I've been trying the following process for B&W image creation.
http://www.greenapple.com/~rcoutant/tut ... 0white.htm
The process seems to work well until I get towards the bottom when I see the line that says
"I then extracted the V channel from the modified color image".
How do I extract the V channel?
Extract V channel
Moderator: jsachs
Re: Extract V channel
Color ---> Extract channel ---> select "V"
The process described there works well for images that have good color contrast but only weak brightness contrast where you'd like to see contrast in the B&W, but there are many other ways of approaching B&W. Sometimes, extracting just the R, G, or B channel makes a good starting point -- I like to look at those plus luminance plus the V channel just to see what's already available within the data. Den gives his preferences on his web site. Best thing is to familiarize yourself with a variety of methods so that you can then make the appropriates choices for a particular image.
The process described there works well for images that have good color contrast but only weak brightness contrast where you'd like to see contrast in the B&W, but there are many other ways of approaching B&W. Sometimes, extracting just the R, G, or B channel makes a good starting point -- I like to look at those plus luminance plus the V channel just to see what's already available within the data. Den gives his preferences on his web site. Best thing is to familiarize yourself with a variety of methods so that you can then make the appropriates choices for a particular image.
Bob Coutant
-
- Posts: 227
- Joined: November 24th, 2009, 2:00 am
- What is the make/model of your primary camera?: Fuji X-Pro 2
- Contact:
Re: Extract V channel
Picture Window Pro | Transformations | Color | Extract channel | HSV Value
-
- Posts: 856
- Joined: April 25th, 2009, 6:33 pm
- What is the make/model of your primary camera?: Canon EOS-350D/Fuji X100T
- Location: Birch Bay near Blaine, WA USA
Re: Extract V channel
Some additional ways to produce a monochrome image based upon its HSV-V channel in addition to the already twice mentioned Extract transform which will produce an 8 or 16-bit BW image...
(1) There are two methods using the Mask Tool that produces an 8-bit BW image:
___(a) Apply/OK the default diagonal Brightness Curve [0,0], [100,100]. Note: the Mask Tool-Brightness Curve is based upon the image's HSV-V histogram distribution.
___(b) Apply/OK the Color Range tool in the HSV Color Space: move the H white pointers facing each other to the extreme left and right ends of the color bar, the black pointers should follow; move the S white pointers facing each other to the extreme left and right ends of the color bar, the black pointers should follow; and locate the V white pointers so that they overlap at the extreme right end of the color bar, drag the right facing black pointer to the extreme left end of the color bar.
The following will produce a BW image, either 24 or 48 bit color depending upon the bit depth of the Input image where R=G=B=HSV-V...
(2) Apply/OK a Levels & Color transform in the HSV Color Space with the Saturation = -100%; Amount = 100%; and Amount Mask: None.
(3) Apply/OK a Color Curves transform in the HSV Color Space where the HSV-S default diagonal curve [100,100] control point has been relocated to [100,0]; Amount = 100%; and Amount Mask: None.
(4) Apply/OK a Saturation transform in the HSV Color Space; Preserve: Neither; Amount = -100%; and Amount Mask: None.
I only mention the above to perhaps provide further understanding as to how other transforms and the Mask Tool in PWP function and provide image data access/manipluation depending upon their Color Space models... ...not so much to say any method is better than another... ...after all the HSV-V tone distribution is the HSV-V tone distribution and if 24 or 48-bit color is preferenced where R=G=B=HSV-V, one could simply Convert transform an 8 or 16-bit BW HSV-V image...
Incidently, there are times I like to use the (3) conversion to BW, and adjustments to the HSV-V curve to obtain minor preference changes to shadows, mid-tones, and highlights... ...all-in-one go.
(1) There are two methods using the Mask Tool that produces an 8-bit BW image:
___(a) Apply/OK the default diagonal Brightness Curve [0,0], [100,100]. Note: the Mask Tool-Brightness Curve is based upon the image's HSV-V histogram distribution.
