User talk:RaboKarbakian

From Wikisource
Jump to navigation Jump to search
Babel user information
en-N This user has a native understanding of English.
Users by language
Here and other places


Fancy art book[edit]

The fancy art book I mentioned a few times finally came in: Index:Brinkley - The Art of Japan, vol. 1.djvu. The files are on IA (my upload), in high-quality (M)TIFF (~150 MB/image). I missed your Dec. 9 messages about the Grammar above—sorry about that. Your work there was very much appreciated, and it looks excellent. Would you mind creating the images for The Art of Japan? It is a bit much, though, so if you’d like to trade (with me doing some of your proofreading or scanning work), just ask. TE(æ)A,ea. (talk) 18:05, 31 January 2023 (UTC)[reply]

From a quick look, those images are beautiful. I will get started today on them. No trade required. But anything you want to pick at or do, this is what I am up to:
For January, I was cleaning up and out Pan. February will be for Grimm. Index:Little brother & little sister - and other tales (IA littlebrotherlit00grim 1).pdf. I started digging up magazine articles from Feb 1905. That has been kind of fun, picking a date and seeing what was going on.
Looking around like that I found the Railroad Gazette and liked it enough to try to get the February 3 edition done for that day due to its not many pages and it being interesting with the very specific mention of local happenings (ifra-structure and business) also national Railroad Law. So I grabbed and uploaded the rest of the pages Index:Railroad Gazette, Vol. 38, No. 5
But "section" transclusion of/by parts is not working for this png Index and the very few pages in the edition are each three columns of very fine print! Also, very weird is Page:Railroad Gazzette-Vol 38-142-000.png which started life as part of Index:File:Railroad Gazzette-Vol 38-p0104 and is also showing up at the newer index for the issue, making this a small but interesting bundle of issues. Weird issues at that. I would not have thought the pages to be so linked to the document file as it is to the proofread page text.
I am not aiming for Feb. 3rd any longer, but I am picking away at that issue. This "picking" has involved some software being installed, the goal being to eventually move the completed png pages to a pdf or whatever that will work for sectioning.
Okay, many words, sorry -- in the middle and at the end of things, some of which are broken messes. When you plopped Grammar into Main, I thought it was a terse but pleasant answer to all my questions: No; and that I was now done with them. One day, I might go back and change the text colors and some of the images from gray back to its muted color, but not today.--RaboKarbakian (talk) 20:09, 31 January 2023 (UTC)[reply]
  • I’m glad you like the images. They are all very pretty, but the plates especially struck me with their brilliance. I saw your Pan—great work, as always. The Grimm book seems to have many pages of text alone, but I will hold off on proofreading until you have a style set up—it’s much easier to follow than to go back and fix errors. On that topic, do you want a style set in place for Brinkley’s work? I’ve only gone over two pages of actual text (plus some preliminary pages), so you can hop in to add some styling if you’re interested. For the Railroad Ga(z)zette, I would recommend deleting (asking through sdelete) the original, single-page index; I think that that should fix the problem. The ProofreadPage system (for main transclusion and linking pages and indexes) is set up with the idea in mind that each page corresponds to one index, and one alone; breaking that rule confuses it quite a bit. (As an aside, it might also be the inclusion of “File” in the index name; and you should pick one spelling of “gazette” for all of the pages.) I will start doing some proofreading of the Gazette soon—the generated OCR is fairly usable, and there are many images taking up the pages, so it should not be too troublesome a task. TE(æ)A,ea. (talk) 20:34, 31 January 2023 (UTC)[reply]
