Support NeoGAF

[PumpkinSpice]

I can understand that. But it's still something subjective, and the "this is what the devs intended" line gets towed out far too often, I feel. I'm not even aware of a dev saying that, just that its been inferred (undoubtedly correctly) by the techniques they employed, such as the waterfalls in sonic you mentioned.

To name one, there was a Double Fine Devs Play video where one of the devs for the Lion King on Genesis talked about how the title screen didnt look right because they intended the colors to blur over composite/RF.

 

[NormalFish]

So then how do you folks feel about dithering filters?

Taken from the OSSC thread on shmups:

**

My main interest was in the bit rate reduction happening on old machines. What looks right or wrong is a largely subjective conversation.

Fair enough.

To name one, there was a Double Fine Devs Play video where one of the devs for the Lion King on Genesis talked about how the title screen didnt look right because they intended the colors to blur over composite/RF.

I hadn't heard this. Interesting.

 

[missile]

I thought it was from 24-bit to 16-bit which caused the dithering. Is it really 16-bit to 8-bit?

He means RGB 24-bit to RGB 16/15-bit. 8-bit is color palette mode, usually.

... Have to be honest, I'm not entirely familiar with the intricate details of what causes this cross hatch artifact aside from "low bit rate fucks it up." If anyone has some reading material worth having a look at, I'd love to consume it. ..."

Start with any gray-scale image and quantize it to 1-bit using Bayer dithering.

I quickly wrote something up. All bugs are mine, mind you! xD

// classic Bayer dither pattern (type: dispersed-dot ordered dithering)
// see: Robert Ulichney; Digital Halfoning
int Bayer4x4[4][4] =
{
  1, 9, 3, 11,
  13, 5, 15, 7,
  4, 12, 2, 10,
  16, 8, 14, 6
};

// dithering of an 8-bit (256) gray-shade into an 1-bit (2) binary shade  
// (usually black and white) using a 4x4 Bayer dither pattern to simulate
// 17 shades (haltones) by placing dots on the screen depending on a threshold
int dither_pixel_8bit_to_1bit(int x, int y, int shade)
{
  int m = (shade*100)/255;  // scale and compute fraction 
    
  // get threshold, Bayer 4x4 gives 4²+1 = 17 halftone shades
  int T = (100*Bayer4x4[y&3][x&3])/17;
 
  if(m > T)
    return 1; // white 
  
  return 0; // black
}

// set pixel either black or white
int pixel_shade = dither_pixel_8bit_to_1bit(x, y, shade);
setpixel_1bit(x, y, pixel_shade);

There are about 10^32 things to improve upon here. But better start simple!

Some results:

A slight horizontal lowpass to fake signal/CRT bandwidth limits.


Edit:
About three years ago I wrote this one for Notch's DCPU-16;

Edit II: For the fun of it;

random

newspaper print

 

[socksfelloff]

He means RGB 24-bit to RGB 16/15-bit. 8-bit is color palette mode, usually.


Snip

Missile is the MVP of this thread. You don't post broscience. Just cold hard facts.

 

[Khaz]

Edit:
About three years ago I wrote this one for Notch's DCPU-16;

The cross pattern feels so weird. I wonder if you could insert a sort of pseudo random distribution of the dots to make it less artificial while keeping a dithering effect? Or I suppose the cross pattern would disappear with horizontal blending as it's only an artefact of the vertical lines visually interacting.

I don't know what I'm talking about. It's just this cube right now feels weird.

 

[Knobiwan]

Any word on the Gamecube HDMI mod that Badass Consoles was going to make more available? I've been following it since the teaser video, and have been dying to get an HDMI cube.

GCN is my favorite console and my TV doesn't have S-Video and the component cables are literally the price of a PS4 right now, so I just can't justify it, but just can't stand how it looks through composite. I probably should just bite the bullet and buy the component cables after saving up, I have never regretted saying "Aw $45 is a bit much for cables" years back more in my life.

 

[missile]

Missile is the MVP of this thread. You don't post broscience. Just cold hard facts.

xD Thx for the flowers! Thought televator has some real interest.

The cross pattern feels so weird. ...

