Color streaks in greyscale photos
marjory
experimenting with it lately. (I know the general advice is to shoot
in color and convert it later, but this gives me better feedback right
away.)
I had some printed as 5x7s after editing in PSP and was quite pleased
with the results. With a more recent batch however, I was surprised
to find some bands of color in the prints. Looking again at the files
I sent, I discovered that the color was in my work and wasn't the
printer's fault as I had assumed. I hadn't noticed it when I was
working with the photos.
I don't remember now just which adjustments I had made in PSP, but it
could have been any combination of Histogram, Clarify,
Brightness/Contrast and Curves, depending on each photo's needs. It
particularly shows in areas of uniform tone, such as the sky and
water. Actually, it may be from jpg compression, though I took the
photos at a high quality setting. It seems odd however that the
streaks have color in them. I don't think I had repeatedly saved and
reopened any of the photos, as I know that creates more jpg
compression problems.
I posted two versions of one ofthe scenes here -
http://community.webshots.com/album/33793835gKkIcM
Any suggestions on what caused this and/or how to avoid it would be
appreciated!
Marjory
Uni
Buy an Epson printer!!!! My guess is, you own an HP printer.
Uni
> Marjory
A Soberon
> experimenting with it lately. (I know the general advice is to shoot
> in color and convert it later, but this gives me better feedback right
> away.)
>
> I had some printed as 5x7s after editing in PSP and was quite pleased
> with the results. With a more recent batch however, I was surprised
> to find some bands of color in the prints. Looking again at the files
> I sent, I discovered that the color was in my work and wasn't the
> printer's fault as I had assumed. I hadn't noticed it when I was
> working with the photos.
>
> I don't remember now just which adjustments I had made in PSP, but it
> could have been any combination of Histogram, Clarify,
> Brightness/Contrast and Curves, depending on each photo's needs. It
> particularly shows in areas of uniform tone, such as the sky and
> water. Actually, it may be from jpg compression, though I took the
> photos at a high quality setting. It seems odd however that the
> streaks have color in them. I don't think I had repeatedly saved and
> reopened any of the photos, as I know that creates more jpg
> compression problems.
You don't mention your camera make or model, so I'll guess at what's
happening.
When you open the "b&w" pictures taken from the camera, do they acutally
open as gray-scale? or are they 16 million color pictures with only
black-gray-white on them? You can tell by the different color palette in
PSP. If they are 16 million color pictures, then a colored palette should
appear, if not it will be a 256 gray scale palette.
My guess is your camera is taking 16 million "false" b&w pictures, or, when
adjusting the pictures you're incresing the number of colors from 256 to
16million. I any case, you can easily fall into putting some color in them.
Why don't you start by going to COLORS-->GREY SCALE and then you'll be sure
to have only black-grey-white shades, and then do all the adjusting you
want?
Antonio
Kris Zaklika
>
> My digital camera has a black and white setting, and I've been
> experimenting with it lately. (I know the general advice is to shoot
> in color and convert it later, but this gives me better feedback right
> away.)
>
> I had some printed as 5x7s after editing in PSP and was quite pleased
> with the results. With a more recent batch however, I was surprised
> to find some bands of color in the prints. Looking again at the files
> I sent, I discovered that the color was in my work and wasn't the
> printer's fault as I had assumed. I hadn't noticed it when I was
> working with the photos.
>
> I don't remember now just which adjustments I had made in PSP, but it
> could have been any combination of Histogram, Clarify,
> Brightness/Contrast and Curves, depending on each photo's needs.
Since Histogram Adjustment and Curves are capable of working
on individual color channels, the color is probably the result
of not applying exactly the same adjustment to all three color
channels. The solution to this problem is to not have three
color channels but, instead, one greyscale channel. Do this
with Colors > Greyscale.
> It
> particularly shows in areas of uniform tone, such as the sky and
> water. Actually, it may be from jpg compression, though I took the
> photos at a high quality setting. It seems odd however that the
> streaks have color in them. I don't think I had repeatedly saved and
> reopened any of the photos, as I know that creates more jpg
> compression problems.
