/vr/ - Retro Games

"What's the best color" is certainly the sort of question that a certain sort of mindset would ask, and the one time nobody but that sort of person was involved in a discussion with real life consequences, the result was the CGA palette kek

>>10084398
>probably the one topic I see comparatively little discussion of is color.
Where were you when all the Mario purple/blue sky threads were going on?

I think a problem when discussing colors is that it depends a lot on a monitor's- and especially a TV's settings or even the model of your TV/monitor so there's a lot of variety in experience from person to person.

With NTSC/PAL there's the whole issue where on the C64 for example the colors looked completely different if you were in the US or Europe. I've actually started taking that somewhat into account when emulating, so I'll try it with NTSC for American C64 games and PAL for European to make it closer to how it looked for the developers. But there too it depends a lot on your TVs settings. Just try playing around with the sliders on this page to see the differences:
https://www.colodore.com/

sega master system had the best color. That is all

>>10084609
NES could never

>>10084398
This is not exclusive to the NES
It goes all the way up to 5th gen
However it’s not as egregious as the NES on most consoles
It’s similar to full and limited RGB in how games look

>>10084609
Pretty pointless because it had worse video output

>>10084421
I've heard there's also considerations to be made between Japan and NA, given that supposedly Japs TVs typically defaulted to a color temperature of 9300k or thereabouts compared to our 6500k, so a case could be made for changing the color temperature to a colder one for playing grorious Nipponese games.

>>10084723
You are confusing this with MD vs SFC. Master system was superior, from a specifications perspective, FC and NES in every way.

reminder that no one actually cares about any of this shit. colors, shaders, input lag, any of it. its just autists who want to flex their technobabble "complaining" about it. they just want people to know that they know what "dithering" is. just emulate your games, have fun, and move on.

>>10084398
>is there a definitive NES palette?
I'd say the closest color to what the North American system outputs would be direct composite (FBX) which is a pure, unmodified direct-capture of the composite output of the front-loader NTSC NES.

For Famicom and PAL units I believe the most authentic palette's would be ones directly translated from the system. While looking cool and being what a kid in the early 80"s might have seen CRT's would be a layer of abstraction from the original system output which would be the systems true colors.

Ultimately NES palette's can be swapped on the fly via emulator and it's ultimately up to your preference. Find what works best with you and have fun.

>>10085130
Makes sense, NTSC has a lot of issues with color. I know some old timers that worked in, non syndicated, local broadcasting. They didn't have the precision tools to ensure color was right, so they just fiddled, had a color card and guessed.

All I know is the sky is purple and the bricks are brown

>>10085185
>>10085150
The problem with NTSC video is that it has to be decoded, so the colors ultimately depend on how that is done by the decoding hardware. That's why even the colors extracted from a direct capture of the video output can't be considered a "pure" palette - it still had to decode the NTSC signal, and so what you're getting isn't the so-called "real" colors, but rather the decoder's interpretation of the NTSC colors.

>>10084398
It's interesting to look at the differences but at the end of the day these discussions should keep in mind that the nature of analog tech means there will always be slight variances. With the NES as an example: different revisions will have subtly different colors because the underlying circuitry is just a little bit different. We do have a "canon" RGB palette for the NES, with the PlayChoice-10 and some other high-end releases like the C1 TV. This palette is a lot more bright and garish than what a NES will typically output, and outright breaks some (pretty obscure) games. It's likely this palette was picked because they thought eye candy mattered more than accuracy. But this palette was used for developing at least some high-profile games, notably Mario 3, so you could argue that this palette is the "proper" way to play that game.
I think these discussions happen because people are so used to modern-day standardization and "canon" that they have trouble wrapping their heads around the idea that video games used to be a wild west, primarily seen as toys and not art. Just look at the drastic differences in artwork and marketing between regions.
>>10084421
It's also doesn't help that most TVs don't come calibrated to be accurate, but rather to look the most bright and colorful. This seemed to happen even back in the CRT age, with Trinitrons having an infamous "red push".

>>10085270
>But this palette was used for developing at least some high-profile games, notably Mario 3, so you could argue that this palette is the "proper" way to play that game.
If that's the case, I want whatever they were smoking.

>>10085281
forgot pic

>>10085283
Huh, looks like it’s from an emulator. If everything was cheap or free it would be cool to have a set of each just side by side for comparison.

