Created attachment 101316 [details] LibreOffice '.odg' file showing recommended color palette The current color palette used by LibreOffice is not reproducible across a range of display monitor brands, types and settings (e.g. LCD vs. CRT), printer brands, types and settings (e.g. laser vs. ink jet) and software type (e.g. native LibreOffice '.odg' vs. 'pdf'). Recommend using the standard color palette shown in the attached file: <RgbColorVectors_18JUN2014.odg> Page 1 shows the color palette; page 2 shows the mathematics used to generate the palette; page 3 shows two representative color wheels. The advantage of this color palette is that it is independent of display monitor, printer and software technology. This is the same color palette that many manufacturers use to test their products. It will never go obsolete so long as 256-bit RGB color vectors are used. The attached document has been placed into the public domain. Operating System: All Version: 4.1.6.2 release
enhancement request. status NEW
Perhaps same request as Bug 42159 or Bug 37480. Or hex color picker: Bug 44366
(In reply to comment #2) > Perhaps same request as Bug 42159 or Bug 37480. Or hex color picker: Bug > 44366 No, it's not the same request. This bug is regarding the default color palette.
(In reply to comment #0) > Created attachment 101316 [details] > LibreOffice '.odg' file showing recommended color palette > > The current color palette used by LibreOffice is not reproducible across a > range of display monitor brands, types and settings (e.g. LCD vs. CRT), > printer brands, types and settings (e.g. laser vs. ink jet) and software > type (e.g. native LibreOffice '.odg' vs. 'pdf'). > > Recommend using the standard color palette shown in the attached file: > <RgbColorVectors_18JUN2014.odg> Page 1 shows the color palette; page 2 > shows the mathematics used to generate the palette; page 3 shows two > representative color wheels. > > The advantage of this color palette is that it is independent of display > monitor, printer and software technology. This is the same color palette > that many manufacturers use to test their products. It will never go > obsolete so long as 256-bit RGB color vectors are used. That's true, and it's also incredibly easy to implement. However, I'm thinking the reason that it's not used is because the colors aren't very pleasant to look at (at least the fully saturated colors can be quite an eyesore). There's also a lack of light options. Those are just my initial observations, though -- it'd be good to see some research on this. Could you make an example document showcasing these colors? I'd be especially curious about how well they work in charts.
Created attachment 104569 [details] Version compatibility problems with color palette on OpenOffice The OpenOffice drawing file showing examples of how the color palettes changed between version 3.4.1 and 4.0.1.
Created attachment 104570 [details] Standard color palette that I've adopted for my documents This is the standard color palette that I've adopted for my documents. My documents are heavily used by electronic engineers, and so it's based on the electronic color code, where the numerals '0' - '9' are assigned a color.
Created attachment 104571 [details] A document example showing how I use colors This is a document in '.pdf' format that shows how I use my color palette.
@ Mirek2 "example document" The color palette I proposed is intended to solve device compatibility and software longevity issues. It's an objective compromise to a solve a problem that is inherently subjective to every person. It will always be necessary to allow a user to make his or her own colors. In fact, it is doubtful that any two graphic artists would agree on the best choice of colors. For example, the color brown is almost impossible to match between devices. If you have to use brown, then you will probably need to adjust the color mixture so that it looks good on your monitor. However, be prepared to have your 'brown' look more orange or red on other devices. I've attached some examples for your consideration: < OpenOfficeColorPalette_08DEC2013.odg > shows color palette compatibility problems between different versions of OpenOffice. < StandardColorPalette_06MAY2014.odg > shows the standard color palette that I've adopted in my line of work. I wouldn't expect others to like it, as it's intended to be used by electronics engineers. < WbBridge_28MAR2012.pdf > shows a '.pdf' example of a document that I created in my line of work. These documents sometimes take two or three years to complete, so any change of color palette during revision changes is annoying.
Created attachment 104623 [details] Spreadsheet for generating color vectors of a specified greytone A LibreOffice Calc '.ods' (spreadsheet) file for generating color vectors of a specified greytone. Instructions are provided inside the spreadsheet. This spreadsheet has been injected into the public domain.
