XNA SharpNES - the first NES Emulator for Xbox 360 and XNA Posted by XanTium | January 27 01:01 EST

'Lone Coder' released the first 'indie' emulator for the Xbox360. No, the Xbox360 is not 'hacked' to run unsigned code (yet) ... 'XNA SharpNES' is made with 'XNA Game Studio Express', the Microsoft dev tool released last month aimed at helping students and hobbyists build games for Windows and the Xbox 360. While the tool is developed to program indie games, and not really made/optimized to make other stuff/applications, it looks like 'Lone Coder' managed to use it to port a NES (8-bit Nintendo) emulator :) Note that to run this on your Xbox360 you will need a 'Creators Club' account ($99/12m or $49/4m) on your LIVE subscription and both Visual C# Studio Express (free) + XNA Game Studio Express (free) installed on your Windows PC. The emulator will also run on Windows, if you have the XNA Framework installed. From the readme/nfo: [QUOTE] XNA SharpNES the first NES emulator for the xbox 360 and XNA. If you'd like to talk about this project my gamer tag is Lone Coder or you can email me at bryanlivingston -AT- gmail.com This is a conversion of SharpNES by Jonathan Turner. Converting it only took a couple of hours. It runs at 60% or 70% of normal speed on the 360, so it's playable but slow. There's probably some very easy optimizations and cleanups to be done still. Right now there is no ROM loading menu, so to switch ROMs you have to include the ROM in the SharpNES360 project and set it to "Copy to Output Directory", then edit the filename in Program.cs. This requires a creators club account which runs $99 a year or $49 for four months. You'll also need Visual C# Studio Express which is free. Buttons: A = A ; B = X or B ; Start = Start ; Select = Back ; Up, Down, Left, Right = D-Pad ; Exit Emulator = RB (Right Shoulder Button) What's Missing: * Second Controller * Sound * Saving or State Saving * Rom Loading Menu Supported Mappers: Mappers 1 (mostly), 2, 3, 4 (mostly), 7 (partial), 9 (mostly), 10, 11 (partial), 22 (partial), 34, 64 (partial), 66 [/QUOTE] Official Site: http://code.google.com/p/xnasharpnes/ Download: here (latest release) (subversion rep. (latest code))

no need for nes or any emus i got psp to do that

dude way to shit on peoples parade..

anyway.. Its cool that people are getting stuff written like this, since the XNA wasnt written to do this.. and they did it anyway.. people just gotta figure shit out and probably get other shit (loaders etc) to run..

Congrats to the author and to whoever has the account to run the shit.. have fun

kudos!:D
im for one happy to see this project alive
as said XNA wasnt designed for this.. but nevertheless

anyhoo - i think imma wait untill theres a proper emu out that does sound and run at 100% speed.. -

PSP? Those haven't died out yet?

sweet! never gonna use it but sweet none the less!

my heart almost skipped a beat... then i saw XNA was involved..

Yo this could be a good start to many things to come....good job..... has neone tried it out yet I hace xna but im getting a bunch of errors when it builds

wow. no love for the xna emus? so sad.

Paying 100 bucks per year to run my own programs on my own console (in a sandboxed environment) is just... 'too bad'. I will still wait until the first true homebrew-allowing mod will be released.

i can confirm that this works threw XNA but many things need to be done to get this to 100% it runs at about 75% on zelda 2 and alil slower wheni tried to use bases loaded but it works .......good job .......... now all thats left is the tweaking.........lol

All we need now is VLC ported or some sort of open source media player so we can start watching x264 on it

i dont know shit about programing but cant you just transfer the program to the hdd and run it from there.

Nice work, but the fact that a straight port of a NES emulator already runs slowly on a 360 tells me enough about XNA already. Sure, XNA wasn't designed for this, but a 360 literally has about 3000 times the processing power that a NES has. Even an unoptimized, naive, 100% interpreted (no dynarec) implementation should easily be able to run at full speed.
'
So I think we can safely rule out PSX/PS2/PSP/NGC emulation using XNA. SNES/GBA/Megadrive might just barely be possible if some really smart guy throws his skills into it. I don't count on it though...

