Support NeoGAF

[Nikodemos]

Yeah no. Sega CD was awesome, well designed, and had a very solid library.

Again this is wrong. There are a lot of games that make use of the additional chips.

The SEGA-CD offered a lot of features. Every single game certainly didn't have to use every single feature. They were there if the developer wanted. They could use Redbook Audio, or they could do what they did in Lunar 2 (PCM). Sonic CD used both Redbook Audio and the additional sound chip for the Past levels.

A lot of games made use of the ASIC chip for special effects. I am not going to start a list, there are simply a ton of them. You could do FMV with the well known encoder (that was grainy yes) but you also had the liberty to code your videos in software as space was not an issue anymore, thanks to the CD format, and many games did this which resulted in superb quality that doesn't age at all. Games had voice overs, and some had elaborated sound design such as Snatcher. The SEGA-CD introduced a lot of people to brand new features. The PC-Engine CD was much more niche in the US and didn't even make it in Europe.

The problem is, you're talking about a product that, at minimum, cost $400 in 1992 money (Sega CD + base Genesis). That is around $870 today. In order to justify that sort of price, pretty much every game needed to be at least one step above anything you could find on the Genesis. And only some of them were. It's the same problem which afflicted the 3DO. Now, of course, part of it was due to bad devtools (expecially in the West), but that's not really an excuse. Sega simply overshot (due to the aforementioned base console limitations) while trying to put out a product competing with the CD-ROM^2. They were too concerned with whether or not they could, they never stopped to consider whether or not they should.
Frankly, if they were so dead-set on getting a CD-based product out, they should've simply gone the whole distance and made the Sega CD a standalone console. The base Genesis would always be a metaphorical millstone around the Sega CD's neck.

Me mentioning PC-Engine was in the context of the Japanese market, where it was enjoying decent marketshare. And was the first console of the main 3 to come with a CD-ROM attachment.

Exactly.
From the leaked confidential document we know that Sega of America were supporting 6 product lines at the same time (Genesis/Nomad, Game Gear, Sega CD, 32X, Pico, Saturn) and all of them were projected to lose money in the FY97!
The expected total loss for FY97 for SoA was projected at $122.6M.

Yes, they were spread disastrously thin.

 

[nush]

Nah, the core issue of Kalinske decision making is that he was hell bent to gain marketshare at expense of profitability and securing long term growth for Sega.

That’s how it works, has worked and even works today. Shareholders like numbers going up by whatever metric because once you’ve secured your area then you build on that. For example Amazon ran at a loss and even Xbox for years but they turned it around because they had that big footprint.

 

[Nikodemos]

That’s how it works, has worked and even works today. Shareholders like numbers going up by whatever metric because once you’ve secured your area then you build on that. For example Amazon ran at a loss and even Xbox for years but they turned it around because they had that big footprint.

This. Kalinske going for a loss-leader strategy was nothing out of the ordinary then, and it wouldn't be today. Especially since they were going against an already-entrenched name, and their previous attempt (Master System) had fizzled out unmemorably.

 

[Celine]

Me mentioning PC-Engine was in the context of the Japanese market, where it was enjoying decent marketshare. And was the first console of the main 3 to come with a CD-ROM attachment.

The CD attachment for the PCE is the only case in history that I can think of an add-on which wasn't just successful but with a transformative impact for the main platform.
By 1991 the majority of game releases for the PCE platforms were on CD-rom and ultimately (for PCE) there were more releases on CD-rom than Hu-card.
Usually this isn't possible because, by their own nature, add-on can at best address a portion of the main platform install-base.
In the end the "CD players" for PCE were near 2M in Japan so below the 4.8M "Hu-card players", the better economics tied to CD-rom production and differentiation factor granted compared to the dominant Famciom/Super Famicom is probably what led the transformative change.
I'm also sure Hudson Soft was the one getting royalties from hu-card games (instead of NEC, the hardware manufacturer!) but I'm unsure if the same applied to the CD-rom games which I suspect wasn't the case, hence why NEC may have had very good reasons for the shift to CD-rom.

