How to get good black & white from digital?

"Matthew Winn" wrote in message
...
The original scene has far more chromatic information than the
recorded image. When using filters you are manipulating the entire
spectrum. If you try to manipulate the image you're working with
three spikes.

Accepting that is only a slight exageration, what makes you think
manipulating data that will not be captured in those "spikes" will have any
affect on them in the final photo? Will the capture abilities of the camera
somehow be increased by using glass filters on the lens? Can you please
explain that process?

Sometimes that may not matter, but other times it might. A tri-colour
image may look similar to the original scene, but it's _not_ the
original scene and it can't respond to digital filtration in the way
that the original scene could. An orange filter, for example, doesn't
do the same thing as cutting out the blue channel, reducing the green
and boosting the red.

So what affect does it have, in your opinion, to the digital image that is
captured? (since we were discussing digital photo manipulation, not just
looking through the eyepiece of your camera without pressing the shutter!

That's not to say that there's no place for digital filtering, but a
discerning photographer should be able to work out the right choice
for each image and know _why_ it's the right choice. Sometimes the
right choice is to filter the light before it's recorded.

And I would not suggest otherwise. Everyone gets to choose for themselves,
whether they can justify it or not.

MrT.

On Thu, 25 Oct 2007 08:33:45 +1000, "Mr.T" MrT@home wrote:


"Matthew Winn" wrote in message
...
The original scene has far more chromatic information than the
recorded image. When using filters you are manipulating the entire
spectrum. If you try to manipulate the image you're working with
three spikes.

Accepting that is only a slight exageration, what makes you think
manipulating data that will not be captured in those "spikes" will have any
affect on them in the final photo? Will the capture abilities of the camera
somehow be increased by using glass filters on the lens? Can you please
explain that process?

Suppose you have two objects that look similar to the eye but which
reflect light in different ways: perhaps vegetation, where some leaves
have a fairly sharp peak in the yellow-green part of the spectrum
while others have a broader reflection that is weakly deficient in red
and strongly so in blue but otherwise includes most of the visible
spectrum.

Both will look yellowish-green, and if you photograph them with a
digital camera you'll find that both are recorded with a strong signal
in the green, a weak one in the red, and almost nothing in the blue.
At this point there's no way to distinguish between them: they're both
recorded the same way and the differences in the way they reflect
light has been lost forever. If you try to create a black and white
image from this all the leaves will come out looking the same shade
of grey.

A filter, however, could separate them and bring out the differences,
adding texture to the foliage. A yellow filter will darken the leaves
with the broader reflection while leaving the narrower peak untouched.
A greenish-blue filter, on the other hand, will turn the narrower peak
dark and leave the broader one lighter.

Working with the image alone denies you that ability. You no longer
have as much control over the tones of the image. All you can do as
adjust three numbers for each pixel, and if reducing the image to
those three numbers has lost useful information there's nothing you
can do to get it back.

--
Matthew Winn
[If replying by mail remove the "r" from "urk"]

Matthew Winn wrote:
Mr.Troll wrote:
Matthew Winn wrote:

The original scene has far more chromatic information than the
recorded image. When using filters you are manipulating the entire
spectrum. If you try to manipulate the image you're working with
three spikes.

Accepting that is only a slight exageration, what makes you think
manipulating data that will not be captured in those "spikes" will have
any
affect on them in the final photo? Will the capture abilities of the
camera
somehow be increased by using glass filters on the lens? Can you please
explain that process?

Suppose you have two objects that look similar to the eye but which
reflect light in different ways: perhaps vegetation, where some leaves
have a fairly sharp peak in the yellow-green part of the spectrum
while others have a broader reflection that is weakly deficient in red
and strongly so in blue but otherwise includes most of the visible
spectrum.

Both will look yellowish-green, and if you photograph them with a
digital camera you'll find that both are recorded with a strong signal
in the green, a weak one in the red, and almost nothing in the blue.
At this point there's no way to distinguish between them: they're both
recorded the same way and the differences in the way they reflect
light has been lost forever. If you try to create a black and white
image from this all the leaves will come out looking the same shade
of grey.

