Alfred Molon
------------------------------
Olympus E-series DSLRs and micro 4/3 forum at
http://tech.groups.yahoo.com/group/MyOlympus/
http://myolympus.org/ photo sharing site
Depends. Full sun, open shade or under a tree?
> And when the sky is overcast,
> what would be the correct WB setting for RAW conversion - around 6000K,
> 6500K or what else?
Depends. Florida overcast or London overcast?
What does the WB picker say to your grey card in the test photo?
-Wolfgang
> Sunny, blue sky: is it 5000K or 5500K? And when the sky is overcast,
> what would be the correct WB setting for RAW conversion - around 6000K,
> 6500K or what else?
Here is a WB measurement shot taken mid-morning in vertually cloudless
conditions with a D300s (Auto WB on, all shots RAW): 5350K
< http://homepage.mac.com/lco/filechute/DNC_7836WBw.jpg >
This one I just took, It is overcast and raining: 5500K.
< http://homepage.mac.com/lco/filechute/DNC_7925WBw.jpg >
--
Regards,
Savageduck
Not shadow - full sun, in the middle of the da.
> > And when the sky is overcast,
> > what would be the correct WB setting for RAW conversion - around 6000K,
> > 6500K or what else?
>
> Depends. Florida overcast or London overcast?
Assume no significant pollution.
> What does the WB picker say to your grey card in the test photo?
That depends on the shot.
Not much difference in colour temperatures, despite the dramatically
different weather conditions.
I'm asking because I'm processing a series of shots taken in the
Malaysian rainforest (on a canopy walkway high above the trees looking
down on the treetops, blue sky at noon with very few clouds) and the
camera WB measurements are all in the 5700-6000K range.
I'm wondering what the "right" colour temperature for RAW conversion
should be - the camera measurements, which are quite high colour
temperatures indeed, or something lower in the 5000-5500K range.
Perhaps the camera measures such high colour temperatures because the
sun is so strong and there is so much blue light coming down from the
sky.
Well, the blue sky itself is VERY cold; up in the 7000K somewhere. So
open shade on a sunny day is really weird lighting. But sunlight is
around 5500K.
Any random number I or anybody else give you is just an approximation
that fits their tastes in the situation it's put forward for; none of
them are in fact "RIGHT". The atmosphere changes day to day,
differently in each location (well, mostly it's the stuff being carried
around IN the atmosphere; dust, pollen, water vapor, and so forth), the
sun angle changes how much atmosphere the light goes through, and so forth.
Might as well try the presets built into most raw conversion software
I've seen.
If you need precise color accuracy, shoot a neutral reference surface
(gray card, sheet of paper) exposed to not clip any color channels under
the same lighting, and then color balance off of that (either by doing a
custom white balance in-camera, or by setting white balance on that
other frame in raw conversion). And you'll still quite possibly have to
fine-tune colors (if for example you're doing catalog photos, and the
paints on various parts of the product have to match their official
numbers exactly).
>On 2011-03-24 13:04, Alfred Molon wrote:
>> Sunny, blue sky: is it 5000K or 5500K? And when the sky is overcast,
>> what would be the correct WB setting for RAW conversion - around 6000K,
>> 6500K or what else?
>
>Well, the blue sky itself is VERY cold; up in the 7000K somewhere. So
>open shade on a sunny day is really weird lighting. But sunlight is
>around 5500K.
>
>Any random number I or anybody else give you is just an approximation
>that fits their tastes in the situation it's put forward for; none of
>them are in fact "RIGHT". The atmosphere changes day to day,
>differently in each location (well, mostly it's the stuff being carried
>around IN the atmosphere; dust, pollen, water vapor, and so forth), the
>sun angle changes how much atmosphere the light goes through, and so forth.
>
As well as the colors of all objects in the immediate vicinity that are
coloring your subject with their own reflected hues. Including the color of
the ground you are on, or even any distant buildings or other large
structures reflecting their own colors.
For example: One of the easier ways to detect an "everyone believes it"
faked-photo is to see if the hatchet-job creator knows enough to include
the tints of any nearby objects on the new items they have introduced into
the image. Or conversely, removed the tints of any nearby objects they
might have removed as well. E.g. a person's white shirt and face has a
strong tinge of magenta on the right side, but there's nobody standing
there dressed in red and there's nothing else in the photo to suggest
another item further away that might be causing that.
These are some of the first things that an artist learns when creating
paintings and drawings.
>Might as well try the presets built into most raw conversion software
>I've seen.
>
>If you need precise color accuracy, shoot a neutral reference surface
>(gray card, sheet of paper) exposed to not clip any color channels under
>the same lighting, and then color balance off of that (either by doing a
>custom white balance in-camera, or by setting white balance on that
>other frame in raw conversion). And you'll still quite possibly have to
>fine-tune colors (if for example you're doing catalog photos, and the
>paints on various parts of the product have to match their official
>numbers exactly).
There's really no such thing as a proper white-balance in any photo. It's
all relative to what's in the photo itself and the lighting in which it is
taken. For example: You don't ever want to remove the overall green hues
that are washing over your subject taken under a dense forest canopy just
to get neutral grays and whites. This will lead to drastically maroon
browns and tans, and turn the foliage into an unnatural looking cyan and
blue.
A good photographer will strive to retain the ambiance within which the
photograph was taken; so as to impart that same realism, mood, and
sensation to the viewer.
I blame "auto white-balance" since the inception of digital photography for
much of the unnatural-looking pure-crap photography that has come into
existence.
The sun isn't what produces the blue light; the sun is actually pretty red.
The blue sky results from the sun's blue light being scattered by the
atmosphere while other colors are radiated away. So sunlit pictures are redder
(lower color temperature of the ambient light) than those taken under overcast
or in shade.
In the Canon world, the answer to your original question is 5200°K. I.e.,
that's the assumed color temperature of the ambient light if you select
"daylight" WB on a Canon camera. I think Nikon uses a slightly higher number.
Bob
This is all related to the reason why event photographers (particularly
wedding photographers) dress in black. They don't want to inadvertently
colorize the bride's dress.
: These are some of the first things that an artist learns when creating
: paintings and drawings.
