Rebuilding the Dragon Ball Z Character Color Design Sheets for Color Correction

[lansing]

With the uncover of the actual color chart (now accurately digitized) that was used to paint Dragon Ball, theoretically it's now possible to color correct the show to its original cel color. However, in order to reach that mature stage, more things needs to be done first. Having the color chart is just the foundation. We'll need to have the color design sheets for the models in order to properly use the chart. The color design sheets are the model sheets of the show that show where each color goes.

It is not common now to find a full set of color design sheets since nobody really has any use of them, but there are tons of douga lying around in people's cel collections that had color code markings on them. So ideally if we collect enough of them, we can rebuild the color design sheets ourselves.

This is likely to be an ongoing project considering the scale of the anime. But as of now, I just want to see how far can this go. We start small and slowly build it up.

This is what a finished color design sheet would look like:


Here are some examples of color correcting Dragon Ball movies using the Goku model above:

movie 1: https://imgsli.com/MTI0Mzgx
movie 7: https://imgsli.com/MTI0Mzgy

The background color are probably wrong since I don't have reference for those yet.

So if you have any of the color design sheet or douga, please share it.

I will be updating here for new sheets or updates.

[johnny1132]

I don't want to insinuate anything, but what source would this be applied to should all required materials be obtained? Would it be kai? We both know the dboxes are undoable with these colors. Thanks

[Someoneimportant]

I have never in my life seen such majestic color correction.

Keep em coming Lansing

[lansing]

I don't want to insinuate anything, but what source would this be applied to should all required materials be obtained? Would it be kai? We both know the dboxes are undoable with these colors. Thanks

I tested on both kai chapters, the HD movies, some screen shot from Dragon Ball OG, they all works. I don't have dragon box to test but it should also work.

[Venny]

With the uncover of the actual color chart (now accurately digitized) that was used to paint Dragon Ball, theoretically it's now possible to color correct the show to its original cel color. However, in order to reach that mature stage, more things needs to be done first. Having the color chart is just the foundation. We'll need to have the color design sheets for the models in order to properly use the chart. The color design sheets are the model sheets of the show that show where each color goes.

It is not common now to find a full set of color design sheets since nobody really has any use of them, but there are tons of douga lying around in people's cel collections that had color code markings on them. So ideally if we collect enough of them, we can rebuild the color design sheets ourselves.

This is likely to be an ongoing project considering the scale of the anime. But as of now, I just want to see how far can this go. We start small and slowly build it up.

This is what a finished color design sheet would look like:


Here are some examples of color correcting Dragon Ball movies using the Goku model above:

movie 1: https://imgsli.com/MTI0Mzgx
movie 7: https://imgsli.com/MTI0Mzgy

The background color are probably wrong since I don't have reference for those yet.

So if you have any of the color design sheet or douga, please share it.

Pretty sure the sky and skin tones aren't meant to be pink like this lol. Also idk why you would try and use exact color charts to correct a Blu-Ray or DVD source. Even in the Movie 1 screenshot you posted you brought out compression artifacts when trying to push out dynamic range in the blacks. I understand you want to have the most accurate colors possible, but the material we have isn't good enough to do this and judging by the last color correction thread it didn't really work there either. Too much color artifacting from tryin to perfectly match colors of static color charts to video and colors come off as too punchy and overblown. Doesn't even really look like cels which give off natural and somewhat mute colors pallets so each color has a specific contrast with one another so they don't overpower each other. This is why groups like SoM are pushing for colors which match broadcasts or color timed TV prints because that not only represents something close to the original viewing of the shows and films, but is doable for sources which have been processed so many times already

[VegettoEX]

This is the exact same content being carried over from a thread that was closed due to an inordinate amount of activity in disregard of any amount of spirit from the community guidelines.

Any instance of similar such activity will result in an immediate account strikes (and/or bans).

Please also remember, topic creator, that you were extensively called out for repeated unrealistic and inaccurate information. Unless you are willing to justify and accurately explain away any additional such (caring, truthful, well-meaning, good faith) callouts, consider that you (general, all-inclusive, plural "you" here) may all be in the wrong business here.

[lansing]

Pretty sure the sky and skin tones aren't meant to be pink like this lol. Also idk why you would try and use exact color charts to correct a Blu-Ray or DVD source. Even in the Movie 1 screenshot you posted you brought out compression artifacts when trying to push out dynamic range in the blacks. I understand you want to have the most accurate colors possible, but the material we have isn't good enough to do this and judging by the last color correction thread it didn't really work there either. Too much color artifacting from tryin to perfectly match colors of static color charts to video and colors come off as too punchy and overblown. Doesn't even really look like cels which give off natural and somewhat mute colors pallets so each color has a specific contrast with one another so they don't overpower each other. This is why groups like SoM are pushing for colors which match broadcasts or color timed TV prints because that not only represents something close to the original viewing of the shows and films, but is doable for sources which have been processed so many times already

The skin color is the correct cel color for goku. The compression artifact would be the job for video filtering, it does not matter at this stage.