A filter, however, could separate them and bring out the differences,
adding texture to the foliage. A yellow filter will darken the leaves
with the broader reflection while leaving the narrower peak untouched.
A greenish-blue filter, on the other hand, will turn the narrower peak
dark and leave the broader one lighter.

Working with the image alone denies you that ability. You no longer
have as much control over the tones of the image. All you can do as
adjust three numbers for each pixel, and if reducing the image to
those three numbers has lost useful information there's nothing you
can do to get it back.

Thanks for that explanation, Matthew. It makes sense to me.

On Oct 25, 7:43 pm, "Wilba" wrote:
Matthew Winn wrote:
Mr.Troll wrote:
Matthew Winn wrote:

The original scene has far more chromatic information than the
recorded image. When using filters you are manipulating the entire
spectrum. If you try to manipulate the image you're working with
three spikes.

Accepting that is only a slight exageration, what makes you think
manipulating data that will not be captured in those "spikes" will have
any
affect on them in the final photo? Will the capture abilities of the
camera
somehow be increased by using glass filters on the lens? Can you please
explain that process?

Suppose you have two objects that look similar to the eye but which
reflect light in different ways: perhaps vegetation, where some leaves
have a fairly sharp peak in the yellow-green part of the spectrum
while others have a broader reflection that is weakly deficient in red
and strongly so in blue but otherwise includes most of the visible
spectrum.

Both will look yellowish-green, and if you photograph them with a
digital camera you'll find that both are recorded with a strong signal
in the green, a weak one in the red, and almost nothing in the blue.
At this point there's no way to distinguish between them: they're both
recorded the same way and the differences in the way they reflect
light has been lost forever. If you try to create a black and white
image from this all the leaves will come out looking the same shade
of grey.

A filter, however, could separate them and bring out the differences,
adding texture to the foliage. A yellow filter will darken the leaves
with the broader reflection while leaving the narrower peak untouched.
A greenish-blue filter, on the other hand, will turn the narrower peak
dark and leave the broader one lighter.

Working with the image alone denies you that ability. You no longer
have as much control over the tones of the image. All you can do as
adjust three numbers for each pixel, and if reducing the image to
those three numbers has lost useful information there's nothing you
can do to get it back.

Thanks for that explanation, Matthew. It makes sense to me.

Nice to see lots of photographic content going on here!

This discussion is quite interesting, and having just played around at
some length with 8 and 16 bit images, I'm getting the hang of some of
the more esoteric problems with b&w. It's tricky!

But I'm feeling brave - here's an attempt at doing a little better
with black/white points, getting a nice spread of tones, avoiding
posterisation, getting the 'feel' right, etc.. Ok, maybe 'snot a work
of art, but it's been a good sample image to experiment with, and I'm
starting to feel a lot more comfortable with the issues:

http://www.marktphoto.com/commuter_bw.jpg

Harsh (but constructive) critique very welcome. FTR, I was looking
for an image that looked a little 'odd' and had a bit of undefined
tension in it, like part of a strange dream. But I poetically
digress*....

* - aka 'wanking'... (O;


If anyone wants the full size (8Mp) colour original so they can play
(and then post a better version!), I'm happy to send it. It's only a
jpeg, but it's a pretty good one.. (And I'm fine with people
reposting my stuff. (O


Thanks for all the excellent contributions to this thread, folks.

mt

In article .com,
wrote:

BIG SNIP



Thanks for that explanation, Matthew. It makes sense to me.

Nice to see lots of photographic content going on here!

This discussion is quite interesting, and having just played around at
some length with 8 and 16 bit images, I'm getting the hang of some of
the more esoteric problems with b&w. It's tricky!

But I'm feeling brave - here's an attempt at doing a little better
with black/white points, getting a nice spread of tones, avoiding
posterisation, getting the 'feel' right, etc.. Ok, maybe 'snot a work
of art, but it's been a good sample image to experiment with, and I'm
starting to feel a lot more comfortable with the issues:

http://www.marktphoto.com/commuter_bw.jpg

Harsh (but constructive) critique very welcome. FTR, I was looking
for an image that looked a little 'odd' and had a bit of undefined
tension in it, like part of a strange dream. But I poetically
digress*....

