Rhonda Week 3 Cathedral 1

This is the discussion thread for the September 2010 Class.
mikemeister_admin
Posts: 4927
Joined: Fri Sep 20, 2013 8:29 pm

Postby mikemeister_admin » Wed Oct 20, 2010 8:31 pm

I dispatched this one pretty quickly.  I see the impact of moving the blue slider from 0/0 in toward the right.  The impact (just in playing around) in moving the top end (255) to the left (thus making the curve more steep) was less intuitive.  It brought more cyan into the image.  What is the description/definition of the effect of sliding the top of the curve from the top right 255 to the left?  I just want to know.  In otherwords, I don't understand how the numbers work in RGB on individual channels.  You'd think this would be a no brainer, but to me it's not quite clear.
Attachments
cathedral-blue-shadow-cast-screenshot-rgb-jpg
cathedral-blue-shadow-cast-screenshot-rgb-jpg (999 Bytes) Viewed 11293 times
cathedral-1-in-photoshop-jpg
cathedral-1-in-photoshop-jpg (999 Bytes) Viewed 11293 times

ggroess
Posts: 5342
Joined: Wed May 24, 2006 2:15 am
Contact:

Postby ggroess » Thu Oct 21, 2010 1:47 pm


I dispatched this one pretty quickly.  I see the impact of moving the blue slider from 0/0 in toward the right.  The impact (just in playing around) in moving the top end (255) to the left (thus making the curve more steep) was less intuitive.  It brought more cyan into the image.  What is the description/definition of the effect of sliding the top of the curve from the top right 255 to the left?  I just want to know.  In otherwords, I don't understand how the numbers work in RGB on individual channels.  You'd think this would be a no brainer, but to me it's not quite clear.


Rhonda,
Any time you move just the ends of the curve you either flatten or steepen the slope of the curve...Slope being the geometry property of the curve related to the math used to calculate the values on the curve line.  A steeper slope = more contrast a flatter slope = less contrast.  This can be global in the case of the master channel in RGB or it can be color specific in the case of an individual channel.  It can even be local in the case of a contrast pin adjusting only part of a curve.

For your case you increased the "slope"...Contrast...of just the blue channel.  This would have the effect of upsetting the color balance.  i.e. more cyan in the image... 

The numbers are for comparison and give you a measured difference.  Most times you move the end point of a curve to set a threshold value.  This is also true for channels.  In some more advanced work flows you can adjust single channels and then copy these channels into each other to adjust the contrast and lighting of the image in a luminosity mode layer...

For your image you have a bit of Cyan in the shadows still but rather than adjust the green or Blue channels you need to remember that CYAN is the color opposite of Red.  A slight tweak to the Red channel will fix the Cyan shadows...

See screen shots...

Please ask more questions if I did not fully answer yours this time...

Greg
Attachments
screenshot001-jpg-38
screenshot001-jpg-38 (78.98 KiB) Viewed 11293 times
screenshot002-jpg-30
screenshot002-jpg-30 (133.98 KiB) Viewed 11293 times

mikemeister_admin
Posts: 4927
Joined: Fri Sep 20, 2013 8:29 pm

Postby mikemeister_admin » Thu Oct 21, 2010 8:46 pm

Thanks Greg,

I didn't know about manipulating the red, because I still don't know how to think of color outside of the specific assignment.  How did you pick the points at which to manipulate the red?  I left my image the way I did, because (not withstanding the red channel) moving the blue any more added too much yell for my brain's eye. I'll mess with the red and see what I can find.

I understand that steepening the curve adds contrast.  I'm having trouble explaining what I want to know.  For instance, in RGB, if you use the threshold slider to find the highlight point, the RGB numbers on hue clocks change as you slide the red line to the left.  However, the curves don't change on the R, G or B channels.  Why don't the individual curves move regardless of what you do on the master curve in CM?