 

[Celine]

That’s how it works, has worked and even works today. Shareholders like numbers going up by whatever metric because once you’ve secured your area then you build on that. For example Amazon ran at a loss and even Xbox for years but they turned it around because they had that big footprint.

But the foundation lay down by Sega were frail which is how they went from their peak to exit the console market in a short period of 7 years.

Look at the sudden rise of the console division operating income and then quick contraction and deep fall into permanent red. That's not how it should work!

 

[Nikodemos]

I'm also sure Hudson Soft was the one getting royalties from hu-card games (instead of NEC, the hardware manufacturer!) but I'm unsure if the same applied to the CD-rom games which I suspect wasn't the case, hence why NEC may have had very good reasons for the shift to CD-rom.

I think they still did, otherwise NEC wouldn't bother re-releasing the CD-ROM^2 unit under so many variants, that were only minorly different from each other. The licensing was probably tied to the very hardware (HudsonSoft processors) used in the unit, meaning it couldn't be escaped.

Look at the sudden rise of the console division operating income and then quick contraction and deep fall into permanent red. That's not how it should work!

Seems to me they expanded quickly, but didn't stabilize, so they collapsed like an overinflated souffle. Them releasing several devices that were, de facto, different platforms (Game Gear, Sega CD, 32X, Saturn) in the span of something like 3 years didn't help with cost control either. Plus the 1994 mini-bubble bursting for 16-bit games.

 

[Astray]

Reading these documents, it's clear to see that Sega's biggest problem by far was the lack of an all-encompassing Global Product & Marketing Strategy.

From the lack of alignment on the Sega Saturn between Sega US and JP to the disastrous launches of the 32X and Sega CD, there's just far too much dysfunction happening here, far too much in-fighting and a ton of money and partner goodwill wasted R&Ding products that ultimately went into a landfill or something.

Sony didn't so much end Sega as much as they happened to walk right when Sega self-immolated, reaping the benefits in the process.

 

[cireza]

The problem is, you're talking about a product that, at minimum, cost $400 in 1992 money (Sega CD + base Genesis). That is around $870 today. In order to justify that sort of price, pretty much every game needed to be at least one step above anything you could find on the Genesis. And only some of them were. It's the same problem which afflicted the 3DO. Now, of course, part of it was due to bad devtools (expecially in the West), but that's not really an excuse. Sega simply overshot (due to the aforementioned base console limitations) while trying to put out a product competing with the CD-ROM^2. They were too concerned with whether or not they could, they never stopped to consider whether or not they should.
Frankly, if they were so dead-set on getting a CD-based product out, they should've simply gone the whole distance and made the Sega CD a standalone console. The base Genesis would always be a metaphorical millstone around the Sega CD's neck.

Me mentioning PC-Engine was in the context of the Japanese market, where it was enjoying decent marketshare. And was the first console of the main 3 to come with a CD-ROM attachment.

You switched to an entirely different argument here. You were stating how the SEGA-CD was a bad product, not well made and its hardware was underutilized. All of which I disagree with.

 

[UnNamed]

It's incredible we had so many leaks from SEGA, from emails, to code, to betas etc. but still no Virtua Fighter 3 or Shenmue on Saturn.
I have a sense we will see Shenmue eventually, but no VF3 because it never existed.

 

[cireza]

VF3 because it never existed

Common sense tells us that it definitely did exist, as Virtua Fighter 2 was the best selling game on Saturn, the console was doing good in Japan and Virtua Fighter 3 was an arcade hit released in 1996. It most certainly had to make a lot of compromises though. You can be sure that SEGA and AM2 at least tried. The game was announced and appeared in plannings as well.

 

[UnNamed]

Common sense tells us that it definitely did exist, as Virtua Fighter 2 was the best selling game on Saturn, the console was doing good in Japan and Virtua Fighter 3 was an arcade hit released in 1996. It most certainly had to make a lot of compromises though.

There is probably some demo somewhere, IIRC even Sega Japan confirmed, somehow, the work of porting VF3, so it's not an invention from magazines. People with better knowledge can report the statements from SoJ on VF3. But i don't think there ever was a playable game. My guess is they used the work on Saturn VF3 to make Fighter's Megamix.

