Sign in with
Sign up | Sign in
Your question

RAW capture & Optimize Image settings on Nikon D70

Last response: in Digital Camera
Share
Anonymous
June 3, 2005 12:03:02 AM

Archived from groups: rec.photo.digital (More info?)

Hi all,

I've been experimenting with my relatively recent purchase of a Nikon D70
and I've been a little disappointed with the colour reproduction when
capturing in RAW/NEF mode. What I am seeking is clarification on the
"Optimize Image" settings in the menu.

From what I understand the RAW mode should simply capture the basic data
from the sensor, so none of the settings in this menus item should make any
difference to the image, at least when it is first imported into Nikon
Capture or Photoshop CS, and any enhancement I would need to perform myself.
Is this correct? In which case it shouldn't make any difference which
settings I leave there as they are not used except for JPEG.

Does this also apply to the Color Mode setting, which is an item lower down
in the menu?

Stan
Anonymous
June 3, 2005 12:03:03 AM

Archived from groups: rec.photo.digital (More info?)

On Thu, 02 Jun 2005 20:03:02 +1000, in rec.photo.digital G Winstanley
<stan@orange.net> wrote:

>From what I understand the RAW mode should simply capture the basic data
>from the sensor, so none of the settings in this menus item should make any
>difference to the image, at least when it is first imported into Nikon
>Capture or Photoshop CS, and any enhancement I would need to perform myself.
>Is this correct? In which case it shouldn't make any difference which
>settings I leave there as they are not used except for JPEG.

PS ignores the default settings. Unless you change the Capture defaults it
starts by using these by default.

----------
Ed Ruf Lifetime AMA# 344007 (Usenet@EdwardG.Ruf.com)
See images taken with my CP-990/5700 & D70 at
http://edwardgruf.com/Digital_Photography/General/index...
Anonymous
June 3, 2005 12:03:03 AM

Archived from groups: rec.photo.digital (More info?)

The different color spaces have radically different sizes, with
prophoto being the biggest. If you adjust your image with prophoto and
then switch to, say, sRGB (the smallest) you'll see your colors go out
of gamut.

Personally I'd rather convert with no profile and sort out the color
space for myself later eg the dcraw method
Related resources
Anonymous
June 3, 2005 4:09:23 AM

Archived from groups: rec.photo.digital (More info?)

On Thu, 02 Jun 2005 06:17:26 -0400, the cup of Ed Ruf <egruf_usenet@cox.net>
overfloweth with the following:

> On Thu, 02 Jun 2005 20:03:02 +1000, in rec.photo.digital G Winstanley
> <stan@orange.net> wrote:
>
> >From what I understand the RAW mode should simply capture the basic data
> >from the sensor, so none of the settings in this menus item should make any
> >difference to the image, at least when it is first imported into Nikon
> >Capture or Photoshop CS, and any enhancement I would need to perform myself.
> >Is this correct? In which case it shouldn't make any difference which
> >settings I leave there as they are not used except for JPEG.
>
> PS ignores the default settings. Unless you change the Capture defaults it
> starts by using these by default.
>
> ----------
> Ed Ruf Lifetime AMA# 344007 (Usenet@EdwardG.Ruf.com)
> See images taken with my CP-990/5700 & D70 at
> http://edwardgruf.com/Digital_Photography/General/index...

So, if I get this right, the selection for the Colour Space in Photoshop's
ACR only affects the translation from RAW/NEF into this colour space, rather
than being a reflection of anything that has already been captured. If I
have the setting on sRGB or Adobe RGB it doesn't matter, and I can select
ProPhotoRGB colour space in ACR to get the best out of the image? I'm
wondering whether by having this setting at sRGB in the camera whether I'm
somehow losing any image quality compared to Adobe RGB in RAW mode.

Stan
Anonymous
June 3, 2005 4:09:24 AM

Archived from groups: rec.photo.digital (More info?)

