Ximagic Graydither
Moderators: helmut, XnTriq, xnview
-
- Posts: 267
- Joined: Sat Apr 25, 2009 9:23 am
Re: Ximagic Graydither
Hello. Glad to hear from you. I agree with 16x16 box = more graylevels the banding dissapears. What I don't know is why when printing normally, there is no banding, even at 600dpi. Maybe the driver does a different and more focused resize/resample algorithm.
I have also found out that the resampling method gives totally different outputs. It seems Lanczos is a good one.
I have also found out that the resampling method gives totally different outputs. It seems Lanczos is a good one.
Intel i7 5960X
Gigabye GA-X99-Gaming 5
8 GB DDR4 (2100)
GeForce GTX 970 (Gigabyte)
Gigabye GA-X99-Gaming 5
8 GB DDR4 (2100)
GeForce GTX 970 (Gigabyte)
-
- Posts: 267
- Joined: Sat Apr 25, 2009 9:23 am
Re: Ximagic Graydither
Good day, Blas. Merry Christmas to you.
I have another question regarding grayscales: Do you know how many levels (shades) of gray do produce error-diffusion methods (floyd-steinberg for instance)?
When using grids, you already pointed the right formulas to me, but how are shades of gray calculated when using error-diffusion filters?
I have another question regarding grayscales: Do you know how many levels (shades) of gray do produce error-diffusion methods (floyd-steinberg for instance)?
When using grids, you already pointed the right formulas to me, but how are shades of gray calculated when using error-diffusion filters?
Intel i7 5960X
Gigabye GA-X99-Gaming 5
8 GB DDR4 (2100)
GeForce GTX 970 (Gigabyte)
Gigabye GA-X99-Gaming 5
8 GB DDR4 (2100)
GeForce GTX 970 (Gigabyte)
-
- Posts: 64
- Joined: Sun Feb 01, 2009 2:15 am
Re: Ximagic Graydither
Merry Christmas to you also.eL_PuSHeR wrote:Good day, Blas. Merry Christmas to you.

AFAIK the concept of shades of gray is not applicable to error-diffusion dithering nor has any equivalent.eL_PuSHeR wrote:I have another question regarding grayscales: Do you know how many levels (shades) of gray do produce error-diffusion methods (floyd-steinberg for instance)?
When using grids, you already pointed the right formulas to me, but how are shades of gray calculated when using error-diffusion filters?
-
- Posts: 267
- Joined: Sat Apr 25, 2009 9:23 am
Re: Ximagic Graydither
Thank you for your prompt reply. Somehow I was guessing it.
Best regards.
Best regards.
Intel i7 5960X
Gigabye GA-X99-Gaming 5
8 GB DDR4 (2100)
GeForce GTX 970 (Gigabyte)
Gigabye GA-X99-Gaming 5
8 GB DDR4 (2100)
GeForce GTX 970 (Gigabyte)
-
- Posts: 267
- Joined: Sat Apr 25, 2009 9:23 am
Re: Ximagic Graydither
I have another question regarding Graydither.
Can Hilbert-Peano curves (i.e. Riemersma) be considered some form of FM-Screening (stochastic) or are they a different thing?
Can Hilbert-Peano curves (i.e. Riemersma) be considered some form of FM-Screening (stochastic) or are they a different thing?
Intel i7 5960X
Gigabye GA-X99-Gaming 5
8 GB DDR4 (2100)
GeForce GTX 970 (Gigabyte)
Gigabye GA-X99-Gaming 5
8 GB DDR4 (2100)
GeForce GTX 970 (Gigabyte)
-
- Posts: 64
- Joined: Sun Feb 01, 2009 2:15 am
Re: Ximagic Graydither
I suppose your question is about hilbert-peano used in error diffusion dithering.eL_PuSHeR wrote:Can Hilbert-Peano curves (i.e. Riemersma) be considered some form of FM-Screening (stochastic) or are they a different thing?
When the halftoning algorithms are classified into FM, AM and hybrid FM-AM, the error diffusion is included into the FM class.
So the error diffusion, with or without hilbert-peano curves, is a kind of FM, at least for those who create this classification