>
> I posted two versions of one ofthe scenes here -
> http://community.webshots.com/album/33793835gKkIcM
>
> Any suggestions on what caused this and/or how to avoid it would be
> appreciated!
> Marjory
--
----------------------------------------------------------------------
Kris Zaklika Jasc Software, Inc. The
Product Ideas: id...@jasc.com Power
Customer Service: customer...@jasc.com To
Technical Support: tec...@jasc.com Create
----------------------------------------------------------------------
Porter
"marjory" <mar...@nycap.not.here.rr.com> wrote in message
news:3de259ad...@news-server.nycap.rr.com...
Run a colors count on both images. You'll find that the grayscale has 256
while the other one is sporting 3147. Apparently you upped the colors count
(in itself not a bad thing), somewhere along the way. Then, on one of your
subsequent adjustments, you altered that colors balance. I see if happen
all the time, and, from all sorts of different operations. Aside from the
pinkish colors there, you also have extra banding in the 3147 version.
These all should have been quite visible at the time, and probably just
slipped past you. That's easy enough to happen. Especially if we aren't
looking for it.
Assuming once all grays, always all grays? Ain't so. Any time we up our
colors counts, and start making adjustments? There are all sorts of
opportunities to introduce extra colors, other than the original 256 grays.
While I prefer to work in 24 bit a lot of the time - also means, I must drop
that saturation (there are assorted ways available) to carefully maintain
those "all grays". Once you take your grayscale image up into full 24 bit
color, you've gotta remain even more diligent about confirming all
colors/controls. Dumping over sixteen million more potentials for color
into our pots of visual soup? Then stirring it all with a big fat spoon?
This offers both some advantages and pitfalls. :) Either be willing to
watch out for them, or simply remain more safely back over there in
grayscale land.
Porter
Ember
(marjory) wrote:
>My digital camera has a black and white setting, and I've been
>experimenting with it lately. (I know the general advice is to shoot
>in color and convert it later, but this gives me better feedback right
>away.)
>
>I had some printed as 5x7s after editing in PSP and was quite pleased
>....
Here is my entry for worst analogy of the year...
If you have a house with a water tap connected to the watermain, the
watermain contains the colour (I'm cheating - you would need three
watermains and three taps for all three colours), then:
Situation 1. Removing the pipe between the house and the main is
converting to greyscale. No matter how little or much you turn the tap
you don't get any colour. Having streaks of colour is technically
impossible, you have a greyscale tap and the only thing that will come
out is 256 levels of grey.
Situation 2. With the pipe put back, you have a potential 16 million
colour image. If you turn the tap on full you get full rich colour
(full saturation). As you turn the tap down you get fainter colour. If
you turn the tap off (zero saturation) you get a grey looking image.
But it's still a 16 million colours image, every colour however is
represented by a shade of grey.
Your camera may be turning colour saturation down to 1% or 2% instead
of 0% in its so-called B&W mode and that would give faint colour
streaks. It's also possible while manipulating your 16 million colour
image to accidentally turn the tap on (increase saturation) a little
to introduce a trace of colour.
Use Color > Greyscale to remove the connecting water pipe for 100%
guaranteed B&W. Also, your file size will be considerably smaller.
Ember
Porter
> Here is my entry for worst analogy of the year...
It wasn't all that bad, I've seen worse. Heck, I've written far worse. Of
course, my subtler ones, usually have just enough vague sordid references,
to hopefully to make 'em kinda fun. Even if they don't present a lot of
sense, techno-graphically that is. While yours wasn't all that bad, Ember -
and, actually even made some sense? It just wasn't interesting. It needed
more sex. :)
Porter
Ember
But water pipes *are* sexy Porter.
Now I'm one of those rare persons who actually stop and look into that
hole in the street when there is water pipe work going on. Once the
foreman realizes I'm not a mental escapee, his eyes light up that
somebody actually cares what his crew is doing. Years of working his
way up to foreman and nobody asks him anything.