The Grimm might also be the usual translation, the one from the complete works lady. A quick round up of my errors here and the summary is "paste errors", although, that is with selection error being a subset of paste. The two z is from a left sided keyboard bounce this poor device has acquired. The = was from my custom terminal font being too small for this decade on this computer and me being too busy and too not caring to fix. Maybe that Index could be kept as a "so many things wrong with this" example if I get my software working. It seems easy for me to add the style after, I think the sheet is probably less mess also.
This picking whatever from any magazine has been a little like sitting right at the salad bar with a fork and the bottle of favorite dressing....--RaboKarbakian (talk) 21:07, 31 January 2023 (UTC)[reply]
  • Isn’t it? It’s so nice, to have so many books (and magazines, etc.) at your disposal. The Internet really is amazing. Although, saying that, of course, I have just been through the trouble of scanning a whole bunch of books, but it’s nice that so much is available anyway. TE(æ)A,ea. (talk) 21:32, 31 January 2023 (UTC)[reply]
Hmm, well, thanks for letting me know that the labs DJVU was done. This morning my script was broken for jp2 and I (and not my computer) was the only one who knew where the tessdata was. I think tonight it works, the title page missing was a problem from this AM.
The problem with that page formatting is that I had not saved the image and therefore, I had not uploaded it. I uploaded those images while my djvu script was running, and really really wanted (at that moment) to get the index working and paste my djvu file at scan lab. So, the page formatting is also suffering from a prioritization error. Really great scan, btw. I am not rotating the images very much and often not at all. So, thanks for the good bad news or the bad good news or better, just the news. I am going to fix it, paste it at the lab and then start asking for deletions.
And, btw, I can make DJVU now. DJVU from ia or hathi. Well, 300ppi scans. I am going to have to rewrite it for GIMP for 600ppi scans as ImageMagick can't handle hathi pbm on this computer. Just ask, regardless the size, because that script will be easy to write.--RaboKarbakian (talk) 00:12, 8 February 2023 (UTC)[reply]
Well, no OCR so no showboating for me. Bleh.--RaboKarbakian (talk) 00:14, 8 February 2023 (UTC)[reply]
Eh. Is that upside down --> Page:Brinkley - The Art of Japan, vol. 1.djvu/59? About djvu: I should have the text layer working soon. Wrestling with anything "sed" is never what I expect. This is a very very pretty book! That carp is great. I am sorry it is not a plate. --RaboKarbakian (talk) 17:55, 8 February 2023 (UTC)[reply]
So, the magic lamp turns into a funny hat! Perhaps you could get the not broken Big Sur into MC? I just thought it should be here, and it should, but I got a bunch of stuff to do, so it just seems like a good MC thing, especially for appeal and all that.--RaboKarbakian (talk) 21:58, 8 February 2023 (UTC)[reply]
  • But now you’ve turned him around! It would be easier to see if his mask wasn’t so obfuscatory. As for MC, I’ll ask; it does seem like a work of some import, and would do well if someone was interested in proofreading it. I might just proofread it myself; the works I have lined up for myself (1, 2, 3, 4, besides one pending at Scan Lab and 5 in the background) I keep putting aside (but I still want to proofread them), so I end up proofreading other side works quite quickly. (I started 1 as a break from fiction, but the formatting for the current article is somewhat painful, so I’ve avoided it. 4 is interesting too, I scanned in it myself, but I just put it aside for a while. Rant rant rant, you get the idea.) TE(æ)A,ea. (talk) 22:35, 8 February 2023 (UTC)[reply]