The pattern can be modified to remove the hatches (losing some halftones).

... I wonder if you could insert a sort of pseudo random distribution of the dots to make it less artificial while keeping a dithering effect? ...

Yes. But the distribution forms patterns or may flicker. Bad for animations.

I worked the last month on a solution to remove all disadvantage. Not ready yet.
(for another project of mine)

... It's just this cube right now feels weird.

That's because the pattern is sticky to each face. But can be made static.

 

[Mega]

Came across this hideous mess while browsing ebay

 

[4444244]

Someone rescue that PVM!

 

[theclaw135]

So then how do you folks feel about dithering filters?

Taken from the OSSC thread on shmups:

**


Fair enough.


I hadn't heard this. Interesting.

Big step forward there. This goes a long way to show how much of an unsung issue the inability for composite to discern what it should or should not affect is.

 

[NormalFish]

Big step forward there. This goes a long way to show how much of an unsung issue the inability for composite to discern what it should or should not affect is.

Yeah it's pretty cool. It requires a frame buffer, but i could see it implemented on real hardware. Those screenshots are running in Retroarch, though.

 

[Barrel Cannon]

Came across this hideous mess while browsing ebay

Lol, god I can't believe he's using an X-arcade. I honestly wouldn't hate on him if he was using some other inputs only solely because he probably designed it himself and used a good monitor. But a horrible input device can greatly hamper the enjoyment of a game.

 

[televator]

He means RGB 24-bit to RGB 16/15-bit. 8-bit is color palette mode, usually.



Start with any gray-scale image and quantize it to 1-bit using Bayer dithering.

I quickly wrote something up. All bugs are mine, mind you! xD

// classic Bayer dither pattern (type: dispersed-dot ordered dithering)
// see: Robert Ulichney; Digital Halfoning
int Bayer4x4[4][4] =
{
  1, 9, 3, 11,
  13, 5, 15, 7,
  4, 12, 2, 10,
  16, 8, 14, 6
};

// dithering of an 8-bit (256) gray-shade into an 1-bit (2) binary shade  
// (usually black and white) using a 4x4 Bayer dither pattern to simulate
// 17 shades (haltones) by placing dots on the screen depending on a threshold
int dither_pixel_8bit_to_1bit(int x, int y, int shade)
{
  int m = (shade*100)/255;  // scale and compute fraction 
    
  // get threshold, Bayer 4x4 gives 4²+1 = 17 halftone shades
  int T = (100*Bayer4x4[y&3][x&3])/17;
 
  if(m > T)
    return 1; // white 
  
  return 0; // black
}

// set pixel either black or white
int pixel_shade = dither_pixel_8bit_to_1bit(x, y, shade);
setpixel_1bit(x, y, pixel_shade);

There are about 10^32 things to improve upon here. But better start simple!

Some results:

A slight horizontal lowpass to fake signal/CRT bandwidth limits.


Edit:
About three years ago I wrote this one for Notch's DCPU-16;

Edit II: For the fun of it;

random

newspaper print

Ah, I see. It's the dithering itself. Which is obviously used to fake more shades of color.

 

[Morfeo]

Does any Framemeister-users use the hdmi pass through feature that was patched in? It doesnt seem to work quite right at my tv, getting lots of black screens and other stuff typical of when the framemeister actually process something and the resolution is changed. Is this feature known to be bugged?

 

[Beer Monkey]

To name one, there was a Double Fine Devs Play video where one of the devs for the Lion King on Genesis talked about how the title screen didnt look right because they intended the colors to blur over composite/RF.

Here's an extreme example of IBM PC CGA developers using composite artifacting to create > 4 colors. Same game, difference is between RGB output (left) and composite (right).

Even if plugging a real IBM into a real composite monitor, to get the desired effect you have to use an old composite monitor, as a comb filter in newer monitors can remove the artifacting!

This type of coding was used most often on title screens and other static screens but some games like Burgertime and Kings Quest use it in the actual game engine.

 

[mdzapeer]

Here's an extreme example of IBM PC CGA developers using composite artifacting to create > 4 colors. Same game, difference is between RGB output (left) and composite (right).