G Winstanley wrote:

> On Thu, 02 Jun 2005 06:17:26 -0400, the cup of Ed Ruf <egruf_usenet@cox.net>
> overfloweth with the following:
>
>
>>On Thu, 02 Jun 2005 20:03:02 +1000, in rec.photo.digital G Winstanley
>><stan@orange.net> wrote:
>>
>>>From what I understand the RAW mode should simply capture the basic data
>>>from the sensor, so none of the settings in this menus item should make any
>>
>>>difference to the image, at least when it is first imported into Nikon
>>>Capture or Photoshop CS, and any enhancement I would need to perform myself.
>>>Is this correct? In which case it shouldn't make any difference which
>>>settings I leave there as they are not used except for JPEG.
>>
>>PS ignores the default settings. Unless you change the Capture defaults it
>>starts by using these by default.
>>
>>----------
>>Ed Ruf Lifetime AMA# 344007 (Usenet@EdwardG.Ruf.com)
>>See images taken with my CP-990/5700 & D70 at
>>http://edwardgruf.com/Digital_Photography/General/index...
>
>
> So, if I get this right, the selection for the Colour Space in Photoshop's
> ACR only affects the translation from RAW/NEF into this colour space, rather
> than being a reflection of anything that has already been captured. If I
> have the setting on sRGB or Adobe RGB it doesn't matter, and I can select
> ProPhotoRGB colour space in ACR to get the best out of the image? I'm
> wondering whether by having this setting at sRGB in the camera whether I'm
> somehow losing any image quality compared to Adobe RGB in RAW mode.


No, you have full control in post-processing. There probably will be
some times you shoot jpegs though so set it the way you want to match
your raw conversion defaults.


--
Paul Furman
http://www.edgehill.net/1
san francisco native plants
Anonymous
June 3, 2005 4:09:24 AM

Archived from groups: rec.photo.digital (More info?)

On Fri, 03 Jun 2005 00:09:23 +1000, in rec.photo.digital G Winstanley
<stan@orange.net> wrote:


>So, if I get this right, the selection for the Colour Space in Photoshop's
>ACR only affects the translation from RAW/NEF into this colour space, rather
>than being a reflection of anything that has already been captured. If I
>have the setting on sRGB or Adobe RGB it doesn't matter, and I can select
>ProPhotoRGB colour space in ACR to get the best out of the image? I'm
>wondering whether by having this setting at sRGB in the camera whether I'm
>somehow losing any image quality compared to Adobe RGB in RAW mode.

No. I was referring to more of the WB and other tone comp and image
adjustment settings. I believe ACR doesn't know what color space your
image is and you need to explicitly tell if which one the image was
taken in if not sRGB. However, since I don't use PS, I may be
misinformed.

________________________________________________________
Ed Ruf Lifetime AMA# 344007 (Usenet@EdwardG.Ruf.com)
See images taken with my CP-990/5700 & D70 at
http://EdwardGRuf.com
June 3, 2005 4:17:46 AM

Archived from groups: rec.photo.digital (More info?)

Maintain your color space as Adobe RGB for your raw capture and when you
open the image in Photoshop. Later in Photoshop you can convert color spaces
to your heart's content.
There are many tutorials on different ways to use the shadow/exposure
adjustments or make adjustments later in Photshop.
You can choose a white point in the RAW converter or balance color later in
Photoshop.
I think it is more useful to learn how to use the options for adjusting
chromatic aberrations and vignetting.
The most controversial issue about how to use the RAW converter is whether
to work in 16bits or 8 bits. There is alot of misinformation out there about
the benefits of 16bit color and how to use it and what happens when the 8
bit printer driver gets hold of the data.
Anonymous
June 3, 2005 10:30:58 PM

Archived from groups: rec.photo.digital (More info?)

On Thu, 02 Jun 2005 12:55:40 -0400, the cup of Ed Ruf <egruf_usenet@cox.net>
overfloweth with the following:

> On Fri, 03 Jun 2005 00:09:23 +1000, in rec.photo.digital G Winstanley
> <stan@orange.net> wrote:
>
>
> >So, if I get this right, the selection for the Colour Space in Photoshop's
> >ACR only affects the translation from RAW/NEF into this colour space, rather
> >than being a reflection of anything that has already been captured. If I
> >have the setting on sRGB or Adobe RGB it doesn't matter, and I can select
> >ProPhotoRGB colour space in ACR to get the best out of the image? I'm
> >wondering whether by having this setting at sRGB in the camera whether I'm
> >somehow losing any image quality compared to Adobe RGB in RAW mode.
>
> No. I was referring to more of the WB and other tone comp and image
> adjustment settings. I believe ACR doesn't know what color space your
> image is and you need to explicitly tell if which one the image was
> taken in if not sRGB. However, since I don't use PS, I may be
> misinformed.
>
> ________________________________________________________
> Ed Ruf Lifetime AMA# 344007 (Usenet@EdwardG.Ruf.com)
> See images taken with my CP-990/5700 & D70 at
> http://EdwardGRuf.com


Ah, now we're getting closer to my intention...
> I believe ACR doesn't know what color space your image is and you need to explicitly tell if which one the image was taken in if not sRGB."

