Discuss alternate realities here!

[Joe Redifer]

Here is my alternate reality:

Sega never stops manufacturing the EXT "modem" port on the back of the Genesis. The 32X is released and plugs into the EXT port. The processor chips are designed by me and can do cool things like scaling and rotation and not just a wimpy 16 million colors, but 17 million because more is better. It has 100 TB of RAM and each capacitor has its own RAM as well (for no reason). Sega goes into the future and steals ALL of the bandwidth from the year 2036. So that means each and every Genesis + 32X has ALL of the bandwidth that is available everywhere/all year from 2036. The processor isn't just a wimpy 30Mhz or whatever it was, but all of the Mhz from the year 2027 is stolen and added to each chip inside both the Genesis and 32X (you want to make sure you have more bandwidth, just in case, ya know?) It could calculate the full value of Pi in less time than it takes to draw a single pixel of any color. Also, there would be a blinking light on the 32X itself because that is how goddamn powerful it would be.

[KnightWarrior]

Have the 32X releaed in 93..

Market it as a Small Arcade System & up to par with the SNES

Better Poly Count for games like VR & VF

[sheath]

Bhahah! I like Joe's alternate history the best!

On Kool Kitty's comments, the Y board and "Turbofied" Sega CD scenarios were specifically designed around historical trends. On the Y Board portion, Sega both tended to design their console hardware using their arcade designs and custom design their console video processors. By 1991 the Y Board had been around for a few years, and more importantly had games being made for it by Sega that they promised would come out for Sega CD in enhanced form.

US magazines, from 1990-92, were harping on a System 32 based console/add-on but everything they heard about involved 68000 CPUs. So, the Y Board is the most logical choice and the most historically likely hardware for Sega to base an additional VDP on for the Sega CD. We've already discussed this, but I'll reiterate that Tomio Takami was originally charged to create a CD-ROM for the Genesis, that didn't load as long or as frequently as the PCE CD did, for $150 cost.

By 1991 Sega's research indicated that low price didn't matter as much probably due to PCE CD sales in Japan mostly. Thus, Sega asked Takami to add additional scaling capabilities to the Sega CD. The result was an additional 68000, all of the RAM and ROM (totaling 984KB), the graphics coprocessor/ASIC, the D/A converter, and the Ricoh sound chip for a total cost of $380.

There is really no way that adding a Video processor and all of the necessary hardware for meshing Genesis AV and outputting Sega CD AV would not have shot the cost over $450. I was being generous and assuming that some cuts might have been made to facilitate that addition and minimize costs. Keep in mind, the NEO GEO was still $700 and we're talking about making the Sega CD *more capable*.

Now, with that historical cost perspective laid out, the 1990 "Turbo CD" style Sega CD should look more realistic as well. The Turbo CD retailed in 1990 for $399 remember, and that was the original Turbo CD with only 64kb of RAM. How in the world was Sega going to come in for less than that with 256Kb of RAM? Similarly, in 1992 the Turbo DUO launched at $299, the same as the Sega CD later launched. Earlier that year the PC-Engine DUO with the 256KB of RAM listed for $460 in Japan but was discounted to $330 [fn]"Electronic Gaming Express... NEC to Bring Out U.S. TurboGrafx DUO in Late 1992...?," Electronic Gaming Monthly, January 1992, 48.[/fn]. In 1992 the Super System Card 256Kb upgrade listed for $90.[fn]"Electronic Gaming Express...New CD-ROM Drives Announced By NEC, Sony And Phillips," Electronic Gaming Monthly, August 1991, 38.[/fn] (I've seen footnotes in this forum, what's the code?)

With that in mind, I say there is no way Sega could have opted for 256Kb RAM in a CD-ROM designed in 1989-90. That's as much as all of the SNES' RAM combined, which launched at the end of 1990 in Japan for $200 with no pack in. Historically in 1989-90 it was either RAM or CD-ROM for consumer priced hardware. Obviously by the middle of 1991 this wasn't so much the case as the real world Sega CD had its 768Kb RAM + 64Kb PCM RAM + 16 KB
CD Cache + 8 KB Backup RAM + 128 KB ROM. We know that RAM and ROM prices dropped dramatically in 1991 and 1992 though.

[roundwars]

The Turbo CD retailed in 1990 for $399 remember, and that was the original Turbo CD with only 64kb of RAM. How in the world was Sega going to come in for less than that with 256Kb of RAM?

By not adding on a ridiculously high markup like NEC did.

[sheath]

By not adding on a ridiculously high markup like NEC did.

Source please.

[KnightWarrior]

In Joes Post, it prob going to end up costing like $400 in 94 time

I like my Idea better

[kool kitty89]

not just a wimpy 16 million colors, but 17 million because more is better. It has 100 TB of RAM and each capacitor has its own RAM as well (for no reason).

32x was only 32,768 colors. 24-bit RGB is 16,777,216 colors anyway, so if you're going to round it, it's already 17 million colors. ;P

blinking light

You mean like this? (5:35)
http://www.youtube.com/watch?v=V6e6r-NOq5U

Have the 32X releaed in 93..

Market it as a Small Arcade System & up to par with the SNES

Better Poly Count for games like VR & VF

No... don't market it as competition for the SNES, the Genesis already does that fine (with trade-offs), market it as more than that. (particularly as anythign liek the 32x is not going to be most impressive at 2D relative to contemporary consoles, but 3D and other software rendering -anythign the competition can't do in hardware) I mean, point out the better 2D as well, but the most dramatic stuff is the polygons/voxels/ray-casting/textures/scaling/rotation/etc and audio, though on the historical 32x they screwed that up anyway. (no documentation of the DMA support and some other things)

Unless you make the 32x totally different, not based on CPU grunt in the same manner it was. (in 1993 you wouldn't be using the SH2s anyway, maybe an SH1 or something else: maybe SH1+SVP if they'd already licensed the SSP-1601 from Samsung, or other things)
I already speculated on a possible 1993 32x (or alternate 32x in general) in previous threads, but already here too:

In the context of the minimalistic CD, a second add-on would be FAR more attractive and efficient. The 32x as it was could have been fine, though with more optimization and catering to specific features, probably something a bit different from the 32x would be preferable. (especially something a bit cheaper in some areas and a bit less limited in others) Pushing such an add-on by 1993 might make more sense too. (probably an SH1 at that point, maybe the SVP too if they had it already, but especially some hardware acceleration -something like the historical CD's ASIC would be really nice for 2D and 3D -especially texture mapping --maybe just the blitter/VDP+SH1 OR blitter/VDP+SVP -in the latter case have the Genesis CPU handle all the game logic)
If RAM expansion hadn't yet been implemented for the CD beyond the onboard 256 kB, definitely have that as a significant feature of such a Mars/32x design. (you've got the framebuffers/word RAM for sure, but then probably 1 MB on top of that, maybe 512 kB) Probably some sound expansion too, for those who didn't have a CD already.