Well, yes, but in photography the effect is very subtle, and the
photographer's reaction to it has to be subtle too. Painters can get away with
color effects that would look silly in a photograph.
: >Might as well try the presets built into most raw conversion software
: >I've seen.
: >
: >If you need precise color accuracy, shoot a neutral reference surface
: >(gray card, sheet of paper) exposed to not clip any color channels under
: >the same lighting, and then color balance off of that (either by doing a
: >custom white balance in-camera, or by setting white balance on that
: >other frame in raw conversion). And you'll still quite possibly have to
: >fine-tune colors (if for example you're doing catalog photos, and the
: >paints on various parts of the product have to match their official
: >numbers exactly).
:
: There's really no such thing as a proper white-balance in any photo. It's
: all relative to what's in the photo itself and the lighting in which it is
: taken. For example: You don't ever want to remove the overall green hues
: that are washing over your subject taken under a dense forest canopy just
: to get neutral grays and whites. This will lead to drastically maroon
: browns and tans, and turn the foliage into an unnatural looking cyan and
: blue.
True, and a lot of good pictures are ruined by photographers trying to make
all their pictures look as though they were taken under standard lighting
conditions.
: A good photographer will strive to retain the ambiance within which the
: photograph was taken; so as to impart that same realism, mood, and
: sensation to the viewer.
:
: I blame "auto white-balance" since the inception of digital photography for
: much of the unnatural-looking pure-crap photography that has come into
: existence.
You have to start somewhere, and auto WB at least provides a (usually)
plausible starting point. The problem arises when a photographer stops
recognizing that a starting point is all it is and starts trusting it even
when his common sense and artistic judgement say it's wrong.
Bob
>The sun isn't what produces the blue light; the sun is actually pretty red.
It's actually pretty green. Our sun is considered a green star. This is why
vision peaks in the green portion of the spectrum. This is why Bayer
sensors have 2 green photosites per group. This is why, in order to better
emulate the luminosity of colors when converting from color to B&W, a
channel-mixer tool is used in the proportions of R-28%, G-61%, B-11%.
Etc.
We only perceive it as a bright "white" because that's the light we evolved
under.
Just in case you or anyone else wasn't aware of the green-sun "why" of all
these things.
It really is Green Eggs & Ham, Sam I Am.
>On Thu, 24 Mar 2011 21:42:50 -0500, Better Info <bi...@address.info> wrote:
>:
>: There's really no such thing as a proper white-balance in any photo. It's
>: all relative to what's in the photo itself and the lighting in which it is
>: taken. For example: You don't ever want to remove the overall green hues
>: that are washing over your subject taken under a dense forest canopy just
>: to get neutral grays and whites. This will lead to drastically maroon
>: browns and tans, and turn the foliage into an unnatural looking cyan and
>: blue.
>
>True, and a lot of good pictures are ruined by photographers trying to make
>all their pictures look as though they were taken under standard lighting
>conditions.
>
After hitting send I realized I left out an all important word in the above
paragraph. It should have read, "You don't ever want to remove *all* the
overall green hues that are washing over ..."
One exception I'll make is when older fluorescent lights are the main
source. In *most* situations, that should be corrected for. Yet I've seen
(and taken) photos in factory and office situations where that lighting is
the only source. In a factory setting, leaving in some of that fluorescent
yellowish-green shift actually seems more appropriate. It immediately gives
the viewer the feeling of the "industrial" aspect of the setting. Anyone
who has ever been in a factory of that nature will immediately sense the
atmosphere in that work-environment in the photo. And all workers may not
know why or even notice it, but they will perceive that green tint in those
photos as an all-too-familiar territory. Twice or more per day for most of
their lives their senses have been bombarded with walking into and out of
that severely color-shifted lighting. They may not consciously be aware of
it, but their minds having learned to rapidly compensate each time have
most definitely learned that color-shift. Leaving some fluorescent-green in
a photo and their minds automatically tell them "Oh yeah, inside THE
FACTORY."
>: A good photographer will strive to retain the ambiance within which the
>: photograph was taken; so as to impart that same realism, mood, and
>: sensation to the viewer.
>:
>: I blame "auto white-balance" since the inception of digital photography for
>: much of the unnatural-looking pure-crap photography that has come into
>: existence.
>
>You have to start somewhere, and auto WB at least provides a (usually)
>plausible starting point. The problem arises when a photographer stops
>recognizing that a starting point is all it is and starts trusting it even
>when his common sense and artistic judgement say it's wrong.
>
>Bob
A starting point, yes. Though sadly, far too many think it's the end-point
and goal. The same is true for the right exposure setting. Not one camera
on earth will ever set the proper overall exposure for you automatically in
about 90% or more of all situations. This leads to people depending on, and
the popularity of, things like RAW to try to fix later what they failed to
set properly to begin with--they are instead relying on a typical
snapshooter's auto-everything.
But... why do different WB settings exist? Should we then use the same
colour temperature on images taken indoors with incandescent light and
outdoors under the sun?
Isn't it so that the human brain to a certain extent applies some white
balance transformation to what the eye is recording?
>In article <vetpo61h62c9n9edv...@4ax.com>, Robert Coe
>says...
>> True, and a lot of good pictures are ruined by photographers trying to make
>> all their pictures look as though they were taken under standard lighting
>> conditions.
>
>But... why do different WB settings exist? Should we then use the same
>colour temperature on images taken indoors with incandescent light and
>outdoors under the sun?
It depends on the camera, the lighting, what's in the scene, and what you
want to do with it.
I laugh at any reviewers who praise a camera that can make all
white-balance settings in their respective lighting situations 100%
neutral. Reviewers like that have never sold one image in their whole
lives. Their perception of, and experience with, photography is no greater
than the idiot tech-heads and hardware-geeks in these forums that rail on
at great lengths about miniscule amounts of softness caused by diffraction,
yet have never produced one image in their lives that anyone wants to look
at for more than a second.
I *want* my camera to leave in some warmness to all colors for the
incandescent setting. I *want* my camera to leave in the colder blues of
heavy-overcast lighting and north-light shadowed scenes if using the cloudy
setting. I even prefer that the daylight/sunlight white-balance setting
leave in some warmness under full sun, instead of it providing a base
neutral tone, as they all do.
