CompuServe Thread

B&W to W&B

10 messages in this thread
#194653From: Stephen GrisezOct 9, 1995 10:23 PM
I have an image that is Black on White and I want to convert it to White on Black. This is usually a simple task if there are only 2 colors. But on the edges there are Shades of GREY. Will any of the Autodesk products invert not only the Black and White colors but also the Shades of GREY. How? Thanks, Stephen
#194811From: will taitOct 10, 1995 9:57 PM
>>>Will any of the Autodesk products invert not only the Black and White colors but also the Shades of GREY. How?<<< Animator Pro: Use the sep tool to turn the grays to either white or black, create a mask for either the white or the black. f mask is over white, use Fill to fill the black with white, invert the mask and again use fill to fill the original white with black. Done. Hope this helps. You could also use something like Paint Shop Pro. It is all in the manuals of both of these products. Will
#194882From: Stephen GrisezOct 11, 1995 9:13 AM
Will, Thanks for the input… but I am dealing with a 24bit image. Animator Pro is only 8 bit. Also switching the 2 black and white colors is easy. I am more concerned about the shades of grey. Stephen
#194974From: James Coulter[Mindscape]Oct 11, 1995 8:21 PM
> Will any of the Autodesk products invert not only the Black and White colors but also the Shades of GREY. How? < Certainly. Enter the 3DS Video Post queue and do the following: 1. Create an IFL (Image File List) of the files you wish to flip. 2. Set the total frames equal to the number of images you need to flip. 3. Make the first entry in the VP Queue a Solid Color: White. 4. Make the second entry in the VP Queue a Solid Color: Black. 5. Edit the Alpha slot for the second entry to be RGB Mask and enter the IFL from step 1 in the file slot. Make sure Intensity is ON and all other transparency options are OFF. Opacity threshold = 255. Negative = OFF 6. Render to files. The resulting files will be perfect grayscale flips of the originals. BTW, save the project file for use as a utility for future needs of this function. Cheers. — James — Choreo Motion —
#194978From: Stephen GrisezOct 11, 1995 8:42 PM
James, Thanks for the info. I'll give it a try. Are you still using OS/2? Stephen
#195012From: James Coulter[Mindscape]Oct 12, 1995 1:45 AM
> Are you still using OS/2? < Nope. Now that I have enough machines to separate office and production matters, I no longer have a need for OS/2. I am aclimating to NT, however, in preparation for 3DS MAX. — James — Choreo Motion —
#195691From: Stephen GrisezOct 16, 1995 11:23 PM
James, Thanks for the conversion tip. It worked great! One more question… I am working with my new white on black with grey edges flc in animator pro (8 bit). Shouldn't I get better compositing results if, when I am done, I re-render my flc as tga files via 3ds? If this is correct then do I have to do anything else besides load the flc in the background and then set my USE ALPHA to ON and tga depth 24? Thanks again, Stephen
#195783From: James Coulter[Mindscape]Oct 17, 1995 3:45 PM
Stephen, > Shouldn't I get better compositing results if, when I am done, I re-render my flc as tga files via 3ds? < Why are you using FLCS to begin with? For previewing? To answer you're question, an 8-bit FLC should _theoretically_ have exactly the same fidelity as a grayscale TGA stream. I would offer that TGA files are more predictable when arranged in IFL files as opposed to the same approach using FLCs. > If this is correct then do I have to do anything else besides load the flc in the background and then set my USE ALPHA to ON and tga depth 24? < You've lost me here. Without knowing exactly what you're trying to accompish, I can't provide an accurate answer. In it's simplist form, a composite reqires three sets of data: An "A" roll of images, a "B" roll, and a "Matte" roll. The Matte roll controls where the portions of the "B" roll overlay "A" roll. If the Matte roll is a series of black & white images, it's referred to as a Luminence Matte. If the Matte data is the embeded in a TGA file as the fourth channel, it's referred to as an Alpha channel matte (or simply Alpha matte). In the 3DS VP Queue, Each line is a "B" roll and the summation of the lines before it is the "A" roll. The Matting of "B" over "A" is controlled by the [ALPHA] slot (of the "B" roll entry) in the VP Queue. Using Queue Alpha uses Alpha data embeded in the TGA file. Using Queue RGB uses a set of images files for the matte roll. — James — Choreo Motion —
#195916From: Stephen GrisezOct 18, 1995 9:20 AM
James, >> Why are you using FLCS to begin with? For previewing? I have a logo that has handwriting. I am animating the text so that the script handwriting is being created over time about 6 seconds. What it amounts to is a lot of erasing pixels and fliping to the previous screen. This is a very simple but tedious task in Animator Pro. Start with the 1 image, the logo. -Erase 'some' of the last letter. -Hit the insert key -Erase 'some more' of the last letter. -Hit the insert key -etc. I believe this task would be much slower in Animator Studio so I chose to work with FLC files in Animator Pro. After a while, you have created a animation of handwritning being written. My concern was that I went from 24bit tga (my original logo) to an (8 bit) flc. I am concerned about the 'pseudo alpha' not working as good as 'alpha'. I understand that 'alpha' (which works on 24bit images) works for better compositing than 'pseudo alpha'(which works on 8bit gif images). This is why I believe that I must change my images back to TGA files, and then composit my handwriting logo onto some cool background. Is this correct? Thanks again, Stephen
#195967From: James Coulter[Mindscape]Oct 18, 1995 2:58 PM
The Alpha vs. Psuedo Alpha is sort of a non-issue for technical reasions. Here's why: A 24 bit TGA file has 8 bits per pixel for each of red, green, and blue colors respectively, hence 24 bits per pixel. Each set of 8 bits is called a Channel. Hence, an 8-bit red channel, an 8-bit green channel, and blue. A TGA file with an Alpha Channel is a 32 bit format. 24 bits per pixel for the three color channels RGB, and 8 more bits per pixel for Alpha data. These 8 bits for Alpha can be thought of as 256 grayscale levels. 0 (black) being transparent; 255 (white) being opaque, with a linear gradient of transparency in between. The information in the 8 bits are identical to an 8-bit grayscale FLC file. 8 bits is 8 bits no matter which way you look at it. As you already know, 3DS can either use the Alpha Channel of the TGA file or a user-supplied Psuedo Alpha as you call it. Providing the data is the same, there will be absolutely no difference in the finished composite. Using the FLC to create the Alpha Channels for a series of TGA files gains nothing in terms of quality. It _does_ make compositing simpler because the alpha data is carried with the TGA file. This is usefull if you expect to re-use the files often. Otherwise, it's an unnecessary additional step in a "one-off" animation. BTW, good erasing technique! You could use it with TGA files too. Erase and save, etc., etc., then create a reverse-numbered IFL file. — James — Choreo Motion —