We sort of did some of this before here too: http://sega-16.com/forum/showthread.php?t=10761 (1992)

And honestly, in the context of the minimalist CD drive in 1990 (JP), you might just jump to a full new (backwards compatible) console in 1993/94, or have an add-on that makes the MD+CD/Duo into the full nextgen platform. (not quite as flexible/efficient as skipping straight to a full compatible evolution of the existing hardware without limits of the interfacing available -let alone a clean slate)

Otoh, had Sega had an expansion bus like Hudson put in the PCE, then there wouldn't even really be any limitations to making it an add-on and full nextgen console. (as it is, they could easily have made the SGX an add-on for the PCE rather efficiently -especially integrated with the CD/duo, and even the PCFX could be handled that way -well not if the expansion bus is missing from the Duo...)


And another thread in the vein of this thread to some extent that I forgot about before: http://sega-16.com/forum/showthread.php?t=11237 (How Ancient was the Genesis really?)

By not adding on a ridiculously high markup like NEC did.

Got that right... NEC seemed inept at marketing in the US... too bad it wasn't until 1992 that Turbo Technologies was created. (better marketing would have made a world of difference for the TG-16)
Not releasing the TG-16/PCE to Europe was another mistake by NEC. (hell they probably should have put priority on Europe given the marketing strategies used in Japan and the poor conversion to the US -could have worked a lot better in Europe/UK, especially in the compact original PCE form factor)

On Kool Kitty's comments, the Y board and "Turbofied" Sega CD scenarios were specifically designed around historical trends. On the Y Board portion, Sega both tended to design their console hardware using their arcade designs and custom design their console video processors. By 1991 the Y Board had been around for a few years, and more importantly had games being made for it by Sega that they promised would come out for Sega CD in enhanced form.

Sega never used directly arcade derived hardware in ANY home console. They did do the opposite though: repurpose their consoles as low-cost arcade boards. (even the rather wimpy, dated, and completely off the shelf SG-1000 hardware)

The Y board would NOT have been realistic cost-wise. What they did in the Sega CD is a low-cost consumer oriented solution, and even more flexible than their sprite scaling hardware in some ways. (it's a flexible blitter, does hardware bitmap to tile/sprite conversion, renders rectangular bitmap objects that can be scaled and rotated, and by extension, can be useful for actual texture mapping and warping/rotation effects -as seen in the bios and some games, be it "mode 7" like stuff or some texture mapped polygons)
None of those arcade boards could rotate in hardware afik, and weren't nearly as flexible. They relied on LOTs of animation (ROM) and straight (albeit very fast) scaling with no rotation or perspective warping. They had no hardware aid for rotation or texture mapping, and such hardware would be totally impractical for a CD based system. (unless you did like the Neo Geo CD and added tons of RAM)

Again, see: http://www.sega-16.com/forum/showthread.php?p=169232
On pretty much that exact topic. (and again in the thread that's suggesting video expansion for the CD)

The CD just needed a simple 256 color (possibly fixed resolution like 32x) bitmap display controller, not even with all the functionality of the extremely bare bones 32x VDP. The framebuffers are there, the ASIC is there just begging for video expansion, etc. (hell, even a 16-color bitmap display would have helped a lot -though given the effort/cost for a 256 color display would barely be any different, there's little point) Support to allow 4-bit graphics data to be used by the ASIC could be useful though (less space taken in RAM), indexed 16-color tiles/objects/textures with 256 color paletized output. (especially if it could have the 16 color palettes indexed anywhere within the 256 color palette, and not just split the 256 colors into 16 fixed subpalettes -the latter would probably be simpler)
Of course that would be on top of supporting plain 8bpp graphics data. (which would take up 2x the space in RAM uncompressed)

US magazines, from 1990-92, were harping on a System 32 based console/add-on but everything they heard about involved 68000 CPUs. So, the Y Board is the most logical choice and the most historically likely hardware for Sega to base an additional VDP on for the Sega CD. We've already discussed this, but I'll reiterate that Tomio Takami was originally charged to create a CD-ROM for the Genesis, that didn't load as long or as frequently as the PCE CD did, for $150 cost.

That's pure speculation I think, and false assumption, like Sam Pettus's claim that the original Saturn/Gigadrive in 1992 was to be directly based on the System 32 with some modest 3D capabilities added for model 1 ports.

Otoh, a home console designed with the goal of system 32 (or model 1) LIKE capabilities makes a ton more sense, but having capabilities roughly equivalent does not have anything to do with being directly related. (and for the practical purposes of a home cart/CD based platform, it would be WAY cut-back in terms of cost than any such arcade board)
As it was the Sega CD has nothing in common with any of Sega's custom arcade hardware: it does share some off the shelf hardware, but that's another issue entirely.

Also, I think almost any magazine "specs" especially "leaked specs" should be considered dubious, and even direct quotes (or especially paraphrasing) can be taken out of context.

Rumor mill outside media can be even worse and sometimes feeds back into media as "official" information.

Full interviews, original documents, etc are th eonly real facts to go on, the rest must be considered more carefully. (and any interviews which are of 2nd hand knowledge -or more- are also to be taken with such consideration)

That's how there got to be so much false history surrounding companies like Atari Inc and Atari Corp.
Unfortunately, there aren't the same sort of opportunities to get to the real truth in the case of Sega in the same manner historians have done with Atari.

By 1991 Sega's research indicated that low price didn't matter as much probably due to PCE CD sales in Japan mostly. Thus, Sega asked Takami to add additional scaling capabilities to the Sega CD. The result was an additional 68000, all of the RAM and ROM (totaling 984KB), the graphics coprocessor/ASIC, the D/A converter, and the Ricoh sound chip for a total cost of $380.

Yeah and that was a great value too, cheaper than the PCE CD at the time I think by a fair margin! Or the PCECD+ Super CD card at least. (NEC had the advantage of vertical integration too... I think they just took higher profits -and in Japan that made sense in their position -they botched that in the west along with the rest of their marketing strategies there) I believe Sega may have been selling their hardware at cost (if not in Japan, at least in the US).


It's also important to note the discrete RAM in the CD, and not lump it all in one as the small chips will be much more expensive than the large ones relative to capacity, not to mention SRAM/PSRAM vs cheap DRAM... ROM is a different category entirely. The 512 kB DRAM chip would eb the cheapest on the board (in terms of $/bit), then the 128 kB chips, then the PSRAM chips for wave RAM and CD-ROM buffer (assuming it's not SRAM -and that sega had a good deal on PSRAM), and then the 8kB SRAM for the game saves would be the most expensive on the mass market. (per bit) Then you have the proportionally higher board space consumed by smaller capacity RAM chips too.
I'd bet the 768 kB DRAM is a good bit less than 1/2 the overall RAM costs for manufacturing the CD. (as it was Sega had an edge over the mass market on the 8 kB SRAM chips and PSRAM chips with their use in the in the Genesis and SMS -and arcade... not that the SMS or Genesis wouldn't have been significantly better off with cheap DRAM instead -a bit more R&D time and more board space on early models, but the benefits were huge: more RAM for less cost and lower overall cost in the long run)

There is really no way that adding a Video processor and all of the necessary hardware for meshing Genesis AV and outputting Sega CD AV would not have shot the cost over $450. I was being generous and assuming that some cuts might have been made to facilitate that addition and minimize costs. Keep in mind, the NEO GEO was still $700 and we're talking about making the Sega CD *more capable*.