Just two weeks ago, four houses away the emergency crew arrived to fix
a broken connection between the street and the house. As it turns out
there had been a small undetectable leak for many years. The leak was
in the threads where the "corporation cock" (a combination tap-in
connector for the water line and cut-off valve) threaded into the 6"
main. The jet of water from the 90 psi line had been squirting over
the heavy duty fitting and had eroded it away to create a beautifully
sculptured piece of bronze. At this point the fitting broke open
because it was so thin, having done its job for 50 years without
complaint.
I also found out that the 6" main was plain cast iron and not
malleable iron as now used. The plain iron is brittle and now after
every earth tremor, even one we cant feel, the waterworks people get a
few leaks, always on the underside of the pipe.
I'll spare you the sordid details of how he patched the main and made
a reconnection. Wouldn't be a sexy story.
As a bonus to this conversation, I was advised by one of the workers
who to vote for, for mayor in the municipal election on Nov. 16. He
said the guy I was going to vote for was a complete a-hole. This
steady no-nonsense trench worker swayed me his way! But neither his
guy or my former choice won.
The lady of the greyscale house went next door to a 16 million house
for a container of water and made coffee for the crew and supplied
cookies. She is a stay-at-home under 30 mom and sexy.
Okay, so water pipes aren't sexy. They just have character :-)
Ember
Kris Zaklika
Are you sure this wasn't a case of electrolysis occurring where
two dissimilar metals were in contact?
> I also found out that the 6" main was plain cast iron and not
> malleable iron as now used. The plain iron is brittle and now after
> every earth tremor, even one we cant feel, the waterworks people get a
> few leaks, always on the underside of the pipe.
>
> I'll spare you the sordid details of how he patched the main and made
> a reconnection. Wouldn't be a sexy story.
>
> As a bonus to this conversation, I was advised by one of the workers
> who to vote for, for mayor in the municipal election on Nov. 16. He
> said the guy I was going to vote for was a complete a-hole. This
> steady no-nonsense trench worker swayed me his way! But neither his
> guy or my former choice won.
And now a bonus question for you. How much bigger must you make
the diameter of your water pipe to double the rate of flow of
water through it assuming, say, 50 psi pressure at the input end
and a smooth interior to the pipe?
> The lady of the greyscale house went next door to a 16 million house
> for a container of water and made coffee for the crew and supplied
> cookies. She is a stay-at-home under 30 mom and sexy.
>
> Okay, so water pipes aren't sexy. They just have character :-)
>
> Ember
>
>
--
Ember
wrote:
>Ember wrote:
>>
>> On Mon, 25 Nov 2002 21:18:48 GMT, "Porter"
>> <campratty...@attbi.com> wrote:
>>
>> >"Ember" <REMe...@operamail.com> wrote in message
>> .....
>> Just two weeks ago, four houses away the emergency crew arrived to fix
>> a broken connection between the street and the house. As it turns out
>> there had been a small undetectable leak for many years. The leak was
>> in the threads where the "corporation cock" (a combination tap-in
>> connector for the water line and cut-off valve) threaded into the 6"
>> main. The jet of water from the 90 psi line had been squirting over
>> the heavy duty fitting and had eroded it away to create a beautifully
>> sculptured piece of bronze. At this point the fitting broke open
>> because it was so thin, having done its job for 50 years without
>> complaint.
>
>Are you sure this wasn't a case of electrolysis occurring where
>two dissimilar metals were in contact?
Watermain connections are always dissimilar metals because the
building feed is copper (formerly galvanised steel -- zinc) and the
pipe is a form of iron. If electrolyses was happening the water
systems of the world would be in continuous chaos. I think the reason
electrolysis is not a practial factor is that both metals are at the
same voltage potential -- being buried in the ground, whereas in a
house, pipe leaks can happen after just a few years at the junction of
copper and steel.