When a reviewer eschews a camera that does this, retains some of the base
color of the respective white-balance setting, and they give it a greatly
lower score as a result, I only think, "Now that's a camera worth buying!
And that's also a total idiot who reviews cameras for a living. As are all
those who make their purchase decisions based on that total idiot's
advice."
When taking a romantic, somber, or melancholic candle-lit portrait or
character-study, do you *really* want it to be just as harshly neutral in
white-balance as any flash image? Think about it.
>Isn't it so that the human brain to a certain extent applies some white
>balance transformation to what the eye is recording?
Yes. "Some" being the key-word there. But few cameras (and few snapshooters
praising RAW) provide or produce that. Instead they strive to make a flash
image as neutral as an incandescent image, or an image under dense clouds
all the same neutral tones if using their respective white-balance
settings. Their concept and comprehension of photography and the art of
photography no greater than nor deeper than the image on a neutral-gray
test card.
I agree with a lot of that, but not with your deprecation of RAW mode. In
general, I think you want to correct for excessively warm or cold light about
as much as the eye would - sometimes a little less, for effect, but rarely
more. RAW gives you much greater freedom to achieve exactly what you want than
going with the manufacturer's standard settings. Even if you can set the WB in
degrees K, it's much easier to set it accurately after the fact than before.
Bob
>Even if you can set the WB in
>degrees K, it's much easier to set it accurately after the fact than before.
Ah, but then we are back at square-one with "how good is YOUR color
memory"?
If you had read that thread "Tweaking Monitor Calibration", most are
convinced (and rightfully so) that they can't even remember colors from one
minute to the next, let alone a subtle color hue-shift from a week ago
until they finally tweak their RAW file.
Then add-in all the problems with people that can't even visually discern
if their monitors are calibrated properly or not, and ....
I guess I'd rather have an immediate and accurate JPG representation of the
scene at the time I have shot it, than a RAW file later where I have no
recollection of what kind of lighting it was taken under. Both, if it's a
very important image. Relying on the JPG first (if the camera is properly
set to begin with) before tweaking the RAW.
Get it right the first time and there's no need for RAW. No monitor nor
paper has as much bit-depth as the JPG, so RAW isn't needed if done right
from the onset. (Do note, that there is a 16-bit-depth JPG format that's
been available for many years now. It's even included in my favorite editor
and I was looking forward to the day it would be widely accepted and
implemented. Do you see much support for that? In *any* camera for that
matter? 16-bit color-depth is *so* highly overrated.)
It's horses for courses, but I'd rather find the best way to have a
recording of the colors taken at the time of the shot than try to recreate
what I may or may not have remembered from one to three weeks ago. Or
worse, when going back through your archives from five years ago and trying
to remember what that scene really looked like.
In truth, it's all relative, and may the best relative win.
You can also shoot RAW+JPEG and then later decide if you want to use the
JPEG or process the RAW.
>In article <2grro6luad3qdlvfp...@4ax.com>, Better Info
>says...
>> Get it right the first time and there's no need for RAW.
>
>You can also shoot RAW+JPEG and then later decide if you want to use the
>JPEG or process the RAW.
You left out the all-important-step for using RAW properly, using the JPG
as reference: IF the camera's white-balance is properly set to begin with.
>On Sat, 26 Mar 2011 05:49:44 -0500, Better Info <bi...@address.info> wrote:
>: On Sat, 26 Mar 2011 10:26:21 +0100, Alfred Molon <alfred...@yahoo.com>
>: wrote:
>:
>: >In article <vetpo61h62c9n9edv...@4ax.com>, Robert Coe
>: >says...
>: >> True, and a lot of good pictures are ruined by photographers trying to make
>: >> all their pictures look as though they were taken under standard lighting
>: >> conditions.
>: >
>: >But... why do different WB settings exist? Should we then use the same
>: >colour temperature on images taken indoors with incandescent light and
>: >outdoors under the sun?
>:
>: It depends on the camera, the lighting, what's in the scene, and what you
>: want to do with it.
--- snipped for brevity ---
>: >Isn't it so that the human brain to a certain extent applies some white
>: >balance transformation to what the eye is recording?
>:
>: Yes. "Some" being the key-word there. But few cameras (and few snapshooters
>: praising RAW) provide or produce that. Instead they strive to make a flash
>: image as neutral as an incandescent image, or an image under dense clouds
>: all the same neutral tones if using their respective white-balance
>: settings. Their concept and comprehension of photography and the art of
>: photography no greater than nor deeper than the image on a neutral-gray
>: test card.
>
>I agree with a lot of that, but not with your deprecation of RAW mode. In
>general, I think you want to correct for excessively warm or cold light about
>as much as the eye would - sometimes a little less, for effect, but rarely
>more. RAW gives you much greater freedom to achieve exactly what you want than
>going with the manufacturer's standard settings. Even if you can set the WB in
>degrees K, it's much easier to set it accurately after the fact than before.
>
Something that seems to have been overlooked in all this discussion is
that 'Color Temperature' as it is conventionally used means the
spectrum of a theoretical black body at the particular temperature.
See http://www.ips-innovations.com/bilder/blackbody_emissive.gif for a
diagram.
Wikipedia says http://en.wikipedia.org/wiki/Color_temperature
"The color temperature of a light source is the temperature of an
ideal black-body radiator that radiates light of comparable hue to
that of the light source. Color temperature is conventionally
stated in the unit of absolute temperature, the kelvin, having the
unit symbol K."
Many of the light sources under discussion in this thread don't have a
typical black body spectrum. Wikipedia makes this clear when further
down it says:
"Many other light sources, such as fluorescent lamps, emit light
primarily by processes other than thermal radiation. This means
the emitted radiation does not follow the form of a black-body
spectrum. These sources are assigned what is known as a correlated
color temperature (CCT). CCT is the color temperature of a black
body radiator which to human color perception most closely matches
the light from the lamp."
Typical spectra are seen in
http://www.bealecorner.org/best/measure/cf-spectrum/index.html
which shows why with many artificial light sources it is not possible
to obtain a daylight (or similar) color balance by mere adjustment of
color temperature. If there is a hole in the spectrum of the light
source used to create an image, the hole will remain no matter how you
adjust things up or down with color temperature.