Nope, were're talking super bare bones video, the main cost would be sega actually designing the chip... if they could cram it into the same ASIC as the graphics coprocessor, that shouldn't have added much cost at all (in terms of manufacturing -R&D is another matter). Though the added RGB inputs and AV out would add some cost, of course. (worst case is you initially have another ASIC on the board for the VDC functionality and analog video mixing -unless you dropped that feature and only allowed genesis OR Sega CD video to be used with simple passthrough -and possibly switching modes in-game- but no overlay like 32x, and you could still have the ASIC render to the Genesis layer too, but the 256 color stuff would be ALL CD/ASIC/CPU rendered)

At least that's the impression going by the previous technical discussions on this...
Again, the Super VDP thread especially, but several others as well. http://www.sega-16.com/forum/showthread.php?t=3402

The fact that Sega had already been designing the CD to be mounted on the limiting expansion port completely removed any possibility of video overlay though. (they COULD have done the all or nothing thing with the added video -the overlay ads a lot more value though)

Now, with that historical cost perspective laid out, the 1990 "Turbo CD" style Sega CD should look more realistic as well. The Turbo CD retailed in 1990 for $399 remember, and that was the original Turbo CD with only 64kb of RAM. How in the world was Sega going to come in for less than that with 256Kb of RAM? Similarly, in 1992 the Turbo DUO launched at $299, the same as the Sega CD later launched. Earlier that year the PC-Engine DUO with the 256KB of RAM listed for $460 in Japan but was discounted to $330 [fn]"Electronic Gaming Express... NEC to Bring Out U.S. TurboGrafx DUO in Late 1992...?," Electronic Gaming Monthly, January 1992, 48.[/fn]. In 1992 the Super System Card 256Kb upgrade listed for $90.[fn]"Electronic Gaming Express...New CD-ROM Drives Announced By NEC, Sony And Phillips," Electronic Gaming Monthly, August 1991, 38.[/fn] (I've seen footnotes in this forum, what's the code?)

NEC most certainly was NOT selling the Turbo CD at cost when they launched it in the US... they should have, but NEC Home Electronics USA had already proven inept in grasping the North American market, so it's no surprise.
The Turbo Duo in 1992 was managed and released by Turbo Technologies Incorporated, NOT NEC Home Electronics USA. (a shame they hadn't had better marketing when they really needed in in 1989-91)
As it was the Sega CD was more expensive in some respects than the Turbo Duo: it had more RAM, but the Duo had 256 kB SRAM, more chips on the board in general, tray loading CD drive, and most importantly the lack of vertical integrated to the degree NEC had it. (mega corp manufacturing their own chips, no 3rd party vendor like Sega/Nintendo, etc) So even then, Sega was almost definitely cutting it closer with the MCD.

Any idea what the PC Engine Duo was priced at in Japan in 1991, or the PCE CD at the time for that matter? (or super CD card)
I'm willing to bet Sega cut the edge razor thin on their pricing. (in the US at the very least)

With that in mind, I say there is no way Sega could have opted for 256Kb RAM in a CD-ROM designed in 1989-90. That's as much as all of the SNES' RAM combined, which launched at the end of 1990 in Japan for $200 with no pack in. Historically in 1989-90 it was either RAM or CD-ROM for consumer priced hardware. Obviously by the middle of 1991 this wasn't so much the case as the real world Sega CD had its 768Kb RAM + 64Kb PCM RAM + 16 KB
CD Cache + 8 KB Backup RAM + 128 KB ROM. We know that RAM and ROM prices dropped dramatically in 1991 and 1992 though.

PCE CD RAM wasn't limited by cost... there were trade-offs like 2 separate RAM blocks, etc, but given the cost of DRAM at the time, NEC really should have launched the CD with at least 256 kB in '88, maybe compromised audio with that too. (maybe swap the ADPCM chip+RAM for a simple DAC+DMA chip reading from program RAM -unless the ADPCM chip could be made to share program RAM) If they kept the ADPCM chip+RAM, 128 kB main memory would have been practical in the very least.
I think it mainly has to do with NEC not realizing how limiting the RAM was, and wanting to keep a certain market model/profit margin on the hardware. (and the CD was NEC designed, not Hudson) Or hell, even make it a single 128 kB DRAM chip shared for audio and video, that would have been the cheap route. (or leave the sound expansion off entirely and sticking with conventional CPU driven PCM playback) Bumping up to 256 kB would likely have been a lot more significant than the sound expansion though (you get CD-DA either way). They could have added the sound expansion later on anyway via the card port -mono audio input on the card slot. (and PCE ADPCM is mono only anyway -could do stereo via the expansion port with a new CD-ROM dock unit -or in the duo of course)
256 kB of DRAM in 1988 wasn't much different than the cost of 768 kB DRAM in 1991. (let alone SRAM like the Super CD used...)

But that design choice on NEC's part when designing the CD. (I do wonder what Hudson might have come up with had they designed it... for that matter I think it was NEC and not Hudson that did the Supergrafx, which would explain the waste of the expansion port and not making it an add-on)

Also: RAM is not RAM is not RAM is not RAM. There's LOTS of types of RAM, slow, mass market DRAM/FPMDRAM was the cheapest, faster FPM DRAM more so, then you had EDO and a bit later SDRAM. SRAM was much more expensive than DRAM, PSRAM a bit of a compromise, but very exotic at the time and not really a performance advantage over DRAM. (just quicker/easier to interface too -shortening development time and board space, but costing a lot more in the long run)

SNES used cheap DRAM for main RAM, expensive fast SRAM for video RAM, and either SRAM or PSRAM for audio. Genesis used PSRAM for main memory and expensive VRAM (dual ported DRAM) for video. PCE used 8 kB of fast SRAM for work RAM and 64 kB SRAM for video.
Using DRAM in the MD could have been VERY significant in terms of long term cost reduction and practically increasing RAM at lower cost. (256 kB work RAM easily at lower cost than PSRAM -especially in the long run) And a good change for the VDP would have been using dual 64 kB DRAM banks rather than VRAM... (cheaper in the long run -maybe even in the short run even- and solving DMA/VRAM issues -which plagued some MD games and the Sega CD: 1/2 the reason for the video expansion is the terrible DMA/VRAM limitations)
But for all such cases, it's more development time and more board space for some things early on. (added interfacing/refresh/banking logic) Assuming they couldn't fit that on one of the ASICs already on the board. (definitely do that in later revisions -eventually integrated all into the VDP)

For that matter it's a shame the Colecovision/SG-1000/SMS didn't use DRAM for main RAM (video RAM is DRAM in all 3 I think -SMS might be VRAM), but I assume that was also due to development time and initial board space. (though the Z80 does have built-in refresh, reducing the added logic needed for DRAM) The MSX used DRAM... well and pretty much all home computers too. (and the 5200 because of that) I think the Astrocade also used DRAM, hence the large amount of RAM for a console at the time.
In some cases you need SRAM for speed that DRAM can't offer, but with a 3.58 MHz Z80 at the time, that wouldn't have been the issue at all. With the SMS and MD keeping SRAM, it might also have been compatibility related as using DRAM could have screwed up SG-1000 compatibility... Except I think the SC-3000 used DRAM too, so I'm not sure at all. (unless they kept the 2 kB SRAM for compatibility)
The need for fast SRAM with the CPU is what limited the TI-99/4 to only 256 bytes of CPU work RAM...