Even if plugging a real IBM into a real composite monitor, to get the desired effect you have to use an old composite monitor, as a comb filter in newer monitors can remove the artifacting!

This type of coding was used most often on title screens and other static screens but some games like Burgertime and Kings Quest use it in the actual game engine.

Mind blown, I used CGA display games back in the late 80s in our computer labs ( I was pretty young). And even back then the monitors did not use composite and always were connected with the D-Sub (VGA) connection.

Seeing the above image is a huge surprise, I used play zaxxon in CGA in the labs...I wonder now what it would have looked over composite.

 

[Shadow Hog]

Personally, the best use of that composite mode quirk would be 8088 MPH, a 2015 demoscene program designed to abuse the shit out of it, resulting in images with 1024-ish different colors on-screen at once. Over CGA, the graphics mode maligned for only having four ugly colors at a time.

Seeing the above image is a huge surprise, I used play zaxxon in CGA in the labs...I wonder now what it would have looked over composite.

Like this, probably.

 

[Beer Monkey]

Seeing the above image is a huge surprise, I used play zaxxon in CGA in the labs...I wonder now what it would have looked over composite.

This list is pretty comprehensive, I think:

http://nerdlypleasures.blogspot.com/2013/11/ibm-pc-color-composite-graphics.html

It does list Super Zaxxon as an artifacting game but not Zaxxon. It can be fun looking up videos and screens of these games being played on real composite hardware or even through emulators that fully simulate real composite artifacting. Hackers are even editing sprites and playfield graphics on some old games to use artifact palettes. This Hard Hat Mack edit looks a lot more like a C64 or Atari 800 game than an IBM PC CGA game, for example.

https://www.youtube.com/watch?v=z3o1pTdsXr0

Here's a patched-for-artifacting Commander Keen played on real hardware.

https://www.youtube.com/watch?v=VL1RcuZsnjg

 

[televator]

Okay, so if certain consoles dithered down to 16/15 bit, couldn't an emulator spit out the original 24/32 bit frames? Anybody know if any emulator currently does that?

Also how much hassle would it be to hack a real console to do the same? I mean, I'm sure it's easily more involved than HDMI mods...

 

[NormalFish]

Okay, so if certain consoles dithered down to 16/15 bit, couldn't an emulator spit out the original 24/32 bit frames? Anybody know if any emulator currently does that?

Also how much hassle would it be to hack a real console to do the same? I mean, I'm sure it's easily more involved than HDMI mods...

wouldn't it be at a software level? Might be able to do something with a hex editor/rom patch/whatever.

 

[Mega]

I was thinking if there were any NES examples, then I remember seeing this one a while back:

http://www.chrismcovell.com/gotRGB/rgb_compare.html

That's pretty crazy.

 

[televator]

wouldn't it be at a software level? Might be able to do something with a hex editor/rom patch/whatever.

I suspect that could be a solution to at least some games on certain consoles. However there's a hard cap due to hardware limitations as well. Frame buffer size obviously can't exceed physical memory.

 

[NormalFish]

I suspect that could be a solution to at least some games on certain consoles. However there's a hard cap due to hardware limitations as well. Frame buffer size obviously can't exceed physical memory.

Of course. I'm not sure a hardware board would be much bigger compatibility wise, though, unless you put a pretty extensive piece of tech in there.

 

[Timu]

I was thinking if there were any NES examples, then I remember seeing this one a while back:

http://www.chrismcovell.com/gotRGB/rgb_compare.html

That's pretty crazy.

It's like looking at 2 completely different games!!!

 

[televator]

Of course. I'm not sure a hardware board would be much bigger compatibility wise, though, unless you put a pretty extensive piece of tech in there.

Well, I got to thinking that upgrading RAM in some consoles is a real thing. The OG Xbox and even the Sega Genesis are known example off the top of my head. Couple that with some software hacks to address the larger memory accordingly and... profit?

 

[NormalFish]

Well, I got to thinking that upgrading RAM in some consoles is a real thing. The OG Xbox and even the Sega Genesis are known example off the top of my head. Couple that with some software hacks to address the larger memory accordingly and... profit?

