Sign in with
Sign up | Sign in
Your question

D70 histogram

Last response: in Digital Camera
Share
Anonymous
April 11, 2005 12:16:31 AM

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

I was taking some pictures of Hibiscus today w/50mm f1.8 for a relative. I
checked the histogram on the LCD on each shot, as well as the "highlights".
Histogram showed a nice high peak left of center, with nothing at all at
each end. But when imported into Nikon Capture, all the shots looked
over-exposed. The histogram in Capture showed almost the opposite of what I
saw on the camera with peaks at each end, and the photos appearance
corresponded with that, a few dark shadows and blown highlights. I have not
seen this before. What could I have done wrong?

More about : d70 histogram

April 11, 2005 7:27:47 AM

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

Regardless of the histogram appearance when first opened are you able to
reasonably normalize the histogram by adjusting the shadow and exposure
controls in the RAW plug-in? If so you can likely still pull a good image
out of the data in Photoshop with some basic work. My practice is that the
only adjustments I make in the RAW plug-in are to try to normalize the
histogram as above. For my purposes PSCS contains better tools, including
the shadow/highlight adjustment tool, than the RAW plug-in itself.
If the highlights truly are blown you are SOL: sic semper digital sensors.
The D70 is engineered to "underexposure" because it is possible to bring
details up out of digital shadows but not down out of blown highlights.
Anonymous
April 11, 2005 10:43:52 AM

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

On Sun, 10 Apr 2005 20:16:31 -0700, in rec.photo.digital.slr-systems "Steve
Gavette" <sgavette@no.cox.spam.net.4me> wrote:

>I was taking some pictures of Hibiscus today w/50mm f1.8 for a relative. I
>checked the histogram on the LCD on each shot, as well as the "highlights".
>Histogram showed a nice high peak left of center, with nothing at all at
>each end. But when imported into Nikon Capture, all the shots looked
>over-exposed. The histogram in Capture showed almost the opposite of what I
>saw on the camera with peaks at each end, and the photos appearance
>corresponded with that, a few dark shadows and blown highlights. I have not
>seen this before. What could I have done wrong?

Somehow changed the default adjustments set?
Tools => Options => General

Or changed the default color space setting to use other than the embedded
profile?
Tools => Options => Color Management

----------
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...
Related resources
Anonymous
April 11, 2005 3:43:22 PM

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

In article <l_l6e.5671$%c1.4902@fed1read05>,
"Steve Gavette" <sgavette@no.cox.spam.net.4me> wrote:

> I was taking some pictures of Hibiscus today w/50mm f1.8 for a relative. I
> checked the histogram on the LCD on each shot, as well as the "highlights".
> Histogram showed a nice high peak left of center, with nothing at all at
> each end. But when imported into Nikon Capture, all the shots looked
> over-exposed. The histogram in Capture showed almost the opposite of what I
> saw on the camera with peaks at each end, and the photos appearance
> corresponded with that, a few dark shadows and blown highlights. I have not
> seen this before. What could I have done wrong?

Like any camera, sometimes you are better off using a hand held meter
to isolate subjects that have a lot of dark around them. The histogram
is always going to be an inaccurate way of judging some subject images.

To me the histogram is just one tool and relying on just one tool
is going to be a problem sometimes.

--
LF Website @ http://members.verizon.net/~gregoryblank

"To announce that there must be no criticism of the President,
or that we are to stand by the President, right or wrong,
is not only unpatriotic and servile, but is morally treasonable
to the American public."--Theodore Roosevelt, May 7, 1918
Anonymous
April 11, 2005 6:47:29 PM

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

Steve Gavette <sgavette@no.cox.spam.net.4me> wrote:

>I was taking some pictures of Hibiscus today w/50mm f1.8 for a relative. I
....
>The histogram in Capture showed almost the opposite of what I
>saw on the camera with peaks at each end, and the photos appearance
>corresponded with that, a few dark shadows and blown highlights. I have not
>seen this before. What could I have done wrong?

Is it possibly due to the on-camera histogram showing only the
green channel, while NC shows you composite [or maybe separate
colour channel] histograms? Hibiscus, being solid blocks of
red, could cause a blown problem that way, I'd have thought.
Then again, if exposure was set by 3D colour matrix metering
I'd have expected the camera to get it pretty good.