* - aka 'wanking'... (O;


If anyone wants the full size (8Mp) colour original so they can play
(and then post a better version!), I'm happy to send it. It's only a
jpeg, but it's a pretty good one.. (And I'm fine with people
reposting my stuff. (O


Thanks for all the excellent contributions to this thread, folks.

mt

Mark, the tonal spectrum appears to be quite good, as are the
composition and subject matter. But (donning asbestos suit) the image
has the plasticky look of digital. You need to add some grain for
interesting texture.

HFL

--
www.pbase.com/hlockwood
Change hlockwood to hflockwood in email address

Mark Thomas wrote:

Nice to see lots of photographic content going on here!

This discussion is quite interesting, and having just played around at
some length with 8 and 16 bit images, I'm getting the hang of some of
the more esoteric problems with b&w. It's tricky!

But I'm feeling brave - here's an attempt at doing a little better
with black/white points, getting a nice spread of tones, avoiding
posterisation, getting the 'feel' right, etc.. Ok, maybe 'snot a work
of art, but it's been a good sample image to experiment with, and I'm
starting to feel a lot more comfortable with the issues:

http://www.marktphoto.com/commuter_bw.jpg

Harsh (but constructive) critique very welcome. FTR, I was looking
for an image that looked a little 'odd' and had a bit of undefined
tension in it, like part of a strange dream. But I poetically
digress*....

* - aka 'wanking'... (O;

Looks fine to me. I can see all the tones in it.

If anyone wants the full size (8Mp) colour original so they can play
(and then post a better version!), I'm happy to send it. It's only a
jpeg, but it's a pretty good one.. (And I'm fine with people
reposting my stuff. (O

Yeah I'd like to have a play with it, just for the chance to be able to
compare results, and to find out how others achieve theirs. Would you put it
with commuter_bw.jpg and then give us the URL?

Harry Lockwood wrote:

Mark, the tonal spectrum appears to be quite good, as are the
composition and subject matter. But (donning asbestos suit) the image
has the plasticky look of digital. You need to add some grain for
interesting texture.

Is that what the question of digital b&w is really about - the "plasticy
look"? That phrase makes me think of Pixar animation, which looks brilliant
for toys but crap for humans. So digital b&w "just don't look right" ...?

On Oct 26, 7:16 am, wrote:
On Oct 25, 7:43 pm, "Wilba" wrote:





Matthew Winn wrote:
Mr.Troll wrote:
Matthew Winn wrote:

The original scene has far more chromatic information than the
recorded image. When using filters you are manipulating the entire
spectrum. If you try to manipulate the image you're working with
three spikes.

Accepting that is only a slight exageration, what makes you think
manipulating data that will not be captured in those "spikes" will have
any
affect on them in the final photo? Will the capture abilities of the
camera
somehow be increased by using glass filters on the lens? Can you please
explain that process?

Suppose you have two objects that look similar to the eye but which
reflect light in different ways: perhaps vegetation, where some leaves
have a fairly sharp peak in the yellow-green part of the spectrum
while others have a broader reflection that is weakly deficient in red
and strongly so in blue but otherwise includes most of the visible
spectrum.

Both will look yellowish-green, and if you photograph them with a
digital camera you'll find that both are recorded with a strong signal
in the green, a weak one in the red, and almost nothing in the blue.
At this point there's no way to distinguish between them: they're both
recorded the same way and the differences in the way they reflect
light has been lost forever. If you try to create a black and white
image from this all the leaves will come out looking the same shade
of grey.

A filter, however, could separate them and bring out the differences,
adding texture to the foliage. A yellow filter will darken the leaves
with the broader reflection while leaving the narrower peak untouched.
A greenish-blue filter, on the other hand, will turn the narrower peak
dark and leave the broader one lighter.

Working with the image alone denies you that ability. You no longer
have as much control over the tones of the image. All you can do as
adjust three numbers for each pixel, and if reducing the image to
those three numbers has lost useful information there's nothing you
can do to get it back.

Thanks for that explanation, Matthew. It makes sense to me.

Nice to see lots of photographic content going on here!

