Scanner/Histogram Question

Scanner/Histogram Question

Post by Ron » Wed, 05 Apr 2000 04:00:00



Hi,
Where is it better to adjust the histogram for B & W points after
doing the pre-scan at the scanner or making the adjustment in
Photoshop? Does it matter?
Thanks...Ron
 
 
 

Scanner/Histogram Question

Post by Ross » Wed, 05 Apr 2000 04:00:00



>Hi,
>Where is it better to adjust the histogram for B & W points after
>doing the pre-scan at the scanner or making the adjustment in
>Photoshop? Does it matter?
>Thanks...Ron

Adjust in your scanner software. You'd like to scan as many different
tonal values as possible, so by setting white and black you get the
full range.

Ross

 
 
 

Scanner/Histogram Question

Post by PFig » Thu, 06 Apr 2000 04:00:00


Quote:>Where is it better to adjust the histogram for B & W points after
>doing the pre-scan at the scanner or making the adjustment in
>Photoshop?

If your scanner software makes those adjustments internally on higher than 8
bit per channel data, then it's better to adjust in the prescan. If your
scanner can output the full data from the scanner (higher than 24 bit), then as
long as you don't clip data from the highlights and shadows, it won't make much
difference where you adjust.
 
 
 

Scanner/Histogram Question

Post by timo.. » Thu, 06 Apr 2000 04:00:00



Quote:> Where is it better to adjust the histogram for B & W points after
> doing the pre-scan at the scanner or making the adjustment in
> Photoshop? Does it matter?

It depends. If you scanner is able to ouput higher bit-depth images then
do so and edit in Photoshop. There is technically no difference in this
case other than that Photoshop has much better tools. Oh, when
converting down to 8-bit Photoshop applies random noise over the image,
you can get around this:
http://www.aim-dtp.net/aim/photoshop/v5/16-bit_noise.htm

If your scanner does not give higher bit-depth images then do a simple
test. Scan something that is very dark, twice:
(1) using range 0... 64.
(2) using the full range of the scanner, that is 0...255, and then scale
the histogram to 0...64 in Photoshop.

Now, if the histograms of the above images are similar then there is no
benefit in doing it in the scanner driver. Important: view the
individual red, green and blue channel histograms.

If the histograms of the image (1) have notably less gaps those of image
(2) then image (1) can be either of better quality or worse quality. The
extra data can be extra image info or it can be just extra noise.

To experiment this: Have the image (1) as a layer over the the image (2)
and then slowly click the eye icon in the layers palette on/off while
evaluating image quality.

Furter evaluation: Start recoding an action and edit one of the image
the best that you can do. Stop recording and then apply that action over
the other image. Again toggle the eye icon.

Cheap scanners apply the scaling over the digitized data, more expensive
devices adjust the exposure according to the scaling. Simple experiment
such as the above reveals this. Adjusting the exposure is beneficial in
that case where the signal to noise ratio of the scanner is high. In
case the S/N is low then it only gives more noise into the image.

Timo Autiokari

Sent via Deja.com http://www.deja.com/
Before you buy.

 
 
 

Scanner/Histogram Question

Post by Chris C » Fri, 07 Apr 2000 04:00:00




> > Where is it better to adjust the histogram for B & W points after
> > doing the pre-scan at the scanner or making the adjustment in
> > Photoshop? Does it matter?

> It depends. If you scanner is able to ouput higher bit-depth images then
> do so and edit in Photoshop. There is technically no difference in this
> case other than that Photoshop has much better tools. Oh, when
> converting down to 8-bit Photoshop applies random noise over the image,
> you can get around this:
> http://www.veryComputer.com/

Warning - the above URL leads to the website of a known net.kook.

Most people reading this newsgroup have put Timo in their killfiles because
of his constant misinformation.  He has been proven wrong many times on
many topics.  Unfortuantely he has enough time on his hands to make web
pages that look nice despite his total lack of knowledge on the subjects on
which he writes.

Mr. Autiokari seems to be of the opinion that anything he doesn't
understand must be a bug and that everyone who tries to explain his
mistakes is part of a vast * against himself.

If you do read his web sites, please consult other authoritative (and peer
reviewed) web sites such as:

    http://www.veryComputer.com/~poynton/GammaFAQ.html

    http://www.veryComputer.com/~poynton/ColorFAQ.html

    http://www.veryComputer.com/~ccox/gamma/index.html

    http://www.veryComputer.com/~poynton/notes/Timo/Concerning_Timo.html

Chris

PS. In his ongoing and every growing * theory, Timo now imagines
that I try to steer people away from his misinformation because of some
Photoshop 'bugs' also posted on his web site.  Say what?  Not only are they
NOT bugs (just more things he doesn't understand) but what the heck does
that have to do with his misinformation?  I just don't want to hear from
more users who have been fooled by Timo into destroying the quality of
their images.

 
 
 

Scanner/Histogram Question

Post by Timo Autiokar » Sat, 08 Apr 2000 04:00:00



>> http://www.veryComputer.com/
>Warning - the above URL leads to the website of a known net.kook.