--
Ken Tough
Anonymous
April 11, 2005 6:47:30 PM

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

"Ken Tough" <ken@objectech.co.uk> wrote in message
news:uthoH$DhHnWCFwUL@objectech.co.uk...
> Steve Gavette <sgavette@no.cox.spam.net.4me> wrote:
>
> >I was taking some pictures of Hibiscus today w/50mm f1.8 for a relative.
I
> ...
> >The histogram in Capture showed almost the opposite of what I
> >saw on the camera with peaks at each end, and the photos appearance
> >corresponded with that, a few dark shadows and blown highlights. I have
not
> >seen this before. What could I have done wrong?
>
> Is it possibly due to the on-camera histogram showing only the
> green channel, while NC shows you composite [or maybe separate
> colour channel] histograms? Hibiscus, being solid blocks of
> red, could cause a blown problem that way, I'd have thought.
> Then again, if exposure was set by 3D colour matrix metering
> I'd have expected the camera to get it pretty good.
>
> --
> Ken Tough

I forgot about the green-only on the camera's histogram, thanks. This was a
very close shot (12"), so in anticipation of overexposing due to the darker
shadows and leaves I was using spot metering on the bloom. Perhaps I should
stay with matrix.
Anonymous
April 12, 2005 5:30:52 AM

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

In message <l_l6e.5671$%c1.4902@fed1read05>,
"Steve Gavette" <sgavette@no.cox.spam.net.4me> wrote:

>I was taking some pictures of Hibiscus today w/50mm f1.8 for a relative. I
>checked the histogram on the LCD on each shot, as well as the "highlights".
>Histogram showed a nice high peak left of center, with nothing at all at
>each end. But when imported into Nikon Capture, all the shots looked
>over-exposed. The histogram in Capture showed almost the opposite of what I
>saw on the camera with peaks at each end, and the photos appearance
>corresponded with that, a few dark shadows and blown highlights. I have not
>seen this before. What could I have done wrong?

I don't know if it's true, but I heard that the D70 histogram only shows
you the green channel; IE, it is not weighted 30:60:10 (R:G:B) for
luminance, but rather 0:100:0.
--

<>>< ><<> ><<> <>>< ><<> <>>< <>>< ><<>
John P Sheehy <JPS@no.komm>
><<> <>>< <>>< ><<> <>>< ><<> ><<> <>><
Anonymous
April 12, 2005 6:17:55 AM

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

<JPS@no.komm> wrote:

> I don't know if it's true, but I heard that the D70 histogram only shows
> you the green channel; IE, it is not weighted 30:60:10 (R:G:B) for
> luminance, but rather 0:100:0.

I don't think it's true. I just tried taking some test shots lit by a
red LED and nothing else, and the histogram seems to have shown me pretty
much exactly what I got. I exposed so the histogram showed just under
clipping, and got no blown highlights in the red channel, but red data
is nearly at peak in the resulting RAW file. (The D70 does have some
headroom in RAW beyond where the histogram shows clipping, so I probably
could have gone a little further.) I did need to dial in more than 2
stops of positive exposure compensation to get there, though, suggesting
that the meter had a little trouble with red-only lighting.

But if the on-camera histogram were showing me only green, I'd expect
it to read very low at that exposure, or else to have completely blown
out the red channel with the histogram showing just under clipping with
the red light. The RAW file shows green dropping off to nothing a bit
below the midpoint on the Photoshop histogram. (The light obviously
isn't pure red.)

--
Jeremy | jeremy@exit109.com
April 12, 2005 6:17:56 AM

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

Jeremy Nixon wrote:

> <JPS@no.komm> wrote:
>
>
>>I don't know if it's true, but I heard that the D70 histogram only shows
>>you the green channel; IE, it is not weighted 30:60:10 (R:G:B) for
>>luminance, but rather 0:100:0.
>
>
> I don't think it's true. I just tried taking some test shots lit by a
> red LED and nothing else, and the histogram seems to have shown me pretty
> much exactly what I got.


Hey that's good news. I might have heard that the blue channel is the
one most affected though I don't recall.