>>.....
>> As a bonus to this conversation, I was advised by one of the workers
>> who to vote for, for mayor in the municipal election on Nov. 16. He
>> said the guy I was going to vote for was a complete a-hole. This
>> steady no-nonsense trench worker swayed me his way! But neither his
>> guy or my former choice won.
>
>And now a bonus question for you. How much bigger must you make
>the diameter of your water pipe to double the rate of flow of
>water through it assuming, say, 50 psi pressure at the input end
>and a smooth interior to the pipe?
>
Pipe flow varies with the pipe diameter to the 2.63 power, the head
loss and water velocity. The head loss is a factor of pipe length and
roughness coefficient. There isn't a pat answer to your question
(need more data as you would say!)
But for 100 feet of new 3/4" copper house connection at 50 psi inlet
pressure with zero pressure at the house end (i.e. water was freely
flowing into a barrel), if that pipe diameter was 3/4" inch the pipe
would have to be increased to 1" diam to double the flow. Actually a
hair less than 1" but that's the next nearest standard size.
Ember
Kris Zaklika
Not bad. I'd say fourth power of the radius. Both my son's
science project and the plumbing handbook agreed. I should,
of course, have specified straight pipe since the effective
size dependence is influenced by bends, joins and junctions
in the pipe. I forget the dependence on pipe material. And,
of course, the flow has to be sufficiently slow to avoid
cavitation. Interesting stuff, plumbing.
> But for 100 feet of new 3/4" copper house connection at 50 psi inlet
> pressure with zero pressure at the house end (i.e. water was freely
> flowing into a barrel), if that pipe diameter was 3/4" inch the pipe
> would have to be increased to 1" diam to double the flow. Actually a
> hair less than 1" but that's the next nearest standard size.
>
> Ember
--
Ember
digital photography.
Ilya Razmanov
<snip>
>
> Watermain connections are always dissimilar metals because the
> building feed is copper (formerly galvanised steel -- zinc) and the
> pipe is a form of iron. If electrolyses was happening the water
> systems of the world would be in continuous chaos. I think the reason
> electrolysis is not a practial factor is that both metals are at the
> same voltage potential -- being buried in the ground, whereas in a
> house, pipe leaks can happen after just a few years at the junction of
> copper and steel.
Hm, actually, copper and iron have quite different potential. There are
other factors, though: one is "sacrificial anodes" Ron mentioned. Another is
possible pipe coverage - don't know about US but here it's made from
"vaporized" zinc. Since zinc has larger potential than iron and copper, this
protective layer work as "sacrificial anode", plus it forms rather dense
oxide layer. Chrome may work this way, too.
Ilyich.
--------------------------------------------------------------------------
Ilya Razmanov (a.k.a. Ilyich the Toad)
http://photoshop.msk.ru/ - Photoshop plug-in filters
"When there's only one candidate, there's only one choice" - Monkey Island
--------------------------------------------------------------------------
Ilya Razmanov
"Kris Zaklika" <kzak...@jasc.com> wrote in message
news:3DE3101E...@jasc.com...
> Ember wrote:
> >
> >
> > Pipe flow varies with the pipe diameter to the 2.63 power, the head
> > loss and water velocity. The head loss is a factor of pipe length and
> > roughness coefficient. There isn't a pat answer to your question
> > (need more data as you would say!)
>
> Not bad. I'd say fourth power of the radius. Both my son's
> science project and the plumbing handbook agreed. I should,
> of course, have specified straight pipe since the effective
> size dependence is influenced by bends, joins and junctions
> in the pipe. I forget the dependence on pipe material. And,
> of course, the flow has to be sufficiently slow to avoid
> cavitation. Interesting stuff, plumbing.
Huh, "cavitation"? Here we normally refer to "turbulence". That is, when
flow changes from laminar to turbulent, the dependence of pipeline
"resistance" from Reynolds number changes and you can't directly extrapolate
old "laminar" data onto it. Cavitation, i.e. partial evaporation, is rarely
discussed because it is avoided by all means possible :-) - before it may
happen in the pipeline, it is expected to happen in the pump, significantly
decreasing its efficiency and probably damaging the "underloaded" engine.
