What's new
I think there is a myth which needs to be cleared out. MiSTer devs are NOT looking to do 100% accurate simulation and unless the specific platform has some sort of a community back for that it will never happen.
well that counts me out then. I don’t see the point of not striving for 100% accuracy. There are plenty of current solutions that are available now for cheaper, that are not 100% accurate. So why would I hold out for a MiSTer again?
And I know it might upset a few, but various devs have clearly mentioned that they arent making cores which can be used to recreate parts of the original hardware - some people feel thats the ultimate test. None of the MiSTer cores currently can pass that test. And most probably it never will since cost will outweight the benefit. But subjectively - playability wise, its very close to the real hardware in most cases. Software emulation on my i9 PC doesnt come close to it.
 
I've got a neogeo darksoft multi and a CPS2 darksoft multi both in their own dedicated cabs, so no MiSTer won't be replacing that kind of setup.

However it's damn great as the next best thing if you can't afford the real hardware.

While we don't have 100% accuracy today, it technically is possible to get there. Especially when some of the cores can be implemented based on decapped original chips. We have for example a good cycle accurate m68k that all cores that use a 68k has benefitted from (Genesis, Amiga, neogeo, upcoming CPS1 etc). Then on the other hand we don't have the same for a Z80 yet so cores that use a Z80 can be improved once we get one.

Also i do believe kevtris did solder his FPGA cores for various SNES chips into a real SNES as a part of his testing/dev process. Unfortunately his cores are closed source though (which makes sense for a commercial product..).
 
Last edited:
And I know it might upset a few
Let me just stop you there. I'm not upset by this, I'm legit asking why would I invest in yet another inaccurate device that emu/simulates the real thing when I can buy something like a Pandora's box, or hack a SNES/NES/PSX Classic and run retroarch and get something equally inaccurate.

I mean, obviously original hardware is king, but if the arguement is these cores are pretty accurate for the handful of consoles (not arcade) that are already well emulated by people like byuu, or FBA emulators, then what makes the MiSTer stand out? Why should I invest? I thought the whole point of going with FPGA emu/simulation is to faithfully recreate the original hardware timings, reducing overhead that software emulation creates, and produce an accurate reproduction of original hardware. Wasn't this all the hoopla about going with FPGA solutions in the first place?

Again, not mad, just asking legit questions.
 
I thought the whole point of going with FPGA emu/simulation is to faithfully recreate the original hardware timings, reducing overhead that software emulation creates, and produce an accurate reproduction of original hardware. Wasn't this all the hoopla about going with FPGA solutions in the first place?
Well, that’s what Analogue would have you believe with their FPGA consoles. Their marketing is continuously stating claims like “zero lag” and “total accuracy”. Now, it may not be fair to hold MiSTer devs accountable to Analogue’s marketing claims, but i do think their marketing has set enthusiast’s expectations for all FPGA implementations. Analogue’s inflated marketing is what prompted Byuu to write his article.

I’m also confused about whether “100% authentic to OG hardware” is the final goal for emulator and MiSTer core developers. Sure, every developer is different and I’m sure some devs are satisfied with just “good enough”, but 100% accurate has certainly been Byuu’s goal with BSNES over the years at least (to name one example). I certainly hope that 100% accuracy is the long-term goal of the MiSTer cores, even if their current devs aren’t targeting it. At least the open-source code of the cores will help ensure that any devs that do have a 100% target in mind will be able to keep at it. Like a relay race - if one dev stops short of the goal, pass the baton and keep running.
 
Analogue are merely pushing marketing hype to sell and justify an overpriced product. They boldly claim "No emulation" which is an outright lie.
 
Analogue are merely pushing marketing hype to sell and justify an overpriced product. They boldly claim "No emulation" which is an outright lie.
how are Analogue’s FPG Consoles overpriced? Serious question..
analog shipping is wayyyyy overpriced.
I agree on the shipping. It's a rort.

I still love the quality kevtris puts out. I like the look of the new GBA console however - don't like the tall GBA crampyness and prefer the wide ones for my stupid adult sized hands. My 13 yr old actually requested a GBA SP for christmas which i did end up getting him (plus a few multi carts and the pokemon games he wanted) this year, which has only highlighted to me how bad that form factor is for adults.
 
Last edited:
Regarding Analogue those are expensive emulators. Same with MegaSD. I prefer MiSTer over those anyday. Performance is nearly identical except nes mappers which lag far behind on Mister.
 
I finally got my MiSTer up and running, even though I still don't have my RAM yet. Took a while to get used to controlling the menu, but I think I've finally gotten a hold of it.

The biggest thing I'm running into is trying to get it to work on my Sony PVM-2030. It doesn't have full BNC connectors on the back and I had been using it with a 25pin to SCART cable. I ordered a 25pin to VGA cable, but I don't think that's going to work. I tried it, but the PVM just makes a loud buzzing noise and shows no picture. I've now ordered a SCART to VGA cable, which will hopefully be more successful.

My question is, what are the settings that I should set in the INI/wherever to make sure I don't damage my PVM? My ultimate goal is to stick this in my candy cab, but I want to make sure it's all working on the PVM before I do that.
 
@Cloggerdude I'd double check CSYNC is <1V before trying to connect the MiSTer to a valuable PVM. The DSO150 is a cheap scope (~20€) suitable for this task.
 
Digital oscilloscopes are very easy to use... just look for DSO150 manual or tutorials, it's not rocket science. Once adjusted a csync signal looks like this:

dso150-csync.jpg


In this photo CSYNC is attenuated to ~0.4V. If the MiSTer IO board attenuates the signal correctly for SCART you should see <1V.
 
Analogue are merely pushing marketing hype to sell and justify an overpriced product. They boldly claim "No emulation" which is an outright lie.
how are Analogue’s FPG Consoles overpriced? Serious question..
analog shipping is wayyyyy overpriced.
$420 AU for an emulated SNES or Mega Drive console. $120 AU extra to retain desired video outputs. Now we're at $540 for a product that I can buy the original of for $30 to $80 AU.
 
Analogue are merely pushing marketing hype to sell and justify an overpriced product. They boldly claim "No emulation" which is an outright lie.
how are Analogue’s FPG Consoles overpriced? Serious question..analog shipping is wayyyyy overpriced.
$420 AU for an emulated SNES or Mega Drive console. $120 AU extra to retain desired video outputs. Now we're at $540 for a product that I can buy the original of for $30 to $80 AU.
I see, I understand..on my side of the world, it is a bit cheaper:)
 
Damn, I'm overwhelmed by the huge wave of common sense in the latest posts. What happened?

Besides, FWIW, jotego is using a logic analyzer to study each board to replicate the exact behavior of the various custom chips which have not been decapped/documented yet.

While it's not perfect down to each single gate, it's pretty damn near.

Also, @opt2not , pandoras, classic mini consoles and FBA are the worst kind of emulation which is out there: they are made with a truckload of hacks to make them run on very limited hardware at full speed so a MiSTer, while not 100% accurate, would be a huge step up from those.

May I also ask, what's people's definition of "100% accuracy", emulation-wise? Genuinely curious.
If someone says their emu is 100% accurate and you pit it against real hardware and you're unable to detect any difference, video, audio and input-wise, does it make "100% accurate"?
 
For me 100% accurate means, apart from not being able to notice differences when playing, being able to use the documentation to make hardware reproductions to fully fix an original board or even recreate it from zero with no noticeably difference from the original when running.
 
Last edited:
Back
Top