I am under the impression that any image taken in RAW/NEF mode does *not*
have an assigned colour space, but just raw values from the sensor. This
would imply that in raw conversion you can select any colour space you want
in order to do processing.

However, if the RAW/NEF file contains colour space information, then it
*definitely* matters which colour space you choose in raw processing, and
should choose the one in which the image was shot in camera.

But...I'm still unsure as to which is correct, and the manual really doesn't
help me with this.

Stan
Anonymous
June 3, 2005 10:30:59 PM

Archived from groups: rec.photo.digital (More info?)

G Winstanley wrote:
>
> I am under the impression that any image taken in RAW/NEF mode does *not*
> have an assigned colour space, but just raw values from the sensor. This
> would imply that in raw conversion you can select any colour space you want
> in order to do processing.


There might be a little exif tag saying what the camera preferences were
but ACR lets you convert the NEF in any color space you wish. You can
change the defaults in ACR & it will open all NEFs with that default.


>
> However, if the RAW/NEF file contains colour space information, then it
> *definitely* matters which colour space you choose in raw processing, and
> should choose the one in which the image was shot in camera.
>
> But...I'm still unsure as to which is correct, and the manual really doesn't
> help me with this.
>
> Stan

--
Paul Furman
http://www.edgehill.net/1
san francisco native plants
Anonymous
June 3, 2005 10:31:00 PM

Archived from groups: rec.photo.digital (More info?)

On Fri, 03 Jun 2005 00:17:46 GMT, the cup of "birdman"
<apquilts@pacbell.net> overfloweth with the following:

> Maintain your color space as Adobe RGB for your raw capture and when you
> open the image in Photoshop. Later in Photoshop you can convert color spaces
> to your heart's content.
> There are many tutorials on different ways to use the shadow/exposure
> adjustments or make adjustments later in Photshop.
> You can choose a white point in the RAW converter or balance color later in
> Photoshop.
> I think it is more useful to learn how to use the options for adjusting
> chromatic aberrations and vignetting.
> The most controversial issue about how to use the RAW converter is whether
> to work in 16bits or 8 bits. There is alot of misinformation out there about
> the benefits of 16bit color and how to use it and what happens when the 8
> bit printer driver gets hold of the data.
>

But does have the setting Adobe RGB actually get used for the RAW file?
Surely a RAW file contains raw sensor values, not values relating to a
colour space. By my thinking the colour space information is applied when
performing the conversion from RAW into Photoshop, so my best option would
be to convert into ProPhotoRGB to give me the widest gamut possible.

This is the crux of the misunderstanding for me. Are the raw values
contained in the NEF file colour space aware, or not? If they are I need to
ensure I choose the correct colour space option in ACR; if they are not I
can choose whatever I like. What you're saying implies they *are* colour
space aware...are you sure, or is this simply whta your current workflow
would imply?

Stan
Anonymous
June 4, 2005 4:13:08 AM

Archived from groups: rec.photo.digital (More info?)

On Fri, 03 Jun 2005 06:27:18 -0700, the cup of Paul Furman
<paul-@-edgehill.net> overfloweth with the following:

> G Winstanley wrote:
> >
> > I am under the impression that any image taken in RAW/NEF mode does *not*
> > have an assigned colour space, but just raw values from the sensor. This
> > would imply that in raw conversion you can select any colour space you want
> > in order to do processing.
>
>
> There might be a little exif tag saying what the camera preferences were
> but ACR lets you convert the NEF in any color space you wish. You can
> change the defaults in ACR & it will open all NEFs with that default.
>
>

This seems like the information I need, and is what I hoped. However, what
still slightly concerns me is this quote from the Photoshop CS help file,
about the colour space setting in ACR:
"Specifies the target color space profile. Generally, this should be
set to the same value as your Photoshop RGB working space. Keep in mind that
the source profile for camera raw image files is usually the camera-native
color space. (Higher-end cameras let you select a color space like Adobe RGB
for the color profile when photographing.) The profiles listed in the Space
menu are built into the Photoshop Camera Raw plug-in. If you want to use a
color space that's not listed in the Space menu, choose ProPhoto RGB, and
then convert to the working space of your choice when the file opens in
Photoshop."

