What games did you want to see most on the 32X?

[sheath]

Many intelligent, fair minded people I know prefer genres currently popular and have limited tolerance for playing "old" games (some more, some less depending on the case). I've got some friends from the retro game community who are the opposite: rather turned off by modern genres. (and some who really dislike the early simple games -ie 1s/2nd gen consoles- but really like the stuff popular in the 3rd/4th gen -I know Tomaitheous tends to get bored playing most games from the 2600's period)

I consider it two different things to have a limited tolerance for a particular generation's games and having no tolerance at all for classic genres. Think about it, what if in the US baseball, football and basketball all got replaced by table tennis because that is all people would pay to watch? This would be a bad development wouldn't it?

Define "classic" . . . almost every generation (or every other generation) ends up leaving some major genres in the dust (or certain art/sound styles), and certain people will inevitably be disappointed by that (but many, many more will become more interested in the new). You've got people who never really went beyond the old 70s/80s style console/arcade games who really dislike stuff from the NES/MD/SNES era, etc, etc.

I'll stick to the automobile definition of anything 20 years or older. I am specifically talking about genres that just don't get made anymore, not platforms or franchises. Now, I know that every generation of hardware had hold overs who never moved on, or lost interest entirely after that. These people aren't responsible for genres falling completely out of production, the masses are. Supply is only created if there is a demand, and the masses simply do_not want anything they consider "old" anymore.

That's the nature of the market, things never stay the same. (the 3rd and 4th gen are perhaps the longest period where genres/interests -and overall market size- stayed unchanged, for consoles at least -PC gaming evolved hugely in that time)

It is, that doesn't mean I have to like it.

[kool kitty89]

I consider it two different things to have a limited tolerance for a particular generation's games and having no tolerance at all for classic genres.

That's not really the case though. "Classic" genres never died, they just became extremely niche. That and there being so many older games in those genres, that much of that niche is already filled (especially since the majority of people who like those genres are also tolerant of playing games from old/outdated hardware -though there are remakes too, so filling that even more -not to mention remakes of more recent games in 3D genres done in updated/modern graphics, thus making them more marketable again -and why not? they were good games to begin with, so why not make them interesting for the masses who prefer up to date graphics? -if some older genres got really popular again, you'd see a lot of remakes/sequels to those too)

You didn't see some of this in past generations because the market wasn't as big or as old, but today you have a massive cumulative back library of almost any genre conceivable.

I'll stick to the automobile definition of anything 20 years or older. I am specifically talking about genres that just don't get made anymore, not platforms or franchises.

Please name a specific genre (and not an extremely obscure sub-genre) that isn't made at all anymore. (including small-time/niche/indie titles, and such)

Now, I know that every generation of hardware had hold overs who never moved on, or lost interest entirely after that. These people aren't responsible for genres falling completely out of production, the masses are.

Well obviously . . . it's those few exceptions of "holdovers" who get left behind, or go to totally different segments of the market (I have one friend who was a PC-only gamer until the dreamcast and then went back to PC games alone after the DC died -he also don't care much for retro games, especially 2D stuff, though he does share interests in some of my favorite genres -like flight/space sims).

Supply is only created if there is a demand, and the masses simply do_not want anything they consider "old" anymore.

This was always true though, nothing changed in recent years. Many genres (or specific sub-genres) were seen as outdated with each passing generation. The bread and butter genres of the 2600 were niche by the time the 3rd gen hit its 1989/1990 peak and almost nonexistent in the heat of the 4th generation. Likewise the bread and butter games of the late 80s and early 90s faded in the 5th gen and became even more niche after that. (and many of those niches were filled with compilations, re-releases, remakes, and basic sequels/clones)

Except today you've actually see better coverage of niche genres via small indie developers and DLC games (including some of the genres that were big on the VCD -like fixed screen arcade style games). Amateur/fan/homebrew games available online are also rather significant. (be it online flash games, PC executables, homebrews/hacks for older systems)

It is, that doesn't mean I have to like it.

