MiSTer FPGA Forum

Some people probably rolled their eyes looking at the title of the thread lol. The PC-FX was NEC's completely botched follow-up to the hugely successful PC Engine in Japan. It completely killed all of the momentum and goodwill they generated with the PC Engine. It's not a very good console and is filled with FMV anime games, but I just wanted to create this thread to discuss the likelihood of a core being created for this obscure Japan-exclusive console.

The Mednafen team has created a very good emulator for the PC-FX, but hardware emulation > software emulation (which is why we're all here ) I also like the idea that if we get a PC-FX core completed, then we'll have the entire NEC catalog preserved and playable on MiSTer, which would be a cool milestone. What are your thoughts on the PC-FX and it's feasibility?

This would be challenging, but may be possible. Lots of work would need to be done first though:
- V810 CPU is needed, but this could also be reused for Virtual Boy, so somebody may put one of these together
- HuC6270 ("7up") video processor already exists, and is implemented in PC Engine core
- Several other chips would need to be implemented, and this could be hard work
(more info at http://daifukkat.su/pcfx/ )

Another major challenge would be the fact that all color is implemented in YUV colorspace, including MotionJPEG, alpha layers, and palettes.
While the DE10-Nano has a ADV7513 HDMI driver on it (which can handle both RGB and YUV, and colorspace conversions), there appears to be resistance to allow YUV colorspace to be used within the system framework. While it may be possible to implement a YUV->RGB colorspace conversion on the core's output, that would consume DSP, memory, and developer resources, making it that much more challenging.

The amount of video processing hardware, and the separate KRAM are the big question mark as to whether this is implementable on the DE10-Nano; hard to say.

Nice breakdown. So it seems like a Virtual Boy core would come before the PC-FX.

I suppose those of who wish to enjoy its library will have to start learning Japanese as well (which is an idea I am toying with). By my count, there's three fully-playable unofficial translations. Looks like the chances of this happening may be pretty slim. I wonder if there are any Japanese people out there who have any interest in video game preservation or the MiSTer...

I'm sure a core is doable but the PC-FX is like the CD-i and 3DO... consoles that bet heavily on FMV and didn't really do that well compared to the competition. So it's a lot of work for cores that may not be used that much. I'll gladly explore the whole library for it if a core is ever made, though.

I wouldn't lose hope, I think there is a good chance of it happening within the next couple of years. As David outlined, it is difficult but possibly doable (and David owns several PC-FX consoles and a complete library). Robert has also mentioned in the past he is interested in the Virtual Boy, so he might do that after PS1 sometime in the future (obviously he will be doing PS1 for a good while yet) and the work done for that would pave the way for PC-FX.

It is such an esoteric console that some devs do seem to have a soft spot for it, so I am optimistic. I also am confident we will see more fan translations in the future.

YUV to RGB is not really an issue. There's plenty of ways to tackle it. The actual blockers:
-It's an unpopular Japanese system and documentation is fairly sparse and in a foreign languages, so a PITA to use
-It uses many very complex, custom, unimplemented chips
-The library is not well liked (to the point of ridicule) and just finding any interest from a motivated developer willing to deal with the other issues is unlikely

While most of those statements are fair to make, it has already been implemented quite well in an emulator (Mednafen), so the chips aren’t as insurmountable a barrier as mentioned.

The library is indeed the issue - heavy Japanese use, and many of the titles are not enjoyable.
So realistically in the same group as 3DO, CD-I and Jaguar, in more ways than are obvious at first.

It is probably more worthwhile to invest effort (at least for the moment) in things like PC-98, FM Towns, and so on.

…and that’s coming from somebody supportive of having a core for this.

That's really interesting insight, thanks guys - it's always interesting to get an understanding of what is involved with building cores. Hopefully it does happen one day, but as you say it's not going to be easy, the library is far from top tier, and it's not English friendly. That being said, it's not without it's merit.

I will say, for anyone who is really interested in PC-FX you do have the option of software emulation but also you could track down the original hardware. I actually bought a PC-FX recently (although I haven't found the time to properly dig into it). I assumed it would be incredibly expensive to get one (especially to the UK) but the prices aren't as crazy as you might think. I paid around £150 boxed and shipped from Japan for one in nice condition. The games however are very expensive now. But if you do want to try the system out they aren't crazy expensive (yet), there is no copy protection so you can burn off the discs, and it has S-Video out so you don't need an exotic cable, just a scaler than can handle S-Video. You will need a voltage converter though.

PC-98 and FM Towns definitely have more interesting libraries, however Puu was working on PC-98 and has moved onto FM Towns. Apparently in order to improve PC-98 it makes sense to tackle the FM towns first and bring work from that over to PC-98, so in a similar fashion to how we should soon have his x68000 and PC-88 cores cleaned up, improved, and into main the same can hopefully happen one day down the road for those two big systems.

As an aside, I do find it quite interesting that there has been big progress of the late 80s and early 90s Japanese PCs but not the predecessors to these machines, the big systems from the early-mid 80s are largely uncovered.

So we have FM Towns being worked on, but we don't have the previous Fujitsu systems like the FM-77 (covering the FM-7 and FM-8).
We have Sharp x68000 but not the precursor the Sharp X-1.
We have PC-88 on the way and PC-98 being worked on but we don't have PC-8000 or PC-6001.
We have an MSX-2 but no official MSX-1 core.

The games for these systems are obviously more basic and not as exciting, but I do find it interesting that nobody has tackled such relatively big computer cores, especially since they seem to have all their chips done already in other cores, and it seems every months we get a new obscure home computer from the era. So fans of obscure Japanese systems and game libraries, there are still a lot of Japanese systems that haven't come to MiSTer yet that from a technical standpoint seem to be very doable. Something to look forward to.

Would love to see this happen, if only for Galaxy Fraulein Yuna FX. Granted, it would need an English translation too, but you know. Dream big.