Actually it leads to the new USA based site of the Accurate Image Manipulation
for Desktop Publishing (AIM), maintained by me, Timo Autiokari.

Quote:>Most people reading this newsgroup have put Timo in their killfiles because
>of his constant misinformation.

That is not so. The sole reason why I moved the AIM to the States is the very
high traffic, growing all the time. It is now very fast, even from Finland.

Quote:>He has been proven wrong many times on many topics.

There is plenty of info on the site, please point out even one single issue
that is wrong there.

Quote:>Unfortuantely he has enough time on his hands to make web pages that
>look nice despite his total lack of knowledge on the subjects on which
>he writes.

Well, I trust that the readers can make their own conclusion about that,
instead of believing a cheeky Adobe programmer, freshman at it. Is it so that
you are now out from the Photoshop team?

Quote:>If you do read his web sites, please consult other authoritative (and peer
>reviewed) web sites such as:

>    http://www.veryComputer.com/~poynton/GammaFAQ.html
>    http://www.veryComputer.com/~poynton/ColorFAQ.html
>    http://www.veryComputer.com/~ccox/gamma/index.html
>    http://www.veryComputer.com/~poynton/notes/Timo/Concerning_Timo.html

Mr. Poynton pages are pure lies. Your pages really do not have much of weight
nor content. If any of the above are peer reviewed, the result of that review
most probably was negative since there is no mention about the reviewers nor
about the result of the review on those pages.

Quote:>PS. In his ongoing and every growing * theory,

I really can not understand what it is that makes you think this, is there
something like that on my pages or in my posts?

Quote:>Timo now imagines that I try to steer people away from his misinformation
>because of some Photoshop 'bugs' also posted on his web site.

So, you would very much like to have this page removed:
http://www.veryComputer.com/
It shows the root cause why there is, always, tonal error between legacy files
made with 4.01 or earlier version compared to v. 5.0 and newer. Images made
with 4.01 or earlier version appear either darker or brighter in 5.0 and
later, when both versions are calibrated exactly into the same gamma space. I
will not remove that page, no matter how much and often you spit over me here.

Quote:>Say what?  Not only are they NOT bugs (just more things he doesn't
>understand) but what the heck does that have to do with his misinformation?  
>I just don't want to hear from more users who have been fooled by Timo into
>destroying the quality of their images.

So please point out any misinformation and it will be corrected at that very
moment. Maybe you could start from:
http://www.veryComputer.com/
That intro page demonstrates one of the most damaging but not well known bug
in the Levels -dialog, but that aside, please explain if you detect any
misinformation. In addition to the main purpose of the gallery, evaluation of
gamma induced errors, you can as well use those demonstrations trying to find
the dreaded "banding issue" that there should be in the linearly calibrated
workflow.

Timo Autiokari

 
 
 

Scanner/Histogram Question

Post by Ross » Sat, 08 Apr 2000 04:00:00


People have already wasted enough time trying to "point out any
misinformation" to you. Anybody interested should follow the links
provided by Chris. Determine for yourself where the lies are.

Ross




>>> http://www.veryComputer.com/

>>Warning - the above URL leads to the website of a known net.kook.

>Actually it leads to the new USA based site of the Accurate Image Manipulation
>for Desktop Publishing (AIM), maintained by me, Timo Autiokari.

>>Most people reading this newsgroup have put Timo in their killfiles because
>>of his constant misinformation.

>That is not so. The sole reason why I moved the AIM to the States is the very
>high traffic, growing all the time. It is now very fast, even from Finland.

>>He has been proven wrong many times on many topics.

>There is plenty of info on the site, please point out even one single issue
>that is wrong there.

>>Unfortuantely he has enough time on his hands to make web pages that
>>look nice despite his total lack of knowledge on the subjects on which
>>he writes.

>Well, I trust that the readers can make their own conclusion about that,
>instead of believing a cheeky Adobe programmer, freshman at it. Is it so that
>you are now out from the Photoshop team?

>>If you do read his web sites, please consult other authoritative (and peer
>>reviewed) web sites such as:

>>    http://www.veryComputer.com/~poynton/GammaFAQ.html
>>    http://www.veryComputer.com/~poynton/ColorFAQ.html
>>    http://www.veryComputer.com/~ccox/gamma/index.html
>>    http://www.veryComputer.com/~poynton/notes/Timo/Concerning_Timo.html

>Mr. Poynton pages are pure lies. Your pages really do not have much of weight
>nor content. If any of the above are peer reviewed, the result of that review
>most probably was negative since there is no mention about the reviewers nor
>about the result of the review on those pages.

>>PS. In his ongoing and every growing * theory,

>I really can not understand what it is that makes you think this, is there
>something like that on my pages or in my posts?

>>Timo now imagines that I try to steer people away from his misinformation
>>because of some Photoshop 'bugs' also posted on his web site.