Another issue--what would I have to do to the RGB/Master channels to get a B&W image?  I know how to do that in LAB, but I assume that, since color and lightness are linked in RGB, there is no way to desaturate an image without it being black, gray or white.  Right?

I've been playing around with the ends of the sliders in RGB for about an hour in both PS and CM and I still can't get it through my head how the up and down numbers relate to the right and left numbers.  If you pick the upper right of a curve and slide it down along the edge, does that effect saturation or contrast or both?  If you pick the upper right of a curve and slide it left, does that effect saturation or contrast or both?  You have hue, saturation, brightness and contrast--all of which vary when changing a curve, right?  The curve is everything, right?  Oh my--out of my league.  Never mind! 

Finally, before I've set any points in RGB on the channels, if I move my mouse over the grids, I see all the numbers changing as I move my mouse.  But once I've set a point on any of the curves, the mouse no longer gives floating numbers in that curve, but rather is stuck at the point that I set.  I can float the numbers again only by setting another point and moving it around.  So if I want to find 63 on the curve I use the grids to guess at a starting point, dive in and set a point and start moving it from there until I get to 63.  This is the design of the program, right?  It works the same way in PS, so must be.

As always, thank you.  Don't take time to try to explain the inner life of RGB.  I'll look it up on the internet some time.

Rhonda

ggroess
Posts: 5342
Joined: Wed May 24, 2006 2:15 am
Contact:

Postby ggroess » Fri Oct 22, 2010 1:38 am


Thanks Greg,

I didn't know about manipulating the red, because I still don't know how to think of color outside of the specific assignment.  How did you pick the points at which to manipulate the red?  I left my image the way I did, because (not withstanding the red channel) moving the blue any more added too much yell for my brain's eye. I'll mess with the red and see what I can find.

I understand that steepening the curve adds contrast.  I'm having trouble explaining what I want to know.  For instance, in RGB, if you use the threshold slider to find the highlight point, the RGB numbers on hue clocks change as you slide the red line to the left.  However, the curves don't change on the R, G or B channels.  Why don't the individual curves move regardless of what you do on the master curve in CM?

Another issue--what would I have to do to the RGB/Master channels to get a B&W image?  I know how to do that in LAB, but I assume that, since color and lightness are linked in RGB, there is no way to desaturate an image without it being black, gray or white.  Right?


B&W is tricky in CM the best solution is to use LAB and slide the saturation slider to the left until the curves are horizontal.  I use a different technique..  In Photoshop I split the channels.  It is a command for PS in the channels pallet.  This creates 3 separate b&W images one from each color channel from the image and you can adjust each image separately and then layer them back together in any way you want.  It is a bit more work but it is much more flexible.



I've been playing around with the ends of the sliders in RGB for about an hour in both PS and CM and I still can't get it through my head how the up and down numbers relate to the right and left numbers.  If you pick the upper right of a curve and slide it down along the edge, does that effect saturation or contrast or both?  If you pick the upper right of a curve and slide it left, does that effect saturation or contrast or both?  You have hue, saturation, brightness and contrast--all of which vary when changing a curve, right?  The curve is everything, right?  Oh my--out of my league.  Never mind! 


I'll have something a bit later on this.it is not out of your league and in fact it might just make a bit more sense after I post more ..I am working on a separate issue that explains your question about the up down left right aspects. It might be Sunday before I answer that one fully.please be patient..As for the threshold question the Master channel is separate from the R, G, And B channels.  It is calculated differently.


Finally, before I've set any points in RGB on the channels, if I move my mouse over the grids, I see all the numbers changing as I move my mouse.  But once I've set a point on any of the curves, the mouse no longer gives floating numbers in that curve, but rather is stuck at the point that I set.  I can float the numbers again only by setting another point and moving it around.  So if I want to find 63 on the curve I use the grids to guess at a starting point, dive in and set a point and start moving it from there until I get to 63.  This is the design of the program, right?  It works the same way in PS, so must be.