The PCE CD already used only DRAM, so that was pure choice on NEC's part.



SRAM is what also made the Atari Panther totally impractical: the Panther video chip needed fast SRAM to work (and on a 32-bit bus, not more/larger SRAM chips in addion to being fast/high-end 30 ns SRAM) and for the price point atari was aiming at, that limited them to 32 kB for the planned 1990 release, 32 kB shared between the Panther and the 16 MHz 68000 (which would have been starved for bus time and thus somewhat slower than the ST/Amiga/MD 68000s). It's a good thing Atari didn't push that onto the market in 1990... a shame they didn't have a good alternative though. (I wonder why they didn't consider the Slipstream/Flare 1 chipset -Flare still owned the IP and could license to anyone and it was a good chipset for 1990, probably could have afforded 512 kB DRAM in that case at a similar cost to the high-end SRAM in the panther)

[sheath]

In regard to Sega *never* using its arcade hardware to design home consoles:

SMS hardware specs:
CPU: 3.58 MHz Z80
Audio: 4 Channels
Co-Processor: VDP
Resolutions: 256x192, 256x224, 256x240 (PAL)
RAM: 8KB
Video RAM: 16KB
Color RAM: 32 bytes5
Colors On Screen: 32
(two 16-color palettes)6
Color Palette: 64

Sega Arcade Systems using the Z80:
General Z80 Hardware (Bank Panic)

G80 Based Hardware (See: Astro Blaster)

VCO Object Hardware (Buck Rogers:/ Zoom 909 compare to Zaxxon 3D)

Zaxxon Board

Champion Baseball Board (Compare to Great Baseball)

System 2 Hardware (See: Choplifter, Wonderboy, Wonderboy in Monsterland)

Only then did Sega develop a "suped up" Master System based arcade board:
System E Hardware (Fantasy Zone II, Hang-On Jr.)

It ought to be clear that Sega developed the Master System around several competing arcade board concepts involving the Z80.

Sega Genesis specs:
CPU: 7.67 MHz 16-bit 68000
Co-Processors: 3.58 MHz Z80 (Audio/Master System) 5
Texas Instruments 76489 (PSG Audio):
Yamaha 2612 (FM Audio):7
Colors On Screen: 61 (30-75 in game, average 50) 15
Color Palette: 512

By the Master System's standard, the Genesis included backward compatibility. So, some of the hardware was redundant versus being totally focused on "cutting edge" capabilities. All the same, compare it to:

Space Harrier Hardware (Super Hang-On)

Outrun Hardware (esp Outrun)

System-16A Hardware (esp Alex Kidd The Lost Stars)

System-16B Hardware (esp Altered Beast, Golden Axe, and note that there are carry overs for older hardware games)

X Board (After Burner II, Super Monaco GP, Thunder Blade)

So, if the Sega CD's development wasn't focused on the System 32, it was most certainly focused on Y Board mentality.

[kool kitty89]

I'm quite well aware of System-16's museum, pretty nice site.



Still, the only commonality with arcade hardware was COMMON off the shelf parts used (earlier) in computers, consoles, etc... The specs have superficial similarities, but not much else. (especially the vague ones you posted) Hell, I used to think that there was some direct relation due to misleading articles on the topic, but after talking with those who actually have programmed and know the hardware, I found out some interesting stuff. (like the SMS's VDP was no related to the Genesis one and the Genesis hardware shares no direct similarities in the custom hardware or system configuration to the System 16) I wouldn't doubt Sega designed the consoles with arcade ports in mind, but the hardware is not directly related.

The ONLY custom sega-developed hardware in most/all of their consoles/arcade boards are I/O and video hardware, in the SG-1000 they didn't even have that. (maybe I/O)


And yeah, backwards compatibility unnecessarily limited the Genesis... it could have been integrated more efficiently/usefully. (the Z80 is crippled with shitty bank switching and locked at the SMS clock speed, VDP was cut back to leave space for the SMS compatible block -as the MD VDP has really nothing to do with the SMS VDP, TMS9928, or the System 16 video hardware either iirc) It could have been designed rather differently to really build on the SMS hardware (the System E is sort of in that vein, albeit much less advanced than a 1988 console would have been), but they went the tacked-on route instead. Really odd in hindsight with an external adapter needed anyway. (makes me wonder if they originally planned compatible cart slots, which was very possible)

We've been through this before recently, but I can find the discussions again if you want.


Also a shame the SG-1000 used the SN76489 rather than the more capable AY-3-8910/YM2149 (especially in the context of the MSX), and didn't add any onboard sound enhancement to the SMS. (not even dual PSGs, though had the YM2149 been used, they could have upgraded to the compatible YM2203 ). The AY/YM chip would also have taken care of I/O.

So, if the Sega CD's development wasn't focused on the System 32, it was most certainly focused on Y Board mentality.

Why not the SNES? That makes a LOT more sense given that was the main competition worldwide. (but ESPECIALLY in Japan -and why not kill 2 birds with 1 stone: PCE's CD and SFC;s technical advantages -granted Sega undoubtedly would have been worried about Nintendo's CD add-on too -of course it was 1991 when Nintendo hit a snag there )

I could certainly see Sega keeping the arcade ports in mind (as they always did), but as usual they home hardware necessarily was in a different vein than the arcade hardware, being in the consumer market. (totally different cost/design goals)

As it was though, the CD"s ASIC is NOT optimized for what any of Sega's scaler boards did (or any other arcade sprite scaling hardware for that matter), the ASIC is a different beast and does things that would not be included if the main idea was arcade scaling type games.

Mode 7 did something no arcade hardware did prior to texture mapped 3D, none had rotation, let alone perspective warping and texture mapping.
Shat Sega seemed to do was make somethign flexible enough to handle stuff like mode 7 but be much more flexible and render many objects on-screen simultaneously (perhaps more like the GBA's scaling/rotation hardware).

If Sega wanted plain scaled objects as in the arcades, and wanted to focus on pushing as many on-screen at once and as fast as possible within the limits of the MD VDP. (and render that to genesis VDP tiles), they'd almost certainly have gone a different route. The Panther and Jaguar object processors were very much in line with that (the Jaguar also had a powerful and flexible blitter and other coprocessors, but the OP made it a sprite/object scaling monster, a shame they didn't push more games like Super Burnout -no rotation or warping, only plain scaled rectangles -rotation/warping required texture mapping using the blitter)
I'd think doing that would have been a rather bad move though compared to the flexible ASIC they did use. (especially given the similar limits imposed by the VDP)



One other particularly odd thing was a total lack of any Sega scaling arcade games on the CD... (not even some system 26 games with modest scaling) and they did have Afterburner 3 (a port from the FM Towns renamed adaptation of Strike Fighter) it had no scaling and generally looked and played worse than ABII on the Genesis.
Even stranger was the fast that SoJ jumped on porting Space Harrier and Afterburner II to the 32x in 1994.