 

[Nikodemos]

You switched to an entirely different argument here. You were stating how the SEGA-CD was a bad product, not well made and its hardware was underutilized. All of which I disagree with.

Well, yes, it wasn't well made, both from a hardware and coding stand-point. JVC conned Sega into using a CD-Audio grade drive, instead of a proper CD-ROM grade one. This meant the spin-up and seek times were shit (we know SoA complained about this when they eventually received the units), and it wore out the read head assembly faster. Plus, Sega had initially wanted to make the add-on look like a premium audio player, so they went for a motorized tray loader, instead of the cheaper and more durable clamshell top-loader. That loading tray was the unit's main point of failure, and they ditched it for Model 2. It couldn't talk directly to the Genesis (the expansion port had no interrupt assigned); it had to rely on 256 kB of buffer RAM, and very carefully constructed game code, in order to send and receive data from the base unit.

And, like I said, a lot of games settled on using Sega CD's otherwise beefy hardware for CD tracks and video files (anime in JP, FMV in ROW), because it was difficult to extract processing power from the Sega CD, due to the base Genesis being so shit at expansion though the databus port. Plus the hardcoded graphics drawbacks like no 128k mode and small color palette (due to the Genesis VDP only having half its originally intended VRAM). To this one can add SoJ's reluctance to share its otherwise underwhelming devtools with SoA in a timely manner.
And the launch window games did nothing to whet the public's appetite (at least in the West): Sol-Feace was a bog-standard shmup, Sewer Shark was a run-of-the-mill rail shooter with some grainy FMVs, and Cobra Command was good, but nearly 10 years old at that point. Not to mention the ensuing software drought caused by poor devtool documentation.
Also, it's worth noting that weebs hadn't been invented yet, so directly localizing anime-heavy games from JP would've likely seen very few sales in the West (possibly not even covering the localization costs).
As I mentioned, in its current incarnation, Sega CD was a poorly thought-out device. Had they gone for a cost-optimized (top-loading clamshell) standalone with a proper CD-ROM grade drive from the beginning, things might've been different.

 

[Celine]

I think they still did, otherwise NEC wouldn't bother re-releasing the CD-ROM^2 unit under so many variants, that were only minorly different from each other. The licensing was probably tied to the very hardware (HudsonSoft processors) used in the unit, meaning it couldn't be escaped.

Hudson Soft got royalties on both hardware and software on hu-card (in a sense it was similar to the 3DO business model with the difference there was only one company that manufactured the hardware).
NEC had to make money from hardware (consoles) and of course the software from NEC Avenue (by paying a royalty to Hudson Soft in the case of games on hu-card, just like any third-party publisher).

The business relationship between the companies was unusual, too. Hudson got a royalty for every PC Engine manufactured, since it designed the CPU and GPU that powered the system; it also got a royalty on each and every game sold in the PC Engine's HuCARD cartridge format, which it also designed. The games were manufactured in Japan by Mitsubishi on behalf of Hudson, while NEC handled hardware production, and made its profits from selling systems.

Stalled engine: The TurboGrafx-16 turns 25

25 years ago, NEC released the video game system that fell first in the 16-bit console wars -- but what an interesting failure it was. The history of the TurboGrafx-16 as told by those who were there.

www.gamedeveloper.com

 

[West Texas CEO]

This thread is amazing.

 

[cireza]

As I mentioned, in its current incarnation, Sega CD was a poorly thought-out device. Had they gone for a cost-optimized (top-loading clamshell) standalone with a proper CD-ROM grade drive from the beginning, things might've been different.

So basically releasing an entirely different console to be added on top on all others in 91, having to support it, explaining to users how the new consoles doesn't play any MegaDrive games etc... Sounds like a great advice...

The mechanical tray wasn't a great idea, but this is an issue outside of the hardware specs we were discussing. Model 2 is a great improvement and what should have been done from the beginning. Good on SEGA for making it. The MegaDrive was built with an external port precisely to receive input from such an add-on. SEGA-CD had a lot of RAM and the loading/seek time issues you speak of only exists in your mind. SEGA-CD games load speeds are perfectly fine considering the time of release and type of content that are on the discs. It became absolutely unbearable during the 32 bits era, where a ton of games have constants loadings, especially RPGs. SEGA-CD had way enough RAM to store a lot of visual detail, which led games such as the first Lunar or Shining Force CD to barely have any load times. It was even enough to get Sam Sho and Fatal Fury Special ports that retained the full size of the Neo Geo sprites.