(Defensively) That artist draws feet that look like cartoon smoke. So embarrassing! Not always, but often, I use math (width/ppi) to determine the image widths, especially with image rich books like this one. But, still, these images are huge! Is this physical book poster-size? So, that is one layout thing that I would like to do. The other layout thing that I would like to do is to spank the type-setter for just throwing images into the middle of text. So there is that also, which I am not going to reproduce in html, even if I knew how to do this. So, maybe you can transclude it and I can work on it from there, where I can see how things are going. Or let me know what you think could be done and/or done better than it is now. Also, at commons, I made one large logic leap that ended with Giokuden being Murase Yoshinori. Maybe if I had finished the key I would know. And I stopped working on the key due to: moodiness and wanting to know where the links go. So, it is not "done" but it is closer than it was this am.--RaboKarbakian (talk) 17:03, 11 February 2023 (UTC)[reply]

  • I was going to measure the physical book, but I forgot about it. I still have the second volume to scan through, though, so I’ll get good measurements then. It’s roughly 18″×18″×1½″, if I remember correctly; it is massive. The images-in-the-middle-of-the-text also turns the OCR into two columns, which meant that I had to manually type out every page where that happened. It has now been transcluded, and it doesn’t break MediaWiki, which is nice. I also finished up the key. I was thinking about the plate for the key having each of the signatures as individual images, but that sounds really annoying, so probably not. You can check The Art of Japan/Volume 1 to see how the images interact with each other. TE(æ)A,ea. (talk) 17:12, 11 February 2023 (UTC)[reply]
It looks pretty good, or pretty and good. My image work has had better edges, but I was careful because you could really see the wood underneath the paint, etc., on some of them. I am blaming the weather here for feeling flighty and insubstantial today. It is clear with no visible clouds or ceiling. We have had about 6 weeks of this gloomy, dark, and solid gray stuff that seemed to hold one down; which sounds bad, but is great for doing annoying things like you suggested for the key plate. I was going to suggest that the book was 16 x 11, but no, bigger! So, weather or not....--RaboKarbakian (talk) 17:24, 11 February 2023 (UTC)[reply]
  • Yeah, the weather around here has just gotten a lot better; going outside is enjoyable, for a change, although I think there’s to be a cold spell this weekend. But that’s not even my biggest book! I have another book of Japanese pictures which is about 12″×24″; that will be interesting to scan. 18 inches may be a bit much, but I do know that it’s pretty much square. The 150 MB/image is a little tiresome, though. TE(æ)A,ea. (talk) 17:47, 11 February 2023 (UTC)[reply]
  • I finally finished scanning the book, and have now uploaded it to Internet Archive: https://archive.org/details/TheArtOfJapanVolTwo! The book actually measures 15″×12½″×1½″—still, quite a sizable item. The second volume (on applied art) only has 41 numbered pages, but has more actual pages because of the number of keys at the end. I will break those up to separate pages, and separate the one from the first volume, as well—but I leave the DJVU and the image extraction to you! I look forward to seeing what you can do! TE(æ)A,ea. (talk) 23:05, 25 February 2023 (UTC)[reply]
TE(æ)A,ea. The computer I was using dropped offline, sorry for the delay. I will get to work on Part II! --RaboKarbakian (talk) 18:58, 27 March 2023 (UTC)[reply]
You know, I had someone more used to "officious" language (try to) read that... so IA has to quit their checkouts? My personal experience with both library and electronic checkouts is that if I really like the book, I buy it. Either tree or e, depending on the purpose. Perhaps the publishers, if they had a questionaire about where the book was first read, might reconsider their stance on this. About the djvu, I got one thing to work out and my script is for hathi pdf (making it work on ia texts was the next thing). It is weird, I felt a little in a rut when the internet ceased here; now, I am trying to get back into the groove. So, rut and groove, the same thing I guess! Frogs and mice sure is pretty!--RaboKarbakian (talk) 22:40, 28 March 2023 (UTC)[reply]
  • They’ve appealed the ruling, but yes, that means they can’t loan books (using CDL) anymore. The tough question (if they do eventually lose the CDL bid) is how much money they owe—the publishers could ask for ~$540 billion, as it stands. Your argument about liking and purchasing is echoed in IA’s arguments. (In my case, it’s all ILL.) The judge made numerous errors (in my non-bar legal opinion), if you’d like to be so regaled. I found the frogs-and-mice while looking for an old version—The crowne of all Homers workes Batrachomyomachia or the battaile of frogs and mise of 1624. TE(æ)A,ea. (talk) 23:23, 28 March 2023 (UTC)[reply]
TE(æ)A,ea. So, I cropped it as closely as I can. The page needs some black borders on the sides and bottom (4em should be nice) and I am not certain how to add that.--RaboKarbakian (talk) 18:00, 3 April 2023 (UTC)[reply]
TE(æ)A,ea.: That title page is very pretty. The border and the P's are in glittery gold and at least some of the other letters are dark red! What about glittery gold drop caps? The other image is stealing my time and enthusiasm. I am going to start over with it and just do some simple for it.--RaboKarbakian (talk) 21:43, 7 April 2023 (UTC)[reply]
  • Unfortunately, the rest of Pippa Passes isn’t so pretty as the cover page—and I never noticed how nice that was until you showed me. It’s a shame that the other image (my Japan one, right?) is so life-stealing—I just wanted the images to be as nice as possible. (Or do you mean the weird half-title image in Pippa Passes?) I might/should have a few other small images in the near future (esp. from here), in addition to the second volume of that (if you want to try to assemble another file). TE(æ)A,ea. (talk) 22:16, 7 April 2023 (UTC)[reply]