Talking points about a color palette organized according to gray scale (a.k.a. gray tone, or luminance contrast): 1) In the human visual system, color recognition is secondary to gray scale recognition. The following quotes from Marmor and Ravin (2009) describe the role of color during visual processing in the brain: “Thus, the color information has already been segregated from the more basic (evolutionarily older) mechanisms of perception, which are effectively color-blind. The bottom line is that we need luminance (brightness) differences to optimally recognize shapes, faces, depth, and movement.” (p. 55) “To summarize, color alone will not function effectively to represent objects (or depth), unless it also provides realistic luminance contrast. Most representational artists recognize light and dark aspects of the paintings accordingly, such as in Raphael's Renaissance painting 'A Lady with Unicorn'.” (p. 57) 2) The most common form of color blindness is a red-green color deficiency, and is typically found in men. These people tend to see colors as shades of blue and yellow. According to Marmor and Ravin (2009): “Red-green color deficiency affects 8 to 10 percent of males, so it is not uncommon.” (p. 88) 3) People with a color deficiency tend to rely on gray scale color recognition. According to Marmor and Ravin (2009): “Because colors are less subjectively intense, the color-deficient person – and artist – tends to be more sensitive to subtle lights and darks which may be obscured by colors that distract our attention (as in camouflage).” (p. 89) “When asked about colors in the countryside, he [Jens, an artist with a red-green color deficiency] responded: 'Well mostly these are known, and my mind tells me one thing and my eyes another. Essentially I am seeing a lot of contrast. I like things with contrast in them.'” (p. 90) 4) Ansel Adams, a well-known American landscape photographer who worked exclusively with black-and-white images, relied on the 'Zone System' for his photographs. This system divided an image into eleven (11) gray tone levels between black and white. 5) Many people print color images on black-and-white printers. 6) Many technical artists design their images so that they can still be useful (recognizable) when printed on black-and-white printers. 7) A color palette organized according to a gray tone scale is the easiest way to mentally convert a color image to a monochrome image. This allows the graphic artist to more easily design color works that will still look good when printed on a black-and-white printer. Cite: Marmor, Michael F. and James G. Ravin. The Artists Eyes : Vision and the History of Art. Abrams, 2009. ISBN: 978-0-8109-4849-5
Regardless of whether this proposal becomes the default palette I do feel it would be good to have this tonal / standardised palette available. The percentage naming (rather than subjective naming) of hues is much clearer and important for the role this palette fulfills e.g., "64% Magenta" vs "Magenta 4". Excellent work Wade. Thank you for sharing. Very interesting. (In reply to comment #4) > There's also a lack of light options. I agree, but attachment 101316 [details] indicates a subtractive rather than additive approach to creating a palette i.e., emulation of CMYK using RGB. In simple terms this is why it works the way it does in practice. This is a specialised palette targeting interoperability that I feel would be a great asset to LO.
Could you create a .soc palette with 10 columns (see color picker design: https://redmine.documentfoundation.org/attachments/download/138/color-picker.png) based on your calculations? That way people could get a feeling for it and decide if it would work well as default palette.
(In reply to comment #12) > Could you create a .soc palette with 10 columns ... The colour picker was changed in v4.2 (bug 67653) to use 12 columns (to bring it in-line with AOO) and under v4.3 and v4.4 it still uses this layout. To get similar tonal values aligning in columns (for this specialised palette) will require additional filler swatches.
@Krisztian, you've been deeply into this code of late for your GSOC project. Any thoughts?
Created attachment 105105 [details] SOC example, using fillers for missing values (In reply to comment #13) > The colour picker was changed in v4.2 (bug 67653) to use 12 columns (to > bring it in-line with AOO) and under v4.3 and v4.4 it still uses this > layout. To get similar tonal values aligning in columns (for this > specialised palette) will require additional filler swatches. Here is a thrown together example SOC file. I have made the additional filler swatches (at bottom) as dark as possible to ensure they are clearly distinguishable from the above palette values. Each filler swatch is also named in an abbreviated manner for clarity. Note that instead of "Gray" I have used "White" for the monochromatic series as a pale grey being labelled as "90% White" seems clearer than "90% Gray". I have also placed the monochrome series at top, in order to make it easier to distinguish Black and the darker grey swatches from the filler swatches.