I'd like to mention that this is quite possibly the greatest thing that has been made for the Xbox 360 in terms of modding considering "Lone Coder" you have created what I would call an amazing feat' and contribution to the scene and everywhere else that would inspire people to do further research and experiments with the Xbox 360 to get this system under our own control even closer by the day!

So I'll also say thank you for getting us closer to softmodding and exploiting the Xbox 360 by releasing this little hint that proves homebrew and money combined still have a chance at hacking this 360!

If applicable at all, this brings us further from softmodding and exploiting the Xbox 360, because many would-be developers will trade the full, albeit less legal, development environment for a sandboxed environment, just because they can't wait until the machine is truly homebrew 'compatible'.

And I doubt that it will be possible to exploit the 360 using the XNA. Remember, you must be connected to Live. EVERYTHING will be monitored (one of the reasons that everything is so slow, perhaps).

everyone said you whont be ale to port emus in xna or sandbox mode and well this guy did it being a direct port yea i can see it being slow a bit of fixing in the code will fix that.

I love the whiners on this page "ooh I want homebrew" Waaah.

$100 per year for 100% legal homebrew is a small price to pay, seriously. What nobody tends to realize is legal homebrew is what's going to actually push the limits, instead of a Warezed XDK and hacked bioses, etc. Everyone loves XBMC and the other original Xbox programs but they're also all illegal considered they were made using a Warezed development kit. This is giving those who want to play a chance to play and see what they can do, perhaps in the future we'll be able to spend points to download or play individual programs without the need for the full subscription. A quick and dirty port that runs at 70% of speed is pretty cool in my book, now once a little optimizing and tweaking is done then it will be pretty sweet.

To combat piracy Microsoft is giving everyone a legal alternative for their homebrew, now that's not good enough? I'd suspect people want a fully exploited Xbox 360 for more than just their homebrew.

Get off the high horses about how this is so uncool or whatnot because people don't desire to hack the Xbox 360 now. I'm going to go pretty far out on a limb and say most of the hacking community isn't interested in homebrew as much as they are getting their "back-up" to play.

If XNA can't even run a NES emulator without optimization, XBMC will NEVER run. This is just more evidence that XNA is just a bullshit PR move by Microsoft that has no real use.

Not to brag or anything, but a friend of mine actually had Nester ported to XNA before 2006 year end. Only Mapper 0 games are supported since he didn't bother porting the other mappers yet.

The project was abandoned however because there is presently no way to get emulated sound working in XNA and he thought that was too big of a missing piece. XNA does not provide a way to access sound buffers directly so one can't properly emulate NES sound. The whole sound bank stuff isn't viable at all for what needs to be done.

I will ask him to contact the author of XNA SharpNES and share the code.

Damn I didn't realize how close in architecture the NES and X360 must be, that's the only way your logic would be even reasonable.

You're basing everything off a quick and dirty NES port? The guy spent what a few hours on it, he didn't even take the time to optimize or whatever else. XNA isn't the ultimate porting application, you do need to do some code-work. I can't just take the source to whatever and expect it to run perfectly.

Given the length of time XBMC has been worked on and available I'd say there's no reason anyone can't make something to blow it out of the water in roughly the same about of time.

Fuck, heaven forbid some of you people ever be a beta tester for a project. "Doesn't do it perfect right out of the box, must be impossible, time to whine like girls".

Considering the XNA kit is based off the actual developer kits I'd say it can do a lot more than we've seen yet.

It's the crybabies on here though about "we need to hack it" because we want homebrew and we don't want to use our legal options. Caustik had the same problem with the OpenXDK, nobody wanted to use it because it was easier to use warezed kits. The Dreamcast scene got it right, everyone else has been off base.

Finally.... bashing developers who do nothing more than something quick for fun sure turns them off to doing much of anything in the future, for the legal or illegal scene. People enjoy a thank you once in a while, even if it's not what you want.

this i s a great start considering its in xna

Dude, the NES has an 8 bit CPU at 1.7 Mhz (megahertz that is) or something like that, an xbox 360 has a 3Ghz triple-core CPU... Even a 486 can easily emulate a NES at full speed, even if the emulator is straightforward and unoptimized. Emulating something like a SNES is already 10 times harder, so don't even think about PSX.