I don't like change either, but it's silly to get worked up about the world not matching your own preferences (we're talking about entertainment here, not talking about social inequities ).

I may not directly share preferences with some of the genres you do, but as I already mentioned I can certainly call up examples of genres that have largely faded away. (like space sim combat games)
I'm just happy I've got so many old games to look into, and that I'm the sort of person who can still enjoy old games rather than being turned off by the primitive nature of them relative to modern games. (especially if that's limited to graphics) Albeit, many of those games fit more with the trends still popular with modern games (3D, cinematic/engaging storyline with substantial cutscenes and voice acting, etc).

Personally, I get a lot angrier over things like Nintendo's treatment of developers in the 80s and 90s (well, to some extent to this day) among other things.
Hell, in that context, Sony was pretty much the polar opposite of that, largely following in Sega's footsteps -especially SoA/SoE- in terms of 3rd party relations, that was one of the keys to Sony's success -on top of good management and tons of money to back it up. (even the developers Sony bought out tended to fare much better than anything close to that with Nintendo or -especially- Microsoft . . . like how RARE was losing a significant amount of talent due to frastration with the way the Nintendo partnership was going -MS's buyout was far worse obviously)
In the context of the PSX, one of the few things that Sony actually detracted from with 3rd parties was the bias towards 3D games (ie favoring developers who pushed 3D more and not giving much interest/support to those pushing 2D games), but even that was nowhere near as harmful as things like Nintendo had done. (probably not much different than Sega of America's promotion of FMV -except 3D ended up being a much better long-term investment)

[Curryman123]

Street Fighter Alpha would have been nice.

[sheath]

Ack, the What's New tab helped me mess this response.

That's not really the case though. "Classic" genres never died, they just became extremely niche. That and there being so many older games in those genres, that much of that niche is already filled (especially since the majority of people who like those genres are also tolerant of playing games from old/outdated hardware -though there are remakes too, so filling that even more -not to mention remakes of more recent games in 3D genres done in updated/modern graphics, thus making them more marketable again -and why not? they were good games to begin with, so why not make them interesting for the masses who prefer up to date graphics? -if some older genres got really popular again, you'd see a lot of remakes/sequels to those too)

I am considering a genre that gets less than one or two quality releases a year as good as dead here.

You didn't see some of this in past generations because the market wasn't as big or as old, but today you have a massive cumulative back library of almost any genre conceivable.

Please name a specific genre (and not an extremely obscure sub-genre) that isn't made at all anymore. (including small-time/niche/indie titles, and such)

Right, I'm exaggerating, very few genres have absolutely completely and totally disappeared. Beat-em ups is what I was talking about originally, and I would argue that with very rare and often poor examples, beat-em ups is a dead genre. Space sims is another, Arcade action in general is frowned upon in today's market and yet these games easily rival today's RPGs for potential total play time.

Well obviously . . . it's those few exceptions of "holdovers" who get left behind, or go to totally different segments of the market (I have one friend who was a PC-only gamer until the dreamcast and then went back to PC games alone after the DC died -he also don't care much for retro games, especially 2D stuff, though he does share interests in some of my favorite genres -like flight/space sims).

My argument can be summed up nicely with "It's THEIR fault THEY don't like what I like."

This was always true though, nothing changed in recent years. Many genres (or specific sub-genres) were seen as outdated with each passing generation. The bread and butter genres of the 2600 were niche by the time the 3rd gen hit its 1989/1990 peak and almost nonexistent in the heat of the 4th generation. Likewise the bread and butter games of the late 80s and early 90s faded in the 5th gen and became even more niche after that. (and many of those niches were filled with compilations, re-releases, remakes, and basic sequels/clones)

I disagree, the mainstay 2600 genres were simply expanded upon with the advent of hardware supported full background scrolling in later generations. Overhead shooters, Side Scrolling shooters, Adventure games, Platformers, all of these were expanded upon. Even stuff like Space Invaders is basically a vertically scrolling shooter without the scrolling.