[lansing]

This is the exact same content being carried over from a thread that was closed due to an inordinate amount of activity in disregard of any amount of spirit from the community guidelines.

Any instance of similar such activity will result in an immediate account strikes (and/or bans).

Please also remember, topic creator, that you were extensively called out for repeated unrealistic and inaccurate information. Unless you are willing to justify and accurately explain away any additional such (caring, truthful, well-meaning, good faith) callouts, consider that you (general, all-inclusive, plural "you" here) may all be in the wrong business here.

No, this thread is an attempt to collect or recreate color design sheets for the show. I will be updating the database collected in the first post. It's a different topic than color correction.

[Venny]

Pretty sure the sky and skin tones aren't meant to be pink like this lol. Also idk why you would try and use exact color charts to correct a Blu-Ray or DVD source. Even in the Movie 1 screenshot you posted you brought out compression artifacts when trying to push out dynamic range in the blacks. I understand you want to have the most accurate colors possible, but the material we have isn't good enough to do this and judging by the last color correction thread it didn't really work there either. Too much color artifacting from tryin to perfectly match colors of static color charts to video and colors come off as too punchy and overblown. Doesn't even really look like cels which give off natural and somewhat mute colors pallets so each color has a specific contrast with one another so they don't overpower each other. This is why groups like SoM are pushing for colors which match broadcasts or color timed TV prints because that not only represents something close to the original viewing of the shows and films, but is doable for sources which have been processed so many times already

The skin color is the correct cel color for goku. The compression artifact would be the job for video filtering, it does not matter at this stage.

idk it seems like your color matching isn't as correct as you say it is.

[lansing]

Pretty sure the sky and skin tones aren't meant to be pink like this lol. Also idk why you would try and use exact color charts to correct a Blu-Ray or DVD source. Even in the Movie 1 screenshot you posted you brought out compression artifacts when trying to push out dynamic range in the blacks. I understand you want to have the most accurate colors possible, but the material we have isn't good enough to do this and judging by the last color correction thread it didn't really work there either. Too much color artifacting from tryin to perfectly match colors of static color charts to video and colors come off as too punchy and overblown. Doesn't even really look like cels which give off natural and somewhat mute colors pallets so each color has a specific contrast with one another so they don't overpower each other. This is why groups like SoM are pushing for colors which match broadcasts or color timed TV prints because that not only represents something close to the original viewing of the shows and films, but is doable for sources which have been processed so many times already

Spoiler:

The skin color is the correct cel color for goku. The compression artifact would be the job for video filtering, it does not matter at this stage.

idk it seems like your color matching isn't as correct as you say it is.

Spoiler:

I think for this image I got an average deltaE of 3.xx or 4 on a 9 color match. For anyone who doesn't know, average deltaE is an accuracy rating, anything under 5 is a really good match under an uncontrolled environment. But again, it does not matter at this stage, someone may come by and get a better result in the future. The images shown here are just to get people interested in sharing their collection.

[Someoneimportant]

This is the exact same content being carried over from a thread that was closed due to an inordinate amount of activity in disregard of any amount of spirit from the community guidelines.

Any instance of similar such activity will result in an immediate account strikes (and/or bans).

Please also remember, topic creator, that you were extensively called out for repeated unrealistic and inaccurate information. Unless you are willing to justify and accurately explain away any additional such (caring, truthful, well-meaning, good faith) callouts, consider that you (general, all-inclusive, plural "you" here) may all be in the wrong business here.

Honestly, and in all seriousness, I have no real issue with this thread and what’s it about. Like I even said this in the last thread, I only had an issue with this method being brought up in a thread that was specifically about correcting the Dragon Boxes. But since this isn’t really related to that or is not specifically about that, I don’t really see the harm.

Seeing as I likely won’t be posting ITT again, I should probably mention that some very knowledgeable people that own plenty of film prints have stated that this method possibly wouldn’t work even on the film reels, without decimating heavy amounts of the grain, or at least it would be very hard to do it without causing artifacts of some kind. So even 4K film scans that are less compressed than anything that’s available for DB, you will still most likely encounter some issues with this. There you have it. Just a bit of information that some might find interesting.

Either way, good luck to anyone who is interested in this method.

[lansing]

The neutrality of the white paint varies way too easy compares to other color and its brightness also varies depending on the size of the painted area. For the purpose of easier manipulation in the color correction stage, I have decided to set the value to be fixed at lightness of 99, which is (252, 252, 252) in sRGB.

For the same reason, the black value is also fixed at lightness of 7, which is (21, 21, 21) in sRGB.

[jjgp1112]

All other reservations aside, Movies 6-9 used a different color palette than the rest of the anime, so even trying to match them to the character design sheet color charts is a pointless endeavor.

