Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss
Photoshop Elements and RW2 colorspace questions
11 views
Skip to first unread message
Paul Ciszek
May 29, 2012, 1:31:07 PM5/29/12
to
I went ahead and bought Photoshop Elements. It turns out that the
ability to separate color channels is present but inaccesible to the
user, so I spent another $12 on an extension called Elements+ which
was recommended by several sites. Elements has a nice, easy-to-use
tool for adjusting the contrast scale of any color channel arbitrarily,
but once I use it, the next time I look at a histogram of that color
channel, it consists of a set of evenly spaced spikes--what I would
expect to see if I stretched the contrast of an 8 bit channel. I
have my doubts as to whether I am really working in a 16 bit per
channel color space. If I were, I should be able to stretch the
contrast by a factor of four as I have been doing and still have too
many levels of brightness to be able to count them in a histogram.
The explanations that seem most likely are:
1) Some of the Photoshop Elements tools operate in 8-bit mode
2) While Photoshop Elements is willing to import RW2 images, they
pass through an 8-bit-per-channel format along the way
3) The RW2 format itself does not have as much dynamic range as I
thought it did.
I have been googling for details on the RW2 colorspace and all I have
been finding is lists of programs and converters that can handle it.
Is there a good description of the RW2 format anywhere?
--
Please reply to: | "We establish no religion in this country, we
pciszek at panix dot com | command no worship, we mandate no belief, nor
Autoreply is disabled | will we ever. Church and state are, and must
| remain, separate." --Ronald Reagan, 10/26/1984
ability to separate color channels is present but inaccesible to the
user, so I spent another $12 on an extension called Elements+ which
was recommended by several sites. Elements has a nice, easy-to-use
tool for adjusting the contrast scale of any color channel arbitrarily,
but once I use it, the next time I look at a histogram of that color
channel, it consists of a set of evenly spaced spikes--what I would
expect to see if I stretched the contrast of an 8 bit channel. I
have my doubts as to whether I am really working in a 16 bit per
channel color space. If I were, I should be able to stretch the
contrast by a factor of four as I have been doing and still have too
many levels of brightness to be able to count them in a histogram.
The explanations that seem most likely are:
1) Some of the Photoshop Elements tools operate in 8-bit mode
2) While Photoshop Elements is willing to import RW2 images, they
pass through an 8-bit-per-channel format along the way
3) The RW2 format itself does not have as much dynamic range as I
thought it did.
I have been googling for details on the RW2 colorspace and all I have
been finding is lists of programs and converters that can handle it.
Is there a good description of the RW2 format anywhere?
--
Please reply to: | "We establish no religion in this country, we
pciszek at panix dot com | command no worship, we mandate no belief, nor
Autoreply is disabled | will we ever. Church and state are, and must
| remain, separate." --Ronald Reagan, 10/26/1984
PeterN
May 29, 2012, 7:50:26 PM5/29/12
to
On 5/29/2012 1:31 PM, Paul Ciszek wrote:
> I went ahead and bought Photoshop Elements. It turns out that the
> ability to separate color channels is present but inaccesible to the
> user, so I spent another $12 on an extension called Elements+ which
> was recommended by several sites. Elements has a nice, easy-to-use
> tool for adjusting the contrast scale of any color channel arbitrarily,
> but once I use it, the next time I look at a histogram of that color
> channel, it consists of a set of evenly spaced spikes--what I would
> expect to see if I stretched the contrast of an 8 bit channel. I
> have my doubts as to whether I am really working in a 16 bit per
> channel color space. If I were, I should be able to stretch the
> contrast by a factor of four as I have been doing and still have too
> many levels of brightness to be able to count them in a histogram.
> The explanations that seem most likely are:
>
> 1) Some of the Photoshop Elements tools operate in 8-bit mode
>
> 2) While Photoshop Elements is willing to import RW2 images, they
> pass through an 8-bit-per-channel format along the way
>
> 3) The RW2 format itself does not have as much dynamic range as I
> thought it did.
>
> I have been googling for details on the RW2 colorspace and all I have
> been finding is lists of programs and converters that can handle it.
> Is there a good description of the RW2 format anywhere?
>
I have seen that effect in CS5 after using levels. I think it has to do
> I went ahead and bought Photoshop Elements. It turns out that the
> ability to separate color channels is present but inaccesible to the
> user, so I spent another $12 on an extension called Elements+ which
> was recommended by several sites. Elements has a nice, easy-to-use
> tool for adjusting the contrast scale of any color channel arbitrarily,
> but once I use it, the next time I look at a histogram of that color
> channel, it consists of a set of evenly spaced spikes--what I would
> expect to see if I stretched the contrast of an 8 bit channel. I
> have my doubts as to whether I am really working in a 16 bit per
> channel color space. If I were, I should be able to stretch the
> contrast by a factor of four as I have been doing and still have too
> many levels of brightness to be able to count them in a histogram.