Now, I will admit that I always thought Atari fans who couldn't get into later 8-bit gen and 16-bit generation games were a bit eccentric. To me it was all just a march of progress in a very real and visible sense, everything about the games improved. Today we have controller lag graphics and cinematic fests stealing the sales of actual games.

Except today you've actually see better coverage of niche genres via small indie developers and DLC games (including some of the genres that were big on the VCD -like fixed screen arcade style games). Amateur/fan/homebrew games available online are also rather significant. (be it online flash games, PC executables, homebrews/hacks for older systems)

Yeah, I have found indie games pretty hard to sort through so far and often times disappointing casual stuff.

I don't like change either, but it's silly to get worked up about the world not matching your own preferences (we're talking about entertainment here, not talking about social inequities ).

I may not directly share preferences with some of the genres you do, but as I already mentioned I can certainly call up examples of genres that have largely faded away. (like space sim combat games)
I'm just happy I've got so many old games to look into, and that I'm the sort of person who can still enjoy old games rather than being turned off by the primitive nature of them relative to modern games. (especially if that's limited to graphics) Albeit, many of those games fit more with the trends still popular with modern games (3D, cinematic/engaging storyline with substantial cutscenes and voice acting, etc).

I do consider this a social iniquity. It is as bad to me as if all Art forms of all kinds were shut down in favor of iPhone etchescetch art due to some modern fad.

Personally, I get a lot angrier over things like Nintendo's treatment of developers in the 80s and 90s (well, to some extent to this day) among other things.
Hell, in that context, Sony was pretty much the polar opposite of that, largely following in Sega's footsteps -especially SoA/SoE- in terms of 3rd party relations, that was one of the keys to Sony's success -on top of good management and tons of money to back it up. (even the developers Sony bought out tended to fare much better than anything close to that with Nintendo or -especially- Microsoft . . . like how RARE was losing a significant amount of talent due to frastration with the way the Nintendo partnership was going -MS's buyout was far worse obviously)

Well yeah, it is essentially Nintendo's sole fault that the Master System, and TG16 "failed" in the US and the Genesis took until 1991 to gain momentum.

In the context of the PSX, one of the few things that Sony actually detracted from with 3rd parties was the bias towards 3D games (ie favoring developers who pushed 3D more and not giving much interest/support to those pushing 2D games), but even that was nowhere near as harmful as things like Nintendo had done. (probably not much different than Sega of America's promotion of FMV -except 3D ended up being a much better long-term investment)

This is why I blame Sony and the megapublishers for the decline of 2D genres in the mainstream. But I really blame the masses for being so easily amused.

[Josh]

I always thought it would be cool to see Art of Fighting or Super Street Fighter II on the 32X.

[Obviously]

This is why I blame Sony and the megapublishers for the decline of 2D genres in the mainstream. But I really blame the masses for being so easily amused.

At least handheld gaming and digital downloads and general nostalgia (which you can thank the same masses for) managed to cause a modern resurgence. They might not be triple-A titles but there's a lot of great new 2D stuff out there.

Also Art of Fighting definitely would've been better served on the 32X. I would've liked to see more fighting games in general.

[Josh]

I've always wondered what upgrades or enhancements would have been possible over the Sega Mega Drive version of Art of Fighting had the game been released on the 32X, or even the Mega CD?

[sheath]

At a surface level, I'm not sure the Sega CD could have done much more than what the PCE-Arcade Card version did. For some reason, scaling in side scrolling titles was virtually non-existent. Also, the sprite sizes for Art of Fighting are huge. Though combining multiple sprites should have been possible for the Genesis or Sega CD, I have no idea how much extra cost this would cause in ROM, RAM or VRAM for animations.

I think even the 32X would have suffered some cuts in this area due to RAM and ROM limitations. The 32X should have been more than capable of acing the SNES version, and may have even provided the character sizes and animation of the PCE-ACD game, but I don't see it turning out to be a perfect port of the NEO GEO game without including the Sega CD in the mix.