This discussion is quite interesting, and having just played around at
some length with 8 and 16 bit images, I'm getting the hang of some of
the more esoteric problems with b&w. It's tricky!

But I'm feeling brave - here's an attempt at doing a little better
with black/white points, getting a nice spread of tones, avoiding
posterisation, getting the 'feel' right, etc.. Ok, maybe 'snot a work
of art, but it's been a good sample image to experiment with, and I'm
starting to feel a lot more comfortable with the issues:

http://www.marktphoto.com/commuter_bw.jpg

Harsh (but constructive) critique very welcome. FTR, I was looking
for an image that looked a little 'odd' and had a bit of undefined
tension in it, like part of a strange dream. But I poetically
digress*....

* - aka 'wanking'... (O;

If anyone wants the full size (8Mp) colour original so they can play
(and then post a better version!), I'm happy to send it. It's only a
jpeg, but it's a pretty good one.. (And I'm fine with people
reposting my stuff. (O

Thanks for all the excellent contributions to this thread, folks.

mt- Hide quoted text -

- Show quoted text -

I decided to attack your image. (In PhotoShop)
1. Just an original for comparison
2. Some grain for Harry (serious)
3. A little selective diffuse glow (moderately serious)
4. Some glass (not at all serious)

http://www.jlkramer.net/Pictures/Mar...rksPicture.htm

Jim :-)

On Oct 26, 10:01 pm, Harry Lockwood wrote:
Mark, the tonal spectrum appears to be quite good, as are the
composition and subject matter. But (donning asbestos suit) the image
has the plasticky look of digital. You need to add some grain for
interesting texture.
Haha!! No asbestos required. I value your comments very much, having
seen your work.

Yes, I understand what you mean. I think this raises an interesting
point, that I have reflected upon before. Why do we like grain? Is
it because we have seen it in every photograph/film frame that we have
ever examined closely so that is simply part of the psyche of
photography? Or is there some aesthetic logic to it? Or is it
because the smooth look reminds us of (bad) digital photography, with
lost tones, false/lack of detail. I mean, let's face facts, the sky
*isn't* grainy... yet I agree, a hint of grain often looks better.

And then how does one get realistic grain out of a digital? - because
imo, added noise seems to always look almost as bad as the 'plastic
look'. Another thing to research... Or I guess I could just drag out
the old X700 and load her up with film.. nah.

I have to say that I don't find the plastic look too objectionable,
*as long as* the image has sufficient resolution to show detail right
down to the limit of close examination, provided the gradients don't
look like CGI, ie too smooth, and provided that genuine variations of
tone (eg skintones, imperfections) are not lost. In this case, it is
just the resampling that smoothed the image somewhat - there was no
added noise reduction. I did experiment with adding grain effects but
ended up with results like Jim's version, and I didn't really like
it. (nothing personal, Jim.. see other reply forthcoming... (O

Thanks for dropping by, Harry.

For those interested in good b&w, I suggest you drop in to HFL's site:
--www.pbase.com/hlockwood

Thanks for these Jim - very good/entertaining.. (O:

http://www.jlkramer.net/Pictures/Mar...rksPicture.htm

On Oct 26, 11:51 pm, JimKramer wrote:
I decided to attack your image. (In PhotoShop)
1. Just an original for comparison
2. Some grain for Harry (serious)
I liek it, but would tone it down a little, and somehow you can tell
this is added noise. Is there a better way to get more realistic
looking grain? Or is it just me 'knowing' it is digital? I can't
quite put my finger on why it doesn't look like film grain, but
something isn't quite right.

3. A little selective diffuse glow (moderately serious)
I like it, but again I would go for a bit less, and possibly apply it
to the boat as well. hmm.

4. Some glass (not at all serious)
Heheh! That's seriously weird. And just a bit freaky.

My sole attempt at selective treatment was just to leave the curving
rail on the boat in colour - it's a bright blue, and it sort of
naturally takes your eye around to the man in the suit.. It was ok..
but I looked at it for a while and decided it didn't really improve it
to my eye.


Feel free to leave the images up for as long as you wish, by the way.

Thanks for taking the time - genuinely appreciated.

mt