I think Sonic CD and Panic were the sole 1st party SoJ games to even used the ASIC significantly. (and Panic was a collaborative effort)

[sheath]

Why not the SNES? That makes a LOT more sense given that was the main competition worldwide. (but ESPECIALLY in Japan -and why not kill 2 birds with 1 stone: PCE's CD and SFC;s technical advantages -granted Sega undoubtedly would have been worried about Nintendo's CD add-on too -of course it was 1991 when Nintendo hit a snag there )

Because Tomio Takami said in, the only interview I've found on the topic, that Sega asked him to add scaling and rotation hardware (paraphrased by EGM) because of all of their scaling arcade games. It was, in fact, bad PR on the part of published reader letters and editor speculation that pinned the idea that the Sega CD was Sega's "SNES Killer."

I could certainly see Sega keeping the arcade ports in mind (as they always did), but as usual they home hardware necessarily was in a different vein than the arcade hardware, being in the consumer market. (totally different cost/design goals)

Now this and your comments above about Z80 and System-16 hardware is just splitting hairs. There isn't a measuring stick that says if X amount of code from Arcade game Y runs on Console Z, they are the same hardware. In order for this to even be possible, the consoles would have to target the same screens as the arcade machines displayed on, have the same ROM sizes, the same controllers and I/O, etc.
Yes I agree that it is one thing to look at one hardware and then make it consumer grade. Doing so generally requires re-engineering and/or scaling down virtually everything. Yes I agree that the similarities between the SMS, Genesis and Sega CD to Sega's contemporaneous Arcade systems are limited. With the Z80 (SMS), 68000 and Z80 (Gen), and 68000 and Ricoh sound chip (SCD) though, to say that Sega only "looked at" some of their arcade hardware and then designed a totally unrelated system is wrong.
Perhaps you would have to take off your "engineer hat" for a second to see it. There is simply too much in common in hardware and software to not be able to see the arcade hardware's influence on the console side of things. Besides all of that, the historical record, because of the software alone, declares the influence of the Arcade machines on the console designs.
The presence of some of the exact same chips in all three consoles would even convince most skeptics.
I can even see this in the PCE/TG16 design and M72 Hardware that ran R-Type and Ninja Spirit among other games. While they actually are totally different design philosophies, I can even see the influence of Model 1 on the 32X, or Model 2 on the Saturn.

The NES, SNES and Playstation systems are actually the ones that have "nothing in common" with pre-existing arcade hardware. Just as Sega's systems have nothing in common with any of them.

Mode 7 did something no arcade hardware did prior to texture mapped 3D, none had rotation, let alone perspective warping and texture mapping. Shat Sega seemed to do was make somethign flexible enough to handle stuff like mode 7 but be much more flexible and render many objects on-screen simultaneously (perhaps more like the GBA's scaling/rotation hardware).

It is easy to see the similarities, sure. The facts are though, that Takami was told to target Sega's arcade systems, and the Sega CD's capabilities blow Mode 7 out of the water. The only way they could have been closer to the arcade system's scaling capabilities would be to have more sprite capabilities, which required an all new VDP as well. That brings us back to the Y Board Sega CD scenario. The fact that the Sega CD can hardware rotate too may be an answer to Mode 7, it might just be a "well duh" shoe in from the design process.

One other particularly odd thing was a total lack of any Sega scaling arcade games on the CD... it had no scaling and generally looked and played worse than ABII on the Genesis.
Even stranger was the fast that SoJ jumped on porting Space Harrier and Afterburner II to the 32x in 1994.

I think Sonic CD and Panic were the sole 1st party SoJ games to even used the ASIC significantly. (and Panic was a collaborative effort)

Yeah, as I don't read Japanese, or have any Japanese print from the time I don't speculate on the motivations of SoJ. It would be like trying to guess the President's intentions based on what the troops are doing in Afghanistan, it just doesn't work. After Burner III is almost identical to the Genesis version of G-LOC, I was tremendously disappointed when I saw it (after I got over the animated cockpit to chase view). It's just one example of Sega's policy to "port the Genesis game and add stuff" to save costs.

[kool kitty89]

Because Tomio Takami said in, the only interview I've found on the topic, that Sega asked him to add scaling and rotation hardware (paraphrased by EGM) because of all of their scaling arcade games. It was, in fact, bad PR on the part of published reader letters and editor speculation that pinned the idea that the Sega CD was Sega's "SNES Killer."

I was basing it totally on hardware design of the CD... not media claims of being an "SNES" killer, though it would certianly make sense to try to address shortcomings against the Super Famicom as well as the PC Engine.

Then the most logical inferences on that (assuming Takami's statments were interpreted accurately) would be that Sega wanted to address both arcade scaling and more flexible rendering like the SNES and some computer games were doing, or Takami took it upon himself to make the design more flexible, or Takami based the ASIC on something he was already working on.

Otherwise the functionality of the ASIC makes relatively little sense in terms of pure Sega arcade conversions, it goes way beyond that. And it's pretty interesting that Sega prominently demoed those very flexible non-arcade-like features in the BIOS on the Sega CD starting on the 1.0 JP version.
And on top of that, the first 1st party Japanese game to really use the ASIC, Sonic CD, exclusively used it in a non-arcade manner, in fact mainly simulating mode-7. (rather inefficiently and simplistically at that -and not taking advantage of the flexibility to do additional scaled sprites/objects)

Now this and your comments above about Z80 and System-16 hardware is just splitting hairs. There isn't a measuring stick that says if X amount of code from Arcade game Y runs on Console Z, they are the same hardware. In order for this to even be possible, the consoles would have to target the same screens as the arcade machines displayed on, have the same ROM sizes, the same controllers and I/O, etc.

That's NOT what I meant at all. What made the systems what they were were the specific implementation of off-the-shelf components (ie RAM/ROM/CPU/sound hardware and how that's mapped), but critically Sega's custom graphics hardware.
The latter is the main point of contention where there's no direct ties to home/arcade.
Again, Sega's first foray into console hardware was the SG-1000, a dated off the shelf ColecoVision clone (which itself was off the shelf) with the only advantage of 2 kB instead of 1 kB main RAM. Also very similar to the slightly earlier MSX as well, though that had more RAM (minimum 8 kB) and a better sound chip. (the TMS9918 and SN76489 were originally designed for the 1979 TI-99/4 computer, which indeed had games rather similar to the others, but it didn't use a Z80, but TI's TMS9900 CPU; I believe Coleco was the first to implement the Z80 in combination with that VDP, the MSX came after, then the SG-1000)

They even re-applied the SG-1000 hardware to the arcade as a low-end system.