I'm not sure earlier consoles would be more receptive to such things. I suppose the NES PPU mods aren't a terribly large stretch so maybe something in that vein could work. Not sure how many consoles would have compatible parts out in the wild, though. Probably varies widely.

 

[Shadow Hog]

I was thinking if there were any NES examples, then I remember seeing this one a while back:

http://www.chrismcovell.com/gotRGB/rgb_compare.html

That's pretty crazy.

Really, a lot of that one boils down to which palette you're running your NES with. I expect something like FirebrandX's "Unsaturated" palette would look much closer to the composite shot. (Like, FCEUX's stock palette really oversaturates its reds, for instance, to the point where I can't distinguish two shades of it where I could distinguish between them with FBX's palette; meanwhile, I think that image is the Playchoice-10 palette, which... well, there's a reason it's the "Garish" palette of the NESRGB, let's put it that way.)

 

[NormalFish]

Really, a lot of that one boils down to which palette you're running your NES with. I expect something like FirebrandX's "Unsaturated" palette would look much closer to the composite shot. (Like, FCEUX's stock palette really oversaturates its reds, for instance, to the point where I can't distinguish two shades of it where I could distinguish between them with FBX's palette; meanwhile, I think that image is the Playchoice-10 palette, which... well, there's a reason it's the "Garish" palette of the NESRGB, let's put it that way.)

Yeah, I agree. I don't think those color differences are something that can't be solved through other, more elegant, methods.

 

[Barrel Cannon]

Well, I got to thinking that upgrading RAM in some consoles is a real thing. The OG Xbox and even the Sega Genesis are known example off the top of my head. Couple that with some software hacks to address the larger memory accordingly and... profit?

It's only useful to that really for emulation and homebrew purposes as far as the original Xbox goes. I've never seen anyone make any crazy mods taking advantage of extra RAM mods.

Also that NES screenshot comparison is pretty crazy. I didn't expect the colors to be that different due to composite

 

[Mega]

Really, a lot of that one boils down to which palette you're running your NES with. I expect something like FirebrandX's "Unsaturated" palette would look much closer to the composite shot. (Like, FCEUX's stock palette really oversaturates its reds, for instance, to the point where I can't distinguish two shades of it where I could distinguish between them with FBX's palette; meanwhile, I think that image is the Playchoice-10 palette, which... well, there's a reason it's the "Garish" palette of the NESRGB, let's put it that way.)

The comparison to pay attention to isn't the color differences; it's the blending caused by the composite signal creating greater complexity in the graphics vs. the simple, raw pixelation of the RGB picture. In fact, I dunno why the site didnt use screencaps with alike color palettes or calibrate them to match. There's no reason they should be that different.

Edit: this isn't apparent if you're viewing it on a small screen (your smartphone).

 

[Peltz]

Well 'incorrect' isn't quite true, but the programmers definitely assumed the games would be played on lower quality connections and designed their graphics around that in many cases. Just like the dithering for fake transparencies and shadows in Mega Drive/Saturn games.

I mean I'm not going to play Sonic over composite just for transparent waterfalls, but it's pretty clear that's how the effect was designed.

It's also on a case by case basis. Some PSX/N64 games have no dithering issues so you literally lose nothing with a higher quality video connection. Some have a lot of full screen dithering (eg Silent Hill 1, Majora's Mask - generally darker games and scenes have issues) and while in my opinion it's still not worth going down to composite, something is lost in RGB.

Meh... I just played Virtua Fighter in my local barcade. It has clearly visible dithered shadows over a lossless RGB signal on the original arcade cabinet. To say that they didn't intend for games to be seen this way when many arcade versions WERE seen this way is a bit of an unfair assumption. There's simply too many variables to know what any developer intended back then without hearing it straight from their mouths.

 

[televator]

It's only useful to that really for emulation and homebrew purposes as far as the original Xbox goes. I've never seen anyone make any crazy mods taking advantage of extra RAM mods.