[Barone]

At a surface level, I'm not sure the Sega CD could have done much more than what the PCE-Arcade Card version did. For some reason, scaling in side scrolling titles was virtually non-existent. Also, the sprite sizes for Art of Fighting are huge. Though combining multiple sprites should have been possible for the Genesis or Sega CD, I have no idea how much extra cost this would cause in ROM, RAM or VRAM for animations.

Sega CD couldn't have done what the PCE CD using the Arcade Card did.
Why? (Please, Chilly, kool kitty and others... Correct me if I'm saying shit.)

1)Arcade Card offered a truckload of RAM... 16Mbit! (+2Mbit from the Duo system - let's consider just the Duo configuration to avoid useless complexity)
2)Sega CD didn't have memory expansions (it didn't have slots for it, HuCard technology was fast enough for that though...) so it's always the stock RAM: 4Mbit (program) + 2Mbit (word) + 0.5Mbit from the Genesis system RAM.

So, at a first look we have: 18Mbit (PCE CD + Arcade Card) VS 6.5Mbit (Sega CD).

3)Sega CD has 512Kbit for PCM (read "uncompressed") samples.
4)PCE CD was designed with 512Kbit for ADPCM (read "compressed") samples (in theory, it could store two times what Sega CD could in terms of samples). BUT I guess that PCE sound architecture can access the RAM in the HuCard slot (like the Arcade Card), so it could use the extra RAM to store lots of samples. OTOH Sega CD's sound chip is limited to the tiny 512Kbit for PCM samples...It can't access the Genesis side of things...

So we have two major problems: Sega CD can't store as much sprites and background details as PCE CD using Arcade Card for a single level.
Sega CD can't store as many samples as PCE CD (even more using Arcade Card).

Sega CD has just the processor advantage and the parallax provided by VDP.
What about the scaling features provided by its ASIC?
I think it could handle things as simple trackside objects and stupid cars in games like Jaguar XJ220 (but remember how much slowdown you take in the two player mode). But I've never saw it scaling big well animated sprites like we need in a fighting game. Heck, it would have to resize the whole background and the fighters while they were moving and possibly using their special attacks (You can talk about SNES but it used much less zoom, just a few sprites of animation and smaller ones than in the original AOF. And a BIG black border... Also it runs @256x224 resolution). It couldn't do it.
IMO Genesis VDP didn't have enough bandwidth, Sega CD didn't have enough ram and its ASIC wasn't fast enough for that.

I think even the 32X would have suffered some cuts in this area due to RAM and ROM limitations. The 32X should have been more than capable of acing the SNES version, and may have even provided the character sizes and animation of the PCE-ACD game, but I don't see it turning out to be a perfect port of the NEO GEO game without including the Sega CD in the mix.

Chilly is the one to answer this...
But I think that 32x could handle an almost perfect port of the Neo Geo, but it would require at least a 64Mbit (if more is possible it would be better, the original game has 102Mbit but it's probably totally uncompressed) cartridge. However, I think the sprites would have to be reduced 'cause even using uncompressed stuff in the ROM (32X RAM is not enough IMO) (or efficient on the fly decompression - ???) its cartridge slot didn't have the same bandwidth of the Neo Geo's one (it's not near the same indeed, but AoF is one of the earlier Neo Geo titles and it's not so scary as those later "Giga Power" titles... So I think 32x could handle it quite well).

32X + CD? I don't think so... It would be so complex to develop for this and I doubt that it would help with the problems that I've described. However, Chilly might have some great ideas for it.

[netpredakonn]

Any RPG would have been great. I can't believe there aren't any for the 32x. Maybe this happenede because it takes a while to develop a rpg, and the 32x was dead before they could be released. I don't know, just a guess, but is really a shame.

Shining Force would have been a great choice to start. They could outdo the graphics of any Snes rpg easily and stablish the 32x as something to look forward to.