And also, none of Sega's previous arcade hardware had used the SN76489 PSG. Most had used discrete/CPU/MCU driven sound, and in one case the AY8910. (same as Intellivision/Vectrex/MSX/ST/CPC/Spectrum/FM-7 and several other arcade boards -a common, off the shelf part like the weaker SN76489 -a shame Sega chose the latter like Coleco did)

Yes I agree that it is one thing to look at one hardware and then make it consumer grade. Doing so generally requires re-engineering and/or scaling down virtually everything. Yes I agree that the similarities between the SMS, Genesis and Sega CD to Sega's contemporaneous Arcade systems are limited. With the Z80 (SMS), 68000 and Z80 (Gen), and 68000 and Ricoh sound chip (SCD) though, to say that Sega only "looked at" some of their arcade hardware and then designed a totally unrelated system is wrong.
Perhaps you would have to take off your "engineer hat" for a second to see it. There is simply too much in common in hardware and software to not be able to see the arcade hardware's influence on the console side of things. Besides all of that, the historical record, because of the software alone, declares the influence of the Arcade machines on the console designs.
The presence of some of the exact same chips in all three consoles would even convince most skeptics.

No, the hardware is very different... we could start another thread on exactly how different (I'd actually like to know if the SMS VDP shares any commonality with the TMS9918 or if compatibility was tacked-on like the MD).

Software similarity is another issue, and you have code level similarity, and then what you see and hear on-screen.
You have arcade ports to totally unrelated hardware that can be very close, closer than hardware with more similarities (if superficial in some areas).
In particularly, take something like Ghouls n' Ghosts on the PC Engine vs the Arcade vs the Megadrive, though there's way more than that.

Of, course, using the 68000 also meant commonality for programmers skilled in that in the arcade. (part of the reason you had some slowdown issues on the SNES was the shift away from that architecture, though any developers heavily catering to the PCE/SNES or directly transitioning from the NES had better experience to deal with the CPU architecture, though the graphics architecture is another matter entirely -Amiga and ST had 68000s too, but that doesn't mean they'd be easy to port to because of that -X68000 is a bit of a different case due to it's arcade-like capabilities, though like other consoles it still had to have games ported to its different architecture with the graphics managed differently)

It's not just about sourcing ports (if you even have the source), but general trends of programmers' capabilities. (many MD programmers seemed to have issues with getting really good Z80 cone down for PCM managing for example)

But citing things really superficial without added context like resolution, CPU architecture, sound, etc, is extremely misleading.
The graphics architecture is the huge factor, but beyond that stuff like "yamaha FM chip" or "3 channel sound" means nothing without further detail. In the case of the YM2612 you did indeed have some good capabilities comparable to the YM2151 common to the arcade, though you had 2 less channels and some other differences (not sure if all the same algorithms are availalbe). They were both 4-op FM (again algos would be a factor still), with the YM2612 being in the same OPN family as the YM2203, YM2608, and YM2610 (neo geo), though a separate family from the original OPM (YM2151), but not nearly as different as the OPL family. (YM3526, YM3812, YM2413, YMF262; some used in the arcade, but most well known for the adoption on PC sound cards as well as the Master System and MSX)
Sega hadn't used the YM2612 before the MD, just as they hadn't used the SN76489 before the SG-1000, though in both cases they later used that sound hardware in arcade machines unrelated to the consoles. (and in the console derivatives)

CPU can be an indicator, but not much else on its own. (presence of a common CPU to another system -and other off the shelf components- is not nearly as big a factor alone as other architectural similarities -like the PCE graphics/sound block in the PC-FX)

The NES, SNES and Playstation systems are actually the ones that have "nothing in common" with pre-existing arcade hardware. Just as Sega's systems have nothing in common with any of them.

No, those had more or less as much in common (in a broad and/or superficial sense) with previous arcade hardware as most other consoles AND home computers. Maybe not Nintendo or Sony's OWN arcade hardware so much, but plenty of stuff on the market: the 6502 in the NES was VERY common throughout Japanese and western arcade machines, and the tile/characer+sprite bases system was a newer arcade feature as well and one that would continue to be a staple of 2D arcade machines long after.
The 256 wide resolution (5.37 MHz dot clock) was a fairly common one for arcades and home computers, the same as the TMS9918 used too. (and many other consoles and computers later on as well)
SNES followed that trend as well as much as the PCE at least, though neither jumped on the common use of the 68k in the arcades. (almost certainly due to cost/licensing advantages of the CPUs they selected, plus common use of the architectures in previous home consoles and older arcade games) But the graphics hardware went along with the arcade. (though both kept using planar bitmap graphics like the NES/SMS, not packed pixel like the MD -I think most later arcade board were using packed pixels too, but I'm not sure -the old Atari 8-bits and CGA used packed pixels, though ST, EGA, and Amiga were planar)

The PlayStation had much general commonality with arcade games and some computer stuff (granted PC games were all software/CPU driven, but the rendering methods were similar -general trend for triangles at the very least).
It shared more or less as much as the Saturn did with existing 3D (or even 2D) arcade hardware, or more than the 3DO for that matter.

In fact the only consoles with real general lack of even superficial similarities to previous arcade/computer hardware would be the very early ones which had nothing before them. You had the Odyssey designed before anything else of its kind. But, going beyond discrete TTL stuff, you had the first microprocessor based consoles shortly following hobbyist use in kit computers and before use in the arcade with several designed in 1975/76, some released in '76. (the VCS in '77)

It is easy to see the similarities, sure. The facts are though, that Takami was told to target Sega's arcade systems, and the Sega CD's capabilities blow Mode 7 out of the water. The only way they could have been closer to the arcade system's scaling capabilities would be to have more sprite capabilities, which required an all new VDP as well. That brings us back to the Y Board Sega CD scenario. The fact that the Sega CD can hardware rotate too may be an answer to Mode 7, it might just be a "well duh" shoe in from the design process.

The flexibility shoots mode 7 out of the water, but the overall performance is limited, namely by the horrible DMA bottleneck of being forced to use the Genesis VDP. (the limited VRAM space of the VDP and color limits didn't help matters though)
Mode 7 can mange full-screen 60 fps 256 color rendering for what it does, though it is limited for what it can do, of course.
If the genesis VDP didn't have the crappy DMA limitations, the ASIC could manage full speed easily (though color and VRAM usage would still be factors), which is why I said even a 16-color bitmap overlay would be a massive help to the Sega CD. Though had the Genesis included better expansion provisions, that would have made options to do that much more open, or if the VDP had had double buffered RAM banks, that would have solved the DMA issues. (both for cart and CD games)

As it is, the one thing you can do to open the DMA limits somewhat is to clip the screen, and that's why you see letterboxing on some Sega CD games (and some genesis and SNES games for that matter -including SFII and Virtua Racing). V-DMA uses VBLANK (the blank lines in the display). The fact that PAL has so many more lines in the display makes it fundamentally less limited there. You also have to double buffer in VRAM for anything less than 60 FPS (50 PAL) or you get screen tearing, so that's 2x the VRAM space used. (same problem with software rendered stuff on the MD and SNES -SNES had the additional issue of limiting tile sets to 32 kB banks in video RAM)

Without getting into the exact details of DMA bandwidth, you *could* have a 60fps NTSC updated 16-color display on the MD if you clipped the screen enough, though also depending on how much of the screen was being rendered to (for the entire clipped screen to be updated, it was going to be somewhat limited). One other thing is that DMA bandwidth increases proportionally in the higher res mode, so it's not any tighter in 320 wide than 256 wide. (but you use more VRAM to fill the whole screen)
And also remember, there's more than just the ASIC stuff to update, but normal sprite/tile updates as well in many cases. (unless you don't have much going on with those)
You could also render to sprite tiles, so updates would be a bit more flexible and specific rather than rendering objects to the BG layers, but then you have to keep the sprite limits in mind too.