The small color palette is a MegaDrive legacy, this has nothing to do with the SEGA-CD, and even with the MegaDrive constraints a ton of games looked gorgeous. ROM size and price was an essential reason behind less ambitious visuals, so at least this was a thing of the past with the SEGA-CD. This is because of this that we got games such as Sonic CD and its unbelievable number of stages and variants, with a ton of backgrounds that change dynamically when you go through the level etc... SEGA-CD unlocked a ton of the potential that was stuck because of the cartridges, and this is precisely what was expected from a CD add-on to begin with.

 

[Ozzie666]

You are aware that SEGA had huge success in arcades, Master System outsold NES in Europe, Game Gear had support until 1996 worldwide, same for the SEGA-CD, Saturn was very strong in Japan etc... ?

We can list SEGA errors and agree on them, but their success certainly wasn't due to luck. They were doing a lot of things right. The overall strategy in Europe was excellent, in the US they marketed the Genesis very well and deserved the success, same for the Saturn in Japan.

With Japan and the Saturn, I view this completely differently. Nothing about the Saturn was done well or easy. But I think the Saturn's success in Japan says more about Nintendo's failures than any smart moves by Sega. Saturn finished second in Japan and a VERY distant second. The European success I can't rationally explain, except I don't recall Nintendo pushing hard there and not sure if their strong-armed tactics and third-party control would have been allowed in Europe? Europe is the only place Sega had 2 somewhat successful generations in a row then. Even to this day I am amazed at some of the Master System and Game gear software well after decline in North America.
Hats off to Sega Europe.

 

[StreetsofBeige]

But the foundation lay down by Sega were frail which is how they went from their peak to exit the console market in a short period of 7 years.

Look at the sudden rise of the console division operating income and then quick contraction and deep fall into permanent red. That's not how it should work!

Thats operating profit, so were talking profit.

By the looks of it, they should had just stuck with arcades and arcade machines. Not sure what amusement facility is. Maybe dedicated Sega arcades or amusement park. Not sure.

Genesis did great in early 90s which makes sense. But holy shit did Saturn and Dreamcast die. People bring up Saturn being clobbered, but it seems Dreamcast era was even worse. Probably because DC games were pretty good so they assume the financials werent as bad as clunky Saturn games. But DC is what really killed home consoles.

 

[mosantos]

It's incredible we had so many leaks from SEGA, from emails, to code, to betas etc. but still no Virtua Fighter 3 or Shenmue on Saturn.
I have a sense we will see Shenmue eventually, but no VF3 because it never existed.

Shenmue was made on "Saturn 2":

Canceled Games of SEGA’s Leaked Fiscal Year ’97 Documents

Editor’s note: This leaked document is very long. We will do our best to ensure this article is at least close to complete upon publication, but please be advised that it may be updated over …

www.segasaturnshiro.com

 

[cartman414]

With Japan and the Saturn, I view this completely differently. Nothing about the Saturn was done well or easy. But I think the Saturn's success in Japan says more about Nintendo's failures than any smart moves by Sega. Saturn finished second in Japan and a VERY distant second. The European success I can't rationally explain, except I don't recall Nintendo pushing hard there and not sure if their strong-armed tactics and third-party control would have been allowed in Europe? Europe is the only place Sega had 2 somewhat successful generations in a row then. Even to this day I am amazed at some of the Master System and Game gear software well after decline in North America.
Hats off to Sega Europe.

It may have lagged 3d-wise, but it was the only system that gen with great 2d. Too bad 2d became passe by then.

 

[Nikodemos]

So basically releasing an entirely different console to be added on top on all others in 91, having to support it, explaining to users how the new consoles doesn't play any MegaDrive games etc... Sounds like a great advice...