Also that NES screenshot comparison is pretty crazy. I didn't expect the colors to be that different due to composite

The black box of doom was already a 32 bit color system, but I was merely using it as an example that it's possible to change RAM capacity in systems. There'd be nothing to gain as far as color depth because it's already great. Either way, I just don't think people really felt the need to make Xbox games address larger memory. Perhaps 720p forcing could take advantage of the added ram. IDK

 

[PumpkinSpice]

Meh... I just played Virtua Fighter in my local barcade. It has clearly visible dithered shadows over a lossless RGB signal on the original arcade cabinet. To say that they didn't intend for games to be seen this way when many arcade versions WERE seen this way is a bit of an unfair assumption. There's simply too many variables to know what any developer intended back then without hearing it straight from their mouths.

This is why a lot of games did a grid dither instead of stripes. Stripes look best over composite but awful on RGB, whereas the grid looks decent on composite and acceptable on RGB.

 

[Madao]

heads up for those that want an Ultra HDMI for N64

they're about to go live with a new batch very soon.

http://www.badassconsoles.com/nintendo-64/

u gotta ready that F5.

 

[Barrel Cannon]

The black box of doom was already a 32 bit color system, but I was merely using it as an example that it's possible to change RAM capacity in systems. There'd be nothing to gain as far as color depth because it's already great. Either way, I just don't think people really felt the need to make Xbox games address larger memory. Perhaps 720p forcing could take advantage of the added ram. IDK

It would have to be programmed to take advantage of the RAM(in regards to games) but even then then I highly doubt a RAM upgrade alone would help much with forcing games in 720p(IMO of course)
I was initially interested years ago in doing the RAM mod for my console. It would have only benefited me for emulation of a few select titles and one homebrew game a group I know from some forum was working on. I did do some research on it and then interest quickly waned when I realized the difficulty in it.

It's awesome though when you think about what the Xbox mod/homebrew scene accomplished with the original xbox. The fact that MS tried to mimic the majority Original Xbox Mod features/developments with the Xbox 360 then you truly realize how accomplished the Original Xbox mode scene was.

 

[D.Lo]

Meh... I just played Virtua Fighter in my local barcade. It has clearly visible dithered shadows over a lossless RGB signal on the original arcade cabinet. To say that they didn't intend for games to be seen this way when many arcade versions WERE seen this way is a bit of an unfair assumption. There's simply too many variables to know what any developer intended back then without hearing it straight from their mouths.

Of course, we know of some cases for sure, and can only speculate in others. We also have ports on both PS1 and Saturn (eg Rockman X games) where they look nearly identical via composite, but the effect was achieved via dithering on Saturn, and real transparency on PS1 - so we know the intention more clearly without official word.

VF doesn't prove much either way except that sometimes dithering was the planned look, since it's on arcade-only Model 1 hardware which couldn't do transparencies, so dithering was the only option anyway.

 

[Timu]

heads up for those that want an Ultra HDMI for N64

they're about to go live with a new batch very soon.

http://www.badassconsoles.com/nintendo-64/

u gotta ready that F5.

When they become more common I may get one. Not a fan of limited batches.

 

[Shadow Hog]

Ignoring that the palette wasn't the intended focus here, here's some of those shots with FBX's palette (taken in FCEUX, nearest-neighbor-scaled to 512x480, then bicubic-scaled to 640x480):

Hebereke definitely looks more accurate, though the original composite shot of Blaster Master still looks remarkably... yellow, for some reason. I mean, there's no such thing as a perfectly-accurate NES palette due to the nature of how it works, but still, that's a pretty big discrepancy.

 

[missile]

Appreciate the recent love for composite video in here. Keep it coming!
(techwise it's even more interesting)

Ah, I see. It's the dithering itself. Which is obviously used to fake more shades of color.

That is why it's called halftone, a tone (for your eye) between two real tones.

Okay, so if certain consoles dithered down to 16/15 bit, couldn't an emulator spit out the original 24/32 bit frames? ...

May only work for 3d games, 2d games have the pattern already applied, usually.
(as NormalFish said, it's a software thing)

Meh... I just played Virtua Fighter in my local barcade. It has clearly visible dithered shadows over a lossless RGB signal on the original arcade cabinet. To say that they didn't intend for games to be seen this way when many arcade versions WERE seen this way is a bit of an unfair assumption. There's simply too many variables to know what any developer intended back then without hearing it straight from their mouths.

Faking shades or transparency are two different things considering dithering.