[Chilly Willy]

Chilly is the one to answer this...
But I think that 32x could handle an almost perfect port of the Neo Geo, but it would require at least a 64Mbit (if more is possible it would be better, the original game has 102Mbit but it's probably totally uncompressed) cartridge. However, I think the sprites would have to be reduced 'cause even using uncompressed stuff in the ROM (32X RAM is not enough IMO) (or efficient on the fly decompression - ???) its cartridge slot didn't have the same bandwidth of the Neo Geo's one (it's not near the same indeed, but AoF is one of the earlier Neo Geo titles and it's not so scary as those later "Giga Power" titles... So I think 32x could handle it quite well).

32X + CD? I don't think so... It would be so complex to develop for this and I doubt that it would help with the problems that I've described. However, Chilly might have some great ideas for it.

The minimum cycle time on reading the cart by the SH2 is 7 clocks. The SH2 clock is 3 times the 68000 clock. The 68000 takes a minimum of 4 cycles to read the cart. So the 68000 takes 4*3 SH2 cycles per word to the SH2's 7 cycles per word, meaning the SH2 is not quite twice as fast reading the cart as the 68000. This is pretty good throughput, and the data can also be cached by the SH2. The SH2 can also use DMA to transfer the data.

Given that, I think uncompressed data from the cart would be fine for most any program. My Yeti demos and Wolf3D both read textures from the cart directly. If compression is used, that would slow things a bit, but you could use compression that was fast to decode, or maybe use the Slave SH2 to decompress into a cache ahead of when you needed the data. So there are ways to work around the issues. Running from CD is more an issue since then you have no cart space to pull data from - all you have is the 256KB of SDRAM, which must now not only hold the data, but ALL the code as well. SEGA REALLY needed more SDRAM in the 32X, or at the least put out a ram cart to plug into the 32X for CD games to have storage for code/data pulled off the CD.

I think SEGA was thinking that (had the 32X continued into the late 90s) devs would use the SEGA mapper (often called the SSF2 mapper) to have games up to 32 MBytes. This would have matched a lot of N64 carts. You get that feeling because the 32X SDK covers the SEGA Mapper, and the 32X BIOS has routines for setting the mapper built into it. SEGA knew certain games would need a BIG cart and they tried to make it available to developers from the start. So you can't fault them on that aspect, even if more ram would have been better for some things.

[sheath]

Most of the 32X library is made up of launch games that were hastily thrown together in less than 6 months, and the same that got delayed a few months. Honestly, considering the time frame, the 32X library is arguably better than previous systems as far as its launch lineup goes.

[kool kitty89]

The minimum cycle time on reading the cart by the SH2 is 7 clocks. The SH2 clock is 3 times the 68000 clock. The 68000 takes a minimum of 4 cycles to read the cart. So the 68000 takes 4*3 SH2 cycles per word to the SH2's 7 cycles per word, meaning the SH2 is not quite twice as fast reading the cart as the 68000. This is pretty good throughput, and the data can also be cached by the SH2. The SH2 can also use DMA to transfer the data.

Didn't you mention that the 32x could be configured to access ROM (or RAM) from the cart slot as fast as SDRAM? (though, obviously, that wasn't going to happen with affordable ROM back in the 90s -granted, NEC had been using quite fast ROMs on the PCE since day 1, but they had in-house manufacturing as a major advantage there)

What speed does the Neo Myth or Everdrive allow? (on that note, it was mentioned that the Skunkboard flash cart for the Jaguar is quite fast . . . fast enough to make texture mapping from ROM to DRAM faster than DRAM to DRAM -due to lack of DRAM page breaks)

Sega CD couldn't have done what the PCE CD using the Arcade Card did.
Why? (Please, Chilly, kool kitty and others... Correct me if I'm saying shit.)

1)Arcade Card offered a truckload of RAM... 16Mbit! (+2Mbit from the Duo system - let's consider just the Duo configuration to avoid useless complexity)
2)Sega CD didn't have memory expansions (it didn't have slots for it, HuCard technology was fast enough for that though...) so it's always the stock RAM: 4Mbit (program) + 2Mbit (word) + 0.5Mbit from the Genesis system RAM.