It's all counterfactual history, so it's all "wibbly-wobbly, timey-wimey". But, thing is, the Mega Drive was a bust in Japan. So, as I see it, not having to buy a console that didn't interest you (I mean, the generic "you", not you-you), in order to play Mega CD games, isn't such a big disadvantage. And, in 1991, they only had (in Japan at least) Mega Drive and Game Gear as parallel offerings (Master System had been EoL'd the same year). One of them (Mega Drive) had already been deemed uninteresting by the public. And, from a purely developmental standpoint, the Mega CD was a different console. SoA's documents list it as separate from the Genesis, and I'm reasonably sure SoJ's did as well. So having it as a completely separate standalone wasn't such a great leap.

The mechanical tray wasn't a great idea, but this is an issue outside of the hardware specs we were discussing. Model 2 is a great improvement and what should have been done from the beginning. Good on SEGA for making it.

Fundamentally, anything that impacts the bill of materials, and can cause customer issues, is part of the hardware design. Especially since somebody has to burn engineer-hours to fix said issues (as they eventually did with Sega CD Model 2). Same thing happened with the Mega Drive's stereo mixer. They had initially gone for an off-the-shelf walkman-grade module, since it was cheaper. But then they realized a lot of people started owning TVs with stereo speakers (and using wired headphones wasn't very popular), so they eventually had to add it to the sound chip preamp circuitry. As an added negative-bonus, Mega Drives with the headphone jack need an extra cable for proper audio mixing to both the Sega CD, and 32X.

The MegaDrive was built with an external port precisely to receive input from such an add-on.

That's the thing, it couldn't. The databus port had initially been intended for a diskette drive, similar to the Famicom Disk System. It had no interrupt (IRQ number), it needed to be told when and how to poll the databus. You couldn't just write a command like "IRQ 8 listen"; you had to painstakingly and manually write code to tell the Mega Drive CPU what to do with that port. Hence my previous remark that 32X was a bad product, but well-designed, since it avoided said pitfalls.

SEGA-CD had a lot of RAM and the loading/seek time issues you speak of only exists in your mind. SEGA-CD games load speeds are perfectly fine considering the time of release and type of content that are on the discs.

SoA did a comparative test with their units versus regular consumer CD-ROM drives. The Sega CD's optical drive noticeably underperformed in all areas (spin-up, seek, random read etc.) except sustained output. This is a problem when having to jump between multiple sectors of a 500+ MB CD.

The small color palette is a MegaDrive legacy, this has nothing to do with the SEGA-CD, and even with the MegaDrive constraints a ton of games looked gorgeous.

The reduced palette wasn't even the main constraint (even though it notably impacted stuff like FMV games, who required wide palettes in order to accurately display minute differences in color gradients or shadows). The VDP couldn't do bank switching to perform full-frame buffering (128K mode), because it physically lacked a second memory bank (i.e. chip) to pre-load frames into. It had to compose all frames line by line in real time. This wasn't an issue with classic 2D sprite games, since the CPU was fast enough to brute-force sufficient speed thanks to its burst DMA (the infamous "blast processing"). However, the moment pseudo-3D effects were added, framerate dropped considerably, since the system couldn't pre-render and buffer the next frame, and the 7.6 MHz CPU wasn't powerful enough anymore to do it on the fly.

 

[cireza]

the Mega Drive was a bust in Japan

Well, too bad for Japan but the situation was entirely different in the rest of the world. Which means that a different, dedicated CD machine would have been a mistake anyway.

Fundamentally, anything that impacts the bill of materials, and can cause customer issues, is part of the hardware design. Especially since somebody has to burn engineer-hours to fix said issues (as they eventually did with Sega CD Model 2). Same thing happened with the Mega Drive's stereo mixer. They had initially gone for an off-the-shelf walkman-grade module, since it was cheaper. But then they realized a lot of people started owning TVs with stereo speakers (and using wired headphones wasn't very popular), so they eventually had to add it to the sound chip preamp circuitry. As an added negative-bonus, Mega Drives with the headphone jack need an extra cable for proper audio mixing to both the Sega CD, and 32X.

I am aware of all of this, but it wasn't the discussion to begin with.