Regards,
Eric Stevens
>On Sat, 26 Mar 2011 15:49:58 +0100, Alfred Molon <alfred...@yahoo.com>
Whether or not the white balance is properly set can only be a value
judgement.
Regards,
Eric Stevens
>On Sat, 26 Mar 2011 15:49:58 +0100, Alfred Molon <alfred...@yahoo.com>
Whether or not the white balance is properly set is a value judgement
- a matter of personal preference.
For a perhaps extreme example see http://www.intermonet.com/colors/
which touches on the problems of the painter Claude Monet as his
cataracts slowly developed.
Regards,
Eric Stevens
If you live nearby, go the the Johnson Sculpture Grounds. While much of
the work is modernistic, you get a chance to do your own interpretations
of Monet's bridge, the lilys (in season of course. There are also life
size statutes reproducing the essence of paintings, such as Renoir's
Luncheon at the Boating Party.
<http://philadelphia.about.com/od/photo_galleries/ig/grounds_for_sculpture_photos/gfs_047.htm>
It's an opportunity to practice your post processing skills to present
these famous works in any manner you choose. the subjects never complain
about your futzing with equipment and photography is encouraged.
--
Peter
> If there is a hole in the spectrum of the light
>source used to create an image, the hole will remain no matter how you
>adjust things up or down with color temperature.
I guess you've never held a piece of diffraction-grating up to your eye and
looked at your monitor displaying a white line on a black-background.
Depending on monitor, it's often more holes than colors. According to the
way you think, no monitor can be used for editing photos nor should even be
used for viewing them.
Isn't it fun to try to sound smart when you're not? All you have to do is
copy and paste things from the internet even though you have no idea what
it all means.
>On 3/26/2011 5:28 PM, Eric Stevens wrote:
>> On Sat, 26 Mar 2011 09:55:59 -0500, Better Info<bi...@address.info>
>> wrote:
>>
>>> On Sat, 26 Mar 2011 15:49:58 +0100, Alfred Molon<alfred...@yahoo.com>
>>> wrote:
>>>
>>>> In article<2grro6luad3qdlvfp...@4ax.com>, Better Info
>>>> says...
>>>>> Get it right the first time and there's no need for RAW.
>>>>
>>>> You can also shoot RAW+JPEG and then later decide if you want to use the
>>>> JPEG or process the RAW.
>>>
>>> You left out the all-important-step for using RAW properly, using the JPG
>>> as reference: IF the camera's white-balance is properly set to begin with.
>>
>> Whether or not the white balance is properly set is a value judgement
>> - a matter of personal preference.
>>
>> For a perhaps extreme example see http://www.intermonet.com/colors/
>> which touches on the problems of the painter Claude Monet as his
>> cataracts slowly developed.
>>
>
>If you live nearby, go the the Johnson Sculpture Grounds. While much of
>the work is modernistic, you get a chance to do your own interpretations
>of Monet's bridge, the lilys (in season of course. There are also life
>size statutes reproducing the essence of paintings, such as Renoir's
>Luncheon at the Boating Party.
>
>
><http://philadelphia.about.com/od/photo_galleries/ig/grounds_for_sculpture_photos/gfs_047.htm>
>
>It's an opportunity to practice your post processing skills to present
>these famous works in any manner you choose. the subjects never complain
>about your futzing with equipment and photography is encouraged.
I will probably never see theactual bridge but I have seen, side by
side, two before and after paintings by Monet of the bridge.
Regards,
Eric Stevens
>On Sun, 27 Mar 2011 10:13:48 +1300, Eric Stevens <eric.s...@sum.co.nz>
That's a non sequitur followed by an unnecessary insult.
For a minute I thought we were about to have an intelligent and polite
conversation. But then, as usual, you shied away, shouting insults. I
feel sorry for you.
Regards,
Eric Stevens
The crucial word in that definition being "ideal". An ideal black-body
radiator is nothing but a hypothetical construct that no actual object will
match.
: Many of the light sources under discussion in this thread don't have a
: typical black body spectrum.
Even when one almost does, it's hard to make out its actual color temperature.
Incandescent bulbs, for example, radiate in a variety of colors, the dimmer
ones usually being much redder than the brighter ones.
: Wikipedia makes this clear when further down it says:
:
: "Many other light sources, such as fluorescent lamps, emit light
: primarily by processes other than thermal radiation. This means
: the emitted radiation does not follow the form of a black-body
: spectrum. These sources are assigned what is known as a correlated
: color temperature (CCT). CCT is the color temperature of a black
: body radiator which to human color perception most closely matches
: the light from the lamp."
Fluorescent lamps are tricky enough, and some oddball sources can be nearly
monochromatic. Assigning a color temperature to a sodium-vapor street light is
futile.
: Typical spectra are seen in
: http://www.bealecorner.org/best/measure/cf-spectrum/index.html
: which shows why with many artificial light sources it is not possible
: to obtain a daylight (or similar) color balance by mere adjustment of
: color temperature. If there is a hole in the spectrum of the light
: source used to create an image, the hole will remain no matter how you
: adjust things up or down with color temperature.
In the U.S., the Energy Police are planning to ban incandescent bulbs in favor
of the screw-in fluorescent kind that fit the same sockets (but on which you
can't perch a lampshade). They claim those bulbs have much better color
characteristics than older fluorescents, but I'll believe it when I see it.
And as Supy Win pointed out earlier in the thread, something would be lost if
all indoor light sources were made to look like daylight.
Bob
Not a non-sequitur. But you wouldn't realize that unless you grasped what
you were trying to introduce into the discussion. So the possibility of you
carrying on an intelligent conversation is already undermined. And I don't
have "polite conversations" with deceptive lying commercial spammers that
send people to their fake newsgroup charters just to get them to visit
commercial sites.
As for how you feel about it, add this into your equation: I don't give a
shit how you feel. I'm not here to provide relief for some desperate
emotional need you might be having.
>
>Fluorescent lamps are tricky enough, and some oddball sources can be nearly
>monochromatic. Assigning a color temperature to a sodium-vapor street light is
>futile.