You can use RAM expansion via the cart slot, though it would be directly accessible only to the Genesis VDP/68k/Z80 (or 32x if attached). That's compared to the PCE CD that only every has RAM on the CPU bus anyway (aside from the ADPCM bank) . . . MD has the whole CD subsystem bus for the ASIC/sub-CPU/etc. (actually several buses due to word RAM, program RAM, PCM RAM, etc -though the MD CPU and VDP can directly access all of word RAM and program RAM, albeit in segments)

RAM on the cart slot actually adds a number of advantages over RAM added to the CD subsystem directly. It means flat/linear addressing of up to 4 MB of RAM for the MD VDP/CPU (and as 32k banks to the Z80), and access to those totally independent of MCD RAM (no compromises of sharing MCD RAM). So that could potentially complement the MCD quite well, with some things loaded into the RAM cart (like most graphics/animation) and other things in CD RAM (like game code/data, PCM samples, etc), plus a limited amount in the MD VRAM/work RAM/Z80 RAM.
There are even greater advantages for the 32x, which is horribly bottlenecked for accessing data from CD memory. (the 32x has no access to the CD subsystem directly, so the MD 68k must manually copy from MCD RAM to 32x RAM . . . that's ~1.5 MB/s tops, and the 68k likely won't have 100% of time available for that, not to mention the added overhead for the CD and 32x having transfers so often)

Of course, whether such a RAM cart would have been marketable is another story. (especially as an add-on for the 32x . . . which really just points to the 32x needing more RAM in the first place -assuming you'd even want to support the 32x at all rather than halting that project it in favor of other things prior to launch)
Maybe a lock-on RAM cart for the Genesis (as well as CD) would have been more practical. (though the CD really had a decent amount of RAM already . . . not a massive amount like the Arcade card, but enough to be competitive with most larger MD games with good optimization -unlike the Super CD's 256k, which was just too limited -NEC really should have pushed for 512k . . . the arcade card came too late and was rather expensive on top of that)

From a realistic standpoint, Sega couldn't have had anything close to the arcade card . . . maybe a 512k RAM cart, but nowhere near 2MB. (which would have cost over $50 just for the components -not counting manufacturing/distribution/etc, and probably close to $100 at retail, unless they sold at a significant loss) RAM prices were also stagnant from 1992 through 1995 (started dropping again in 1996), and actually, common (ie not high-end/niche) DRAM grades were actually a bit cheaper in 1992 than 1995.

http://phe.rockefeller.edu/LogletLab/DRAM/dram.htm
(note figures are in bits/kbits/Mbits not bytes -so 256k is 32kB and 4M is 512kB; the MCD used 2 1Mbit DRAMs and 1 4Mbit DRAM for word and program RAM -the rest was SRAM/PSRAM)
So a RAM cart in 1992 actually wouldn't have been cheaper than one in 1994 or '95. (or, even in 1991, for the PC Engine, having the Super CD card use 1MB of DRAM would have been significantly cheaper than the Arcade card in 1994 . . . let alone 512k in 1991)

3)Sega CD has 512Kbit for PCM (read "uncompressed") samples.
4)PCE CD was designed with 512Kbit for ADPCM (read "compressed") samples (in theory, it could store two times what Sega CD could in terms of samples). BUT I guess that PCE sound architecture can access the RAM in the HuCard slot (like the Arcade Card), so it could use the extra RAM to store lots of samples. OTOH Sega CD's sound chip is limited to the tiny 512Kbit for PCM samples...It can't access the Genesis side of things...