___(b) Apply/OK the Color Range tool in the HSV Color Space: move the H white pointers facing each other to the extreme left and right ends of the color bar, the black pointers should follow; move the S white pointers facing each other to the extreme left and right ends of the color bar, the black pointers should follow; and locate the V white pointers so that they overlap at the extreme right end of the color bar, drag the right facing black pointer to the extreme left end of the color bar.
The following will produce a BW image, either 24 or 48 bit color depending upon the bit depth of the Input image where R=G=B=HSV-V...
(2) Apply/OK a Levels & Color transform in the HSV Color Space with the Saturation = -100%; Amount = 100%; and Amount Mask: None.
(3) Apply/OK a Color Curves transform in the HSV Color Space where the HSV-S default diagonal curve [100,100] control point has been relocated to [100,0]; Amount = 100%; and Amount Mask: None.
(4) Apply/OK a Saturation transform in the HSV Color Space; Preserve: Neither; Amount = -100%; and Amount Mask: None.
I only mention the above to perhaps provide further understanding as to how other transforms and the Mask Tool in PWP function and provide image data access/manipluation depending upon their Color Space models... ...not so much to say any method is better than another... ...after all the HSV-V tone distribution is the HSV-V tone distribution and if 24 or 48-bit color is preferenced where R=G=B=HSV-V, one could simply Convert transform an 8 or 16-bit BW HSV-V image...
Incidently, there are times I like to use the (3) conversion to BW, and adjustments to the HSV-V curve to obtain minor preference changes to shadows, mid-tones, and highlights... ...all-in-one go.
-
- Posts: 856
- Joined: April 25th, 2009, 6:33 pm
- What is the make/model of your primary camera?: Canon EOS-350D/Fuji X100T
- Location: Birch Bay near Blaine, WA USA
Re: Extract V channel
Just curious...couman wrote:
...but there are many other ways of approaching B&W. Sometimes, extracting just the R, G, or B channel makes a good starting point -- I like to look at those plus luminance plus the V channel just to see what's already available within the data...
...Has anyone considered using the Stack Images transform as a channel mixer for black/white "Extract"-ed Luminance, R, G, B, V, and L channels of an image?...
Extraordinary subtleties and nuances are possible with the Brightness Curve/Density Mask adjustments and Amount [with/without a mask] that can be made to each Input image... ...then weighted pixel averaged with the other Input images... ...then a final Brightness Curve adjustment...
Of course, you do not need to use all six Input images, just 0% Amount the non-preference channels after "Preview"-ing them... ...it is wise to take notes as you go or one can possibly loose track of which Input image is which and what has been done to it... ...speaking from experience!
Edit1: a resulting BW conversion example using StackImages if of interest: http://www.ncplus.net/~birchbay/12/IMAG ... _12_91.htm
Last edited by den on September 10th, 2012, 9:49 am, edited 1 time in total.
Re: Extract V channel
:):) Even with 2-3 channels, the variations in output can be confusing.one can possibly loose track of which Input image is which and what has been done to it... ...speaking from experience!
As noted above, I usually look at all channels to see what's available within the starting image. Frequently, two channels are close to what I want, and then it is a simple matter to "combine" them with the stack images transformation with adjustments for the preferred output.
Bob Coutant
Re: Extract V channel
Den,
Can you elaborate on your mountain scene workflow?
Thanks,
Bruce
Can you elaborate on your mountain scene workflow?
Thanks,
Bruce
-
- Posts: 856
- Joined: April 25th, 2009, 6:33 pm
- What is the make/model of your primary camera?: Canon EOS-350D/Fuji X100T
- Location: Birch Bay near Blaine, WA USA
Re: Extract V channel
For the mountain scene [http://www.ncplus.net/~birchbay/12/IMAG ... _12_91.htm ], a StackImages BW conversion:
(1) the Luminance, R, G, B, V, and L channels were "Extract"-ed from a preferentially toned/colored/sharpened 48-bit image and were respectively the "StackImages"' Image 1; Image 2; ... Image 6 input images...
![Image](http://www.ncplus.net/~birchbay/tutorials/ImgBlend-ToneMap01/Extract_channels.jpg)
(2) the StackImages transform initial settings were 'default' with Auto Preview checked and the Final Shaping Curve dialog open with its default diagonal curve... ...this dialog's histogram will update along with the Preview as changes are made, aiding in analysis and setting/curve selections.
