Does anyone understand the composite settings for GTU?
I know that the settings are Signal Resolution I and Y, but what do they actually do. Seems like the Y setting has to do with the color bleed but idk what Resolution I is for.
Any help would be be appreciated, trying to figure out how to get S-Video out of it, lol.
Y is “luma” or brightness, I and Q are “chroma” or color. It doesn’t do any of the funky crosstalk artifacts, etc. that we associate with composite, so it’s fairly close to svideo already, tbh.
So the composite mode should be labeled S-Video?
Also why does Y the Luma seem to blur the image at lower settings?
@hunterk How would we introduce cross-talk artifacts into GTU?
Because as it stands that’s the only thing it really needs to be able to pull off composite, which ideally means you’d be able to to get three video modes out of it.
Default settings with no composite set is RGB -SCART. (My opinion, lol)
Composite on, adjust Y res down some for blurring, adjust the Q res to about 40-ish to get S-Video.
Composite on, adjust Y res down some for blurring, is Composite (basically, it just needs some cross-talk, imo.)
For bonus points take the composite with cross-talk mode and add some static (noise) adjust the I and Q res a little (probably) making them stronger, and you have a ghetto RF mode.
The artifacts are where the real NTSC simulation comes in, and I don’t really think it’s worth it here. You have to do a whole modulation/demodulation hokey-pokey instead of the current thing, which is just switching to YIQ and then reducing the bandwidth of Y/IQ.
Ok then the inverse of that question then lol.
How hard would it be to do the Y bandwidth thing GTU does in ntsc?
You could potentially just put them in front and then chop out the beginning of ntsc-pass1-encode-demodulate.inc, where it samples the image and then changes it to yiq (that is, leave it in yiq after the GTU passes).
Again, probably not worth the effort here, as I doubt they’ll jive well together without some fairly extensive modification. GTU, specifically, is pretty macro-heavy, making it difficult to break down.
Well that’s disappointing, lol.
Was really liking my idea for a minute there, this is way out the scope of my coding ability…
If you ever get bored, please remember this idea, lol.
Yeahh, it’s alright lol.
Though I do have to say I haven’t seen any of the other ntsc shaders do the diagonal pattern on the waterfall… From what I remember (read it somewhere) that’s something RF or Composite actually does in some situations.
Yes that’s right the rainbow effect among others. When my parents bought me the amiga 500 at the time of its release, I couldn’t believe the quality of the display compared to everything else. I can understand the nostalgia about rf and composite but at the time we all complained, impossible for me to play again in such conditions haha
Even the NES looks better through RGB; many of the magazine screenshots you see from back in the day were taken off of a Sharp NES TV that had RGB, with the camera slightly defocused.
Then, starting with the SNES, all consoles supported RGB without modification, or at the very least had S-video… manufacturers wouldn’t have bothered to add that unless it was an obvious improvement.
Yeah, I completely understand this.
But honestly for someone that talks about how CRT games “should” have scanlines and a mask to be viewed “properly” it’s slightly weird that you bring this up because some games graphics were made with RF and Composite in mind.
I think even that’s kinda iffy, though. I don’t think there’s really a clear-cut answer, here.
The commonly cited example is the Sonic waterfall, but the Genesis supports RGB without any modification, so clearly Sega thought it was an improvement worth including. You lose the dithering effects but what you gain in terms of picture quality more than makes up for it. Artists at Sega (in the early days) had the dual monitor setup with one RGB monitor and a TV displaying a composite signal, but then they saw fit to support RGB, so the example kinda cuts both ways.
I think it’s a lot more clear-cut when you’re talking about really old computer games where they were working with a severely limited color palette (Apple II, Tandy, Atari, etc) and relied on composite artifacts to generate more colors than the system was actually capable of. I don’t really care about those games, though 
Honestly there’s no answer for how these games “should” be viewed mainly because wildly varying image quality depending on video connection, as well as for the longest time CRT’s had no exact standard for color (sRGB, etc; at least from my understanding) so manufacturers just kind of did whatever they wanted to within reason.
If we really went down this rabbit hole we’d have to find out which brand and setting they had for CRT’s for every individual game (so as to know how the artist originally intended the graphics to look; colors and video connections).
One thing we can say for certain is that these games were intended to be viewed on a CRT. 
Anything beyond that is kinda iffy and involves a lot of speculation.
The artists themselves probably didn’t even intend for the game to be viewed on one type of CRT/connection, since they knew that actual viewing conditions would vary wildly.
Here’s an interesting take, which is making me reconsider the whole dithering on 8-bit computers thing.
copied from: https://www.reddit.com/r/retrogaming/comments/97m7hz/dithering_on_the_sega_genesis_with_composite_video/
The idea that dithering was intended to be blurred to get the effect is largely a modern invention, and more of an accidental side-effect than something ever intended. Dithering was extensively used on computer platforms such as MS-DOS and Macintosh, which typically used tack sharp RGB or monochrome monitors and video with no discernible blurring or color blending. The idea that the picture quality could be “too good” for dithering never crossed the mind of anyone making games at the time, as dithering was just a technique to create the illusion of more shades with limited color values, much like crosshatching with pen, or halftones in printing. You absolutely were intended to be able to see the dithering.
Dithering serves its purpose perfectly well when sharp. The examples on the Wikipedia page for dithering do a pretty good job of illustrating why dithering was used, and how it’s effective even without any blending. Without dithering you end up with large areas of solid shades and obvious color banding. By dithering you are able to smooth out the color transitions.
Beyond that, the composite output on the Genesis is so incredibly awful , that I consider it almost criminal to suggest using it over the RGB output. I genuinely think the awful composite video on the Genesis is a major contributor to the platform’s reputation for being graphically inferior to the SNES. Over composite you get a hideously soft image with a bunch of unintended artifacts such as rainbow banding. Over RGB you get an absolutely gorgeous image that jumps right off the screen.
Yeah, I think it’s presumptuous to talk about “artists’ intention” etc. Dithering was definitely used in places where it wouldn’t be blended (e.g., the monochrome Apple IIe we had in my school’s computer lab), but I also think it’s a stretch to say “because Apple IIe exists, dithering-based pseudo-transparency is supposed to look like a screen door.”
For Gen/MD RGB, the existence of terrible, often very visible jailbar artifacts over RGB seems like a vote against it being the “right” way to play it, not to mention many games’ extensive use of jailbar dithering, which can look downright atrocious over RGB (see: Ecco: the Dolphin and Lion King title screens).
Yeah, it seems like this is really something that you have to consider on a per-game basis. Some Genesis games looked downright terrible through composite. In other cases, like when pseudo-transparency is used in Genesis games, I think it’s reasonable to assume that the artist intended that particular effect. However, even in those cases, the overall picture quality is pretty garbage compared to RBG. In the case of Ecco the Dolphin etc, I think it’s still kind of debatable because without the dithering you get color banding, so it’s a pick your poison situation.
agreed. I think it all comes down to preference and which games you play. I have both RGB and composite cables for my Genesis and I never use composite, but the games I play most often don’t rely heavily on dithering effects, so I don’t get bothered by it.
OTOH, I usually play my toaster+composite NES instead of my Nt+RGB. Maybe it’s because I played a lot of NES over composite as a kid, but that’s what makes the image look “right” to me, so I go with it.
Composite can also look better or worse depending on the TV. On a circa 2000-2005 Sony Wega, composite is damn near identical to S-video. I don’t mind composite on one of those