> I exposed so the histogram showed just under
> clipping, and got no blown highlights in the red channel, but red data
> is nearly at peak in the resulting RAW file. (The D70 does have some
> headroom in RAW beyond where the histogram shows clipping, so I probably
> could have gone a little further.) I did need to dial in more than 2
> stops of positive exposure compensation to get there, though, suggesting
> that the meter had a little trouble with red-only lighting.
>
> But if the on-camera histogram were showing me only green, I'd expect
> it to read very low at that exposure, or else to have completely blown
> out the red channel with the histogram showing just under clipping with
> the red light. The RAW file shows green dropping off to nothing a bit
> below the midpoint on the Photoshop histogram. (The light obviously
> isn't pure red.)
>
Anonymous
April 12, 2005 9:43:39 AM

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

On Mon, 11 Apr 2005 14:47:29 +0200, in rec.photo.digital.slr-systems Ken
Tough <ken@objectech.co.uk> wrote:

>Is it possibly due to the on-camera histogram showing only the
>green channel, while NC shows you composite [or maybe separate
>colour channel] histograms? Hibiscus, being solid blocks of
>red, could cause a blown problem that way, I'd have thought.
>Then again, if exposure was set by 3D colour matrix metering
>I'd have expected the camera to get it pretty good.

According to Thom Hogan's e-book it is the luminance channel. In NC the
choices are RGB( couldn't find anything about weighting) or individual R,G,
B.
----------
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
April 12, 2005 3:35:13 PM

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

On Mon, 11 Apr 2005 20:35:23 -0700, paul <paul@not.net> wrote:

>Jeremy Nixon wrote:
>
>> <JPS@no.komm> wrote:
>>
>>
>>>I don't know if it's true, but I heard that the D70 histogram only shows
>>>you the green channel; IE, it is not weighted 30:60:10 (R:G:B) for
>>>luminance, but rather 0:100:0.
>>
>>
>> I don't think it's true. I just tried taking some test shots lit by a
>> red LED and nothing else, and the histogram seems to have shown me pretty
>> much exactly what I got.
>
>
>Hey that's good news. I might have heard that the blue channel is the
>one most affected though I don't recall.

If someone wants to screw around doing more tests like this, an LED is
not required.

Open a Photoshop file, fill it with pure red and display it
full-screen, fill the D70 frame with that. Repeat for any color you
want to test. Neither the phosphor or the LED are *pure*, but you
should be able to test the rough sensitivity of the histogram.

--
Owamanga!
http://www.pbase.com/owamanga
April 12, 2005 3:35:14 PM

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

Owamanga wrote:

> On Mon, 11 Apr 2005 20:35:23 -0700, paul <paul@not.net> wrote:
>
>
>>Jeremy Nixon wrote:
>>
>>
>>> <JPS@no.komm> wrote:
>>>
>>>
>>>
>>>>I don't know if it's true, but I heard that the D70 histogram only shows
>>>>you the green channel; IE, it is not weighted 30:60:10 (R:G:B) for
>>>>luminance, but rather 0:100:0.
>>>
>>>
>>>I don't think it's true. I just tried taking some test shots lit by a
>>>red LED and nothing else, and the histogram seems to have shown me pretty
>>>much exactly what I got.
>>
>>
>>Hey that's good news. I might have heard that the blue channel is the
>>one most affected though I don't recall.
>
>
> If someone wants to screw around doing more tests like this, an LED is
> not required.
>
> Open a Photoshop file, fill it with pure red and display it
> full-screen, fill the D70 frame with that. Repeat for any color you
> want to test. Neither the phosphor or the LED are *pure*, but you
> should be able to test the rough sensitivity of the histogram.


OK I tried that. I have no idea how to interpret the results. I set the
WB to shade for all. All look about the same in the camera histogram
with a lump at the left-center.

Red screen photoshop shows that same lump in the green channel,
blue to the left, red lump to the right.

Green screen photoshop shows that same lump in the red channel,
green in the middle and blue jammed to the left.