There's a dedicated DMA channel for updating PCM samples though, so paging in added sample data from program RAM isn't that big of an issue. (far, far less problematic than updating the SNES's sample RAM)

ADPCM is limited too, aside from the PCE being limited to 1 (mono) channel of ADPCM (vs a full 8 channel sample synth chip with stereo panning), there's the actual limits of the ADPCM format. You can only play compressed samples vs PCM allowing software decompression (or even mixing) in a variety of formats, and the sub-CPU provided a good amount of resource for that (assuming it's not doing other intensive effects -like 3D/pseudo 3D, graphics/FMV decompression, etc). Plus, compression does degrade the sample, so uncompressed PCM can sound better. (albeit usually at higher bitrates -which would include streaming audio from CD, like some games do for music/speech to save space over CD-DA, including Pier Solar's entire soundtrack done in 32 kHz stereo 8-bit PCM)

Sega CD has just the processor advantage and the parallax provided by VDP.

And a more powerful sprite engine, and a more powerful sound system. (FM+PSG and a full 8 channel sample synth chip vs 6 PCE chip channels and mono 1 channel ADPCM playback -rarely even used in music and unable to play simultaneous speech/SFX -though you could use software PCM playback like plain ol' hucard games )

What about the scaling features provided by its ASIC?
I think it could handle things as simple trackside objects and stupid cars in games like Jaguar XJ220 (but remember how much slowdown you take in the two player mode). But I've never saw it scaling big well animated sprites like we need in a fighting game. Heck, it would have to resize the whole background and the fighters while they were moving and possibly using their special attacks (You can talk about SNES but it used much less zoom, just a few sprites of animation and smaller ones than in the original AOF. And a BIG black border... Also it runs @256x224 resolution). It couldn't do it.
IMO Genesis VDP didn't have enough bandwidth, Sega CD didn't have enough ram and its ASIC wasn't fast enough for that.

You can scroll the scaled layer . . . but the issue is DMA bandwidth not to much ASIC rendering speed. (especially if you want near-full screen -the more you letterbox, the more DMA bandwidth the VDP has to copy animation to VRAM -and that's what the ASIC is doing, rendering animation frames in realtime -the VDP would be able to copy prerendered animation no faster)


For something like Neo Geo games using BG and sprite panning effects, you'd probably do something like this: render scaled BG frames to the tile layer (perhaps with an unscaled 2D layer for the far BG) and scroll/pan the scaled layer normally using the VDP (so zoom with the ASIC, pan with the VDP), and then use hardware sprites for the sprites and scale animation for those too. (probably limit the zooming range to reduce overhead and prevent sprites from getting too big and having flicker/drop-out issues)

You'd also need to limit the scaled BG to a single set of 15 or 16 colors since you'd be using it like a framebuffer (per-tile color attributes wouldn't be practical to use) . . . then again, you usually limit BG to 2 palettes for fighting games anyway (with 1 palette dedicated to each fighter -and optimized for each fighter), but you also can't really use dithering without weird artifacts. (Samurai Showdown on the MCD already uses a 16 color BG, implying that they may have intended to use the ASIC for scaling but dropped the idea)

You'd really be best off to sacrifice a bit of screen size for more DMA time. (SFII already cut it to 200 lines, and going to 192 lines like Virtua Racing doesn't look much worse -and nearly doubles bandwidth over 224 lines) PAL has much more vblank time per frame by default, so less clipping would be necessary for similar bandwidth. (albeit clipping the screen also means less VRAM space to buffer the ASIC rendered bitmap layer) At Virtua Racing's resolution, in NTSC, you might manage 20 FPS scaling/animation, but it would be tight. (184 lines would be more realistic -which is about what BC Racers uses -for the whole screen inside the letterboxing, not the gameplay window) If optimized for PAL, 25 FPS should be realistic at 256x192.

In any case, the MCD is still a lot better off for doing realtime scaling than the PCE+Arcade card would be trying to do pseudo scaling with animation.

[Chilly Willy]

Didn't you mention that the 32x could be configured to access ROM (or RAM) from the cart slot as fast as SDRAM? (though, obviously, that wasn't going to happen with affordable ROM back in the 90s -granted, NEC had been using quite fast ROMs on the PCE since day 1, but they had in-house manufacturing as a major advantage there)

Well, you obviously can't set the cart region to burst mode since it is not wired for it, but you could drop the mandatory wait states. SEGA, of course, tells you to leave the timing at whatever the BIOS sets it to.