This seems to imply the same as what you said, except the part in
parentheses which seems to also imply that some cameras have a colour space
assigned, which would then imply an explicit choice of that space! Each time
it seems straightforward at first glance, then something crops up to
slightly contradict my thinking. So what do you think about this quote from
the help file and how it has a bearing on things?

Stan
Anonymous
June 4, 2005 4:13:09 AM

Archived from groups: rec.photo.digital (More info?)

G Winstanley wrote:

> On Fri, 03 Jun 2005 06:27:18 -0700, the cup of Paul Furman
> <paul-@-edgehill.net> overfloweth with the following:
>
>
>>G Winstanley wrote:
>>
>>>I am under the impression that any image taken in RAW/NEF mode does *not*
>>>have an assigned colour space, but just raw values from the sensor. This
>>>would imply that in raw conversion you can select any colour space you want
>>>in order to do processing.
>>
>>
>>There might be a little exif tag saying what the camera preferences were
>>but ACR lets you convert the NEF in any color space you wish. You can
>>change the defaults in ACR & it will open all NEFs with that default.
>>
>>
>
>
> This seems like the information I need, and is what I hoped. However, what
> still slightly concerns me is this quote from the Photoshop CS help file,
> about the colour space setting in ACR:
> "Specifies the target color space profile. Generally, this should be
> set to the same value as your Photoshop RGB working space. Keep in mind that
> the source profile for camera raw image files is usually the camera-native
> color space. (Higher-end cameras let you select a color space like Adobe RGB
> for the color profile when photographing.) The profiles listed in the Space
> menu are built into the Photoshop Camera Raw plug-in. If you want to use a
> color space that's not listed in the Space menu, choose ProPhoto RGB, and
> then convert to the working space of your choice when the file opens in
> Photoshop."
>
> This seems to imply the same as what you said, except the part in
> parentheses which seems to also imply that some cameras have a colour space
> assigned, which would then imply an explicit choice of that space! Each time
> it seems straightforward at first glance, then something crops up to
> slightly contradict my thinking. So what do you think about this quote from
> the help file and how it has a bearing on things?


Well yes that does confuse things. I believe that the raw file is not
really in any color space at all though. I thought that doesn't happen
till you convert it. At least the raw data has a much wider gamut than
the converted image, that's the reason it's useful to make adjustments
in the raw converter before it's smushed down into a regular image format.

Ideally I guess you would use the ProPhoto 16 bit then do any
adjustments like curves then convert to adobeRGB or sRGB 8 bit but
that's just a lot more hassle.


--
Paul Furman
http://www.edgehill.net/1
san francisco native plants
Anonymous
June 4, 2005 11:16:10 AM

Archived from groups: rec.photo.digital (More info?)

yes you can show out of gamut colours.

it would seem to me that the camera has a gamut which will be the
triangle on the chromaticity graph from the 3 bayer filters. So whn you
convert from raw to photoshop you convert from that space to your
chosen space
Anonymous
June 4, 2005 1:52:17 PM

Archived from groups: rec.photo.digital (More info?)

On Fri, 03 Jun 2005 07:45:56 -0700, the cup of Paul Furman
<paul-@-edgehill.net> overfloweth with the following:

> G Winstanley wrote:
>
> > On Fri, 03 Jun 2005 06:27:18 -0700, the cup of Paul Furman
> > <paul-@-edgehill.net> overfloweth with the following:
> >
> >
> >>G Winstanley wrote:
> >>
> >>>I am under the impression that any image taken in RAW/NEF mode does *not*
> >>>have an assigned colour space, but just raw values from the sensor. This
> >>>would imply that in raw conversion you can select any colour space you want
> >>>in order to do processing.
> >>
> >>
> >>There might be a little exif tag saying what the camera preferences were
> >>but ACR lets you convert the NEF in any color space you wish. You can
> >>change the defaults in ACR & it will open all NEFs with that default.
> >>
> >>
> >
> >
> > This seems like the information I need, and is what I hoped. However, what
> > still slightly concerns me is this quote from the Photoshop CS help file,
> > about the colour space setting in ACR:
> > "Specifies the target color space profile. Generally, this should be
> > set to the same value as your Photoshop RGB working space. Keep in mind that
> > the source profile for camera raw image files is usually the camera-native
> > color space. (Higher-end cameras let you select a color space like Adobe RGB
> > for the color profile when photographing.) The profiles listed in the Space
> > menu are built into the Photoshop Camera Raw plug-in. If you want to use a
> > color space that's not listed in the Space menu, choose ProPhoto RGB, and
> > then convert to the working space of your choice when the file opens in
> > Photoshop."
> >
> > This seems to imply the same as what you said, except the part in
> > parentheses which seems to also imply that some cameras have a colour space
> > assigned, which would then imply an explicit choice of that space! Each time
> > it seems straightforward at first glance, then something crops up to
> > slightly contradict my thinking. So what do you think about this quote from
> > the help file and how it has a bearing on things?
>
>
> Well yes that does confuse things. I believe that the raw file is not
> really in any color space at all though. I thought that doesn't happen
> till you convert it. At least the raw data has a much wider gamut than
> the converted image, that's the reason it's useful to make adjustments
> in the raw converter before it's smushed down into a regular image format.
>
> Ideally I guess you would use the ProPhoto 16 bit then do any
> adjustments like curves then convert to adobeRGB or sRGB 8 bit but
> that's just a lot more hassle.