[lansing]

All other reservations aside, Movies 6-9 used a different color palette than the rest of the anime, so even trying to match them to the character design sheet color charts is a pointless endeavor.

It's not a different palette but a different brand of paint. They used the Sun Color instead of STAC. It's not impossible to get samples of the color, the hardest part is however, to get a hand on the color design sheet set of the movies because they don't come by often.

[razielinflight]

This is the exact same content being carried over from a thread that was closed due to an inordinate amount of activity in disregard of any amount of spirit from the community guidelines.

Any instance of similar such activity will result in an immediate account strikes (and/or bans).

Please also remember, topic creator, that you were extensively called out for repeated unrealistic and inaccurate information. Unless you are willing to justify and accurately explain away any additional such (caring, truthful, well-meaning, good faith) callouts, consider that you (general, all-inclusive, plural "you" here) may all be in the wrong business here.

Honestly, and in all seriousness, I have no real issue with this thread and what’s it about. Like I even said this in the last thread, I only had an issue with this method being brought up in a thread that was specifically about correcting the Dragon Boxes. But since this isn’t really related to that or is not specifically about that, I don’t really see the harm.

Seeing as I likely won’t be posting ITT again, I should probably mention that some very knowledgeable people that own plenty of film prints have stated that this method possibly wouldn’t work even on the film reels, without decimating heavy amounts of the grain, or at least it would be very hard to do it without causing artifacts of some kind. So even 4K film scans that are less compressed than anything that’s available for DB, you will still most likely encounter some issues with this. There you have it. Just a bit of information that some might find interesting.

Either way, good luck to anyone who is interested in this method.

Are you crazy? Paint is lossless! How can you even compare legitimate real world paint to something copied like forty times. Nothing can be better than that.

[Someoneimportant]

This is the exact same content being carried over from a thread that was closed due to an inordinate amount of activity in disregard of any amount of spirit from the community guidelines.

Any instance of similar such activity will result in an immediate account strikes (and/or bans).

Please also remember, topic creator, that you were extensively called out for repeated unrealistic and inaccurate information. Unless you are willing to justify and accurately explain away any additional such (caring, truthful, well-meaning, good faith) callouts, consider that you (general, all-inclusive, plural "you" here) may all be in the wrong business here.

Honestly, and in all seriousness, I have no real issue with this thread and what’s it about. Like I even said this in the last thread, I only had an issue with this method being brought up in a thread that was specifically about correcting the Dragon Boxes. But since this isn’t really related to that or is not specifically about that, I don’t really see the harm.

Seeing as I likely won’t be posting ITT again, I should probably mention that some very knowledgeable people that own plenty of film prints have stated that this method possibly wouldn’t work even on the film reels, without decimating heavy amounts of the grain, or at least it would be very hard to do it without causing artifacts of some kind. So even 4K film scans that are less compressed than anything that’s available for DB, you will still most likely encounter some issues with this. There you have it. Just a bit of information that some might find interesting.

Either way, good luck to anyone who is interested in this method.

Are you crazy? Paint is lossless! How can you even compare legitimate real world paint to something copied like forty times. Nothing can be better than that.

Send me DM. This is not the place.

[Lorium_O]

Hoo boy, it's been a while since I've gone to this website.

Are you crazy? Paint is lossless!

Paint is NOT LOSSLESS. There is a reason why paintings conservators exist.

A paintings conservator is an individual responsible for protecting cultural heritage in the form of painted works of art. These individuals are most often under the employ of museums, conservation centers, or other cultural institutions. They oversee the physical care of collections, and are trained in chemistry and practical application of techniques for repairing and restoring paintings.

How can you even compare legitimate real world paint to something copied like forty times. Nothing can be better than that.

Depends on what paint you're talking about. The cels you see today aren't the same as from 1989. (As I said, paint isn't lossless.)

I hope this doesn't cause a drama.

Edit: RIP Color Correcting the Dragon Box thread.
Also Edit: Can someone create a thread similar to the Color Correcting one? I got disappointed when I came back to the site and saw it locked.
Yet another Edit: Can you send other color palette sheets other than the Goku one? I'm interested.

[ikaos]

Also Edit: Can someone create a thread similar to the Color Correcting one? I got disappointed when I came back to the site and saw it locked.

Our group will make a thread here once we complete one of our color-corrected series releases since it'll be more useful than comparing 50 different variations of single shots.

[lansing]

Yet another Edit: Can you send other color palette sheets other than the Goku one? I'm interested.

The shared link was in the first post, I created 11 so far. It took a lot of time to rebuild one. If you have any scattered color code douga, please share it.

[Lorium_O]

The shared link was in the first post, I created 11 so far. It took a lot of time to rebuild one. If you have any scattered color code douga, please share it.

Now that I think about it, didn't some guy give his own douga in the cc thread?

Reminding you just in case you missed that.