The NES version of The Immoral used a "forbidden black" that technically the NES could output but you was not supposed to set a color level that dark. It gave it more greytones for shading/shadows but almost no emulator displays this properly without fine-tuning the pallets (if the emulator you are using even supports that) and it even causes the game to not display properly on some CRTs back in the time because it messed with the sync signal. No idea if any of the FPGA/clone consoles display it correctly.

Almost nobody seems to talk about this, has anyone here actually adjusted whatever shaders or device or filters they use to get it to display correctly?

>>10085328
Some TV:s - even old ones, may mistake color $0D for a blanking signal since the value of that color is so low it goes outside of what the standard considers to even be a color. This can lead to warping distortions. So it really just depends on your CRT. Most newer models (90's and onwards) have issues rendering it but most older CRT's work just fine. If your having issues rendering it you can always turn your system brightness up to max which might help you better see your near black greys.

>>10085564
What about those FPGA devices or using a software emulator?

>>10084398
Anon just let them live in blissful ignorance instead of them worrying over whether games are supposed to be played at 6500k or 9200k.

>>10085148
You are incorrect anon and you are not in touch with the retro gaming community.

>>10085668
Can't unlearn. Must sperg.

>>10084398
C64 color palettes are really confusing. There's some video modes that quickly blend two different colors into a new one that's not part of the main palette. It's the most hacky machine around.

>>10085668
I wonder if color autism is even compatible with shader-autists since I assume you lose a lot of extra brightness with CRT-shaders.

>>10085720
Brightness is something else, but CRT shaders can absolutely fuck with colors depending on which parameters you fiddle with. Guest-Advanced has a plethora of parameters for the purposes of color correction, balancing, and so on. It even has profiles for attempting to replicate the colors of CRT phosphors based on widely used industry standards like SMPTE-C. And of course shit like color temperature control, gamma correction, saturation... and that's before we even get into fucking NTSC shaders. There's a wide world of unexplored autism in these shaders.

>>10085708
Tbf the extra colors through fast blinking wasn't an official graphics mode, the C64 just had a lot of autists who got good at wringing every last drop of performance out of it.
You can see that effect here if you set the quality to 720p50:
https://youtu.be/S_9Q0uBAvL0?t=776
Notice how if you pause it the dragon has one of the two colors blended.

I've never gotten a CRT with "accurate" color settings without tweaking. My JVC in the picture and a Trinitron I have both had service menu settings that I had to change, but at default blew out the reds hard and made yellows look orangeish. This is my most recent attempt at calibrating my CRTs to be close in color to my OLED

>>10085720
High NIT displays counteract the loss of brightness.

>>10085783
All I see is different sections of the dragon rapidly flickering, which was generally used to simulate transparency in older games on a CRT.

>>10085801
The enemy dragon, not the player's dragon, though the player dragon flickering is similar.

>>10085668
I don’t think there’s much difference
White is white when playing games

>>10085270
Why can't I remember where this thing is from? I know I have seen it a million times, but I can't remember where it's from. Is it Super Mario RPG?

>>10085912
Yes.

>>10085783
That's neat.

>>10085147
Master System had inferior audio chip (only 1/6 of games were compatible with FM module, and the FM was more primitive than Genesis FM), inferior framerate (30 instead of NES's 60), and inferior controller (less buttons, Dpad and buttons feel worse and control worse).

>>10085147
not really, it had superior colours sure
but the actual video output was extremely noisy and ugly

>>10085734
Oh, it's getting explored.
Fuck original hardware, I have a powerful machine and monitor, there is NO REASON why I shouldn't be able to achieve perfect handcrafted synthesis of ideal CRT phosphors in a CRT filter custom fit to each retro game.
It took me a week to make it happen for my last game.

As >>10085270
points out authentic OG CRT age hardware wasn't standardized especially with colors, glow, etc. so it's a confusing journey when you become obsessed with colors.