You have selected a point and CM is waiting for you to move it...if you click on the grid... off the curve line... you will again see the numbers changing..yes it is by design and yes it is not obvious....

As always, thank you.  Don't take time to try to explain the inner life of RGB.  I'll look it up on the internet some time.

Rhonda

mikemeister_admin
Posts: 4927
Joined: Fri Sep 20, 2013 8:29 pm

Postby mikemeister_admin » Fri Oct 22, 2010 3:58 pm

I am more than happy to be patient--you have been so very patient with me and I am so grateful.  I love learning this stuff, but want to know more than a bunch of "tricks."  I don't mean that bad--it's my personality.  I had to take the Master Gardener course because I needed more than just the pieces.  I'm a "why" person.  A blessing and a curse!

Thanks, as always!
Rhonda

ggroess
Posts: 5342
Joined: Wed May 24, 2006 2:15 am
Contact:

Postby ggroess » Fri Oct 22, 2010 7:59 pm

Asking Why is not a curse...

I am happy to put more into it than the "tricks" level.  Most people want to know how more than they want to know why...

Greg

mikemeister_admin
Posts: 4927
Joined: Fri Sep 20, 2013 8:29 pm

Postby mikemeister_admin » Sat Oct 23, 2010 5:08 pm

You are so right about people wanting to know how rather than why.  I was a choral director in my former life and would often do workshops at conventions, etc.  One of my favorite topics was on the subject of using vocal warm-ups to improve tone quality.  Everyone always wanted handouts to add to their bag of tricks.  Even though I stressed what was happening with the breath or vocal cords that made the exercise effective, all they wanted was the activity.  They'd go home and try it once and, but since they didn't pay attention to the "why," they had no idea how to use the exercise in a program of overall voice building. No idea of how to go from specific (macro) to the global concept (macro).  Very frustrating.

I have loved learning the "how to" specifics, which DO generalize to other photos when the other photos have very similar problems.  I'm pretty sure that if I had taken the time to print out everything (which I still intend to do) and keep re-reading, I may have more of a gestalt that I feel like I do now.  I also know that there is only so much that can be covered in a six-week overview class.  We have covered a LOT and it is all great stuff.  But I'm still not able to connect the dots.  With you being at the master level, it may be obvious to you to mitigate the effects of one color by working on its opposite.  But to me, that concept is WAY out there off the chart of my realm of thinking.  So that's why I'm so persistent.  I want to gain the tools to work on color like a chef that knows to add onion rather than garlic to a recipe.  I have had the Margulis LAB for several years since one of our camera club members did a LAB workshop (that totally did not make sense).  I have not yet read it, but in glancing through it now, I can see that it will really be great in filling in the gaps in my LAB concepts.  I am glad to do outside reading, but I'm afraid that I won't comprehend anything that is too too heavy into math.  I'm not one of those musicians who has an affinity for numbers!

I don't want to take advantage of your kindness and generosity, so please just tell me that enough is enough when you've reached your tolerance!

Thanks so much.

Rhonda

ggroess
Posts: 5342
Joined: Wed May 24, 2006 2:15 am
Contact:

Postby ggroess » Sun Oct 24, 2010 2:26 pm

Rhonda,
one thing a teacher loves is a willing student...
The class and the forum are always there for you.  I hope you have read the PDF I posted earlier.  It does explain the Curve moves you asked about. 

As for the numbers...Take Art's advice and do not relate the numbers to anything "actual" until it is required.  For instance; If I tell  you to make a red dot but do not give you exact numbers how would you know what red is?? 

If I give you exact red values; and expect the dot to be that red; then you should hit that mark..otherwise they are just guideposts.  The big idea about the numbers is they tell you when things are right and when things are wrong...it is not the actual number but rather the relationship between them. 

As for the color opposites.  Think back to a color wheel and how we mix colors together in school   Red + Yelolow = Orange..that kind of stuff.  Well if you look at an RGB color wheel you will learn the opposites for RGB are CMY in the same order.  SO...