I'm glad I'm not the only that sees the apparent confusion here then. I
agree that this workflow via ProPhotoRGB seems the right way to go. I still
have a slight niggly feeling that somehow I maybe ought to be open in ACR
with Adobe RGB, then converting to ProPhotoRGB in PS. I'll try to get some
test done when I get a chance, but I really need some scenes which are
likely to stretch near the boundaries of the aRGB gamut.

Stan
Anonymous
June 4, 2005 1:52:18 PM

Archived from groups: rec.photo.digital (More info?)

G Winstanley wrote:

> On Fri, 03 Jun 2005 07:45:56 -0700, the cup of Paul Furman
> <paul-@-edgehill.net> overfloweth with the following:
>
>
>>G Winstanley wrote:
>>
>>
>>>On Fri, 03 Jun 2005 06:27:18 -0700, the cup of Paul Furman
>>><paul-@-edgehill.net> overfloweth with the following:
>>>
>>>
>>>
>>>>G Winstanley wrote:
>>>>
>>>>
>>>>>I am under the impression that any image taken in RAW/NEF mode does *not*
>>>>>have an assigned colour space, but just raw values from the sensor. This
>>>>>would imply that in raw conversion you can select any colour space you want
>>>>>in order to do processing.
>>>>
>>>>
>>>>There might be a little exif tag saying what the camera preferences were
>>>>but ACR lets you convert the NEF in any color space you wish. You can
>>>>change the defaults in ACR & it will open all NEFs with that default.
>>>>
>>>>
>>>
>>>
>>>This seems like the information I need, and is what I hoped. However, what
>>>still slightly concerns me is this quote from the Photoshop CS help file,
>>>about the colour space setting in ACR:
>>> "Specifies the target color space profile. Generally, this should be
>>>set to the same value as your Photoshop RGB working space. Keep in mind that
>>>the source profile for camera raw image files is usually the camera-native
>>>color space. (Higher-end cameras let you select a color space like Adobe RGB
>>>for the color profile when photographing.) The profiles listed in the Space
>>>menu are built into the Photoshop Camera Raw plug-in. If you want to use a
>>>color space that's not listed in the Space menu, choose ProPhoto RGB, and
>>>then convert to the working space of your choice when the file opens in
>>>Photoshop."
>>>
>>>This seems to imply the same as what you said, except the part in
>>>parentheses which seems to also imply that some cameras have a colour space
>>>assigned, which would then imply an explicit choice of that space! Each time
>>>it seems straightforward at first glance, then something crops up to
>>>slightly contradict my thinking. So what do you think about this quote from
>>>the help file and how it has a bearing on things?
>>
>>
>>Well yes that does confuse things. I believe that the raw file is not
>>really in any color space at all though. I thought that doesn't happen
>>till you convert it. At least the raw data has a much wider gamut than
>>the converted image, that's the reason it's useful to make adjustments
>>in the raw converter before it's smushed down into a regular image format.
>>
>>Ideally I guess you would use the ProPhoto 16 bit then do any
>>adjustments like curves then convert to adobeRGB or sRGB 8 bit but
>>that's just a lot more hassle.
>
>
> I'm glad I'm not the only that sees the apparent confusion here then. I
> agree that this workflow via ProPhotoRGB seems the right way to go. I still
> have a slight niggly feeling that somehow I maybe ought to be open in ACR
> with Adobe RGB, then converting to ProPhotoRGB in PS. I'll try to get some
> test done when I get a chance, but I really need some scenes which are
> likely to stretch near the boundaries of the aRGB gamut.