(3) "Shift-left mouse click" on each input Image's 1 thru 6 "Select Amount Mask" button while observing Preview and Final Shaping Curve histogram to begin the analysis and Image settings...
Mentally select the scene's visual elements of interest [sky/clouds; mountain/beams; mid-distance trees/lake; foreground grass/flowers] and observe the effects of setting/curve changes have in the Preview and histogram for these visual elements.
(4) for the mountain scene after several itierations of (3), final StackImages settings were:
___Image 1 Luminance: BrightnessCurve=smooth [0,0], [40,35], [80,90], [100,100]; Amount 50% with no mask.
___Image 2 R: Amount 75% with no mask.
___Image 3 G: Amount 100% with no mask.
___Image 4 B: Amount 25% with no mask.
___Image 5 V: Amount 100% with no mask.
___Image 6 L: Amount 0% with no mask.
___Final Shaping Curve=smooth [0,0], [15,15], [50,55], [100,100].
![Image](http://www.ncplus.net/~birchbay/tutorials/ImgBlend-ToneMap01/Final_SI_settings.jpg)
(5) click Apply or OK [note if you believe the final settings are generic for other images, the SI settings can be 'saved', then 'load'-ed at a later time].
Nothing here is really all that technical... ...not like using StackImages for manual image exposure blending and 'tone range' mapping... ...usually rather rudimentary mid-tone brightening or darkening contrast-S curves and Image Amounts without masks; all with out any Density Masks...
Have fun exploring... the more one uses Stack Images, the less confusing the remarkable, "all-in-one-go" image data access and manipulation becomes...
...den...
(1) the Luminance, R, G, B, V, and L channels were "Extract"-ed from a preferentially toned/colored/sharpened 48-bit image and were respectively the "StackImages"' Image 1; Image 2; ... Image 6 input images...
![Image](http://www.ncplus.net/~birchbay/tutorials/ImgBlend-ToneMap01/Extract_channels.jpg)
(2) the StackImages transform initial settings were 'default' with Auto Preview checked and the Final Shaping Curve dialog open with its default diagonal curve... ...this dialog's histogram will update along with the Preview as changes are made, aiding in analysis and setting/curve selections.
(3) "Shift-left mouse click" on each input Image's 1 thru 6 "Select Amount Mask" button while observing Preview and Final Shaping Curve histogram to begin the analysis and Image settings...
Mentally select the scene's visual elements of interest [sky/clouds; mountain/beams; mid-distance trees/lake; foreground grass/flowers] and observe the effects of setting/curve changes have in the Preview and histogram for these visual elements.
(4) for the mountain scene after several itierations of (3), final StackImages settings were:
___Image 1 Luminance: BrightnessCurve=smooth [0,0], [40,35], [80,90], [100,100]; Amount 50% with no mask.
___Image 2 R: Amount 75% with no mask.
___Image 3 G: Amount 100% with no mask.
___Image 4 B: Amount 25% with no mask.
___Image 5 V: Amount 100% with no mask.
___Image 6 L: Amount 0% with no mask.
___Final Shaping Curve=smooth [0,0], [15,15], [50,55], [100,100].
![Image](http://www.ncplus.net/~birchbay/tutorials/ImgBlend-ToneMap01/Final_SI_settings.jpg)
(5) click Apply or OK [note if you believe the final settings are generic for other images, the SI settings can be 'saved', then 'load'-ed at a later time].
Nothing here is really all that technical... ...not like using StackImages for manual image exposure blending and 'tone range' mapping... ...usually rather rudimentary mid-tone brightening or darkening contrast-S curves and Image Amounts without masks; all with out any Density Masks...
Have fun exploring... the more one uses Stack Images, the less confusing the remarkable, "all-in-one-go" image data access and manipulation becomes...
...den...
Re: Extract V channel
Den,
Thanks once again for a detailed post of what appears to be a really good technique. I can think of several images of mine that I want to experiment with...thanks again!
Bruce
Thanks once again for a detailed post of what appears to be a really good technique. I can think of several images of mine that I want to experiment with...thanks again!
Bruce