All in all, I feel like I have to find my old "Three piglets" engineer
handbook (students call it "Three piglets" because full title takes ca. 1
minute to read, and so do three full names of the authors <g>) to keep
reading this NG...
Ilyich.
--------------------------------------------------------------------------
Ilya Razmanov (a.k.a. Ilyich the Toad)
http://photoshop.msk.ru/ - Photoshop plug-in filters
"Sometimes, you just wake up in trouble" - Ben, Full Throttle Demo
--------------------------------------------------------------------------
Bill Schnakenberg
>On Tue, 26 Nov 2002 05:38:23 GMT, Ember <REMe...@operamail.com>
>wrote:
>
>
>
>>I think the reason
>>electrolysis is not a practial factor is that both metals are at the
>>same voltage potential -- being buried in the ground, whereas in a
>>house, pipe leaks can happen after just a few years at the junction of
>>copper and steel.
>>
>>
>
>In fact it is a factor which is why we now use sacrificial anodes at
>main stops with cast iron water mains. The trend today is to use
>plastic pipe for most new water mains though.
>
It depends upon local building codes and how much pull the local
plumbing unions have with the building departments. Plastic piping is
easy enough for a home owner to install, thereby not needing a plumber.
Copper requires a little more expertise in soldering which not all home
owners have.
--
Bill - PSP and Media Center Plus Private Beta Tester
PSP Terrorist - D'Lanok de Caresk chapter - Anti-Troll Unit 235
"If you're not making waves, you're not underway!"
______________________________________________________
The Paint Shop Pro 7 Style Palette:
http://www.frontiernet.net/~willshak/style_palette/
______________________________________________________
The USS Salem, CA-139. The World's only preserved Heavy Cruiser
http://www.frontiernet.net/~willshak/salem/
Ember
wrote:
>On Tue, 26 Nov 2002 05:38:23 GMT, Ember <REMe...@operamail.com>
>wrote:
>
>>I think the reason
>>electrolysis is not a practial factor is that both metals are at the
>>same voltage potential -- being buried in the ground, whereas in a
>>house, pipe leaks can happen after just a few years at the junction of
>>copper and steel.
>
>In fact it is a factor which is why we now use sacrificial anodes at
>main stops with cast iron water mains. The trend today is to use
>plastic pipe for most new water mains though.
>
You are probably correct that sacrificial anodes are now used.
Technical standards (and taxes to support them) always creep upwards.
I don't have any municipal experience but have a gut feel the practice
might vary with municipality. I live in an area of large rainfall and
acidic soil where to obtain true zero ground potential is common. If
two dissimilar metals are surrounded by a common zero potential
conducting medium, then the flow of electric current between them,
causing electrolysis, must be greatly diminished. Just from observing
my area for 25 years I know that failure of the corporation cock right
at the pipe is rare. At least that's the first time I've ever seen the
crews excavating right at the main itself other than for a hookup to a
re-construction. Usually the failure is between the main and the
house.
But this thread has got me curious. I'm going to drop into the
engineering department and see ask about their current practises.
Ember
Ember
<ily...@cacr.ioc.ac.ru> wrote:
>
>"Kris Zaklika" <kzak...@jasc.com> wrote in message
>news:3DE3101E...@jasc.com...
>> Ember wrote:
>> >.....
>>
>> Not bad. I'd say fourth power of the radius. Both my son's
>> science project and the plumbing handbook agreed. I should,
>> of course, have specified straight pipe since the effective
>> size dependence is influenced by bends, joins and junctions
>> in the pipe. I forget the dependence on pipe material. And,
>> of course, the flow has to be sufficiently slow to avoid
>> cavitation. Interesting stuff, plumbing.