Well for starters, XNA does not have full network access, so forget internet radio, media sharing etc. AFAIK sound and video access is also severely limited, so forget accelerated scaling or decoding. Writing codecs in XNA is simply impossible, so forget about decent video, let alone HD. Also, do you really think MS would allow a homebrew MCE competitor that's as good as XBMC on the 360??



That's just utter BS



You, sir, should try reading up on emulation and homebrew a little more. Comparing dreamcast or OpenXDK homebrew development, ie: native code running directly on the hardware, with a sandboxed interpreted environment like XNA shows that you know shit of what your talking about. Funny thing btw: there actually was a pretty good PSX emulator for the dreamcast that ran at near full-speed, so maybe that makes it a little easier for you to realize the difference between native code and sandboxed bytecode like XNA.

Great start, now we need a C64 emulator .

If XNA can not emulate a 1.79Mhz 6502 CPU at full speed using unoptimized code then XNA as a homebrew platform is pretty much garbage.

Kudos to the author for taking for the time and effort to port this over to XNA. It is really is a step in the right direction, do not lose sight of this because of Microsoft's imposed limitations..

I just wish Microsoft would allow us to run this our 360 for free. Then it would be with a bit of optimization.

Wow, such a heated argument. I just came to the comments section to say good job to lone coder. Truthfully we would have never known the extent of xna's limitations till someone actually had the time to try. I dont think XNA is gods gift to homebrew or what not, but at least its a push in the right direction. Anyways, why are we fighting about this, essentially we all want the same thing out of the 360. So instead of arguing about stupid comments, we should focus on teaming up, and doing the best we can with what we got. -Zenofex-

Are there any plans that anyone knows of for an XNA Express kit or something that would just launch programs that other people make? I don't want to pay 100 freaking dollars.

The official answer of MS engineers in xna official forums to those who cry about low speed of their code under xna reveals everything.

To prevent hacking of the console any code is just virtual (interpreted, jited or not, accessing virtual hardware, surveyed by hypervisor, etc...).

Any "safe" (for MS) code running on CPU is crippled, crippled, crippled...
They don't even trust developpers who purchased official sdk, signed non disclosure agreements, etc...

So their answer is... think "shader". Why? Because there is no crippling (security) at all on GPU side.

That means if you have your code ported to shader language, your dreams will become true on xbox360.

Sound problem remains though... No hope that a gpu shader can interact with sound outputs (I think...)
To be able to upload micro-code into a DSP would have been the solution (like on soundblaster live, ES, ZS) if DSP was able to access memory like shader can do on xbox360... But MS engineer may realize it and do something in future.

Another problem is... once you made the effort to master shaders, you realize, more expensive and more recent graphic cards are more interesting, i.e what you just learnt open new opportunities for you. And somehow you realize MS is a two headed-dragons having his heads fighting each other...
Financially, the game division is in the red. The Windows division will probably disrupt the console market with Vista+DirectX10 compatible card. They realized it and announced recently they will try to release simultaneously games for both "heads". But the hardware of one head can evolve, the other won't.

Also personally, sometimes I surprise myself playing PC, XB1 or PS2 just because I want some "silent" game session... i.e I limit the time I play with xbox360 because of the noise.
So... for me the best usage of xbox360 is to play xbox360 exclusives. I'm no longer interested in homebrew on the raging thing (porting to shaders complex code requires really a lots of brain energy and I'm a bit lazy...)

i tend to agree with most people on this. while its awesome that someone managed to port an NES emulator through XNA...xna is weak man. you'll never see XBMC running through XNA, regardless of how long its developed for. Unless MS really REALLY opens up the system to us, then homebrew is a no no without somesort of hacking to allow us full access to the system. Im starting to wonder if MS realizes just how much they could make by selling an actual XBMC created by them. a Vista media center port would be awesome (espcially since IPTV is coming, and the vista media center interface is freakin slick). sell that as a disc or on live and they'll make mad money (provided it allows all codecs, w/updates for codecs, and will work with external drives or through network access). thats my opinion on that one.

