shmups.system11.org

That's what I was hoping for. Yeah probably would have been a $400 machine since the Sega CD pretty much has a Neo Geo inside it.

There is a better chance that a Turbografx/ PC Engine clone could be compatible with any old USB drive. I don't think it had any custom hardware.
Also, is it actually smaller than a Genesis model 2? I'd assumed they made it the exact same size so it'd properly seat next to a Sega CD 2.

The renders show it as roughly the same size as the Super NT. So it'll be a good bit smaller than the Genesis 2 (maybe by ~40%?). I'm thinking it will look kind of goofy with a Sega CD attached.

But are the people working on it as talented as Kevtris? Also noticed theres no SNES core

ETA Prime has done a quick video on one....

https://youtu.be/1jb8YPXc8DA

the audio on the Gameboy core sounds awful however

Rather than talent it is important to note that kevtris is being remunerated for his work. I assume this is his full time job? That means he can throw himself fully into the work without having to worry about other things. I imagine he is also provided with funding for any expensive equipment he might need.

Compared to a someone doing the work open source, for free and in their free time, talent is less of an issue than just resources (time, money) in general.

Great for Kevtris of course and there is nothing wrong with getting paid for working as hard as he obviously does, but it is kinda sad to think there is presumably a great 68000 core out there and it belongs to just one company.

Not getting into a dick swinging competition but in a bare knuckle fight id put my money on Sorgelig not Kevtris..

How about on an FPGA core writing? Are you saying that Sorgelig is a better coder than Kevtris?

It's not, I recall during one of Bob's (RetroRGB) interviews with him he mentioned that his work with Analogue is on a freelance basis. His dayjob is in cryogenics.

There isn't anything wrong with paying people for their work. Retroarch has that bounty system which seems to work well. Personally, I'd be more than happy to compensate Sorgelig or Kevtris for there efforts in a similar fashion.

It's probably not likely, but maybe the community could woo Kevtris away from Analogue to an open source project like the Mister?

I think you'll find Retroarch is pretty keen to get on the bandwagon too.

2019 will be interesting.

What are you saying?, possibly an FPGA board from RetroArch?

In analogue's case this probably could be someone's day job, but I don't really think so and think it's much more of a significant (paying) passion project. Imho it's less about pure talent and more about enthusiasm for the subject/work and bringing that in together to deliver an end product. They're taking their time and completing nice products in an appreciably precise manner that has all the hallmarks of a clearly defined partnership where each party is in control of their domain, but who also seem to appreciate and be a good match for each other's needs/opinions/ethos... and I don't think they're going out of their way to exploiting their situation - just doing a good practical job of getting it out there and hopefully getting rewarded for their efforts.

By definition an open source project is something of a different proposal - none-the-less intriguing - but will bring with it a certain amount of flexibility and choice that is less channelled. This tends to result (often quite purposefully) in a wider remit, and a less finite goal, often with the expectation this maybe maintained/re-purposed much further down the line. The flip side of this is that it welcomes contributions in all forms, and arguably does not make the project success/failure quite so dependant on the availability of set personnel/materials.

In each case, I think it's vital that the people being relied upon are happy in their role and the position that puts them in with regards to the wider project, and just for that reason I don't *think* you'll likely see much cross-pollination of individuals across these two approaches... unless they simply fancy a change

It's not just being a good programmer per se, but understanding how the original hardware worked and reproducing each single behavior meticolously.

You might be a bad coder and write badly optimized/structured code, but thanks to a better understanding of how the original console's circuit work, you'll have a more precise emulation.

Being UK based and having a mostly PAL MD/SMS collection makes this an easy purchase for me. I play at 60hz where possible but have to use 50hz at times and a PAL megadrive’s refresh is so far off spec it leads to judder and even tearing through both an OSSC and FM. Hell my switch modded JP MD is actually closer to 50hz than a bloody PAL MD!

Game gear support is also huge for me and will probably be the best way to play GG games on a TV.

I'd like to ask those with real understanding of this: how are people such as Kevtris getting so close (at least as we all seem to agree) to the original hardware's behavior? I mean, to my understanding he is a talented hobbyist - not someone involved in the original design process 30ish years ago, not even someone working with those who were. So where is the detailed understanding coming from? We're obviously talking about things specific to Sega's design like the bus design, VDP, timings, etc. How does one get so deep into these thirty years later and several steps removed from the original creators?