You can make a color cast conversion table if you like...
i.e.  Magenta Cast  = Green adjustment

How much adjustment??  that is where the numbers can help.

Greg

mikemeister_admin
Posts: 4927
Joined: Fri Sep 20, 2013 8:29 pm

Postby mikemeister_admin » Tue Oct 26, 2010 11:05 pm

Neat how you do the quotes in shadow.  I tried it, but it quoted the entire reply... so I use italics.

Here is my second try at this image. The first image is after shadow and highlight thresholds set and just blue channel moved to right.  The hue clocks show mostly green. 

The second image is when I lowered the upper right corner of the Red channel.  I like it better, but it did put more cyan back in the image. 

From your suggestion above, you placed control points on the Red Channel--how did you decide where to put those points?  It would be a shot in the dark for me.  Your points go from midpoint to somewhat high on the curve.  Is that a clue?

"You can make a color cast conversion table if you like...
i.e.  Magenta Cast  = Green adjustment"

How could I do this to experiment?  Sorry to not know.

Thanks!
Rhonda
Attachments
cathedral-1b-rgb-screenshot-jpg
cathedral-1b-rgb-screenshot-jpg (280.65 KiB) Viewed 11293 times
cathedral-1b-rgb-screenshot-lowering-red-jpg
cathedral-1b-rgb-screenshot-lowering-red-jpg (275.85 KiB) Viewed 11293 times
cathedral-1-rgb-in-photoshop-red-right-lowered-jpg
cathedral-1-rgb-in-photoshop-red-right-lowered-jpg (251.69 KiB) Viewed 11293 times

ggroess
Posts: 5342
Joined: Wed May 24, 2006 2:15 am
Contact:

Postby ggroess » Wed Oct 27, 2010 1:04 am


Neat how you do the quotes in shadow.  I tried it, but it quoted the entire reply... so I use italics.


You have to set the end quote at the end of the part you want to quote Shot1.
Then you have to re-add the first line of the quoted message with the message numbers back in front of the next part you want to quote.  Shot 2.


Here is my second try at this image. The first image is after shadow and highlight thresholds set and just blue channel moved to right.  The hue clocks show mostly green. 

The second image is when I lowered the upper right corner of the Red channel.  I like it better, but it did put more cyan back in the image. 

That is correct...and that is because...you lowered the overall red channel values making the cyan stand out more.


From your suggestion above, you placed control points on the Red Channel--how did you decide where to put those points?  It would be a shot in the dark for me.  Your points go from midpoint to somewhat high on the curve.  Is that a clue?


The selection of the points is not a guess.  I set the hue clocks on each flat of the church steeple.  The idea is that the hue should be consistent regardless of the brightness.  I noticed that the panel in the shadow was really different from the shadow parts of the the sunlit sides.  All things being equal the shadow should be near the same values as a shadow on the sunlight side.  So I had to move the curve near the shadow value to add more red so that it was closer to the sunlight values and less blue.  When you move the curve it does affect more than just the point you have moved.  I added the other points to bring the red curve back closer to the center line and keep the adjustment from bleeding across the image too far.


"You can make a color cast conversion table if you like...
i.e.  Magenta Cast  = Green adjustment"

How could I do this to experiment?  Sorry to not know.


See Shot3.  This color wheel shows all the printing and display primary colors. 
Colors that are directly across from each other are opposites. 
If an image is too much of one color you can adjust that color or it's opposite to get desired results.

Thanks!
Rhonda
Attachments
screenshot001-jpg-37
screenshot001-jpg-37 (12.13 KiB) Viewed 11293 times
screenshot002-jpg-29
screenshot002-jpg-29 (31.85 KiB) Viewed 11293 times
shot3-jpg-4
shot3-jpg-4 (236.16 KiB) Viewed 11293 times


Return to “Curvemeister 101 September 2010”

Who is online

Users browsing this forum: No registered users and 5 guests