Blue screen photoshop shows that same lump in the green channel,
red channel is nearly absent,blue is way over to the right side.
Anonymous
April 12, 2005 8:03:59 PM

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

On Tue, 12 Apr 2005 08:19:15 -0700, paul <paul@not.net> wrote:

>Owamanga wrote:
>
>> If someone wants to screw around doing more tests like this, an LED is
>> not required.
>>
>> Open a Photoshop file, fill it with pure red and display it
>> full-screen, fill the D70 frame with that. Repeat for any color you
>> want to test. Neither the phosphor or the LED are *pure*, but you
>> should be able to test the rough sensitivity of the histogram.
>
>OK I tried that. I have no idea how to interpret the results. I set the
>WB to shade for all. All look about the same in the camera histogram
>with a lump at the left-center.

From this (presuming each shot looked 'correctly exposed') I'd say the
results are pretty conclusive. The D70 histogram isn't just based on
the green channel. If it was, the results would have been more like:

Red screen on camera Histogram: small lump on far left.
Green screen on camera Histogram: healthy lump near center.
Blue screen on camera Histogram: small lump on far left.

However, this flies in the face of everything I've read.

>Red screen photoshop shows that same lump in the green channel,
>blue to the left, red lump to the right.
>
>Green screen photoshop shows that same lump in the red channel,
>green in the middle and blue jammed to the left.
>
>Blue screen photoshop shows that same lump in the green channel,
>red channel is nearly absent,blue is way over to the right side.

Okay, this is confusing, and I think points out a problem with my
suggested test. The phosphor color bandwidths are too broad, or due to
color settings on your PC, some of the other phosphors are still lit
when a pure Red, Green, or Blue is being rendered.

Splash a drop of water on the screen, use it as a magnifying glass and
check the other color sub-pixels are dark when they should be.

Looking at just the green channel, do these lumps match up to the D70
histograms better than the sum of all channels?

Generally, given this evidence, I think the LED-based tests are going
to be much more accurate.

--
Owamanga!
http://www.pbase.com/owamanga
April 12, 2005 8:04:00 PM

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

Owamanga wrote:

> On Tue, 12 Apr 2005 08:19:15 -0700, paul <paul@not.net> wrote:
>
>
>>Owamanga wrote:
>>
>>
>>>If someone wants to screw around doing more tests like this, an LED is
>>>not required.
>>>
>>>Open a Photoshop file, fill it with pure red and display it
>>>full-screen, fill the D70 frame with that. Repeat for any color you
>>>want to test. Neither the phosphor or the LED are *pure*, but you
>>>should be able to test the rough sensitivity of the histogram.
>>
>>OK I tried that. I have no idea how to interpret the results. I set the
>>WB to shade for all. All look about the same in the camera histogram
>>with a lump at the left-center.
>
>
> From this (presuming each shot looked 'correctly exposed') I'd say the
> results are pretty conclusive. The D70 histogram isn't just based on
> the green channel. If it was, the results would have been more like:
>
> Red screen on camera Histogram: small lump on far left.
> Green screen on camera Histogram: healthy lump near center.
> Blue screen on camera Histogram: small lump on far left.
>
> However, this flies in the face of everything I've read.
>
>
>>Red screen photoshop shows that same lump in the green channel,
>>blue to the left, red lump to the right.
>>
>>Green screen photoshop shows that same lump in the red channel,
>>green in the middle and blue jammed to the left.
>>
>>Blue screen photoshop shows that same lump in the green channel,
>>red channel is nearly absent,blue is way over to the right side.
>
>
> Okay, this is confusing, and I think points out a problem with my
> suggested test. The phosphor color bandwidths are too broad, or due to
> color settings on your PC, some of the other phosphors are still lit
> when a pure Red, Green, or Blue is being rendered.
>
> Splash a drop of water on the screen, use it as a magnifying glass and
> check the other color sub-pixels are dark when they should be.


I just see the color and vertical black stripes in the original colored
screens, the photographs show different colored bands mixed up. The D70
photos do not look very close to the original. Blue is pretty close,
green is muddy, red is muddy. The monitor is a Sony trinitron G500,
manually calibrated without hardware spyder or anything.


>
> Looking at just the green channel, do these lumps match up to the D70
> histograms better than the sum of all channels?


No. I franky don't understand why they are so different. Very strange.
The composite B&W photoshop histogram shows various arrangements of
multiple humps but the camera LCD shows the same single hump for all
colors so I guess it is not considering the channels separately & is
looking at a B&W version (luminance?).