>>10086507
Based color autist

>>10086507
In another few years crt shaders will have completely outclassed the real thing. I'm sure there will always be crt autists who assert to their dying day that the $1500 medical monitor they bought from a hospital auction is the "real" way Paperboy was meant to be played, but I just want my games to look like that one wizardry screenshot that gets posted everywhere without having a giant ugly set in my living room with a cord stretched out all the way from a decaying gray brown plastic box to my couch

>>10087445
>In another few years crt shaders will have completely outclassed the real thing
Lmao

>>10087445
I've been to the far corners of the esoteric internet to discuss CRT shaders and there's just three people there
>Me, seeking colors
>Another CRT shader enthusiast with the same purpose
>The guy who sought us out to tell us that our CRT shader could never replicate the real thing

>>10089128
>CRT shader
>seeking colors
mate, a colour palette has nothing to do with CRT shaders

>CRT shaders can absolutely fuck with colors depending on which parameters you fiddle with. Guest-Advanced has a plethora of parameters for the purposes of color correction, balancing, and so on. It even has profiles for attempting to replicate the colors of CRT phosphors based on widely used industry standards like SMPTE-C. And of course shit like color temperature control, gamma correction, saturation... and that's before we even get into fucking NTSC shaders.

>>10089168
one uses the other
A CRT shader is not necessary to mess with the palette

>>10084398
>So, are there "optimal" colors for consoles
Who fucking cares
Literally kys for giving a shit about this irrelevant garbage

>>10089709
>Willingly enters a topic he has no interest in

>"Who Cares? Kys!"

Is there anything more pathetic?

>>10089184
Apparently the go-to shader for color fuckery these days is Dogway's Grading shader. It has an insane amount of parameters to fiddle with.

>>10089184
Speaking of which, apparently this is the "real" output of the NES before it's decoded? I can't say I understand it all that much, but you're supposed to use this with a "decoding" shader to get actual usable colors.

>>10091490
No no, clearly that's what the game was intended to look like, you're playing it wrong otherwise.

>>10089184
BMF2

>>10091826
Looks way too dark IMO, almost as dark as the Wii VC palette.

>>10089184
here's the best lot of palettes
I used to use the sony CXA one
but now I just use real hardware on a CRT

>>10092893
It's a horrorish game. I don't know about other games but for Castluvania it seems appropriate. I like how the foreground stuff pops against the darker backgrounds.

>>10092895
>>10089184
For me it's FCEUX

>>10086236
the sms runs at 60 fps, not sure where you've got the idea it runs at 30 from

>>10092961
Eh, I suppose there is a case to be made for picking and choosing on a per-game basis if you wish.

>>10093946
Your terms are acceptable.

>>10092895
>>10093946
>>10094045
https://stickfreaks.com/colour/nes-colour-palette-comparisons/faq
There was a guy who used a colorimeter to figure the most accurate NES palette. He came away with the NES Classic being the most accurate, which is pretty dark

>>10092895
Is your CRT color calibrated?

>>10094175
Yeah, used SMPTE

>>10094175
how do you calibrate a CRT for color anyway?

>>10094658
https://www.youtube.com/watch?v=waUahh8DG4M

>>10094045
NES classic is the only acceptable one here

>>10094790
Does it have blue skies in smb?
If not, then no

>>10094879
It's more on the purple side but that's correct
Miyamoto said so + the arcade version of SMB1 used that color

>>10094929
The arcade version is literally secondary
And no, he did not say that
It’s indigo which comes up as blue through an NTSC signal

>>10095247

>>10095293
unlike you, I actually have a Famicom on a CRT, guess which colour it is, not purple I can tell you that

>>10084609
Nope, PC-Engine had the best color palette of any console of *both* 4th and 5th gen

>>10085148
Dithering is sorely needed for modern games to combat color banding and 8-bit color artifacting, it may be autistic but once you show it to normies they get it too, unlike framerate/framepacing which normies never get.

Dithering ain't that important for 4th gen retro games though, their pixel resolution meant they made middling or poor use of it most of the time.

>>10095982
Sounds like you need to re-calibrate your TV

>>10096121
nah

>>10085210
Most of the Never Twice Same Color meme was about broadcast though. Because of how the colour channel was encoded there was no good way to tell if it had become shifted between the tower and the antenna. This manifested as a shift to green or red and why there was a tint control to manually correct for it. This wasn't important for consoles because you hooked them up directly and didn't have to worry about buildings and hills messing with the signal.
PAL didn't have this flaw because it was made later and they realised they better do something about it.

>>10096129
If you take a direct photo of a CRT it's gonna distort the colors. That's why a lot of CRT shaders look bad as they're using photos as a reference. You've been shown quotes by the people who made the game, just admit you were wrong.