Check in alt.graphics.photoshop if you want better advice, that's a
really tricky question. My thought is PS CS isn't going to dump any info
if you convert direct into ProPhotoRGB. I don't think it would let you
make that mistake if it was a mistake.

It is not illogical for someone to convert raw to an ordinary RGB
colorspace then convert to a wider gamut so they could increase
saturation with color adjustments for something really wierd or
artificially tweak to match something important but that's post
processing induced changes, not caring about the source data. I'm in way
over my head though.

Um, if you want to test, there is a way to indicate out-of-gamut colors
in PS graphically. Check the help files for that.

--
Paul Furman
http://www.edgehill.net/1
san francisco native plants
Anonymous
June 4, 2005 6:08:22 PM

Archived from groups: rec.photo.digital (More info?)

In message <KkNne.1686$wy1.857@newssvr13.news.prodigy.com>,
"birdman" <apquilts@pacbell.net> wrote:

>There is alot of misinformation out there about
>the benefits of 16bit color and how to use it and what happens when the 8
>bit printer driver gets hold of the data.

Printers aren't 8 bit either; only the data fed to the drivers is.
Printers are often only 3 or 4 bits. It's dithered, and if you have
16-bit data dithered to 8-bit, fed to the driver, you have a deeper
effective color space than if you worked in 8-bit.

Anyway, the main reason for using 16-bit post-processing is to avoid
posterization, not to *SEE* 16-bit output. Most of the critics of
16-bit processing miss the point; it's not about what you see so much as
it is about what happens to the data.

How would you like a bank account that rounded your balance down to the
nearest dollar after applying interest every day?
--

<>>< ><<> ><<> <>>< ><<> <>>< <>>< ><<>
John P Sheehy <JPS@no.komm>
><<> <>>< <>>< ><<> <>>< ><<> ><<> <>><
Anonymous
June 8, 2005 10:44:34 AM

Archived from groups: rec.photo.digital (More info?)

if you lok at Nikon web site you see that NEF is RAW 12 bit data. The
conversion process can either 'bulk' this out to 16 bits, or shrink it
down to 8.

you should work in 16 bit because it avoids rounding errors which lead
to quantization / posterization

DB
Anonymous
June 8, 2005 6:29:21 PM

Archived from groups: rec.photo.digital (More info?)

with respect I disagree with you :) 

I cannot see that taking a 12 bit number and expressing it in 16 bits
is going to lead to data loss. The 16 bit realm will not be fully
populated, but I can't see their being data _loss_.

Your MP3 analogy goes the wrong way. Its like saying take a 12 bit
image, convert to 8 bit and then back out to 16. Of course then you
only have 8 bits of data in your 16 bit space, but even then your
calculations will be more accurate.

The gamut thing is not relevant. The no of bits is like using a
calculator that has more decimal places available: it doesn't change
how many numbers are avilable, whichis what gamut is.

Why not work in the biggest space to preserve accuracy? To follow your
argument, as most printers are only 4 bits, why not convert to 4 bit?
Anonymous
June 8, 2005 7:01:29 PM

Archived from groups: rec.photo.digital (More info?)

Consider: if you ever print your images your 16 bit colors may be out of 8
bit gamut and all colors will go through two additional (the first was
placing your 12 bit raw data into a 16bit space) data conversions that are
out of control: into 8 bit color space and then into the printer 8 bit
driver gamut. Unless you really look at what happens when you print you can
actually be causing more problems working in 16 bit than you are solving
simply because parts of a 16 bit histogram seem smoother than their 8 bit
counterpart. Although not a totally accurate analogy consider what happens
when you rip an MP3 track into CD format and then back into MP3 format.
Anonymous
June 9, 2005 7:18:06 AM

Archived from groups: rec.photo.digital (More info?)

I think Roger Clarke had a good paper on this ie effect of gamma on how
much information there is?

But I think you and I agree that converting raw to 16 bit to work on is
best (normally)?
Anonymous
June 9, 2005 8:41:36 AM

Archived from groups: rec.photo.digital (More info?)