>
> Generally, given this evidence, I think the LED-based tests are going
> to be much more accurate.
>
> --
> Owamanga!
> http://www.pbase.com/owamanga
Anonymous
April 13, 2005 12:51:53 AM

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

paul <paul@not.net> wrote:

>OK I tried that. I have no idea how to interpret the results. I set the
>WB to shade for all. All look about the same in the camera histogram
>with a lump at the left-center.
>
>Red screen photoshop shows that same lump in the green channel,
>blue to the left, red lump to the right.
>
>Green screen photoshop shows that same lump in the red channel,
>green in the middle and blue jammed to the left.
>
>Blue screen photoshop shows that same lump in the green channel,
>red channel is nearly absent,blue is way over to the right side.

So it sounds like the green is always in the middle. That would
correspond with the camera always attempting to centre the green
channel in the histogram. When you shoot red or blue, it gets
"over exposed" over to the right. When you shoot green, it
gets exposed properly.

Why does green screen in photoshop show red lump on the right?
That's a good question, but must mean your green isn't a
good green.

--
Ken Tough
April 13, 2005 12:51:54 AM

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

Ken Tough wrote:

> paul <paul@not.net> wrote:
>
>> All look about the same in the camera histogram
>>with a lump at the left-center.
>>
>>Red screen photoshop shows that same lump in the green channel,
>>blue to the left, red lump to the right.
>>
>>Green screen photoshop shows that same lump in the red channel,
>>green in the middle and blue jammed to the left.
>>
>>Blue screen photoshop shows that same lump in the green channel,
>>red channel is nearly absent,blue is way over to the right side.
>
>
> So it sounds like the green is always in the middle. That would
> correspond with the camera always attempting to centre the green
> channel in the histogram. When you shoot red or blue, it gets
> "over exposed" over to the right. When you shoot green, it
> gets exposed properly.


I just don't know. Good theory.


>
> Why does green screen in photoshop show red lump on the right?


Noise & reflections I guess. The green looked probably the worst/muddiest.



> That's a good question, but must mean your green isn't a
> good green.


It was clean on the screen, the camera captured it wrong (I think).
Anonymous
April 13, 2005 2:12:32 AM

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

In message <nncn5152kcv8trg45n9rf72ioe8osvonl2@4ax.com>,
Owamanga <owamanga(not-this-bit)@hotmail.com> wrote:

>Open a Photoshop file, fill it with pure red and display it
>full-screen, fill the D70 frame with that. Repeat for any color you
>want to test. Neither the phosphor or the LED are *pure*, but you
>should be able to test the rough sensitivity of the histogram.

You'd be surprised how little difference there is in the three channels
of a digital capture of 255,0,0, 0,255,0, and 0,0,255 on a CRT. CRTs
don't come anywhere even close to saturation. Last time I tried, the
"0" channels registered about 60% of the level of the "255" channel,
IIRC.
--

<>>< ><<> ><<> <>>< ><<> <>>< <>>< ><<>
John P Sheehy <JPS@no.komm>
><<> <>>< <>>< ><<> <>>< ><<> ><<> <>><
April 13, 2005 2:12:33 AM

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

JPS@no.komm wrote:
> In message <nncn5152kcv8trg45n9rf72ioe8osvonl2@4ax.com>,
> Owamanga <owamanga(not-this-bit)@hotmail.com> wrote:
>
>
>>Open a Photoshop file, fill it with pure red and display it
>>full-screen, fill the D70 frame with that. Repeat for any color you
>>want to test. Neither the phosphor or the LED are *pure*, but you
>>should be able to test the rough sensitivity of the histogram.
>
>
> You'd be surprised how little difference there is in the three channels
> of a digital capture of 255,0,0, 0,255,0, and 0,0,255 on a CRT. CRTs
> don't come anywhere even close to saturation. Last time I tried, the
> "0" channels registered about 60% of the level of the "255" channel,
> IIRC.