> The explanations that seem most likely are:
>
> 1) Some of the Photoshop Elements tools operate in 8-bit mode
>
> 2) While Photoshop Elements is willing to import RW2 images, they
> pass through an 8-bit-per-channel format along the way
>
> 3) The RW2 format itself does not have as much dynamic range as I
> thought it did.
>
> I have been googling for details on the RW2 colorspace and all I have
> been finding is lists of programs and converters that can handle it.
> Is there a good description of the RW2 format anywhere?
>
with compressing the channel and expanding it. I find afterword, I am
still working in 16 bit mode. The effect you are seeing is a partial
narrowing of the color space and then an expansion of it. That's why it
looks similar to an 8 bit space.
--
Peter
Paul Ciszek
Jun 1, 2012, 12:54:28 PM6/1/12
to
In article <4fc560ce$0$27331$8f2e...@news.shared-secrets.com>,
PeterN <pete...@nospam.verizon.net> wrote:
>
>I have seen that effect in CS5 after using levels. I think it has to do
>with compressing the channel and expanding it. I find afterword, I am
>still working in 16 bit mode. The effect you are seeing is a partial
>narrowing of the color space and then an expansion of it. That's why it
>looks similar to an 8 bit space.
Is there any way to force Photoshop Elements to explicitly state the
>
>I have seen that effect in CS5 after using levels. I think it has to do
>with compressing the channel and expanding it. I find afterword, I am
>still working in 16 bit mode. The effect you are seeing is a partial
>narrowing of the color space and then an expansion of it. That's why it
>looks similar to an 8 bit space.
color depth of the image you are looking at?
PeterN
Jun 1, 2012, 8:25:04 PM6/1/12
to
On 6/1/2012 12:54 PM, Paul Ciszek wrote:
> In article<4fc560ce$0$27331$8f2e...@news.shared-secrets.com>,
> PeterN<pete...@nospam.verizon.net> wrote:
>>
>> I have seen that effect in CS5 after using levels. I think it has to do
>> with compressing the channel and expanding it. I find afterword, I am
>> still working in 16 bit mode. The effect you are seeing is a partial
>> narrowing of the color space and then an expansion of it. That's why it
>> looks similar to an 8 bit space.
>
> Is there any way to force Photoshop Elements to explicitly state the
> color depth of the image you are looking at?
>
Sorry, I am not familiar with Elements. You might want to try the Adobe
> In article<4fc560ce$0$27331$8f2e...@news.shared-secrets.com>,
> PeterN<pete...@nospam.verizon.net> wrote:
>>
>> I have seen that effect in CS5 after using levels. I think it has to do
>> with compressing the channel and expanding it. I find afterword, I am
>> still working in 16 bit mode. The effect you are seeing is a partial
>> narrowing of the color space and then an expansion of it. That's why it
>> looks similar to an 8 bit space.
>
> Is there any way to force Photoshop Elements to explicitly state the
> color depth of the image you are looking at?
>
forums for that answer.
--
Peter
otter
Jun 5, 2012, 9:44:17 AM6/5/12
to
On May 29, 12:31 pm, nos...@nospam.com (Paul Ciszek) wrote:
> I went ahead and bought Photoshop Elements. It turns out that the
> ability to separate color channels is present but inaccesible to the
> user, so I spent another $12 on an extension called Elements+ which
> was recommended by several sites. Elements has a nice, easy-to-use
> tool for adjusting the contrast scale of any color channel arbitrarily,
> but once I use it, the next time I look at a histogram of that color
> channel, it consists of a set of evenly spaced spikes--what I would
> expect to see if I stretched the contrast of an 8 bit channel. I
> have my doubts as to whether I am really working in a 16 bit per
> channel color space. If I were, I should be able to stretch the
> contrast by a factor of four as I have been doing and still have too
> many levels of brightness to be able to count them in a histogram.
> The explanations that seem most likely are:
>
> 1) Some of the Photoshop Elements tools operate in 8-bit mode
>
> 2) While Photoshop Elements is willing to import RW2 images, they
> pass through an 8-bit-per-channel format along the way
>
> 3) The RW2 format itself does not have as much dynamic range as I
> thought it did.
>
> I have been googling for details on the RW2 colorspace and all I have
> been finding is lists of programs and converters that can handle it.
> Is there a good description of the RW2 format anywhere?
I'm not an expert in using PSE in the way you describe, but you ARE
> I went ahead and bought Photoshop Elements. It turns out that the
> ability to separate color channels is present but inaccesible to the
> user, so I spent another $12 on an extension called Elements+ which
> was recommended by several sites. Elements has a nice, easy-to-use
> tool for adjusting the contrast scale of any color channel arbitrarily,
> but once I use it, the next time I look at a histogram of that color
> channel, it consists of a set of evenly spaced spikes--what I would
> expect to see if I stretched the contrast of an 8 bit channel. I
> have my doubts as to whether I am really working in a 16 bit per
> channel color space. If I were, I should be able to stretch the
> contrast by a factor of four as I have been doing and still have too
> many levels of brightness to be able to count them in a histogram.