There was an interesting experiment that I saw done on one of those Nature
or Nova shows on PBS. A special on light and color, done many years ago.
They took a scene, a simple still-life of some colorful fruit arranged on a
table. Photographed it, then separated it into its RGB layers. Then
re-projected that scene through their respective filters from 3 different
projectors. They turned one off, effectively removing that whole section of
the color spectrum from the scene. They eye/brain connection was still able
to recreate the missing color. Now retained in a person's memory of what
colors *should* be there, so mind adds them in.
I only mention this because of you bringing up the thin bright line of
sodium lights. Only one line of the spectrum does make it impossible to
recreate the colors of the scene (and therefore any semblance of
temperature). But not if you have two. It may not be perfect, but the
eye-brain system can still recreate it and even assign a color temperature
to it by varying the intensity of those two colors. With 3 well spaced
color-frequency references, then the eye-brain can provide a nearly
complete representation of the original scene.
If you ever checked the broken spectrum of a CRT monitor's phosphors or LCD
filters, you'd be amazed that they even work as well as they do for the
task required of them. I first noticed this as a child when viewing a
color-TV's display through a diffraction grating. It was just 3 thin lines
of colors. No yellows, no magentas, no cyans, etc. Things haven't changed a
whole lot.
>
>Fluorescent lamps are tricky enough, and some oddball sources can be nearly
>monochromatic. Assigning a color temperature to a sodium-vapor street light is
>futile.
There was an interesting experiment that I saw done on one of those Nature
or Nova shows on PBS. A special on light and color, done many years ago.
They took a scene, a simple still-life of some colorful fruit arranged on a
table. Photographed it, then separated it into its RGB layers. Then
re-projected that scene through their respective filters from 3 different
projectors. They turned one off, effectively removing that whole section of
the color spectrum from the scene. They eye/brain connection was still able
to recreate the missing color. Now retained in a person's memory of what
colors *should* be there, so the mind adds them in.
But how about your own desperate emotional needs?
Regards,
Eric Stevens
>On Sat, 26 Mar 2011 19:19:11 -0400, Robert Coe <b...@1776.COM> wrote:
The conversion isn't all done in the brain. Part of it is physical in
that the the closely packed cone cells detect the light from adjacent
RBG pixels and the overlap causes them to interpret it as an
intermediate color.
Regards,
Eric Stevens
Now that IS a non-sequitur, TROLL. The only emotional response I get from
reading this group is peals of laughter. Like watching any good comedy show
on TV that's full of buffoons. There's more clowns in this one little
newsgroup than can be fit in any Volkswagen Beetle.
It's called the "Land effect" after its discoverer, Edwin Land, the inventor
of the Polaroid camera. The demonstration doesn't even require color film. It
goes something like this. (I hope I get this right. My brother and I
demonstrated it once, but it was more than fifty years ago, when I was in
college and he was in high school.)
You take two pictures of the subject (something colorful, like the bowl of
fruit) on B&W film. The two pictures are identical, except that one of them is
taken through a dark filter. (We used red, although green might also work.)
You then process the pictures into B&W slides. I forget how we did that, but
you can either expose onto another piece of film in a printing frame or do a
reversal exposure like a color slide. We may even have had a processing lab do
it. Then you put each picture in a projector and align them on the screen.
They're in B&W, of course, but when you put the red filter on the projector
containing the picture taken through the red filter, all the colors show up.
If instead you put the red filter on the other picture, the colors are
reversed (red becomes green, blue becomes yellow, or whatever). We thought it
was pretty spectacular.
One problem that comes up is that after the filter is applied, the brightness
of the two projectors is badly out of balance, which detracts from the effect.
But fortunately we had a modern, bright projector and an older, much dimmer
one belonging to our grandfather. We used the dimmer projector for the copy
without the filter, and it worked like a charm.
: I only mention this because of you bringing up the thin bright line of
: sodium lights. Only one line of the spectrum does make it impossible to
: recreate the colors of the scene (and therefore any semblance of
: temperature). But not if you have two. It may not be perfect, but the
: eye-brain system can still recreate it and even assign a color temperature
: to it by varying the intensity of those two colors. With 3 well spaced
: color-frequency references, then the eye-brain can provide a nearly
: complete representation of the original scene.
I think a sodium lamp does have more than one line. Potentially you have one
line for each possible energy state transition of each electron, and sodium
has several electrons. But some of the transitions are much more probable, so
their lines are much stronger. And because some of the electrons occupy
identical ground states, not all of the transitions produce a different line.
Bob
Funny, we were talking about a trip which would include a visit to the
original, but it probably will not happen. A trip to see our Western
National Parks is more probable. We have been to SF several times and it
is one of our favorite cities. Point Reyes WOULD NOT BE HARD TO TAKE. i
would like to go to Yellowstone and the Jackson Hole area, though my
wife has a strong preference for Yosemite. Knowing us, we may very well
end up somewhere else, that is completely different.
--
Peter
Don't feel insulted. He was probably talking about his own activities
and preferences.
--
Peter
>
>It's called the "Land effect" after its discoverer, Edwin Land, the inventor
>of the Polaroid camera.
Yes, that was one of the demonstrations they had shown in that show.
There's some good websites and papers online demonstrating this,
retrievable now that you remind me, and all, the name of it.
googling for: color light "land effect"
Here's one explaining the "Subtractive Land Effect", which is nearer to
what I remember being demoed in that show--using a two-color separation.
<http://www.perceptionweb.com/abstract.cgi?id=v090360>
Aside: This probably inspired the widespread use of duo-tone printing
processes/chemistries back then, which then evolved into the garish
posterizations that were popular in the photo/darkroom communities during
the 60's and 70's.
Europe has already gone down this route and our indoor lives are a lot
less bright as a result. With the benefit of hindsight, my advice
would be to buy and store a lifetime's supply of incandescent bulbs,
or convert to halogen rather than compact fluorescent.
