-05-26, 7:14 pm

Darryl Okahata <darrylo@soco.agilent.com> writes:

> westin*nospam@graphics.cornell.edu (Stephen H. Westin) writes:
>
>
> True, but the "headroom" is in the camera, not PS CS. By shooting
> RAW, you gain a little headroom in the camera, to (hopefully) avoid
> blown highlights.

Right. And the controls in the raw importer let you retrieve the extra
highlight information.

> If you can avoid blowing the highlights in the RAW
> data (the camera's LCD histogram may still say that the highlights are
> blown, though), then none of the values will be converted to 255/32768
> in PS CS.

By default, some of the values *will* be converted to the
maximum. That's one big reason why there are adjustments possible
during the conversion process.

>
> [ I don't have PS CS, so take the following with the appropriate chunk
> of salt. ]
>
> While this may be true of the PS CS converter, others seem to be
> using a different approach to RAW & PS CS: they do not use the RAW
> converter in PS CS. Instead, they use dcraw (which is free, third-party
> software -- not Canon's) to do a plain linear conversion (no brightness
> or gamma correction) of the RAW data to 48-bit PSD data, and then read
> the PSD file into PS CS.
>
> Note that, even though the *camera* may say that the highlights are
> blown, they may only be blown in the camera's LCD jpeg preview, and may
> not be blown in the RAW data. Because of this, there is no clipping (by
> dcraw -- I have no idea what PS CS does) or loss of information during
> the dcraw conversion process. This is because, assuming that nothing in
> the RAW data is blown, all of the RAW data are below the maximum value
> (and because no gamma or brightness correction is done).

Yup. I'm pretty sure dcraw, at least when asked for linear output,
won't try to emulate the default tone mapping of the camera firmware
or manufacturer's software. But that, of course, demands a fair amount
of postprocessing to create a viewable image.

--
-Stephen H. Westin
Any information or opinions in this message are mine: they do not
represent the position of Cornell University or any of its sponsors.

Stephen H. Westin

2004-05-26, 7:14 pm

nomail@please.invalid (Johan W. Elzenga) writes:

> Stephen H. Westin <westin*nospam@graphics.cornell.edu> wrote:
>
>
> No, indeed you don't and that is not what I meant either. I realize that
> I haven't been clear enough. What I meant is this:
>
> If you open a RAW image in 16 bits rather than 8 bits, you can do all
> the *FURTHER* corrections in Photoshop, without losing data by doing so.
> Why would you want to do that? The RAW opener does not have anything
> like 'Shadow/Highlight' or Adjustments Layers with layer masks, for
> example, so there can be many reasons why someone may prefer to use
> Photoshop for fine tuning the image. Personally, I make as many
> corrections as I can in the RAW opener, but I still use 16 bits output,
> because I may indeed want to do some extra work involving Adjustment
> layers and/or the Shadow/Highlight correction. Only when I'm really
> satisfied with the results, I convert to 8 bits. I believe that gives
> the best possible results. It's not either/or, it's and/and.

Absolutely. But some folks were insisting that anything that the
raw importer does can be done just as well later on. Wait, it was you:

> From: Johan W. Elzenga (nomail@please.invalid)
> Subject: Re: Opening raw files in CS
>
>
> View this article only
> Newsgroups: comp.graphics.apps.photoshop, alt.graphics.photoshop, rec.photo.digital
> Date: 2004-05-23 06:10:46 PST
>
>
> David Kilpatrick <iconmags3@btconnect.com> wrote:
>

<snip>

>
> Not neccessarily. You can also output the data in 16 bits format, in
> which case you haven't lost anything and you could make your corrections
> in Photoshop if you like.
>
>
> --
> Johan W. Elzenga johan<<at>>johanfoto.nl
> Editor / Photographer http://www.johanfoto.nl/

See, that's the focus of the misunderstanding:

"You can also output the data in 16 bits format, in which case you
haven't lost anything and you could make your corrections in Photoshop
if you like."

I think you meant something like

"You can also output the data in 16 bits format, in which case you
have the capability to preserve everything and you could make your
corrections in Photoshop if you like."

See the difference? Just clicking the "16-bit" box isn't enough: you
have to override the default settings to preserve all the extra
information.

<snip>

> I think it's a matter of semantics. What is 'the default conversion'
> anyway? I don't think there is anything like that. There is only the
> conversion using default values,

.... which is what I meant...

> but we agree that those aren't always
> the best settings for every image. You should obviously convert the RAW
> file without causing any unnecessary clipping in the first place.
> Usually, if your image was correctly exposed, that simply means you set
> the correct white balance and you can leave everything else at zero.
> However, if your image was overexposed, you may indeed have to set the
> exposure at -1 or -2 in order not to cause any EXTRA clipping. So be it.

Right. But where I came in was the start of the thread:

> From: Jytzel (jytzel@netscape.net)
> Subject: Opening raw files in CS
>
>
> View this article only
> Newsgroups: comp.graphics.apps.photoshop, alt.graphics.photoshop, rec.photo.digital
> Date: 2004-05-22 18:32:57 PST
>
>
> I was wondering if there is any benifit of using CS RAW dialogue for
> correcting raw images (contrast, exposure, etc.) Would it jus be
> better, or at least the same, not to use any of those corrections and
> applying the ordinary corrections (curves, levels, etc.) in Photoshop?

Those of us who mentioned operations that require access to the raw
sensor data seemed to get an argument. I'm glad that you and I agree.

--
-Stephen H. Westin
Any information or opinions in this message are mine: they do not
represent the position of Cornell University or any of its sponsors.

Johan W. Elzenga

2004-05-26, 7:14 pm

Stephen H. Westin <westin*nospam@graphics.cornell.edu> wrote:

> See, that's the focus of the misunderstanding:
>
> "You can also output the data in 16 bits format, in which case you
> haven't lost anything and you could make your corrections in Photoshop
> if you like."

I should have said "in which case you do not HAVE TO lose anything". As
you may have noticed, English is not my mother tongue. The point is that
I was reacting to someone who suggested that the output was 8 bits and
that you shouldn't do corrections in Photoshop because of THAT. This
person said: "the moment it opens in the CS main window, it's Photoshop
data, 8-bit usually, and all the information you had access to in the
raw file is no longer there". It's still there if you know how to get it
there!

>
> I think you meant something like
>
> "You can also output the data in 16 bits format, in which case you
> have the capability to preserve everything and you could make your
> corrections in Photoshop if you like."
>
> See the difference? Just clicking the "16-bit" box isn't enough: you
> have to override the default settings to preserve all the extra
> information.

Wait a second, I never said that you should just hit "OK" when the RAW
dialog comes up. I specifically said that you should NOT use the default
settings. Anyway, I think we agree that you CAN make all the corrections
in Photoshop, providing that you make sure that:

A: You do not unnecessarily clip any data during the conversion;
B: You convert to a 16 bits file.

Now where did I read that before?... ;-)

--
Johan W. Elzenga johan<<at>>johanfoto.nl
Editor / Photographer http://www.johanfoto.nl/

mail@albumsfamily.com

2004-05-27, 11:14 pm

Download AlbumFamiy software at http://www.albumsfamily.com to help you

Jack-of-the-Dust

2004-05-28, 12:14 pm

Simple do not make a purchase from this spammer.
<mail@albumsfamily.com> wrote in message
news:c6e0f644.0405271712.1e2d56f9@posting.google.com...
> Download AlbumFamiy software at http://www.albumsfamily.com to help you