In message <1118238274.052720.69200@z14g2000cwz.googlegroups.com>,
digiboy@mailinator.com wrote:

>if you lok at Nikon web site you see that NEF is RAW 12 bit data. The
>conversion process can either 'bulk' this out to 16 bits, or shrink it
>down to 8.

Mainly true, but there is so much change in gamma that the shadows don't
lose much going to 8 bit; only the midtones and highlights.
--

<>>< ><<> ><<> <>>< ><<> <>>< <>>< ><<>
John P Sheehy <JPS@no.komm>
><<> <>>< <>>< ><<> <>>< ><<> ><<> <>><
Anonymous
June 9, 2005 8:42:48 AM

Archived from groups: rec.photo.digital (More info?)

In message <dLDpe.1473$Z44.972@newssvr13.news.prodigy.com>,
"bmoag" <aetoo@hotmail.com> wrote:

>Consider: if you ever print your images your 16 bit colors may be out of 8
>bit gamut and all colors will go through two additional (the first was
>placing your 12 bit raw data into a 16bit space) data conversions that are
>out of control: into 8 bit color space and then into the printer 8 bit
>driver gamut. Unless you really look at what happens when you print you can
>actually be causing more problems working in 16 bit than you are solving
>simply because parts of a 16 bit histogram seem smoother than their 8 bit
>counterpart. Although not a totally accurate analogy consider what happens
>when you rip an MP3 track into CD format and then back into MP3 format.

8-bit and 16-bit use the same gamut.
--

<>>< ><<> ><<> <>>< ><<> <>>< <>>< ><<>
John P Sheehy <JPS@no.komm>
><<> <>>< <>>< ><<> <>>< ><<> ><<> <>><
Anonymous
June 9, 2005 8:47:28 AM

Archived from groups: rec.photo.digital (More info?)

In message <1118266161.618904.70140@g14g2000cwa.googlegroups.com>,
digiboy@mailinator.com wrote:

>I cannot see that taking a 12 bit number and expressing it in 16 bits
>is going to lead to data loss.

Well, it is not a padding with zeros, so there are some rounding off
errors, especially in the highlights, and also, not all 12 bits is used
in RAW conversion, by default. By default, only 11 bits of green, 10.5
bits of blue, and 10 bits of red are used in a typical conversion.
--

<>>< ><<> ><<> <>>< ><<> <>>< <>>< ><<>
John P Sheehy <JPS@no.komm>
><<> <>>< <>>< ><<> <>>< ><<> ><<> <>><
Anonymous
June 9, 2005 8:47:29 AM

Archived from groups: rec.photo.digital (More info?)

JPS@no.komm wrote:
> In message <1118266161.618904.70140@g14g2000cwa.googlegroups.com>,
> digiboy@mailinator.com wrote:
>
>
>>I cannot see that taking a 12 bit number and expressing it in 16 bits
>>is going to lead to data loss.


The reason for padding with zeroes is to be able to post-process more
smoothly, then it can be safely printed with rounding to match our eye's
perception. The tiny bit more captured might be useful if emphasized in
post-processing also. (as I understand)


>
>
> Well, it is not a padding with zeros, so there are some rounding off
> errors, especially in the highlights, and also, not all 12 bits is used
> in RAW conversion, by default. By default, only 11 bits of green, 10.5
> bits of blue, and 10 bits of red are used in a typical conversion.

--
Paul Furman
http://www.edgehill.net/1
san francisco native plants
Anonymous
June 10, 2005 6:51:13 AM

Archived from groups: rec.photo.digital (More info?)

In message <ma2dne1vDYzkfzrfRVn-gg@speakeasy.net>,
Paul Furman <paul-@-edgehill.net> wrote:

>The reason for padding with zeroes

There is no padding with zeros. Even if you take an 8-bit image, and
convert it to 16-bit, it is usually multiplied by 257, not 256.

12 bit RAW to 16-bit gamma-corrected generally is not a multiplication
at all. 65535 in the output is generally about 2200 in the green RAW
data, 1600 in the blue, and 1200 in the red, with about 125 in each RAW
channel becoming 0 in the output. The 16-bit numbers, when raised to
the 2.2 power, are proportional to the RAW data (after subtracting the
125). Of course, this applies most directly to grey renders, and some
colors will vary from this a bit for color profiles and spaces.
--

<>>< ><<> ><<> <>>< ><<> <>>< <>>< ><<>
John P Sheehy <JPS@no.komm>
><<> <>>< <>>< ><<> <>>< ><<> ><<> <>><
!