Neo Geo HDMI A/V upgrade board

[Kyuusaku]

I'm thinking of working on a fully-digital Neo Geo HDMI A/V upgrade board. Back when I first mentioned the feasibility of this project I didn't have a HDTV, now I have a few, and a better handle on the digital signal processing for the job.

Since that post I've been working here and there on a complete Neo Geo clone, but I've had to step back since developing accurate CPU cores is very tedious, plus I haven't had the funds to prototype the ridiculously expensive hardware. This project should be a lot easier to pull-off while at the same time is probably as desirable if not more so to the community.

The mod would output 720p60 and only 720p60. The actual game would occupy a 960 x 672 window within the 1280 x 720. There wouldn't be any plans to implement full-screen scaling via interpolation since it's lossy and would require significantly more hardware. 720p doesn't always look so hot on today's 1080p TVs, but it will scale perfectly on new 4K TVs (720p x 3 = 2160p). Likewise I couldn't go down to 480p because the standard 480p format doesn't have square pixels and would require interpolation.

The 720p output would use industry standard frame timing and a literal 60 Hz frame rate, so there would be no question about its compatibility with displays. To achieve this timing the Neo Geo's clock will need to be dynamically controlled to keep synchronization with the HDTV frame, and unfortunately it isn't economically feasible to do this in a way which keeps analog video working. HDTVs are ubiquitous now, so hopefully that isn't a dealbreaker...

Digital audio over HDMI is also the plan, but I'm not sure whether it will fit into my target device, so it might come at an increased cost. This is "special" because the YM2610's SSG component (the beep-boop YM2149/AY-3-8910 sound chip) natively only provides an analog output, so it will be emulated.

The install will require soldering many wires (40+ with audio), but the points are all through-hole leads (on AES consoles), so it's quite easy if you work carefully.


My question to you is: what could I sell this for and how many do you think I could sell? 500, 1000? Is the lack of analog video a dealbreaker for you? Is the Neo Geo still valuable enough to warrant expensive video mods?

To get ready for the project I will need to buy equipment I can't afford for prototyping, so I need to know if this is a non-starter. It seems to me like the market has been losing enthusiasm for the Neo Geo as it ages, but I could be wrong about that.

 

[awbacon]

there is always a market, but it's a limited crowd. You'd be selling to the same group of people that buy anything...so there would def be a finite limit to the units you could move, and I don't know if that would be profitable. You'd really have to do your overhead numbers and understand how many units at X price you would have to move to recoup investment, and whether or not you'd have enough buyers at X price to justify the work. Basically, you need to make sure the price is right, and the potential units moved is right as well. Too expensive = won't sell, and you'll lose money. Right price = should move enough units to at least make back your investment (and you'll have the equipment for further projects)

I'd certainly entertain buying one, dependent upon price

 

[city41]

Sounds like really cool tech. But if I was looking to play my Neo on an HDTV I'd probably opt for an XRGB instead, as then I could spread the love to all my systems.

 

[kahel]

What would be the advantages if this vs buying a xrgb-mini and a scart wire? It less expensive then 500$ and doesn't require soldering...

 

[Kyuusaku]

The advantage over an XRGB would be that it's a lossless digital connection. The picture quality would be optimal, whereas upscanners at the very least require tweaking on a console-by-console basis to get the full dynamic range during the A/D process (this is generally the fault of the consoles or RGB mods). The RGBS signals out of Neo Geos aren't ideal:

-the default signals through the video encoder are bandwidth-limited, so the pixel edges aren't terribly sharp, then there's also noise picked up through the circuitry, and TV compatibility issues due to SNK's weird sync driver circuitry.
-"bypass mod" or MVS signals are not band-limited, but without an amplifier circuit they aren't suitable for driving industry standard 75 ohm video inputs, which leads to impedance matching, noise, distortion, and signal-level (color accuracy) issues.

Also the XRGB-mini, while rather-well designed, (or assembled, Micomsoft didn't develop any of the advanced video technology in the unit) contains just another TV front-end "super chip" to do the A/D conversion, and another video processing "super chip" to do the scaling. In other words XRGBs are redundant if your TV has a good analog front-end and scaling unit.

So although the XRGB-mini looks great due to advanced signal processing, the signal is still rather lossy, and laggy. Using an all-digital approach is lossless for video, next to lossless for audio, and would cut the lag down to a handful of lines instead of frames.

Of course this doesn't matter for most people's uses. I think the XRGB-mini is a good product, but it's not perfect because it's not flexible enough to handle all sorts of non-standard video, and also (AFAIK) the display attached to the XRGB-mini must be able to sync to the vertical sync frequency output by the console since the mini actually synchronizes to the incoming video. Digital TVs aren't always (generally aren't?) flexible enough to handle arbitrary pixel and frame rates, so even a Neo Geo + XRGB-mini I believe could be inoperable on some TVs.

What this HDMI project would do is force the Neo Geo itself to conform to industry standard video timing, while at the same time giving it a digital connection.

 

[BIG BEAR]

I'm not trying to be funny here BUT It's my belief that someone could just adopt the old technology and manufacture new analog RGB monitors. This way not to disturb the integrity of ANY original retro hardware.
BB

 

[Kyuusaku]

For the most part I feel the same way, but there are countless non-ideal video mods being installed on hardware all the time which compromise the console's integrity without even bringing new functionality to the table. I figure why not put in some effort and make one that truly shines. Also since the mod would take the analog video out of service, one could remove the DIN connector, string the HDMI through the hole and at least keep the NG's plastic intact.

As an aside, I have designed most of an analog CRT chassis (the non-dangerous bits), and it's a dream of mine to build a high-end tri-sync arcade monitor. Fun to think about, but certainly not a feasible project for small-scale manufacturing.

 

[Skips]

After buying all the stuff you need it sounds like you would have to sell it at a fairly high price to make your money back. This sounds like it would be targeting a very small customer base. Once you factor in price of the board + installation fees (for those customers who lack the skills) it sounds like it could get VERY expensive. I am not sure how much this will cost but if you sell the kits for $90 - $100 (no clue what it would cost) or so the customers without the skills (and there will be more of them than ones with I am sure) will be looking at around $150-$200 after you factor in the installation fees modders will charge. For only $100-$200 more they could buy an upscaler that works with all their consoles (price varying based on the unit they buy). It sounds like it would cost you more money than would be worth the hassle in my opinion, even though it does sound cool.