[As an aside, one of the key assumptions that underpins the change to
compact fluorescent bulbs is that their service life is much longer
than that of incandescent bulbs because the bulbs themselves are much
more energy intensive to make. Having converted most of the house to
compact fluorescent, I am now replacing failed bulbs at a faster rate
than I ever did with incandescent bulbs. Top brand bulbs appear to be
no better in terms of service life than the cheapest, and no brand
achieves the claimed 5X equivalent brightness.]
> Not shadow - full sun, in the middle of the da.
Good for portraits!
On the green grass, on the blue sea or on grey asphalt?
>> > And when the sky is overcast,
>> > what would be the correct WB setting for RAW conversion - around 6000K,
>> > 6500K or what else?
>>
>> Depends. Florida overcast or London overcast?
> Assume no significant pollution.
Well, there *is* light overcast and dreary overcast ...
>> What does the WB picker say to your grey card in the test photo?
> That depends on the shot.
See? There is no general rule that's exact to a few 100K, you'd
need to use a grey card on location to determine the exact value.
-Wolfgang
Yesterday, for "downright miserable overcast & soaking wet stuff" I
ended up with this:
< http://homepage.mac.com/lco/filechute/screenshot_04B.jpg >
5450K & +2 tint
Allowing me to get a silly shot like this:
< http://homepage.mac.com/lco/filechute/DNC_7930w.jpg >
--
Regards,
Savageduck
> I agree with a lot of that, but not with your deprecation of RAW mode.
You've got to understand Supy: most superzoom cameras can't
produce RAW. A few are able to produce RAW out of the box, but
are mostly very slow then, and a couple are only able to produce
RAW with an inofficial camera hack.
No wonder that a superzoom camera promoter will not like RAW ---
since that is an area where DSLRs are much better equipped and
often much better supported by RAW converters, especially 3rd
party ones.
-Wolfgang
Wrong.
Like every one of your posts. Wrong.
Good point; I was so focused on "actual" sources that I didn't even
think about apparent sources. But they affect the color also, as you
say.
[snip]
> >If you need precise color accuracy, shoot a neutral reference surface
> >(gray card, sheet of paper) exposed to not clip any color channels under
> >the same lighting, and then color balance off of that (either by doing a
> >custom white balance in-camera, or by setting white balance on that
> >other frame in raw conversion). And you'll still quite possibly have to
> >fine-tune colors (if for example you're doing catalog photos, and the
> >paints on various parts of the product have to match their official
> >numbers exactly).
>
> There's really no such thing as a proper white-balance in any photo. It's
> all relative to what's in the photo itself and the lighting in which it is
> taken. For example: You don't ever want to remove the overall green hues
> that are washing over your subject taken under a dense forest canopy just
> to get neutral grays and whites. This will lead to drastically maroon
> browns and tans, and turn the foliage into an unnatural looking cyan and
> blue.
Yep, it's in fact an aesthetic choice, not a question of technical
correctness. (Same thing with exposure, too. People keep asking what
the "right" exposure for something is, and there ain't none; there are
exposures that allow you to produce a final image that meets your
goals, and exposures that don't.)
> A good photographer will strive to retain the ambiance within which the
> photograph was taken; so as to impart that same realism, mood, and
> sensation to the viewer.
>
> I blame "auto white-balance" since the inception of digital photography for
> much of the unnatural-looking pure-crap photography that has come into
> existence.
Possibly; but did machine prints from ProEx do any better?
> In the U.S., the Energy Police are planning to ban incandescent
> bulbs in favor of the screw-in fluorescent kind that fit the same
> sockets (but on which you can't perch a lampshade). They claim those
> bulbs have much better color characteristics than older
> fluorescents, but I'll believe it when I see it. And as Supy Win
> pointed out earlier in the thread, something would be lost if all
> indoor light sources were made to look like daylight.
Where do you live that you haven't seen and photographed the light
from CFs? They're ubiquitous here in Minnesota. I converted most of
the lights in my previous house to CF more than 15 years ago.
And they are in fact a lot better for photography than old-style
fluorescents. That's why you're seeing "lighting equipment" consisting
of arrays of sockets filled with CF bulbs sold as photographic
equipment.
They are also better for indoor videography due to their high-frequency AC
step-up transformers in them. Old ballasted fluorescents pulse with a 60Hz
frequency. This would often "beat" (align partially out of sync) with the
frame-rates of video-cameras (dedicated or otherwise) causing gradual
darkening then lightening in a scene in a rhythmic manner. This could not
only alter the luminosity of the scene in lengthy waves, but the
white-balance drastically as well. As anyone who has ever watched a
spinning fan under ballasted fluorescents. On each turn of polarity, the
light will pulse with a strong orange-yellow light, then on the return a
bluish light. The camera trying to sync a white-balance on any part of
these pulses or constantly changing intensity and color of light doesn't
know what to make of the next one that's also differently bright or
differently tinted.This plays havoc with most systems that are trying to
monitor the source-light color and brightness and adjust a white-balance
and exposure for it. Especially in the digital age, where instead of an
analog monitoring system that's always taking an average, only small
snippets of shared CPU time in digital cameras will be dedicated to this
task during the frames.
It is interesting to view this in any camera that has a shutter-speed
preview option (most all compact and superzoom cameras have this feature by
default). Point the camera at a ballasted fluorescent, then half-depress
the shutter-release (if you need to do that) to engage the shutter-speed
preview. Then slowly ramp-up the shutter-speed. At certain speeds you'll
see the fluorescent light turn totally dark in the EVF or LCD viewfinder
the closer they sync in phase, as well as change colors as it slowly
brightens or darkens again. Using your camera like a fast analytic strobe.
In CHDK capable cameras, if you have the RGB histogram enabled, you can
even watch the blue and red peaks slowly swap positions in the histogram as
the color-balance of the light changes during its cycles.
You can't do this with CFLs because the AC frequency is much too high to
beat against (partially sync with) the top shutter-speeds in most cameras.
On Mon, 28 Mar 2011 11:19:07 -0700 (PDT), David Dyer-Bennet
<illeg...@gmail.com> wrote:
They are also better for indoor videography due to their high-frequency AC
step-up transformers in them. Old ballasted fluorescents pulse with a
50/60Hz frequency. This would often "beat" (align partially out of sync)
with the frame-rates of video-cameras (dedicated or otherwise) causing
gradual darkening then lightening in a scene in a rhythmic manner. This
could not only alter the luminosity of the scene in lengthy waves, but the
white-balance drastically as well.