>
>Huh, "cavitation"? Here we normally refer to "turbulence". That is, when
>flow changes from laminar to turbulent, the dependence of pipeline
>"resistance" from Reynolds number changes and you can't directly extrapolate
>old "laminar" data onto it. Cavitation, i.e. partial evaporation, is rarely
>discussed because it is avoided by all means possible :-) - before it may
>happen in the pipeline, it is expected to happen in the pump, significantly
>decreasing its efficiency and probably damaging the "underloaded" engine.
>
True, but to clarify: Cavitation and the metal erosion / pitting it
causes is most likely to happen in the pump, or specialised device
like a flow-measuring venturi. It is a localised phenomenon. Elsewhere
in a closed system the water pressure is high enough to keep the water
in a liquid, non-gaseous state. Cavitation is avoidable by correct
system design.
Turbulence in pipe flow is the norm for pipes over about one inch
diameter and does not damage the pipe. Turbulence is cross eddies in
the water stream as opposed to parallel laminar flow. The cross eddies
use up pump energy but that energy loss is simply built into the
system, just accepted, because turbulence can't be avoided.
Ember
wrote:
>On Tue, 26 Nov 2002 11:15:00 -0500, Bill Schnakenberg
><will...@frontiernet.net> wrote:
>......
>(typically Blue Brute up to 30"ID) is being used more and more. AFAIK
>copper is still the product of choice for running the burried line
>from the main stop to the house ( I have never seen plastic pipe used
>for buried service lines), plumbers don't work on these as a rule but
>Ron
An advantage of copper pipe is that it also provides very good
electrical grounding. The house service neutral is connected to the
incoming copper pipe and in many homes this provides the only true
grounding, since ground rods or ground plates in dry soil aren't very
effective. They often end up under the eaves or the homeowner builds a
shed or something over the grounding so there is no moisture.
At least the copper house feed is going to be more than the length of
two ground rods, and if the copper connects to a metallic water line,
that's a bonus. Copper may even be a code requirement for this safety
reason but I don't know.
marjory
Thanks for all the helpful suggestions - I've been looking at my
original photos and the things I did that contributed to the problem.
First, the photos from the camera really are greyscale, with 256
colors, but when I open them in PSP I'm working in 16 million colors
so it opens the door for more to be added.
Here are some things that happen if I don't first convert a photo to
greyscale in PSP:
- using FotoCrop to crop and resize to a 5x7 reduces the color count
(ie, to 64)
- simply "saving as" the FotoCrop image as a jpg increases the count
substantially (ie, to 6218)
- saving as a .tif file after that results in the same larger count.
(I used it for the final version that was uploaded to a printing
service.) If I save as a tif first, it keeps the lower color count.
(Using Save Copy As here would help.)
- using Clarify adds colors
A big part of the problem with this photo is the large uniform area of
water and sky. The bands are due to jpg compression in the original
photo, so in addition to taking higher resolution photos for special
prints, I need to set it for lower compression. The ones I had done
earlier this year had more detail, so this wasn't as noticeable. Even
in greyscale, the bands look colored on my monitor, so I probably need
to adjust it. On a laptop it looks better. The JPEG Artifact Removal
tool helps a lot to smooth out the bands, by the way.
Probably working on a photo while zoomed out to 1:3 doesn't help
either. I do it because it's easier to see the preview while using
various adjustment tools that way, and the image on my screen is about
the 5x7" size that the final print will be, so I get a better idea of
the effect I'm creating.
Thanks again for your help - I think I'll be able to do much better
the next time I want black & white prints from my photos.
Marjory
Uni
It seems a lot of people struggle achieving decent looking B&W prints
from HP printers.
Try an Epson, and you'll be the next Ansel Adams :-)
Uni
> Marjory
Uni
1. Buy a real color printer.
2. Get the most recent drivers for your printer.
3. Use the GIMP instead.
Uni
mar...@nycap.not.here.rr.com (marjory) wrote in message news:<3de259ad...@news-server.nycap.rr.com>...