The colors look clean on the screen when magnified in a powerful loupe,
it's after being photographed that they got muddy. There was black but
no other colors. Black being un-illuminated screen space. Maybe
reflections contributed, I took them out of focus at 1/15 sec to avoid
moire.
Anonymous
April 13, 2005 2:36:51 AM

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

In message <2_ydnS8A5bH_msHfRVn-ug@speakeasy.net>,
paul <paul@not.net> wrote:

>No. I franky don't understand why they are so different. Very strange.
>The composite B&W photoshop histogram shows various arrangements of
>multiple humps but the camera LCD shows the same single hump for all
>colors so I guess it is not considering the channels separately & is
>looking at a B&W version (luminance?).

It's your brain at work again, making you think that the CRT screen has
high saturation of R, G, and B.

The brain never lets you see things as they actually are; you are always
getting a second-hand interpretation of the sensory data. The brain
knows through past experience that the CRT has low saturation, and
boosts it in its "RAW conversion".
--

<>>< ><<> ><<> <>>< ><<> <>>< <>>< ><<>
John P Sheehy <JPS@no.komm>
><<> <>>< <>>< ><<> <>>< ><<> ><<> <>><
Anonymous
April 13, 2005 4:50:44 AM

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

In message <6v-dnej5nu0RxMHfRVn-3g@speakeasy.net>,
paul <paul@not.net> wrote:

>It was clean on the screen, the camera captured it wrong (I think).

I just remembered; with my 20D, a shot of a CRT has a *very* unstable
exposure. With the shutter button half-depressed, in Av mode, the
shutter speed will flicker and vary over a few stops. Any comparison of
exposure of different color screens should look at the aperture and
shutter speed to see if they are a factor in the exposure (or use
manual). Might be different with the Nikon, though.
--

<>>< ><<> ><<> <>>< ><<> <>>< <>>< ><<>
John P Sheehy <JPS@no.komm>
><<> <>>< <>>< ><<> <>>< ><<> ><<> <>><
April 13, 2005 4:50:45 AM

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

JPS@no.komm wrote:

> In message <6v-dnej5nu0RxMHfRVn-3g@speakeasy.net>,
> paul <paul@not.net> wrote:
>
>
>>It was clean on the screen, the camera captured it wrong (I think).
>
>
> I just remembered; with my 20D, a shot of a CRT has a *very* unstable
> exposure. With the shutter button half-depressed, in Av mode, the
> shutter speed will flicker and vary over a few stops. Any comparison of
> exposure of different color screens should look at the aperture and
> shutter speed to see if they are a factor in the exposure (or use
> manual). Might be different with the Nikon, though.


Shutter priority at 1/8 sec. f/20 for green, f/11 for red & blue.
April 13, 2005 4:50:46 AM

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

Here's a pic I took with all three colors on the same screen:
<http://www.edgehill.net/1/?SC=go.php&DIR=Misc/photograp...;
Pretty close to the original description but easier to see.

>>>>

All look about the same in the camera histogram with a lump at the
left-center.

Red screen photoshop shows that same lump in the green channel,
blue to the left, red lump to the right.

Green screen photoshop shows that same lump in the red channel,
green in the middle and blue jammed to the left.

Blue screen photoshop shows that same lump in the green channel,
red channel is nearly absent,blue is way over to the right side.
Anonymous
April 13, 2005 4:53:02 AM

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

In message <uLydnRLorYAUyMHfRVn-pw@speakeasy.net>,
paul <paul@not.net> wrote:

>The colors look clean on the screen when magnified in a powerful loupe,
>it's after being photographed that they got muddy. There was black but
>no other colors. Black being un-illuminated screen space. Maybe
>reflections contributed, I took them out of focus at 1/15 sec to avoid
>moire.

I'm not talking about accidental illumination of other primaries; I'm
talking about the low saturation of each primary. Black isn't even
black; it's dark grey.
--

<>>< ><<> ><<> <>>< ><<> <>>< <>>< ><<>
John P Sheehy <JPS@no.komm>
><<> <>>< <>>< ><<> <>>< ><<> ><<> <>><
Anonymous
April 13, 2005 6:44:36 AM

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

In message <-8SdnfjUwMYx5cHfRVn-uQ@speakeasy.net>,
paul <paul@not.net> wrote:

>Here's a pic I took with all three colors on the same screen:
><http://www.edgehill.net/1/?SC=go.php&DIR=Misc/photograp...;

Looks like the metering is actually most sensitive to red, and least
sensitive to blue, but as I said in an earlier post, it is possible that
the camera is making these assessments in a very short time interval
compared to the refresh rate.
--