> The explanations that seem most likely are:
>
> 1) Some of the Photoshop Elements tools operate in 8-bit mode
>
> 2) While Photoshop Elements is willing to import RW2 images, they
> pass through an 8-bit-per-channel format along the way
>
> 3) The RW2 format itself does not have as much dynamic range as I
> thought it did.
>
> I have been googling for details on the RW2 colorspace and all I have
> been finding is lists of programs and converters that can handle it.
> Is there a good description of the RW2 format anywhere?
seeing the effects of posterization/destructive editing. I don't
believe the problem is with the RW2 format, it is photoshop, itself.
Some/many PSE (and PS) tools/filters certainly operate only in 8-bit
mode. What makes you think you were in 16-bit mode? Did you
explicitly set it? I believe 8-bit is the normal default. I believe
the tool generally will prevent you from doing 8-bit operations if you
are in 16-bit mode, and usually will ask if you want to convert to 8-
bit.
ACR (the raw converter) is the non-destructive part of PSE (and PS).
You should do as much work as possible in ACR, especially things like
white balance or adjusting contrast, etc. After you go into
photoshop, operations are generally destructive, so it is important to
use layers to protect the original. "Smart objects" allow you to
build a non-destructive flow in PS, but only works with a limited
number of tools/filters. Not sure if smart objects have made it to
PSE, yet.
Unfortunately, the capabilities available in ACR in PSE are not as
extensive as what you get in ACR in PS or the ACR engine built into
Lightroom. Actually, I think the engine may be the same in PSE, but
the necessary info in the side-car is missing in PSE (that is what an
Adobe support person told me). For example, the noise tool and
perspective adjustment aren't available in ACR in PSE (or at least
weren't a couple years ago when I tried it).
One of the reasons I use Lightroom most of the time is it is
completely non-destructive. When I have to pop over to CS5, I use
smart objects to preserve the non-destructive pipeline as much as
possible. If I have to use tools that are destructive in CS5, I save
those steps for last.
As far as 8-bit vs 16-bit is concerned, I have found that 8-bit for
the destructive edits are usually fine, as long as you have already
made the white-space and histogram adjustments non-destructively in
ACR or LR first.
Paul Ciszek
Jun 5, 2012, 10:10:51 AM6/5/12
to
In article <22210940-e3bb-4c4e...@x39g2000yqx.googlegroups.com>,
otter <bighor...@hotmail.com> wrote:
>
>I'm not an expert in using PSE in the way you describe, but you ARE
>seeing the effects of posterization/destructive editing. I don't
>believe the problem is with the RW2 format, it is photoshop, itself.
>
>Some/many PSE (and PS) tools/filters certainly operate only in 8-bit
>mode. What makes you think you were in 16-bit mode? Did you
>explicitly set it? I believe 8-bit is the normal default. I believe
>the tool generally will prevent you from doing 8-bit operations if you
>are in 16-bit mode, and usually will ask if you want to convert to 8-
>bit.
It turns out that the colorspace *is* identified in one corner of
>
>I'm not an expert in using PSE in the way you describe, but you ARE
>seeing the effects of posterization/destructive editing. I don't
>believe the problem is with the RW2 format, it is photoshop, itself.
>
>Some/many PSE (and PS) tools/filters certainly operate only in 8-bit
>mode. What makes you think you were in 16-bit mode? Did you
>explicitly set it? I believe 8-bit is the normal default. I believe
>the tool generally will prevent you from doing 8-bit operations if you
>are in 16-bit mode, and usually will ask if you want to convert to 8-
>bit.
the image, and the answer is that you *can* import a raw file as
RGB/16 if you specifically ask for it when you convert. Unfortunately,
the color splitting operation provided by Photoshop Elements Plus will
only produce 8 bit deep grayscale images, but at leat it warns you of
this. So: I use the levels tool to mess with the Green channel while
ignoring how magenta the object looks; *then*, when I split the colors,
the altered green channel is reduced to 8 bit greyscale, but those
256 levels of grey are now being used the way I wanted them too, so
it's not too bad. Here is an extreme contrast image of last month's
eclipse of the sun:
http://www.panix.com/~pciszek/xtreme_contrast.gif
It works really well at the bright end, letting you see the structure
of the sunspots and the granularity of the sun, but I need to tweak
the low end of the brightness curve somehow so that the edge of the
sun doesn't just fade away like that.
This picture was taken with a Lumix FZ35 and teleconverter, BTW--small
aperture isn't so bad when there enough light is available.
>One of the reasons I use Lightroom most of the time is it is
>completely non-destructive. When I have to pop over to CS5, I use
>smart objects to preserve the non-destructive pipeline as much as
>possible. If I have to use tools that are destructive in CS5, I save
>those steps for last.
>As far as 8-bit vs 16-bit is concerned, I have found that 8-bit for
>the destructive edits are usually fine, as long as you have already
>made the white-space and histogram adjustments non-destructively in
>ACR or LR first.
messing with the brightness curves.
0 new messages