That's the thing, it couldn't. The databus port had initially been intended for a diskette drive, similar to the Famicom Disk System. It had no interrupt (IRQ number), it needed to be told when and how to poll the databus. You couldn't just write a command like "IRQ 8 listen"; you had to painstakingly and manually write code to tell the Mega Drive CPU what to do with that port.

And yet it worked well enough it seems otherwise we would not have gotten so many great games on the system.

SoA did a comparative test with their units versus regular consumer CD-ROM drives. The Sega CD's optical drive noticeably underperformed in all areas (spin-up, seek, random read etc.) except sustained output. This is a problem when having to jump between multiple sectors of a 500+ MB CD.

This is a purely technical point of view, but this is still not a problem for the users since the loading times are perfectly fine on SEGA-CD.

The VDP couldn't do bank switching to perform full-frame buffering (128K mode), because it physically lacked a second memory bank (i.e. chip) to pre-load frames into. It had to compose all frames line by line in real time. This wasn't an issue with classic 2D sprite games, since the CPU was fast enough to brute-force sufficient speed thanks to its burst DMA (the infamous "blast processing"). However, the moment pseudo-3D effects were added, framerate dropped considerably, since the system couldn't pre-render and buffer the next frame, and the 7.6 MHz CPU wasn't powerful enough anymore to do it on the fly.

Sounds like this is the expected way for the VDP to work, I don't really see any issue here. Of course you are going to stream from the ROM in real time, you don't need "blast processing" for this. All games do this.

Pseudo-3D effects done by the SEGA-CD were not 60fps but anyway, no console was doing this back then. The SEGA-CD had its own CPU, a lot of RAM, so elements that have to be drawn will be prepared in that RAM and streamed to the MegaDrive VRAM when needed. Sounds logical to me, and obviously it worked perfectly fine when you look at Core's output, to name only one developer.

 

[violence]

I know I never trusted Sega’s after buying the 32X

 

[Fatal Chuck Yeager]

Exactly.
From the leaked confidential document we know that Sega of America were supporting 6 product lines at the same time (Genesis/Nomad, Game Gear, Sega CD, 32X, Pico, Saturn) and all of them were projected to lose money in the FY97!
The expected total loss for FY97 for SoA was projected at $122.6M.

And that was SoA, where those items were actually selling in comparison to the home region (excluding the Saturn but that was diving at the time even domestically as well).

Sega of Japan and America both has no profit plan for the Saturn, so pivoted to the internet TV fad, then entered into a bunch of amusement, toy, clothing, and sports/music partnerships and were even sponsors and they were bringing zero money from any of it and it wasn't well executed to attract new players so you can't even say it was loss leading to for that reason. They bundled the Genesis so hard there was no way to keep the thing alive in key countries.

I can't think of a single product or idea Sega had post-1993 that was meant to actually make money instead of paying for another previous bad product or idea.

It's really amazing how small both Atari and Sega were in the home console space and how short lived their 'lucky' success was. Both failed far more than they actually succeeded. Both run badly, actually both run by morons with little cash. the 2600 and the Genesis were just lucky. Never duplicated.

The Atari 52000 was pulled out before we could even see it's performance. Atari 7800 was a success that generated profit despite not being in first place, the Atari Xegs was another product that was a success generating profit despite not being in first place, the Atari Lynx was another product that was in the same boat until Atari pulled it, The Atari ST, STE, were all profitable products that also saved the company and gave it revenue to pursue other interests. The Atari 2600 Jr was a big success until Nintendo made selling products a problem in the US by having retailers push off the competition even if the sold what they sent in.

Really only the MegaST, Falcon, and Jaguar were clear failures that lost a crap ton of money and were poorly executed, or in the case of the Jaguar, launched on credit without any bug testers. Oh, and that laptop thing.

This is the problem about using total number to determine success. PC Engine was know down a big profit for Hudson and a minor profit for NEC but it only sold slight more than 5M consoles there which puts it pretty low on the totem poll for sales in Japan. Coleco sales are unknown though the max is 3M estimated, yet it shook up the industry, had high software sales, and was profitable all the way to its discontinuation. PS3 sold 87M consoles yet was a write-off, Xbox got 2nd place with over 24M sold, yet was a write-off.