-
- Posts: 267
- Joined: Sat Apr 25, 2009 9:23 am
Re: Ximagic Graydither
Again, thanks for clarifying things up.
Intel i7 5960X
Gigabye GA-X99-Gaming 5
8 GB DDR4 (2100)
GeForce GTX 970 (Gigabyte)
Gigabye GA-X99-Gaming 5
8 GB DDR4 (2100)
GeForce GTX 970 (Gigabyte)
-
- Posts: 267
- Joined: Sat Apr 25, 2009 9:23 am
Re: Ximagic Graydither
Howdy. Another question.
When creating 45º halftones from the built-in list, what mesh motive is used? I mean is the halftone made up of round dots, elliptical, square, or what?
Photoshop CS3's built-in halftone mode allow you to choose forms from Round, Elliptical, Square, Cross, Lines...
An on this page there is some info that might interest you.
Best regards.
When creating 45º halftones from the built-in list, what mesh motive is used? I mean is the halftone made up of round dots, elliptical, square, or what?
Photoshop CS3's built-in halftone mode allow you to choose forms from Round, Elliptical, Square, Cross, Lines...
An on this page there is some info that might interest you.
Best regards.
Intel i7 5960X
Gigabye GA-X99-Gaming 5
8 GB DDR4 (2100)
GeForce GTX 970 (Gigabyte)
Gigabye GA-X99-Gaming 5
8 GB DDR4 (2100)
GeForce GTX 970 (Gigabyte)
-
- Moderator & Librarian
- Posts: 6501
- Joined: Sun Sep 25, 2005 3:00 am
- Location: Ref Desk
Re: Ximagic Graydither
Thank you for the link! Very interesting indeedeL_PuSHeR wrote:An on this page there is some info that might interest you.

-
- Posts: 267
- Joined: Sat Apr 25, 2009 9:23 am
Re: Ximagic Graydither
You are welcome!.
Intel i7 5960X
Gigabye GA-X99-Gaming 5
8 GB DDR4 (2100)
GeForce GTX 970 (Gigabyte)
Gigabye GA-X99-Gaming 5
8 GB DDR4 (2100)
GeForce GTX 970 (Gigabyte)
-
- Posts: 64
- Joined: Sun Feb 01, 2009 2:15 am
Re: Ximagic Graydither
The halftones in the built in list, are "designed" patterns which comes from several sources, and it uses different design methods, usually with hand-made adjustments. The point shape is not selectable in those patterns.eL_PuSHeR wrote:When creating 45º halftones from the built-in list, what mesh motive is used? I mean is the halftone made up of round dots, elliptical, square, or what?
Photoshop CS3's built-in halftone mode allow you to choose forms from Round, Elliptical, Square, Cross, Lines...
Otherwise, the photoshop halftones screens what you talk about are calculated patterns, like the ones in the plug-in.
Very good reference, thanks.eL_PuSHeR wrote: An on this page there is some info that might interest you.
-
- Posts: 267
- Joined: Sat Apr 25, 2009 9:23 am
Re: Ximagic Graydither
Thank you very much. Currently I prefer CS3 because it allows me to select Lines Per Inch (Screen frequency).
Intel i7 5960X
Gigabye GA-X99-Gaming 5
8 GB DDR4 (2100)
GeForce GTX 970 (Gigabyte)
Gigabye GA-X99-Gaming 5
8 GB DDR4 (2100)
GeForce GTX 970 (Gigabyte)
-
- Posts: 267
- Joined: Sat Apr 25, 2009 9:23 am
Re: Ximagic Graydither
Has anybody tried this BKonvert program? It's in some weird language to me (Polish?) but it seems it has a lot of dithering algorithms.
Intel i7 5960X
Gigabye GA-X99-Gaming 5
8 GB DDR4 (2100)
GeForce GTX 970 (Gigabyte)
Gigabye GA-X99-Gaming 5
8 GB DDR4 (2100)
GeForce GTX 970 (Gigabyte)
-
- Moderator & Librarian
- Posts: 6501
- Joined: Sun Sep 25, 2005 3:00 am
- Location: Ref Desk
Re: Ximagic Graydither
Ximagic GrayDither Changelog wrote:
- v3.0.0 (29/Mar/2011)
- New UI design
- Added automatic saved previews
- Added 64 bit support for Mac
-
- Moderator & Librarian
- Posts: 6501
- Joined: Sun Sep 25, 2005 3:00 am
- Location: Ref Desk
Re: Ximagic Graydither
New: ColorDither (split from Quantizer)Changelog wrote:
v 3.1.0 28-Apr-2011v 3.0.2 19-Apr-2011
- Added smart filter support (CS3/CS4/CS5)
- Win- Patch- Editing numeric text field doesn't update associated updown control
- Win- Patch- Crash when closing the plug-in, using OK or Cancel, detected on several editors (PS CS5, GIMP, ...)
v 3.0.1 11-Apr-2011
- Win- Patch- Crash when closing the plug-in, if a progress bar appears after hitting OK.(detected on PS CS3 + XP)
- Win- Patch- Disable plug-in window when opening a child window (f.i. progress bar)
- Win- Patch- Disable double click on title bar
- Win- Patch- Catch plug-in errors and avoid host crash whenever possible
- Win- Use XP style for UI even if the plug-in host don't use it (f.i. GIMP)
- Win- Patch - Error in preview paint which could crash the plug-in
- Win- Tested with new hosts Paint.NET, NiGulp, RealWorld Paint.COM, ImageN.
- Win- Patch - Plug-ins without product version crash(f.i. Magix Photo Designer)
- Win- Patch - Workaround for bug in Magix PhotoDesigner when used with a selection mask
- Mac- Patch - Page scroll don't work as expected where smooth scroll is active