Edit:
Btw; Dithering kills resolution, it's sort of a lowpass filter on its own.

Look again

Bayer4x4 = 17 halftone shades

The more halftone shades, the lower the resolution becomes.

The trading equation is;
shades * resolution = constant
(amplitude * frequency = constant)

Spoiler

Did you know that?

Edit II:

37 halftone shades

Hence, the resolution of this image is lower compared to the Bayer4x4 one.
(that's why the patterns were kept rather small, i.e. 2x2 to 4x4)


Edit III:
IIRC old newspapers used about 65 shades for their b/w graphics.

 

[NormalFish]

May only work for 3d games, 2d games have the pattern already applied, usually.
(as NormalFish said, it's a software thing).

tangentially, does anyone know how the skies are done in PS1/N64/Saturn 3D titles? Are they modern skyboxes or static flat backgrounds or what?

 

[missile]

^ Depends on the game, i.e. how advanced the 3d engines was.

The skyboxes in old games have some distortions and are a lil floody at times.

 

[NormalFish]

^ Depends on the game, i.e. how advanced the 3d engines was.

The skyboxes in old games have some distortions and are a lil floody at times.

Well that makes sense. I admit I haven't spent a lot of time paying attention to various games' skyboxes. Lots of gradients and such used in PS1 games, though. I've always assumed they were skyboxes as I'm used to them in modern games, but I do wonder how many of them are just a single static texture backdrop. It's interesting the little quirks of 5th gen.

 

[missile]

Well that makes sense. I admit I haven't spent a lot of time paying attention to various games' skyboxes. Lots of gradients and such used in PS1 games, though. I've always assumed they were skyboxes as I'm used to them in modern games, but I do wonder how many of them are just a single static texture backdrop. It's interesting the little quirks of 5th gen.

The rock-solid ones are usually flat textures, pretty sure.
Btw; I've even seen skybox clipping (the corners of) in some old 3d games. xD

Here's an extreme example of IBM PC CGA developers using composite artifacting to create > 4 colors. Same game, difference is between RGB output (left) and composite (right).

Even if plugging a real IBM into a real composite monitor, to get the desired effect you have to use an old composite monitor, as a comb filter in newer monitors can remove the artifacting!

This type of coding was used most often on title screens and other static screens but some games like Burgertime and Kings Quest use it in the actual game engine.

Part of my work is applying these artifacts to low-res 3d graphics.

Spoiler

(don't hit me, Khaz! xD)

 

[KC-Slater]

What makes you think it's going to die soon?

I don't necessarily think it's going to die soon, (it's actually in fantastic shape both cosmetically and operationally) but I have come to terms with the fact that one day it will die, and it just got me thinking about replacement solutions, is all.

 

[Mega]

I don't necessarily think it's going to die soon, (it's actually in fantastic shape both cosmetically and operationally) but I have come to terms with the fact that one day it will die, and it just got me thinking about replacement solutions, is all.

Funny timing to have this come up. My Ikegami TM14-17R may effectively be non-operational when only a couple of days earlier it was perfectly fine.

I was doing something unrelated (setting up a second audio output), fired up some of the CRTs to compare glare and noticed the Ikegami had a very dim picture. Now it needs manual brightness turned up to max level to get a visible but dark picture. I think increasing the RGB levels via the front pots may brighten up the image, but this will only accelerate tube wear. I think something inside must've given up and the monitor is on its way out. It is my oldest and most worn monitor but just goes to show some of these old screens with aging parts may not last too long and suddenly crap out. They may still turn on and show a picture for a long time but the screens won't forever be in a state fit for flawless RGB gaming.

I read Sony monitors were serviced and had the picture tube replaced every 30K hours, about every 3-5 years if they were constantly on. For one of us that's more like 20~ years if the tube has moderate hours on it and with regular gaming use... and assuming nothing else fails. Because of all this, I am glad to own several CRTs in really good shape. If this one has truly crapped out, off to the recycling center it goes... and maybe I'll pick up one of those 15" JVCs I've been eyeing forever to round out the bunch.

 

[KC-Slater]

Funny timing to have this come up. My Ikegami TM14-17R may effectively be non-operational when only a couple of days earlier it was perfectly fine.