Created attachment 105106 [details] Screenshot of color picker using the provided SOC.
@ Owen Genat "screenshot of color picker" (In reply to comment #16) > Created attachment 105106 [details] > Screenshot of color picker using the provided SOC. I think you're close. It would work better if the grey scale tones between BLACK (0% grey) and WHITE (100% grey) were aligned with the color grey scale values. If you choose the same values that were in my chart it would mean 12 rows and 13 columns (or, alternatively 13 rows and 12 columns if you do in the other direction). My diagram uses twelve (12) color vectors plus grey, but you could go with 3, 6, 12, 24, 48, etc. You can see the idea better if you print out your color picker on a black and white laser printer. That's something akin to how a person with a color deficiency might see them, too. The lighter tones for all the colors should line up with the corresponding grey tone. That will give the color deficient person, as well as the graphic artists designing colors for B&W printers, a fighting chance to make their documents look good. In fact, I use the color palette to evaluate laser printers. For example, a Xerox Phaser 5500 produces a substantially different grey tone pattern (and color rendition) than a Lexmark M1145. I don't know for sure, but I suspect this might be because of their toner powder or their color calibration. Also, most of the laser printers are weighted toward the blue end of the scale. That means that blues turn out black. That's not done on purpose; its a problem for them -- a remnant of their technology. I'm looking at the standard color palette in much the same way as television color bars. In fact, I'll most a comparison so that you can see how that works.
Date: 22 AUG 2014 From: Wade D. Peterson, Silicore Corportion To: File Subj: Comparative analysis using television color bars Subject: comparative analysis between television color bars and the proposed grey scale color palette. Description: this is a comparative analysis. It describes how a television color bar standard is applied in the broadcast television industry. This is not a proposal; rather it's a way to compare and contrast the proposed grey scale color palette to a similar technology used in another field. Talking points: 1) Color bar test patterns are used in the broadcast television industry as a quality control tool for the purpose of calibrating people, equipment and systems. For example, they are used to align color television sets. 2) There is a Wiki at: http://en.wikipedia.org/wiki/Color_bars 3) An example is the EG 1:1990 “Alignment Color Bar Test Signal for Television Picture Monitors”. It is published by The Society of Motion Picture and Television Engineers. A free pdf archive document can be obtained by accessing: http://standards.smpte.org/, and then doing a search for “color bars”. [I can't post it myself on this thread because of copyright violations]. Find “EG 1:1990” in the list; you should be able to download it at no cost. Note: this is an older analog TV standard that they now make available at no charge. 4) Figure 2 of the EG 1:1990 shows the how the alignment color bars are documented in a standard. This standard is used for the NTSC television signal. 5) The exact colors such as “YELLOW”, “GREEN” or “RED” are not explicitly shown. Instead, they represent saturated electronic signals. 6) The color bars are generally broadcast in conjunction with a 1,000 Hz audio tone. This allows the volume level and audio distortion to be calibrated too. 7) In preparation for a television show, the 'bars and tone' (as they are known) are first broadcast from the studio. Technicians monitor the signal at various points along the signal path. This includes camera equipment; amplifiers; microwave links; cabling; television transmitters; receivers and TV sets. 8) The bars and tone are used in two different ways, depending on the equipment. A color bar ‘card’ is used to align television cameras. The camera is focused on the standard (printed) card, and the output signal is inspected with an oscilloscope to see if it’s capturing the correct colors in its output signal. A color bar ‘pattern generator’ is used to align the rest of the television network. The pattern generator produces a standard color signal, which is then connected to the TV network. Technicians throughout the network inspect their signals with an oscilloscope and calibrate their equipment accordingly. 9) The EG 1:1990 is copyrighted by The Society of Motion Picture and Television Engineers. 10) Colors themselves are not offered copyright protection in the United States, or the rest of the world under the Berne convention. 11) A pattern of colors is copyrightable (much like a color photograph) so long as it expresses an idea, and it takes some level of originality to create it. The Courts sometimes use the “sweat of the brow” test to determine this. That is, if it took some “sweat of the brow” to make it, then copyright protection is afforded. 12) The Society of Motion Picture and Television Engineers (SMPTE) has an intellectual property policy statement at: https://www.smpte.org/sites/default/files/SMPTE_IP_Policy_2013-08.pdf 13) Section 9.3 of the SMPTE policy is the Copyright Policy. It says: “The Society shall own the copyrights of all Engineering Documents and Registered Disclosure Documents, whether in draft or published form.”