>>10096486
lol, ok mate
you do you

>>10096486
Color representation between RGB screenshots and CMYK print can vary drastically, especially in the 80s. A screenshot of a vintage game packaging isn't a reliable reference.

>>10096486
Are we sure these are direct photographs? Looks more like a screenshot/hardcopy from a devkit that was processed by early desktop publishing tools at the time.
Compare these to screenshots in the back of N64 games that were clearly taken by an SD capture card or genlock.

>>10096790
Kek, this level of schizophrenia is pretty new for this topic
I hope one day you can fight your affliction and actually are able to play a game before losing your mind over the colour

>>10095293
I don't actually care about this at all, but I just want to say that something about the file name is killing me. "miyamotopurplish" had me dying a bit ago and even now I can only smile when I see it.

>>10096885
Well, this a thread about correct colour representation in video games. A rather nitpicky topic to begin with and I like nitpicking.

>>10096005
Dithering is either there or it isn't for 4th gen retro games since it was hand pixelled into the sprite. So it doesn't matter what normies think about that.

>>10096878
>>10096885
NTA but there's all sorts of ways colors can get messed up when it comes to old scans. Recently it was discovered that the Gen 1/2 Pokemon art circulating online had very off colors
Granted arguing over NES palettes is pointless for reasons >>10085270 said. If you're playing on real hardware you're just gonna use whatever your TV spits out. If you're RGB modding/emulating then "developer's intent" has gone out the window anyway and you should pick what you think looks best for each game. If you must feel "correct" then the most objective answer would be the RGB PPU, as that was what Nintendo consistently used for RGB screens back in the 80s and makes the SMB1 sky purplish like Miyamoto said, indicating Nintendo probably used that to make their games. But that's not a popular option which speaks to how this is just people using "authenticity" or whatever to justify their personal taste.

>>10097527
>as that was what Nintendo consistently used for RGB screens back in the 80s
This has consistently been proven false
How do you think these people made games on the famicom

I like the NTSC-Colors shader. It makes the colors really pop. That said, it does oversaturate blues and especially reds. Maybe they're out of gamut on an sRGB display?

>>10097712
Regular colors for comparison.

>>10097675
Sharp X1?

>>10097675
They made Famicom games on a RGB monitor and tested how they'd look on a composite TV. What exactly are you arguing here? Even that TV they're testing on seems to be using the RGB PPU, considering how bold that yellow is (pic related gives a clearer image). Though it's also possible that the saturation on that TV is cranked up.
>>10097717
FM R-50HD, actually.

>>10097752
I've also heard that Jap TVs typically had more saturated yellows for some reason, but I've never been able to verify if this is true.

>>10097752
>using the RGB PPU
Sure anon

>>10097764
Nah, that's just an urban legend that caught on because people thought NES games look better if you make them more yellow (which is what the RGB PPU does). The guy who first programmed it into NES emulators was later like "woops yeah that was bullshit". I'm not sure if it's even in any modern emulators anymore.
>>10097778
>I can't refute him so I'll just act passive aggressive

>>10097804
Ah yes anon, clearly the RGB PPU was the intended look here >>10094045
you can play like that all you want

>>10085148
>nobody cares
You care enough to post.

>>10097821
I'm arguing the opposite. I doubt NES devs had a rigid idea of an "intended look." You can plainly see them testing Mario 3 on a RGB NES (or a TV with cranked up saturation). They likely just picked whatever looked good on their setups and rolled with it.

>>10097821
I'm arguing the opposite. I doubt NES devs had a rigid idea of an "intended look." You can see them testing Mario 3 on whats either a Famicom modded for RGB or a TV with really cranked up saturation. They likely just picked whatever looked good on their setups and rolled with it.

>>10097934
I'm beginning to think this may be the case, at least for something like NES. I do wonder about later consoles, though. A lot of times the colors you get out of an emulator feel like they lack vibrancy, almost like they expected the display to pick up the slack somehow. See >>10097712 >>10097714 for example. Makes me wonder if we were at all intended to see it like the latter pic.