As it is, for a lot of stuff it's practically limited to 12-20 FPS for NTSC games making heavy use of the ASIC. However, that's just the update speed of the VDP, and if the ASIC can't max it's bandwidth bushing things fast, it can still do it pushing a LOT of things at slower speeds. (batman and robin supposedly maxes out the bandwidth at some points)
That's when it looks most impressive: rendering a TON on-screen to the point where the rendering speed is the limit, not the VDP bandwidth. (so with a separate video layer the advantages would be mainly added color and more flexibility over the use of the genesis layer in the BG -less compromise over VRAM/sprite/BG usage)

So "sprites" aren't a limitation as it is, as the ASIC doesn't render sprites at all: it renders rectangular bitmap objects that can be scaled and rotation. The ASIC has hardware bitmap to tile conversion allowing it to render those graphics in genesis tile format and DMA them to the VDP either as background tiles or sprites. (interestingly rebel assault uses no hardware sprites in-game at all, it's all rendered to a BG layer -others have some hardware sprites as well as BG rendered objects) But don't think of the ASIC rendering "scaled sprites" or being limited to outputting it's scaled graphics to sprite tiles alone.

Again, Sega could have put the emphasis on just scaling/zooming for such, and pushed even more on-screen, if still limited by the Genesis VDP. You don't need hardware sprites for any of that, just rendering power. (again, you're not limited to genesis sprites and you don't have to overlay sprites either: you can render composited objects onto single sprites/sprite clusters or tile layers) Ie the same sort of scaling used by many 32x games. (in that case, you see more/faster straight scaling than rotation/texture mapping as it's FAR less intensive to do in software, same thing with some software scaling on the genesis and SNES -the rare REAL software scaling, NOT just animation)

As it is, many scaled games don't often render onto sprite tiles. (CORE's stuff tends to be all rendered on BG layers -in fact some, like BC Racers, cram almost all onto a single BG layer with no use of the 2nd layer and sprites just used for some parts of the status bar though SoulStar does a mix of sprite and BR rendering with both bg layers, while Wing Commander uses sprites for a similar purpose but also uses one bg layer for rendering and the other for the cockpit overlay, and the Sega CD BIOS demo is also all rendered to a single BG layer, no sprites, no 2nd BG)

Yeah, as I don't read Japanese, or have any Japanese print from the time I don't speculate on the motivations of SoJ. It would be like trying to guess the President's intentions based on what the troops are doing in Afghanistan, it just doesn't work. After Burner III is almost identical to the Genesis version of G-LOC, I was tremendously disappointed when I saw it (after I got over the animated cockpit to chase view). It's just one example of Sega's policy to "port the Genesis game and add stuff" to save costs.

Is ABIII even the same game as G-LOC? I thought ABIII was a port of Strike Fighter (also called ABIII on the FM-Towns), which was the Sequel to G-LOC.

Some of the scrolling ground effects on G-LOC actually look a lot more impressive for the Genesis than the Sega CD.

There's nothing wrong with porting MD games and adding content (should have done more of that... and in key areas like RPGs and certain high-profile games), though doing so in some cases is rather odd. Plus, scaling effects could indeed be one of the enhancements, especially if you're talking modest/occasional use in System 16/18 games. (plus better voice samples/sfx, etc) And, of course, lower prices for many such conversions vs the cart releases. (especially the simpler conversions and compilations with less cost overhead -though SoJ tended to avoid compilations in favor of multiple separate re-releases in Japan for some things including the Saturn)
Again, that doesn't seem to be the case with G-LOC in particular, but maybe I haven't paid close enough attention.

[sheath]

I was basing it totally on hardware design of the CD... not media claims of being an "SNES" killer, though it would certianly make sense to try to address shortcomings against the Super Famicom as well as the PC Engine.

Then the most logical inferences on that (assuming Takami's statments were interpreted accurately) would be that Sega wanted to address both arcade scaling and more flexible rendering like the SNES and some computer games were doing, or Takami took it upon himself to make the design more flexible, or Takami based the ASIC on something he was already working on.

EGM, who followed the SNES's development slavishly from the summer of 1989 on, speculated that the Sega CD was the culmination of the System 32 based rumors. Magazine reader letters, with obvious SNES envy, began speculating in Spring 1992 that the Sega CD would have everything the SNES had that the Genesis did not. If EGM's SNES bias was going to twist the Sega CD into Sega's answer to the SNES it would have done it when they published the Takami interview in Fall 1991.

Otherwise the functionality of the ASIC makes relatively little sense in terms of pure Sega arcade conversions, it goes way beyond that. And it's pretty interesting that Sega prominently demoed those very flexible non-arcade-like features in the BIOS on the Sega CD starting on the 1.0 JP version.
And on top of that, the first 1st party Japanese game to really use the ASIC, Sonic CD, exclusively used it in a non-arcade manner, in fact mainly simulating mode-7. (rather inefficiently and simplistically at that -and not taking advantage of the flexibility to do additional scaled sprites/objects)

The Graphics Coprocessor makes sense as an "off the shelf" consumer grade solution to boost the effects of Sega CD close enough to Y-Board capabilities without adding another VDP. I'm really not sure what all of this "no rotation in scaler games" is all about. After Burner, G-Loc, Galaxy Force II rotates the entire screen, Power Drift tilts the screen, as does Gale Racer/Rad Mobile. After Burner animates rotation of the jet as it crashes. Crackdown animates sprite rotation as they die. Aurail was an over head game designed around full screen rotation. It seems pretty simple to me, the arcade machines could depend on excessive ROM for effects and the Sega CD could not. Perhaps SoJ didn't port all of these to Sega CD because they had already figured out that arcade ports didn't make enough money?

That's NOT what I meant at all. What made the systems what they were were the specific implementation of off-the-shelf components (ie RAM/ROM/CPU/sound hardware and how that's mapped), but critically Sega's custom graphics hardware.

That's why I said: "Sega both tended to design their console hardware using their arcade designs and custom design their console video processors." What is your point of contention, that the rest of the "off the shelf" components also had nothing to do with Sega's arcade configurations? Sega tended to use components of their arcade hardware, especially the CPU, alongside custom video processors.

The latter is the main point of contention where there's no direct ties to home/arcade.
Again, Sega's first foray into console hardware was the SG-1000, a dated off the shelf ColecoVision clone (which itself was off the shelf) with the only advantage of 2 kB instead of 1 kB main RAM. Also very similar to the slightly earlier MSX as well, though that had more RAM (minimum 8 kB) and a better sound chip. (the TMS9918 and SN76489 were originally designed for the 1979 TI-99/4 computer, which indeed had games rather similar to the others, but it didn't use a Z80, but TI's TMS9900 CPU; I believe Coleco was the first to implement the Z80 in combination with that VDP, the MSX came after, then the SG-1000)

SG-1000 came out first as a PC, then as a console last I checked. Whether or not it was patterned after Arcade tech doesn't change the three systems that followed it. If Sega wanted to, they could have focused on making the most powerful home console they could each generation, be it custom of off the shelf. They could have focused on making the cheapest console they could, or a slightly better console than whatever else was already on the market. Instead, they looked at their arcade hardware, and then did whatever it took to release a console that could come as close as possible to duplicate the games on a standard Television.