As anyone who has ever watched a spinning fan under ballasted fluorescents
might have noticed. On each turn of polarity, the light will pulse with a
strong orange-yellow light, then on the return a bluish light. The camera
trying to sync a white-balance on any part of these pulses or constantly
changing intensity and color of light doesn't know what to make of the next
one that's also differently bright or differently tinted.This plays havoc
with most systems that are trying to monitor the source-light color and
brightness and adjust a white-balance and exposure for it. Especially in
the digital age, where instead of an analog monitoring system that's always
taking an average, only small snippets of shared CPU time in digital
cameras will be dedicated to this task during the frames.
It is interesting to view this in any camera that has a shutter-speed
preview option (most all compact and superzoom cameras have this feature by
default). Point the camera at a ballasted fluorescent, then half-depress
the shutter-release (if you need to do that) to engage the shutter-speed
preview. Then slowly ramp-up the shutter-speed. At certain speeds you'll
see the fluorescent light turn totally dark in the EVF or LCD viewfinder
the closer they sync in phase, as well as change colors as it slowly
brightens or darkens again. Using your camera like a fast analytic strobe.
In CHDK capable cameras, if you have the RGB (Blend Option) histogram
<edited for brevity>
> : There was an interesting experiment that I saw done on one of those Nature
> : or Nova shows on PBS. A special on light and color, done many years ago.
> :
> : They took a scene, a simple still-life of some colorful fruit arranged on a
> : table. Photographed it, then separated it into its RGB layers. Then
> : re-projected that scene through their respective filters from 3 different
> : projectors. They turned one off, effectively removing that whole section of
> : the color spectrum from the scene. They eye/brain connection was still able
> : to recreate the missing color. Now retained in a person's memory of what
> : colors *should* be there, so mind adds them in.
>
> It's called the "Land effect" after its discoverer, Edwin Land, the inventor
> of the Polaroid camera.
<edited>
Correction: Edwin Land invented instant film.
--
Cordially,
John Turco <jt...@concentric.net>
Marie's Musings <http://fairiesandtails.blogspot.com>
That's very odd, "Bruce"...here in the U.S.A., my experiences with
CFL "bulbs" have been completely different from yours. I've found
them >far< more reliable than incandescents.
An assortment of "Great Value" (Wal-Mart brand) CFL products and a
General Electric 3-way puppy (50W-100W-150W equivalent), have all
proven to be quite good values (and every one was made in China).
Thus, quit diving in dumpsters, if you want better results from
your merchandise.
Of my three RAW-capable cameras, only one is a DSLR (Pentax
"K100D"). The others are "super zoom" models, by Kodak ("P850"
& "Z980").
> No wonder that a superzoom camera promoter will not like RAW ---
> since that is an area where DSLRs are much better equipped and
> often much better supported by RAW converters, especially 3rd
> party ones.
>
> -Wolfgang
I don't use RAW, because my current computer is too old and
slow, to adequately run the required software.
Is there a difference?
Regards,
Eric Stevens
Actually his first invention was a process for manufacturing
inexpensive polarizing filters in 1929 while he was still a university
student. Hence the Polaroid Corporation.
The Land camera and film were pretty much developed together in 1947.
--
Regards,
Savageduck
Polaroid's marketing materials almost always referred to it as a
"Polaroid Land Camera".
Actually, "John", our experience here in the USA has matched Tony's a bit
better than it's matched yours. Martha has been griping recently about the
high price and short life of the compact fluorescent bulb she bought for our
back hall last year. As I recall, it was indeed a "name" brand (Philips or GE
or whatever). Which is one of the reason I wrote what I did.
Bob
Turco isn't speaking from experience, he's just posting a contrary
view because he wants to. Like most of the nonsense he posts.
>Martha has been griping recently about the
>high price and short life of the compact fluorescent bulb she bought for our
>back hall last year. As I recall, it was indeed a "name" brand (Philips or GE
>or whatever). Which is one of the reason I wrote what I did.
Philips and GE, despite being top brands (and selling at higher
prices) offer CFLs that perform no better than cheap no-name bulbs.
I chose Philips because they have products that look more like a
traditional bulb and will support a shade. But they fail with
monotonous regularity. In the UK, GE are the cheapest of the well
known brands. They are no worse than Philips but no better than the
cheapest CFLs either.
Given that the energy used to make CFLs is very high compared to
incandescent bulbs, long life is critical to ensuring that their whole
life 'carbon footprint' is kept to a minimum. If, as it would appear,
they don't achieve their claimed long life, then the assumptions made
about saving energy (and carbon) may not be valid.
Interesting article, Supy. Thanks.
Have you (or David, for that matrter) used them enough to have an opinion
about their reliability and lifespan?
Bob
I've used them (the small fluorescent lamps) a lot.
I use a lot of the 4000K ones. They are hard to find. Most
are "2650K" simulator ones.
They are very reliable and their lifespan is what they claim.
The DO sometimes stop working, especially Philips ones (not surprising, of
course, given the crap that Philips usually is), but that is rare.
They just get dimmer and dimmer and eventually you notice and
try a new one.
Even new, the 100 watt ones put out less useful light than
a regular 100 watt one, but not by much. An old one is much less.
Hint: everybody has a trivial tool to tell whether the spectrum is
near-continuous. All you need is a CD or DVD. Have just one on
in a room, walk to the other end of the room, and look at the spectrum
diffracted by the CD or DVD. The smoother the better. Theoretically
you should put a slit in front of the lamp, but at 20 feet
its not really needed.
Doug McDonald
> Have you (or David, for that matrter) used them enough to have an opinion
> about their reliability and lifespan?
I don't label them when I install them, or keep a log. And they last long enough that I generally can't remember when that particular bulb was put in.
I *think* we're just starting to replace bulbs we put in when we moved into the current house at the start of 1996, in general.
<edited for brevity>
Why, yes...there >is<. The "Land" camera, itself, has never been anything
special.
Polaroid's amazing, self-developing film was always the real revelation.