>>10099534
Even after the NES you can't pin down a rigid "intended look" due to the nature of analog and lack of documentation. Similar debates happen with old TVs and movies. Here's an almost 100-page spergfest over the colors in Dragon Ball Z: https://www.kanzenshuu.com/forum/viewtopic.php?t=19448
>Makes me wonder if we were at all intended to see it like the latter pic.
Most devs probably didn't expect consumers to get colors 1:1 to the raw image yea, but that doesn't tell us much. Without hard evidence like that Mario 3 image, it's basically just a matter of "I think it looks better this way", which is highly subjective and can lead to BS spreading and even being programmed into emulators, like so >>10097804

>>10100609
https://www.youtube.com/watch?v=z6pcfINCB4o
How about this

>>10100743
This is a pretty neat find, where's it from?
Anyway, doesn't seem like they used the RGB PPU. The video quality is shit but the RGB PPU's garish colors are real noticeable. Goes to show how much variance there was in the NES days. There actually is a book involving the FF pixel artist from way back when where she talks about designing the sprites, and shows off some design documents.

>>10100864
Found it here, got some good info about development
It’s got some other videos linked too
https://www.retroreversing.com/famicom-nes-development-kit/

>>10100864
I remember reading an interview with Ishii where he claimed the reason the mountains were white with black lines in FF1 was to create a shimmering rainbow prism-like look, similar to the waterfall effect in Sonic, to show the mountains contained crystal shards. If that's true I assume they had the composite output in mind. It could just be Ishii making shit up after the fact too I guess.

>>10100889
>Many sources hint that there was no official Nintendo development kit for the NES and in fact there is very little information that has been leaked about these development kits. It looked like Nintendo themselves used EEPROM writers and custom made cartridges for development.
>However in the NintendoAge article from issue 2 13 it mentioned that in 1987 BEAM software knew they couldn’t get an official development kit from Nintendo and so made their own. When Nintendo found out about their custom development kit they wanted the sale of it stopped. Why did they want the custom development kits to cease production if they didn’t have officially supported ones?
>Did they really rely on every single studio creating their own hardware based development kit? It seems like quite a barrier for entry, but then again Nintendo only wanted quality software so this could be intentional.
lel
What a mess of a console, autistic color calibration would've been the least of their worries

>>10100932
And yet, it had the best games of the time
Pretty wild

>>10101018
Oh definitely, I dig the NES. It also has probably one of the most autistic (in a good way) communities out there. People on NESDev will cite academic articles in their posts.

>>10101037
Yeah, they do a good job of documenting just about everything in their wiki, too.

>>10084398
the main argument RE1 fanboys have against the remakes is also "where's muh colorz!!"

>>10095293
i'm not sure i've ever seen such a dishonest image

>>10103442
What's dishonest about it?

I'm extremely autistic about colors, and my tip is... don't be like me.

Just boot up your game and have fun. Don't worry about colors unless you actually have a device to calibrate your screen properly.
If the colors are so bad/inaccurate that it bothers you, then change your screen.

>>10100932
Wouldn't need to worry about calibration with a good ole CRT desktop monitor

>>10104270
Impart some of your wisdom on us anyway, o color autist.

>>10104652
Lolwut? Those things still had to be calibrated, they aren't magic.

>>10105813
Wish I had a colorimeter.

>>10107145
what?
buy one if you want, they're cheap

>>10084609
Funky

>>10109169
Head Boxers?

>>10084398

>>10109764
A lot of arcade games do seem like they're overbright when displaying the raw colors.

>>10087445
>that the $1500 medical monitor they bought from a hospital auction is the "real" way Paperboy was meant to be played
look buddy play how you want to play but i have 9 CRTs and I've paid about $100 TOTAL for them. im in autism heaven

>>10094045
fceux and classic are perfect

>>10110301
I am the anon you responded to, and I'm happy that you're playing games the way you want to play

>>10092895
I like Nestopia YUV the most. The sky texture on the bottom appears more like a gradient and the dimly lit bricks in the middle panel look more realistic with that hue. I always found visible dithering to be ugly.

>>10084412
I prefer bright colors
I always set the tv like that

this page talks about nes color
https://www.nesdev.org/wiki/NTSC_video

>>10095996
pc engine has two color palettes, one for rgb and one for composite/rf

>>10084412
CGA colors are peak soul. I always set my terminal emulator's colors to those.

>>10111515
Only problem is there isn't really a good color for human skin. Oftentimes characters will look straight-up white like Commander Keen or pink as fuck.

>>10113018
There's 15, the only one you need.