XNA isn't as crippled as people think.. Just look at XNA racer or Rocket Commander, they look nearly as good as Store bought games.

A good NES port at 100% speed is definately possible. People can't spend the time to do any research immediately bad mouth a great first effort. Optimizing can lead to 100%, even 1000% gains. You can't learn Chinese and expect to become Major of Japan.

He still needs to:

- Code cleanup (Garbage collection, iterration, etc. work a little differently on XNA)
- Optimize code for 360 hardware
- Optimize graphics for DX9
- Weight the benefits of managed code

Great work on the emulator! Homebrew is alive and well on the 360..

-dp

Exactly, people are saying XNA is shit because he did nothing but throw the code at the port in an hour or two tops.

I think 99% of the whiners just want a hacked console for piracy. It's pretty pathetic if you ask me.

Btw, the next asshole who says I don't know shit about emulation or emulators such as the Dreamcast scene where they used a 100% legal development environment vs. the Xbox scene needs to produce their own NES emulator for the Xbox.

To the whiners. Put up or shut up. If you don't like XNA, then hack the Xbox 360 yourself, don't troll the forums though.

The point is, even if the emulator was written completely crappy, it should easily play at full speed. Its hard to fathom, even without optimizations, that an NES emulator would run at anything but full speed. This just goes to show how horrible the XNA framework environment really is.

Its unfortunate, but we may never see N64 or PlayStation emulation using the XNA framework.

then your no coder. unoptimized code and run relly relly slow. even a 8 bit emulator. i have done 2d games in my time and unoptiized it runs like shit on my 1.8 mhz pc but optmized it runs fine in a 486. why couse it was memery leaking like crazy cousing everything to slow down to a crawl cpu usage 99% when i was done it was down to like 2%. the point is raw code like this is very very slow no matter what it is running on. i bet the emu is even slow on a windows pc.

weather it is ust the raw code or the xna i still think is guess work. beingi have seen some compleated xna apps i would say its couse its a raw port once someonepicks up on it and cleans it up it would get gains in the thousands when it comes to speed. as for sound that might be a bit harder.

It's a start at least! It shows that the "sandbox" enviroment may not be so limited. Now, i'm dying to get ready to learn some C# and see how I can possibly contribute with these kinds of projects. This is definitely a good sign in the meantime! It has begun!

=Xtreme=

Doh...

Unoptimized it runs like shit on 1.8Mhz, and optimized it runs ok on a 486 .

Let me put this in perspective: if your emulating a NES on a 360, you literally have about 3000 CPU cycles to waste for every SINGLE cycle your emulating. In 3000 cycles you can do *a lot*, not even considering the 360 ISA has instructions that handle stuff that takes 10s of cycles on a NES in a single cycle on a 360. Now what does this mean: if I would port an *existing* NES emu to a 360 using XNA, without optimizing or tweaking it, and it doesn't run at full speed while it DOES run at full speed on a low-end x86, then XNA apparently add so much overhead that you can forget about emulating anything remotely interesting using it.

Sure, you might be able to speed it up like 10 times if you spend enough energy on it, but you'll still end up with a freaking NES emulater. Like I said, emulating SNES is already 10 times harder, emulating PSX maybe 1000 times. Considering the fact that a dreamcast can emulate a PSX, and XNA already has problems with an 8-bit 1.7 Mhz machine from 20 years back says me it is useless for emulators.

It's still pretty cool and all that someone did this, kudos to him. But some people should just see the whole thing in perspective... Sure you'll be able to do some nice stuff like the XNA racer, but if it gets even a tiny little bit more demanding than straightforward polygon pushing, XNA is pretty much useless..

acully i spoke to a xna coder today after i posted that. he says it has a garbage collection system and it can be very slow. bascily if it memery leaks it toses it in garbage. bascily couse of this system its cousing some slowdowns unfortanly he does not have a 360 to acully compile test runs to see whats cousing it to slowdown its just his knolage of the ce net system and xna. we also talked abought acully getting a rom loader working so we where reading some docs on how it acesses the hd and memery cards. he beleves getting it at fullspeed whont be to hard we didnt relly do talk to mutch abought sound.

my frends is a very good coder if i can get him to work on it we can acully probly get everything working hes done some nifty stuff with limited code befor.