Or, have you assumed the Land camera had a tiny darkroom inside it, replete
with miniature technicians, chemicals and equipment? (Admittedly, I'd once
considered something to that effect, as a child.)
<edited for brevity>
> : > [As an aside, one of the key assumptions that underpins the change to
> : > compact fluorescent bulbs is that their service life is much longer
> : > than that of incandescent bulbs because the bulbs themselves are much
> : > more energy intensive to make. Having converted most of the house to
> : > compact fluorescent, I am now replacing failed bulbs at a faster rate
> : > than I ever did with incandescent bulbs. Top brand bulbs appear to
> : > be no better in terms of service life than the cheapest, and no brand
> : > achieves the claimed 5X equivalent brightness.]
> :
> :
> : That's very odd, "Bruce"...here in the U.S.A., my experiences with
> : CFL "bulbs" have been completely different from yours. I've found
> : them >far< more reliable than incandescents.
> :
> : An assortment of "Great Value" (Wal-Mart brand) CFL products and a
> : General Electric 3-way puppy (50W-100W-150W equivalent), have all
> : proven to be quite good values (and every one was made in China).
> :
> : Thus, quit diving in dumpsters, if you want better results from
> : your merchandise.
>
> Actually, "John", our experience here in the USA has matched Tony's
> a bit better than it's matched yours. Martha has been griping recently
> about the high price and short life of the compact fluorescent bulb
> she bought for our back hall last year. As I recall, it was indeed a
> "name" brand (Philips or GE or whatever). Which is one of the reason
> I wrote what I did.
>
> Bob
Actually, "Bob" -- "Bruce" (a.k.a., "Tony Polson") is an obvious
troll.
Why do you trust this twerp, anyway?
<edited for brevity>
> >: > Having converted most of the house to compact fluorescent, I am
> >: > now replacing failed bulbs at a faster rate than I ever did with
> >: > incandescent bulbs. Top brand bulbs appear to be no better in
> >: > terms of service life than the cheapest, and no brand achieves
> >: > the claimed 5X equivalent brightness.]
> >:
> >:
> >: That's very odd, "Bruce"...here in the U.S.A., my experiences with
> >: CFL "bulbs" have been completely different from yours. I've found
> >: them >far< more reliable than incandescents.
> >:
> >: An assortment of "Great Value" (Wal-Mart brand) CFL products and a
> >: General Electric 3-way puppy (50W-100W-150W equivalent), have all
> >: proven to be quite good values (and every one was made in China).
> >:
> >: Thus, quit diving in dumpsters, if you want better results from
> >: your merchandise.
> >
> > Actually, "John", our experience here in the USA has matched Tony's
> > a bit better than it's matched yours.
>
> Turco isn't speaking from experience, he's just posting a contrary
> view because he wants to. Like most of the nonsense he posts.
<edited for brevity>
Look who's typin'!
Incidentally, "Bruce" -- Robert Coe referred to you, above, as
"Tony" (which probably means, he knows that you're really the
poster, formerly known as "Tony Polson").
A troll by any other name, eh?
Are you telling me it doesn't? I was going to leave mine inside the
refrigerator to they could meet the little man inside who turns the
light on and off when I open and close the door.
--
Peter
>Eric Stevens wrote:
>
><edited for brevity>
>
>> > On Wed, 30 Mar 2011 21:28:59 -0500, John Turco <jt...@concentric.net> wrote:
>> >
>> >> Robert Coe wrote:
>> >>
>> >> It's called the "Land effect" after its discoverer, Edwin Land, the inventor
>> >> of the Polaroid camera.
>> >
>> > <edited>
>> >
>> > Correction: Edwin Land invented instant film.
>>
>> Is there a difference?
>>
>> Regards,
>>
>> Eric Stevens
>
>
>Why, yes...there >is<. The "Land" camera, itself, has never been anything
>special.
>
>Polaroid's amazing, self-developing film was always the real revelation.
>
>Or, have you assumed the Land camera had a tiny darkroom inside it, replete
>with miniature technicians, chemicals and equipment? (Admittedly, I'd once
>considered something to that effect, as a child.)
Some of the chemicals and equipment were outside of the camera. A
print-wide swabber held in a glass vial was included with each pack of
film. Soaked with some pinkish-colored strong-smelling stop-bath, fixer,
and sealer. It was wiped over each print after the perforated print layer
was pulled away from the developer layer through a large door on the back
of the camera.
[The prints weren't just ejected from a camera and then developed later
before your eyes in daylight like all color Polaroid prints did in the
latter part of the century. There were no motors and batteries to do so in
the first Polaroid cameras. These were only added later as a necessity to
ensure a proper even speed of getting the film through the pinch-rollers
that spread the developer. Too many people could not get the hang of
pulling out the film carrier backing at the right speed and smoothness so
that their prints were coated evenly with the developer.]
>Eric Stevens wrote:
>
><edited for brevity>
>
>> > On Wed, 30 Mar 2011 21:28:59 -0500, John Turco <jt...@concentric.net> wrote:
>> >
>> >> Robert Coe wrote:
>> >>
>> >> It's called the "Land effect" after its discoverer, Edwin Land, the inventor
>> >> of the Polaroid camera.
>> >
>> > <edited>
>> >
>> > Correction: Edwin Land invented instant film.
>>
>> Is there a difference?
>>
>> Regards,
>>
>> Eric Stevens
>
>
Our difference is partly one of terminology: we really understand it
to mean much the same thing.
The Polaroid self-developing film was nothing without a camera
designed to use it.
When Robert Coe (it was Robert Coe?) wrote " Edwin Land, the inventor
of the Polaroid camera" I took him to mean the whole system, camera
film and all.
>Why, yes...there >is<. The "Land" camera, itself, has never been anything
>special.
>
>Polaroid's amazing, self-developing film was always the real revelation.
>
>Or, have you assumed the Land camera had a tiny darkroom inside it, replete
>with miniature technicians, chemicals and equipment? (Admittedly, I'd once
>considered something to that effect, as a child.)
Regards,
Eric Stevens
Yes, I remember all of those details...thanks, for their inclusion.
All right, I concur (and, it >was< Robert Coe).