Piff, no no! Not the Japanese book. The other Pippa Passes image. Those childrens poems I am working on; I started to do those at WD years ago. And when it went up as finished, I grabbed the sound files with the intentions of finishing that. It is taking so long due to the lack of wikidata on things. The Japanese book needs me to fix one line on my script; after some time and having the internet disappear, and perhaps thinking that there are some winds of change about, that is delayed until I need the break from the poetry or until the winds actually do change.... One very pesky line! Working on that script for Big Sur, I did not throw this computer....--RaboKarbakian (talk) 22:24, 7 April 2023 (UTC)[reply]
TE(æ)A,ea. YES! I am interested! Truly, the hairs on the back of my neck have been prickling since seeing it in the 1928 lists.--RaboKarbakian (talk) 16:48, 31 January 2024 (UTC)[reply]
First, upload the Satyagraha image to commons (which I am doing right after I save this)? Second, sure, just let me know the where when the images are available.--RaboKarbakian (talk) 20:23, 10 February 2024 (UTC)[reply]
Sure to both. Grabbing the hathi images now. Give me a heads up when the scans are there.--RaboKarbakian (talk) 13:56, 4 March 2024 (UTC)[reply]
Your pages were all present and in order (mine, however were not in a different project). Let me know when you have an Index: for it. I am wondering if IAUpload will take a nice pack of tiffs like that? Also, in the thumbnails, while reviewing page numbers, there seemed to be duplicates of the pictures. Perhaps a real review of them will show differences but I have a habit (due to dedicated ebook readers with limited storage) to not have separate files of duplicates. Perhaps this book is an exception for this habit of mine -- it is your call. And, the Page: header -- I am a little giddy about styling the image bearing page header, might I?--RaboKarbakian (talk) 10:47, 4 April 2024 (UTC)[reply]
  • I don’t think that there are any duplicates, but some of the images look similar (two copies of the same piece with minor variations, for example). I actually scanned the first page with the header in high quality, in the hopes that you could use it to get the header: go ahead. I just made the request at the Scan Lab for the files to be collated. TE(æ)A,ea. (talk) 13:10, 4 April 2024 (UTC)[reply]
I am having a difficult time getting to these images. This is the wrong computer for this work. I have it downloaded. I wrote the image pages as I counted, using fwmv2_image viewer. It was easy to do some scrolling to see that very small page number. It would take me twice as long to do each image. This computer makes me sit and wait because it was supposed to be a processing plant. Millions of Cats was a beautiful pawse but the rest are going to be torture. Probably funny and laughable torture, but torture. Also, all of my software used to resize with the image. This software all resizes with largest layer center and not where the image window/view was. I am scrolling a lot more than I need to be and that too is a time grab. Is the software becoming adobe (weird at the layers) or MDI which was what people called the all in one window thing? And that was on the old gimp. The new one is super duper terrible. It has restricted toolbox size. It is stuck. I am on Jakub Steiner's side of whatever line is there.
The other thing: this computer has a file system such that I don't have a way to back up just the files. If any of my dongles fail, and they have been, I will have to write over everything so I need to find storage. "Move my phone off from the 1T" and I don't know how much of the stuff here to keep. This computer (my little brother got me a gadget computer) wants to process things and I am asking it instead to continually update its display and work on a file. It should be for processing books (without the need to display) and maybe run some software as a weather station and ntp server (something like a well-behaved software pet).
So, if excused from getting this done "right away", should I store it or should I store something else (Millions of Cats has been saved, for instance)? It might be interesting to save this image set and grab the thumbnails from the software run here and put them with it. Then get back to it when I get a different computer. Or let it go with the reinstall?
More words when saying no. "Yep, right on it" used to be easier to say. Maybe I should change my wiki name to Gidgit.--RaboKarbakian (talk) 14:54, 6 April 2024 (UTC)[reply]
  • RaboKarbakian: The file has been created here, and I can create the text as I go along. There’s a lot of text, so you can take your time to create the images, if that’s what you’re looking for. You can also add them within the text as you create the images and I create the pages, if you would like. The only image I would really like rushed is the image that appears in the page header, so I can go ahead and start proofreading the main body of the text. (If you also create a sample for the header, I can copy it over as I proofread.) TE(æ)A,ea. (talk) 20:02, 6 April 2024 (UTC)[reply]
Since less than a third of the pages have the fancy header, I stuck it in the index in header and proceeded to pink all of the pages where it is. Then I encountered KOINYŌBŌ SOMEWAKE TAZUNA here spelled with macrons (I think). I only made one with this "utf-8 guess" at Page:The Surviving Works of Sharaku (1939).djvu/84. I stopped here at my spelling. I wanted to just draw them all and then get it out of the index form, for less deleting of wrong headers, but the probably spelling error is stupid to keep pasting.
My style for {{rh}} didn't work in two ways. I set the rule to be dotted and I made the title vertically centered. It still looks good and maybe my usage gets fixed or the template gets fixed.--RaboKarbakian (talk) 01:14, 7 April 2024 (UTC)[reply]