I was doing something unrelated (setting up a second audio output), fired up some of the CRTs to compare glare and noticed the Ikegami had a very dim picture. Now it needs manual brightness turned up to max level to get a visible but dark picture. I think increasing the RGB levels via the front pots may brighten up the image, but this will only accelerate tube wear. I think something inside must've given up and the monitor is on its way out. It is my oldest and most worn monitor but just goes to show some of these old screens with aging parts may not last too long and suddenly crap out. They may still turn on and show a picture for a long time but the screens won't forever be in a state fit for flawless RGB gaming.

I read Sony monitors were serviced and had the picture tube replaced every 30K hours, about every 3-5 years if they were constantly on. For one of us that's more like 20~ years if the tube has moderate hours on it and with regular gaming use... and assuming nothing else fails. Because of all this, I am glad to own several CRTs in really good shape. If this one has truly crapped out, off to the recycling center it goes... and maybe I'll pick up one of those 15" JVCs I've been eyeing forever to round out the bunch.

That's too bad [/pours one out for your Ikegami] and hopefully you can still squeeze a little more life out of it with the tweaks you mentioned. I have no idea what the display in my Naomi has been through, or the hours it may have on it. As an arcade machine, I can only assume it was typically treated like shit, and left on for long periods of time lol. (If anyone knows how to access the service menu on an arcade display, that may be a place to start?)

It's in storage right now, so I can't even say that I'm just enjoying it while it lasts. (No room for it.) At least it is being pampered in it's twilight years?

 

[KiteGr]

I've just got a Wii2HDMI and I've run into some problems.

Whenever I open certain option window, weather that's a game I'm about to run, or some times some options within the game, I get a temporary black screen and a cut in the sound as if the resolution just changed. Also the screen seems a bit zoomed.
I toyed a bit with the TV setting but I couldn't find any way to get rid of it.

 

[Vespa]

I've just got a Wii2HDMI and I've run into some problems.

Whenever I open certain option window, weather that's a game I'm about to run, or some times some options within the game, I get a temporary black screen and a cut in the sound as if the resolution just changed. Also the screen seems a bit zoomed.
I toyed a bit with the TV setting but I couldn't find any way to get rid of it.

Does the black screen happen when backing out of the Wii channels? I had that issue with a vga cable that didn't have resistors applied to the signal so it would desync on peak rgb levels (usually white flashes).

Not sure about the zoomed in picture, is this in widescreen? how does 4:3 look? Just wondering how it's handling the overscan.

 

[televator]

Appreciate the recent love for composite video in here. Keep it coming!
(techwise it's even more interesting)



That is why it's called halftone, a tone (for your eye) between two real tones.


May only work for 3d games, 2d games have the pattern already applied, usually.
(as NormalFish said, it's a software thing)



Faking shades or transparency are two different things considering dithering.


Edit:
Btw; Dithering kills resolution, it's sort of a lowpass filter on its own.

Look again

Bayer4x4 = 17 halftone shades

The more halftone shades, the lower the resolution becomes.

The trading equation is;
shades * resolution = constant
(amplitude * frequency = constant)

Spoiler

Did you know that?

Edit II:

37 halftone shades

Hence, the resolution of this image is lower compared to the Bayer4x4 one.
(that's why the patterns were kept rather small, i.e. 2x2 to 4x4)


Edit III:
IIRC old newspapers used about 65 shades for their b/w graphics.

Yeah I can see that my fur around the edges isn't resolved. Heh... Man, I am so glad I brought up this topic here. Thanks for all the info missle!

 

[KiteGr]

Does the black screen happen when backing out of the Wii channels? I had that issue with a vga cable that didn't have resistors applied to the signal so it would desync on peak rgb levels (usually white flashes).

Not sure about the zoomed in picture, is this in widescreen? how does 4:3 look? Just wondering how it's handling the overscan.

It's When going in the channel (without launching it). Not backing out. It also occured during a play of Mario Galaxy when I chose a galaxy and the "choose a star" option came up.

Besides the case being reversed when opoening the channel, the problem seems similar to the video minus the clicking noises.