What speed does the Neo Myth or Everdrive allow? (on that note, it was mentioned that the Skunkboard flash cart for the Jaguar is quite fast . . . fast enough to make texture mapping from ROM to DRAM faster than DRAM to DRAM -due to lack of DRAM page breaks)

I've not tried it. There are a variety of tests on my list for when I get the time. Just when it seems I have a little free, something comes up - for example, today I spent most of my free time replacing the guts in a toilet... that's a LOT more work than you would think.

[tomaitheous]

Sega CD couldn't have done what the PCE CD using the Arcade Card did.
Why? (Please, Chilly, kool kitty and others... Correct me if I'm saying shit.)

1)Arcade Card offered a truckload of RAM... 16Mbit! (+2Mbit from the Duo system - let's consider just the Duo configuration to avoid useless complexity)
2)Sega CD didn't have memory expansions (it didn't have slots for it, HuCard technology was fast enough for that though...) so it's always the stock RAM: 4Mbit (program) + 2Mbit (word) + 0.5Mbit from the Genesis system RAM.

So, at a first look we have: 18Mbit (PCE CD + Arcade Card) VS 6.5Mbit (Sega CD).

3)Sega CD has 512Kbit for PCM (read "uncompressed") samples.
4)PCE CD was designed with 512Kbit for ADPCM (read "compressed") samples (in theory, it could store two times what Sega CD could in terms of samples). BUT I guess that PCE sound architecture can access the RAM in the HuCard slot (like the Arcade Card), so it could use the extra RAM to store lots of samples. OTOH Sega CD's sound chip is limited to the tiny 512Kbit for PCM samples...It can't access the Genesis side of things...

So we have two major problems: Sega CD can't store as much sprites and background details as PCE CD using Arcade Card for a single level.
Sega CD can't store as many samples as PCE CD (even more using Arcade Card).

Sega CD has just the processor advantage and the parallax provided by VDP.
What about the scaling features provided by its ASIC?
I think it could handle things as simple trackside objects and stupid cars in games like Jaguar XJ220 (but remember how much slowdown you take in the two player mode). But I've never saw it scaling big well animated sprites like we need in a fighting game. Heck, it would have to resize the whole background and the fighters while they were moving and possibly using their special attacks (You can talk about SNES but it used much less zoom, just a few sprites of animation and smaller ones than in the original AOF. And a BIG black border... Also it runs @256x224 resolution). It couldn't do it.
IMO Genesis VDP didn't have enough bandwidth, Sega CD didn't have enough ram and its ASIC wasn't fast enough for that.

I'd say it's pretty much just the huge amount of ram that gives the PCE ACD the advantage. As far as scaling, the Genesis also has multiple resolutions like the PCE, so I don't see why they couldn't have done something very-very similar. On the SNES, since you don't have multiple 'real' resolutions to use, you can use a trick that involves folding/moving the sprites cells in on each other to condense the sprite size. Compact it. The SNES large sprite table size fits nicely for this and the BG layer already has the ability to be mode 7. Although the BG details are lacking because the BG can only be made up of a total of 256 8x8 tiles (I believe only 8x8 tiles, I don't think 16x16 tile mode works for mode 7 but I could be wrong). The scaling of the title screen in the PCE ACD version uses the same sprite compact trick, although in reverse motion to give the illusion of scaling. But in game the sprites themselves are individually updated.

Are there missing samples on the SegaCD version during the game? The ADPCM ram is I/O port based, not linear mapped on the PCE. You can read/write the port, but access needs a delay of like 20 something cycles if the ADPCM is also currently playing a sample. Most games just upload the cache of samples it needs and leaves it there for the level. Though some games treat it like far/slow ram and store game data there. Given the large amount of ACD ram, I doubt this is the case with any of the SNK ports.