Please use {{class block/s}} and {{class block/e}}[edit]

For your use case it would be appreciated if you could use {{class block/s}} and {{class block/e}} in preference to raw HTML, the reason is a plan to move HTML out of content namespaces, at some point. 14:00, 12 February 2024 (UTC) ShakespeareFan00 (talk) 14:00, 12 February 2024 (UTC)[reply]

The only major change is that class lines should no longer need the quotes. (You also gain an @ parameter for an anchor, without needing an additional template.) ShakespeareFan00 (talk) 14:02, 12 February 2024 (UTC)[reply]
The anchor part might be a problem because, if I can, I reuse the (in this case, Table) class.--RaboKarbakian (talk) 14:07, 12 February 2024 (UTC)[reply]
@ is optional. It doesn't get added as an "id" attribute, unless you supply it directly.
Please take a look at the underlying code if you think there will be conflicts. ShakespeareFan00 (talk) 14:09, 12 February 2024 (UTC)[reply]
Seems simple enough. My margins got screwed up tho'. Page:Journal of the Optical Society of America, volume 30, number 12.pdf/24--RaboKarbakian (talk) 14:57, 12 February 2024 (UTC)[reply]
That's because the table is bigger than the margins that get set-up, remember that they nest! ShakespeareFan00 (talk) 15:10, 12 February 2024 (UTC)[reply]

Graphs in JOSA.[edit]

I would suggest grabbing these directly from the JP2 /JPEG scans at IA if you were not already doing so. Not sure if IA has really hi-res TIFF's but those are also a possibility. Thanks for the efforts BTW :) ShakespeareFan00 (talk) 12:58, 14 February 2024 (UTC)[reply]

ShakespeareFan00: Thanks for the reminder. Also, look what I found: https://www.colour-science.org/api/0.3.3/html/colour.colorimetry.luminance.html there are source links!.--RaboKarbakian (talk) 16:57, 14 February 2024 (UTC)[reply]
Noted.
I will also add a rather useful page reference in a later JOSA - https://archive.org/details/sim_optical-society-of-america-journal_1957-07_47_7/page/619/mode/1up?q=Ridgway , If the relevant articles are out of copyright, and in scope for Wikisource, we may have a way to not just recover the original Munsell , but possible Ridgway(1912) and 2 other systems!! (And a substantial chunk of the colors mentioned in NBS Circular 355 :) ).
Of course converting Python to Lua is beyond me. ShakespeareFan00 (talk) 17:37, 14 February 2024 (UTC)[reply]
One of the articles (assuming it's no longer in copyright might prove very useful - https://archive.org/details/sim_paper-trade-journal_1947-11-06_125_19/page/218/mode/1up). It cross references a number of standards in use at publication date. At the very least we have the original Munsell to re-map, but it looks like there might be ways to use the 1943 as a crib into other systems to RGB, and potentially to measure scan fade for a number of early 20 the century works (potentially). And this all started as needing a way to get the colors in an SVG of some charts in one work :rofl: ShakespeareFan00 (talk) 17:37, 14 February 2024 (UTC)[reply]
There is this one matrix conversion: Y'; all color conversions use this algorithm (iirc). In use, you have a color (r,g,b), you convert it to its X, Y, Z and then use the Y to get Y' and with that, you can get Y_munsell and then X_munsell and Z_munsell which will give you (r,g,b)_munsell. If this is not exactly the right method, it is fairly close to it; almost ten years since I wrote my own conversion scripts. I copied those scripts from that site, but have yet to look at them. What I wrote above is what I will be looking for. Luminance is a different thing and (iirc) photoshop had luminance and lightness mixed up for several years.
The article had quite a bit of X, Y, Z in it. Maybe X', Y' and Z' are later. I was very interested in that article from Volume 6, as it might have been the first definition of X,Y,Z. "Cooperation" for nerds means mathematical standards.--RaboKarbakian (talk) 17:53, 14 February 2024 (UTC)[reply]
I also have an idea for a new color encoding though :
Proposed encoding scheme: Whole value is a 40 bit wide value (little endian).
Colorplane F (F 000 000 000-00 FF FF FF) is standard RGB (as for web)- with higher codes being control (such as specific codes for "Transparent","Out of range" etc..
Colorplane E (E 000 000 000-E 000 FFF FFF) is a digital quantisation of Munsell coding (to be determined)
Colorplane D (D 000 000 000-D 000 FFF FFF) Refers to the ISCC-NBS color numbers (Or if space other 'Spot' Color sytems)
Colorplane B-3 are currently unallocated. (Though I'm tempted to use plane 2 for 'compressed' data.
Colorplane 2 (2 000 000 000-2 FFF FFF FFF) is RGB (but 12 bit quantization)
Colorplane 1 (1 000 000 000-1 FFF FFF FFF) is XYZ. (12 bit quantization)
Colorplane 0 (0 000 000 000-0 FFF FFF FFF) is xyY. where the x y and Y are quantised to a 12 bit representation of the xy and Y values ( As these are typicaly scaled to between 0 and 1, the values are essentially a 3 figure fixed point value. This still gives a LOT of unused space though. :)
There may of course be overlaps in simmilar colors having codes in different color planes, There may be other systems that can be included, but I'm not sure how to encode CYMK(possibly it get's converted to sRGB) perhaps.
ShakespeareFan00 (talk) 21:05, 14 February 2024 (UTC)[reply]