For 50 usd I'll sell u my nes and star wars cart. Good work none the less.

That's true, though you wont be able to eliminate all gc overhead by fixing memory leaks. You might be able to tune it and keep the overhead to a minimum... For some purposes gc can actually be faster than doing all memory management in your code, but not for emulators I think.



I'm pretty sure he can get it at full speed if he ported it in a few hours like you said. Not sure about the sound though, I'm not an XNA/C# expert but I'd think it proper sound emulation using XNA would be very hard, even harder than the PCM-type audio that modern consoles use. It's cool stuff anyway, but I'm still holding my breath for real, unrestricted homebrew...

Yet you are willing to pay all that money or all that work to illegally mod your 360.What about 2 years of xbox live? 100 dollars a year is less than paying monthly for an MMO and you can do alot morehtings through XNA.

100 dollars is nothing for most of the people on this forum, they just want something to bitch at Microsoft about.

I haven't done ANY modifications whatsoever to my 360. And I haven't had Xbox Live for 2 years either. Also I don't play any MMOs. I just want an express version because I don't want to program anything on the 360, I just want to be able to run the programs. Shoot I'd pay MS 50 bucks or whatever for it, just not 100 per year.

d-range,

CPU speed really doesn't have anything to do with it.. I'm sure if he totally uncorked it, it would run at 400FPS with no further tweaks.

The point is to sync the framerate and game speed to your platform. That NES emulator is intentionally slow.. it's just not properly sync'd yet to the Xbox yet.

I don't know how far you go back, but at one point they had PCs slower than 5mhz. When you play these games today, they are uncontrollably fast. Why? Not sync'd.


Again, people are bad mouthing something they know little about. The XNA framework is incredibly robust, you just have to be able to code for it. Nobody says PS2 sucks because it's harder to code for than the (PC based) Xbox 1. Last time I checked, it whupped it pretty good.

Whether or not the framework is INTENTIONALLY crippled for some purposes (Video,etc.) is Microsoft's perrogative and still up for debate. Seeing how seriously they take DRM, It wouldn't be a shock.

-dp

For sure. Most of the SharpNES code is incredibly inefficient in terms of memory usage. It has nothing to do with XNA, it has everything to do with starting with unoptimized code that runs poorly on a PC to begin with. He can easily get it to full speed.

-h

Pretty pricey for just playing the NES.

Great work though!

i would take this port as a framework of things to come. if we can get this 100% working maybe we can get other things working to.

Itll be good to get true homebrew running........and by that i mean not paying the gay $99 fee just to be able to run the programs (im also cheap aswell lol)

AFAIK a properly coded emulator does not sync to CPU speed and is not intentionally slow to compensate for the speed of the emulated system. It syncs using a cycle counter or timer. But for a NES maybe it's different because video and sound output is harder to sync using timers or interrupts because of the primitive hardware, I don't know enough about NES emulation to judge on that. Anyway if SharpNES would have been properly coded in the first place I think emulating on a 360 in C# or a random PC in C/C++ should not really matter in terms of emulated speed, or it wouldn't run properly on a high-end PC either.

quote from TwistedSymphony's blog ThoughtHead, commenting on why PS3 homebrew is not important.



That's one reason why people would rather have native hombrew instead of the XNA. The other? Obviously price.

@Mike96sc2, I have the firmware hack and 35 blank DL discs. Blank. You know why they're blank? Because I don't enjoy playing games any more. I have more fun hacking the system itself than I do playing it. I haven't been on Live since December 10th. If I wanted piracy like you claim, wouldn't I be content with the firmware hacks? I had fun learning about the drive firmware, but I want more. I wan to be able to run native homebrew, and mostly because it's something that we're not supposed to be able to do. It's more for me about the actual "doing it" then it is "using it".

this is great maybe more ppl will make programs and stuff easier cuz it works on the pc too, the olny this that obviosly sucks is you could get another live subscrition (wich im already agianst) for that money. its just another way to nickel and dime us to death.