Even the Dreamcast is a cheaper derivative of Model 3 capabilities, and it is the only time Sega released better tech than its previous arcade machines as a console and arcade system. Sony did this first with the PS1, though the arcade tech was inferior to stuff Sega had on the market previously. The only time Nintendo tried to do it was with the faked Ultra 64 arcade machines like Killer Instinct and the torrid Cruisin' USA.

The 256 wide resolution (5.37 MHz dot clock) was a fairly common one for arcades and home computers, the same as the TMS9918 used too. (and many other consoles and computers later on as well)
SNES followed that trend as well as much as the PCE at least, though neither jumped on the common use of the 68k in the arcades. (almost certainly due to cost/licensing advantages of the CPUs they selected, plus common use of the architectures in previous home consoles and older arcade games) But the graphics hardware went along with the arcade.

Exactly, NEC and Nintendo focused on similarities with NES and older tech and cost saving.

Mode 7 can mange full-screen 60 fps 256 color rendering for what it does, though it is limited for what it can do, of course.

But what game(s) actually manage that even in cutscenes? Also, you can't leave out the fact that the Sega CD scales sprites at the same time as the ground, that's a whole different level of capability.

If the genesis VDP didn't have the crappy DMA limitations, the ASIC could manage full speed easily (though color and VRAM usage would still be factors), which is why I said even a 16-color bitmap overlay would be a massive help to the Sega CD. Though had the Genesis included better expansion provisions, that would have made options to do that much more open, or if the VDP had had double buffered RAM banks, that would have solved the DMA issues. (both for cart and CD games)

Looking at the RE docs for the Genesis doesn't expose any "crappy" DMA limitations. From what I can find the VDP's DMA was 7.2 KB per 1/60th of a second to the PPU's 5.7 KB. I've documented the sources for these numbers here. I'm sure you'll find it too simplified, but the numbers are documented. Your whole argument sounds like you just don't like the way the Sega CD worked with the Genesis, more than the approach actually being demonstrably bad.

Without getting into the exact details of DMA bandwidth, you *could* have a 60fps NTSC updated 16-color display on the MD if you clipped the screen enough, though also depending on how much of the screen was being rendered to (for the entire clipped screen to be updated, it was going to be somewhat limited). One other thing is that DMA bandwidth increases proportionally in the higher res mode, so it's not any tighter in 320 wide than 256 wide. (but you use more VRAM to fill the whole screen)

For what the Sega CD can do it does better than the SNES could without the FX chip. Textured Line scrolled ground with scaling sprites is more like Sega's arcade games than Mode 7. That happens to be what the Sega CD's graphics coprocessor does relatively well. You're trying to shoe horn the Sega CD into the "Mode 7" killer model, and as you are demonstrating, the facts just don't line up.

And also remember, there's more than just the ASIC stuff to update, but normal sprite/tile updates as well in many cases. (unless you don't have much going on with those)
You could also render to sprite tiles, so updates would be a bit more flexible and specific rather than rendering objects to the BG layers, but then you have to keep the sprite limits in mind too.

From what I have seen, the Sega CD's "sprite scaling" limit occurs before the VDP's sprite limit. BC Racers, specifically designed to throw as many sprites around as possible, chugs it doesn't flicker.

That's interesting about the BG layer being used for scaling though. I always thought that the VDP sprite size limit would cause the sprites to "hit" the screen much smaller than they do. I never bothered to look at anything in Kmod to see how it was done.

Is ABIII even the same game as G-LOC? I thought ABIII was a port of Strike Fighter (also called ABIII on the FM-Towns), which was the Sequel to G-LOC.

Some of the scrolling ground effects on G-LOC actually look a lot more impressive for the Genesis than the Sega CD.

I haven't played either in over a decade, I just remember the sprites and levels looking the same, though slightly enhanced in AB III.

There's nothing wrong with porting MD games and adding content (should have done more of that... and in key areas like RPGs and certain high-profile games), though doing so in some cases is rather odd. Plus, scaling effects could indeed be one of the enhancements, especially if you're talking modest/occasional use in System 16/18 games. (plus better voice samples/sfx, etc) And, of course, lower prices for many such conversions vs the cart releases. (especially the simpler conversions and compilations with less cost overhead -though SoJ tended to avoid compilations in favor of multiple separate re-releases in Japan for some things including the Saturn)
Again, that doesn't seem to be the case with G-LOC in particular, but maybe I haven't paid close enough attention.

Oh I agree, in fact I think it was smart for Sega to cheaply port Genesis games that were already great and add stuff that CD-ROM makes possible. I wish they spent a little bit more time making Sega CD games from the ground up, but the sales never would have justified the expense.

[chrisbid]

none of these options mention software, so they all fail equally


sega approach was too focused on hardware. new hardware does not move off the shelf without introducing new software that mass market wants to play.

no space invaders, no super mario bros, no sonic the hedgehog, no street fighter ii... no hardware sales

[sheath]

That is just incorrect. Sega the hardware manufacturer released higher quantity of well reviewed *and* fairly unique titles every year than Nintendo ever did.

In fact, Sega proved that releasing well designed highly playable games, that also receive great press, in high quantity is not what the masses want.

I predicted way back when the Dreamcast was canceled that it was the beginning of the end for gameplay innovation and new IP, and I was right. One of the top conceptual comments from publishers today is excusing the fact that the industry won't take chances on new ideas today. Sega did that almost constantly as a hardware manufacturer, which is probably one of the top reasons why they ultimately had to stop.

The masses want simple, familiar, and especially popular brand names. Art doesn't sell, uniqueness doesn't sell, innovation definitely doesn't sell. Quality can make a game successful among gamers though.

[Chilly Willy]

Looking at the RE docs for the Genesis doesn't expose any "crappy" DMA limitations. From what I can find the VDP's DMA was 7.2 KB per 1/60th of a second to the PPU's 5.7 KB.

The DMA on the Genesis is "crappy" because it can't update an entire screen worth of tiles in one frame. It doesn't matter that the SNES is worse. Fast scrolling games like Sonic work by having most of the tiles in vram, and only changing the table that determines which tiles will be shown. The table (the pattern name table) is small and can be changed quickly enough for 60 Hz games. If you wish to change the tiles themselves, the DMA rate on the Genesis limits you to 15 to 20 Hz games. that's why games like VR are only 15 to 20 Hz.

The SEGA CD graphics ASIC is designed to generate rotated/scaled gaphics by generating all new sets of tiles for the rotated/scaled graphics. Do see what that means? The way the ASIC generates graphics is the SLOWEST way the Genesis graphics work. If SEGA had added a video layer to the CD, then the CD could update that layer at 60Hz, not the Genesis layer at 20 Hz. So CD games with lots of scaling and rotation (or FMV) would have been faster.