Reverse engineering basically. Not all that far removed from emulator development. Some emulators nowadays are also extremely accurate. The idea that FPGAs can achieve what software emulators can't is something of a myth. It can run on cheaper hardware and with generally less latency, but no FPGA machine is 100% accurate. Higan is at least on par with the Super Nt in terms of accuracy.

Anyway I digress, it comes down to collecting a ton of data. Using logic analyzers, oscilloscopes, whatever else needed to take measurements of the inputs and outputs of each chip, scouring patents and trial and error.

Doesn't the Mister not even do proper integer scaling yet with HDMI output? From what I read on the forums, this isn't a thing yet. Which pretty much makes it useless in my eyes for HDMI output.

Unless they have cheap, easily available controller and cartridge adapters specifically for it. Just get an Odroid or Pi.

but then again it offers native resolution in RGB for all its cores.

^ This. A million times.

Hopefully people will realize this sooner or later.

EDIT: unless, as I stated in a previous post, we'll reach the point where boards will let us put real hardware and disable the emulation, so that all the FPGA does is hooking the various signals together.

I have something like 4-5 Mega Drive I don't use and each have a perfectly working Motorola 68000, Zilog Z80 and Yamaha YM2612. I'd happily pull all those chips out and dump them on a FPGA board to achieve perfect emulation.

So the FPGA 'hardware emulation' is only ever as good as the person programming it and their knowledge of the real hardware then?.

Begs the question if you put the Super NT up against Retroarch bsnes Accuracy or higan Accuracy cores (with run ahead feature enabled) over say 10 games would you ever be able to pick out the FPGA 10 times in a row?......im guessing not

It obviously depends on the game and the person playing it.

There are some games which have their own quirks and people who have played them for years can easily pick those quirks out every single time.

Let's say that you put me in front of two black boxes, one with the Super NT and one with the worst SNES emulator we have today (dunno, guess ZSNES or Snes9x?) and let me play a game I've never played in my life... yeah, without even using Higan I wouldn't be able to tell a difference. So YMMV.

We're not talking about that here. We're talking about emulator accuracy, in spite of the user able of telling the difference.

Yeah so an FPGA core is only as good as the person programming it so its bound by the same laws as software emulation, i wasn't aware of this up until yesterday i just thought that an FPGB was identical to real hardware but by the sounds of it theres a lot of variables that effect how accurate an FPGA really is....which im kind of disappointed by as i was under the impression FPGA was the holy grail but that would appear to be the case

Think of the FPGA as a "single-purpose computer". It's a programmable chip which you can code to your liking* with the only upside that, since it's doing only one thing and it doesn't have an OS upon which it runs, it can theoretically bring lower latencies and, of course, be way more efficient cpu power-wise than a normal PC.

EDIT: and instant boot time. From the moment you switch it on, it's a matter of seconds before you can jump into the action.

I specified "theoretically" because today you can get normal PCs to emulate old consoles with absolutely zero lag, if you know what you're doing which, imho, make all this FPGA fuss a complete nonsense. FPGA is just becoming a new buzzword and everyone is capitalizing on it.

* this is a broad generalization and not completely correct, but gives you an idea of what an FPGA is.

@Fudoh - the add on I/O which has a VGA out, is that actually for RGB? Does it do 15khz? Can the MiSTer be connected to a PVM with a similar cheap cable? Does the MiSTer do components/composite/RGB over that VGA out put on the I/O?

I don't care much of it is emulated, I just want something that looks (15khz) and play (no traditional OS) like the real thing with a small foot print and easy to setup.

mine's still in the mail (base FPGA board + add on boards), but from what I've see read, 15khz output is fully supported. Whether you can set the output to component as well - I don't know yet.

Given that all the cores can run at their native resolution my idea was to use VGA plus an OSSC instead of utilizing the DE10-Nano's HDMI output.

@Fudoh - thanks for the quick reply... I look forward to your review of the unit please:)....

Can i ask why you'd add another device in the chain, you are going to a HDMI screen right?

Because the OSSC can line multiply the native outputs (240p, 288p, etc) and produce a better image than the native HDMI.

Perhaps there should be a MiSTer thread at this point.
EDIT: MiSTer thread here

Agreed.

Also, it will remove some load and free processing power for the more requesting cores.

To me the appeal of this is good emulation at the flick of a switch without boottine, dicking around with drivers, inages, updates, etc etc etc. As long as i can still pick up crt’s for free and megadrives for a few tenners i don’t think i’ll use this soon but its good that the option is there!

Wish they'd keep on going with premium materials such as anodized aluminum and wood, despite extra cost. It really adds to warranting having a physical machine taking up space.