>So, you would very much like to have this page removed:
>http://www.veryComputer.com/
>It shows the root cause why there is, always, tonal error between legacy files
>made with 4.01 or earlier version compared to v. 5.0 and newer. Images made
>with 4.01 or earlier version appear either darker or brighter in 5.0 and
>later, when both versions are calibrated exactly into the same gamma space. I
>will not remove that page, no matter how much and often you spit over me here.

>>Say what?  Not only are they NOT bugs (just more things he doesn't
>>understand) but what the heck does that have to do with his misinformation?  
>>I just don't want to hear from more users who have been fooled by Timo into
>>destroying the quality of their images.

>So please point out any misinformation and it will be corrected at that very
>moment. Maybe you could start from:
>http://www.veryComputer.com/
>That intro page demonstrates one of the most damaging but not well known bug
>in the Levels -dialog, but that aside, please explain if you detect any
>misinformation. In addition to the main purpose of the gallery, evaluation of
>gamma induced errors, you can as well use those demonstrations trying to find
>the dreaded "banding issue" that there should be in the linearly calibrated
>workflow.

>Timo Autiokari

 
 
 

Scanner/Histogram Question

Post by Alan Walk » Sun, 09 Apr 2000 04:00:00


I hope the original subject of this post has not gone cold - I'd like
to ask a related question -

I have a Minolta Scan Dual film scanner and can't figure out what I'm
supposed to do with the histogram adjustment thingy. It has an input
scale with  3 sliders and an output scale with 2 and there 3 sets of
these - 1 for each colour. Now I think you're supposed to squash the
sliders up to cover in any gaps at the shadowy or highlights ends [you
see I'm well up with the jargon here] but what do you do with the
midtone slider?
But what I really can't work out is when do you use the input sliders
and when the output? What do the 2 things do differently? How should
they be set for optimal scanning?

The other thing with Timo's posts is their glorious
incomprehensibility - which, it seems, is probably just as well.

Cheers, Alan.

 
 
 

Scanner/Histogram Question

Post by Ross » Sun, 09 Apr 2000 04:00:00


Not everybody scans the same way. I've known knowledgeable,
experienced people who differ.

I try to get the full possible dynamic range by setting the left and
right Input sliders so that the darkest tone becomes black, and the
lightest tone becomes white. For color, I first determine if I believe
that white and black exist in the source, and then set each RGB
component to get maximum range. This gives a rough color correction
that can be fine-tuned in PS.

The middle Input slider is a subjective decision. You need to evaluate
your source and decide how much you want to boost (it's almost always
an increase for the stuff I scan) the midtones.

I never mess with the Output sliders. I do all other color correction
in PS, where I use Curves. The Output sliders in Levels can be used to
set your black and white points for printing. There is a minimum
highlight dot and a maximum shadow dot that is reproducible on a given
press. You should ask your printer what those values are. You can set
them in PS either in Levels, or by how you define your white and black
points in Curves.

Ross



Quote:>I hope the original subject of this post has not gone cold - I'd like
>to ask a related question -

>I have a Minolta Scan Dual film scanner and can't figure out what I'm
>supposed to do with the histogram adjustment thingy. It has an input
>scale with  3 sliders and an output scale with 2 and there 3 sets of
>these - 1 for each colour. Now I think you're supposed to squash the
>sliders up to cover in any gaps at the shadowy or highlights ends [you
>see I'm well up with the jargon here] but what do you do with the
>midtone slider?
>But what I really can't work out is when do you use the input sliders
>and when the output? What do the 2 things do differently? How should
>they be set for optimal scanning?

>The other thing with Timo's posts is their glorious
>incomprehensibility - which, it seems, is probably just as well.

>Cheers, Alan.

 
 
 

Scanner/Histogram Question

Post by Chris C » Sun, 09 Apr 2000 04:00:00





> >> http://www.aim-dtp.net/aim/photoshop/v5/ps5_rgb_calibration.htm

> >Warning - the above URL leads to the website of a known net.kook.

> Actually it leads to the new USA based site of the Accurate Image Manipulation
> for Desktop Publishing (AIM), maintained by me, Timo Autiokari.

As I said: the website a well known net.kook.

Chris

 
 
 

1. OpenGL Convolution & Histograms question

I've got a quick question concerning the convolution & histogram
extensions under the SGI implementation of OpenGL.  I'm using RE3
hardware under IRIX 6.2.  The routines don't seem much faster (if at
all) than the pure CPU counterparts that I coded up.  Does anyone know
if these two extensions actually use the graphics hardware or are they
implemented using the CPU instead.

Thanks in advance.

--
 Regards,
   Loren Knutson

2. Kalman Filter

3. Image histogram manipulation question...

4. drop shadow problem

5. Histogram Question

6. Converting Maya pictures to OpenGL

7. Histogram follow-up question for Rick Simon

8. Drivers??

9. PSP 6.0 Histogram question

10. CS Histogram Palette question

11. Histogram questions

12. Histogram question