If we can get this encoding down to a 32bit one.. Hmm... 21:20, 14 February 2024 (UTC)

Haha! Colorspace visions! All in "this can be hacked into tuples"! The closest conversation I had to this here was about how transparency increases file sizes exponentially -- as I was trying to make illustrated books for my reader. Envisioning colorspaces with any degree of accuracy seems just out of my reach, always; trusting the math being my fallback.
The key to that article is here: https://archive.org/details/sim_optical-society-of-america-journal_1922-08_6_6/page/527/mode/1up The language, the math, the method. Also, how the approach has changed throughout the decades between then and now. Most of those changes (if not all) are not yet into the public domain. It is stuck in upload here, I logged out of OAuth and (maybe) should have only logged out of a few privileges.
In the first few years I was getting to know computers and its hacking fans, there was a device you could stick on your screen that determined what color you were really seeing. Greta or Gretl something. United States patent 1682572 looked like the great-grand mammy of this device.
I envy you your colorspace visions.--RaboKarbakian (talk) 12:49, 15 February 2024 (UTC)[reply]

Index:Elementary Color (IA gri c00033125012656167).djvu[edit]

You had this linked as a PDF, I've uploaded a higher scan quality as Djvu. If you can figure out how to update the Wikidata linkages :). ShakespeareFan00 (talk) 15:31, 18 February 2024 (UTC)[reply]

BTW in respect of NBS Circular 553. It would be nice if there was a way to translate it into Wikidata at some future date.. ShakespeareFan00 (talk) 15:31, 18 February 2024 (UTC)[reply]

wikidata is easy for this. Add the file to the same field as the pdf, toggle "Preferred Rank" at the left. Thank you for uploading this!
Believe it or not, I started with the National Bureau of Standards text; should end with it as NBS Circulars are kind of a new thing to me and also here and all of the theres. Also, I looked at Portal:Physics to place this growing list and -- well, put that off until later also.--RaboKarbakian (talk) 15:39, 18 February 2024 (UTC)[reply]

Paul Centore[edit]

I assume you've already found his pages, but if you were wanting to make a big efforts, trying to get him to Open Access some of his papers would be nice. (He also has some Matlab code for doing conversions it seems) ShakespeareFan00 (talk) 20:17, 19 February 2024 (UTC)[reply]

Page:Color standards and color nomenclature (Ridgway, 1912).djvu/61[edit]

The dual tile approach was because I was intending we had the 'captured' tiles alongside those made by doing a blend based on his methodology detailed in the book. I wasn't sure what approach you were planning .ShakespeareFan00 (talk) 17:40, 25 February 2024 (UTC)[reply]

ShakespeareFan00: Those plates give the best information for getting to the color as it was in the book that was scanned that I have ever encountered. "Information" being white and black tiles to Level with. So, I was really in to this. There are still problems though, as each tile has a multitude of colors within. If I continue, I will upload the color adjusted tiles I worked with.
The "if I continue" is if you want me to continue or not. And, it is really great (to me) to have the text already proofed but it is also really sad (to me) that the text is already proofed. More sad is that I don't take the advantage given and just read it so that is clearly on me.--RaboKarbakian (talk) 11:21, 26 February 2024 (UTC)[reply]
Please continue :)
The approach I would have used is to find the sRGB for the intial wavelengths (given) and then using a blend algorithim to generate the intermediates, which in turn would be blended with white black and then grey...
I think I put some values for the RGB values generated by this approach in comments in the relevant tables? but I had not got that far. The reason for wanting 'generated' values alongside 'sampled' values, was to act as a way back into being possibly able to correct in other works based on likely pigment drift :) ShakespeareFan00 (talk) 11:35, 26 February 2024 (UTC)[reply]
Retrieved from "https://en.wikisource.org/w/index.php?title=User_talk:RaboKarbakian&oldid=14026326"