<>>< ><<> ><<> <>>< ><<> <>>< <>>< ><<>
John P Sheehy <JPS@no.komm>
><<> <>>< <>>< ><<> <>>< ><<> ><<> <>><
April 13, 2005 12:53:53 PM

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

JPS@no.komm wrote:

> In message <-8SdnfjUwMYx5cHfRVn-uQ@speakeasy.net>,
> paul <paul@not.net> wrote:
>
>
>>Here's a pic I took with all three colors on the same screen:
>><http://www.edgehill.net/1/?SC=go.php&DIR=Misc/photograp...;
>
>
> Looks like the metering is actually most sensitive to red, and least
> sensitive to blue, but as I said in an earlier post, it is possible that
> the camera is making these assessments in a very short time interval
> compared to the refresh rate.


I had to shoot slower at 1/8 sec to avoid strange patterns from the
refresh rate. I just now corrected the camera histogram for what the
composite shot looked like (2 humps). The individually metered colors
had the same single hump I was showing before. The camera was just
trying to get the darkness even with individual metering since red is
quite dark as the composite shows. Best I can tell the histogram is just
using B&W (luminance?). The lower left camera histogram for the
composite shows a darker hump for red & blue & a lighter hump for green,
it does not show a composite of the various colors which have humps all
over the map. The photoshop black histograms along the top show multiple
humps, the camera does not.


I don't know if the histogram is based on the same thing as the metering
(color matrix?) but the individually metered exposures all look
identical when converted to greyscale so I don't think it's giving
priority to any particular color. Blue & red are the same dark shade
when converted to geyscale so they share the same hump on the camera
histogram.
Anonymous
April 13, 2005 2:58:36 PM

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

JPS@no.komm wrote:

> In message <6v-dnej5nu0RxMHfRVn-3g@speakeasy.net>,
> paul <paul@not.net> wrote:
>
>
>>It was clean on the screen, the camera captured it wrong (I think).
>
>
> I just remembered; with my 20D, a shot of a CRT has a *very* unstable
> exposure. With the shutter button half-depressed, in Av mode, the
> shutter speed will flicker and vary over a few stops. Any comparison of
> exposure of different color screens should look at the aperture and
> shutter speed to see if they are a factor in the exposure (or use
> manual). Might be different with the Nikon, though.

It's like that on all my minolta's as well (varying reading) I guess the
sampling period is too short.

Taking a reading by holding a Minolta incident meter against a white
screen works. From there, just open up 1.7 to 2 stops.

Cheers,
Alan
--
-- r.p.e.35mm user resource: http://www.aliasimages.com/rpe35mmur.htm
-- r.p.d.slr-systems: http://www.aliasimages.com/rpdslrsysur.htm
-- slr-systems FAQ project: http://tinyurl.com/6m9aw
-- [SI] gallery & rulz: http://www.pbase.com/shootin
-- e-meil: Remove FreeLunch.
Anonymous
April 13, 2005 5:52:02 PM

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

Just a few thoughts:

The camera is getting raw intensity values off the CCD for every single
pixel. The CCD has RGB filters over the individual pixels, but the CCD
itself and the values coming off it are just monochrome. You build the
histogram from that and shove it in the exif data. It'd actually be
slightly more complex to get just R, G, or B histograms (which would be
nice to have sometimes!) Though not too hard - just throw out every other
pixel for G or 3/4 pixels for R or B.

I take IR pictures with my D70. These are very very red - nothing but red
unless I let the exposure go on a long time (it turns a bit orange - the
green response curve does overlap the red and blue response curves a
bit), and demonstrate all the blockiness you'd expect from having 75% of
the pixels (B and G) just twidding their thumbs. But the histogram looks
as I'd expect.

The idea that the histogram is green only may be because when you do
demosaic-ing (filling in the missing RG and B pixels) you quite often get
the luminance from the green channel, for several good reasons: There are
twice as many green pixel sites to interpolate from; the ccd is most
sensitive to green wavelengths; so are your eyes; when you do RGB->grey
conversion you give green about 70% of the weight. So building the
histogram from green only would work pretty well in most cases, I think -
but I tend to think it's not being done here.
!