(In reply to comment #17) > I think you're close. It would work better if the grey scale tones between > BLACK (0% grey) and WHITE (100% grey) were aligned with the color grey scale > values. If you choose the same values that were in my chart it would mean > 12 rows and 13 columns (or, alternatively 13 rows and 12 columns if you do > in the other direction). My diagram uses twelve (12) color vectors plus > grey, but you could go with 3, 6, 12, 24, 48, etc. Thanks for clarifying the naming of the monochrome scale. It is not going to be possible to change the layout of the colour picker from 12 to 13 columns, so we are going to have to use a row-per-hue (rather than column-per-hue) layout for this palette if all the scales are to run in the same direction. This is actually simpler for palette creation as a hue > tone (rather than tone > hue) order in XML entries makes more sense IMO. Most palettes tend to have a greater number of hues than tones. (In reply to comment #18) > 2) There is a Wiki at: http://en.wikipedia.org/wiki/Color_bars Here in Australia we use the PM5544 (PAL) test pattern: http://en.wikipedia.org/wiki/Testcard ... in place of the NTSC one, which you indicate later under point (4). > 3) An example is the EG 1:1990 “Alignment Color Bar Test Signal for > Television Picture Monitors”. [...] Find “EG 1:1990” in the list; you should > be able to download it at no cost. Note: this is an older analog TV standard > that they now make available at no charge. I can see the entry in the search list, but it does not appear to be free (US$50.00 showing here). Standards that are not freely available make it more difficult for others to implement. Many documents have author / publisher copyright but are still made freely available. This applies to points (9) and (12) also. A patent is a much more significant issue than copyright, although I sincerely doubt a palette in different arrangement. > 10) Colors themselves are not offered copyright protection in the United > States, or the rest of the world under the Berne convention. LO can use generic colour names, e.g., "58% Magenta" or "Pinkish-purple 123". Branded colour names are however only likely to be usable under license, e.g., "Pantone 630 C". This issue would not appear to affect this palette but IANAL.
Created attachment 105128 [details] SOC palette using rows instead of columns. Here is a reorganised version of the previous SOC file. Each column now aligns with a percentage saturation value. The monochrome scale has been renamed as requested but still sits at the top. Filler swatches have been renamed for greater consistency, with a new initial entry (0%) that is dark and higher entries (50%+) that are now pale.
Created attachment 105129 [details] Screenshot of color picker using row-based SOC. Relates to attachment 105128 [details].
(In reply to comment #19) > > 3) An example is the EG 1:1990 “Alignment Color Bar Test Signal for > > Television Picture Monitors”. [...] Find “EG 1:1990” in the list; you should > > be able to download it at no cost. Note: this is an older analog TV standard > > that they now make available at no charge. > > I can see the entry in the search list, but it does not appear to be free > (US$50.00 showing here). Standards that are not freely available make it > more difficult for others to implement. Many documents have author / > publisher copyright but are still made freely available. This applies to > points (9) and (12) also. A patent is a much more significant issue than > copyright, although I sincerely doubt a palette in different arrangement. That’s interesting. I was able to download the EG 1:1990 standard for free. The copy I received had a copyright notice (plus surplusage) that reads: “Downloaded by guest on 2014-08-20 from IP 71.216.229.113 @ SMPTE All Rights Reserved.” I did a reverse lookup on the IP address from mxtoolbox.com, and it showed my correct location of 44.98 -93.2638 (Hennepin County Library in Minneapolis, Minnesota USA). -Wade
(In reply to comment #21) > Created attachment 105129 [details] > Screenshot of color picker using row-based SOC. > > Relates to attachment 105128 [details]. Greytone organized from left to right is okay with me. Congratulations Owen! From what I can tell, you're now the proud parent of the world's first internationally recognized color palette standard which is organized by greytone! -Wade
(In reply to comment #19) > A patent is a much more significant issue than > copyright, although I sincerely doubt a palette in different arrangement ... would be affected. I forgot to finish this sentence.
Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.