Emulators runs games faster VS Real Hardware?

I was talking with a Friend discussing why I emulate some consoles VS the Real hardware. And he mentioned something to me that Struck me as weird as I never noticed before.

He thinks that emulators seem to run games faster VS the Real thing. I have NEVER observed this before in all my years of playing emus or real hardware.

Now since Roms are a touchy issue with some and I've struck some nerves on this topic I'll only cite examples from games I actually own that I emulate and BEFORE SOME PEOPLE (I'm looking at you jamtex) get on my case about emulation YES I do play games on real hardware just as much as emulated.

Einhander (US Version) I've emulated 2 different ways through Bleemcast which obviously wasn't a perfect emulation but it sure as hell looked gorgeous running on my Dreamcast. I have played it from disc on PSX
and never noticed any speed issues slower or faster VS playing it though my PS2.

My main NES emulator is Nestopia I play using a Buffalo USB famicom pad.
Playing Mega Man 3 a game I can clear blindfolded, Know inside out backwards and forwards and can ALMOST clear in 1 life (DAMN YOU GEMINI MAN) I've never noticed any differences vs playing it on the NES, the ports/emulations on XBOX/PS2 playing it emulated on Dreamcast or through Nestopia or JNES (my back up for the few mappers Nestopia doesn't support) they all play identical to me besides of course the graphical differences in the anniversary collection versions.

Dreamcast emulation isn't perfect (though Demul) but Soul Calibur plays the same as does
Border Down although Border down has some sound issues that annoy me.

So those are my two examples. Is there any weight to this speed thing?

Can anyone cite some differences they've noticed Emulated VS Real Hardware?

Edit-
OH DUH...I forgot to point out, I'm aware they aren't running at native resolution although I know ATI makes a special graphics card just for emu nuts that supports lower resolutions.

 

MSR plays better on Emulated hardware vs real hardware, Has a nice FPS boost.

 

I never tried MSR emulated Demul seems very biploar when it comes to running games for me like Rez and Sonic Adventure make my computer have a seziure (loud beeping noise that I can't tell is coming from internally on my computer or the speakers)

Then again I never really go the hang of MSR to begin with.

 

If you have only ever used pal 50 technically using the NTSC option on an emulation it would be faster.

 

It depends on the emulator. Sometimes games will run slightly faster on various emulators. Infact ZSNES for a long time atleast ran SNES games "faster" than the real hardware, particularly because it didn't emulate the different CPU timings based on memory region accessed. I also don't think it emulated the DRAM refresh period.

Super FX games certainly might run better than on real hardware since most emulators if not all do not emulate the delay cache of the chip.

The thing is unless you look for it, most of the time the average person will not notice the difference between emulation and the real thing, unless something is blatantly wrong.

Personally I do not use emulation much anymore except for testing, debugging, hacking, or maybe for MAME. Emulation is handy but I'd rather have the real thing.

 

Some games do run faster on emulators but that is usually due to them not emulating the hardware 1:1 accurate and the game won't have slowdown where it would normally occur. Used to see sections of Mega Man X for the SNES have massive slowdown on real hardware but next to none on older versions of ZSNES. It was a nice feature but wasn't accurate.

The same can be said for Goldeneye for the N64 which is helped greatly by overclocking the N64. It runs "faster" by virtue of not having to slow down the framerate to handle everything on screen due to being overburdened.

However, in terms of the real hardware running at X fps and the emulator running at X+Y fps? I'm sure it happens a bit due to the emulation not being 100% perfect but I sincerely doubt it runs fast enough to really be that noticeable across the board. I guarentee you'll find a few choice example roms that run at slightly different rates on choice versions of select emulators. You'd have to be of poor logic to think that rom+emulator=faster than the real thing 100% of the time no matter what.

Besides, if timing was off by even a few kilohertz (or less) on a CPU core you'd see some interesting problems with sound, controllers, etc. BSNES aims for 100% accuracy and he has some interesting blog posts about such matters. Everything would have to be working in sync to be faster than the real thing.

EDIT: seems in the 10 minutes it took to write this MottZilla beat me to the punch.

 

Yeah in some cases (like the Dreamcast and Tekken 3 in MAME) I notice some games run SLOWER like unplayablly show not faster.

 

I know PSone (PlayStation) games are emulated on the PS3 using software.

Crash Team Racing has huge FPS issues, and I can not play it at all because you need to have precision timing to make some shortcuts, etc... If it was on the hardware, then I am certain the game would run fine.

 

Crash Team Racing runs great on Dreamcast through bleem....if you can forgive the fact that it paints the polygons as you drive like it was De Blob or Okami

 

I always thought of emulators being very capable of going far beyond the limitations of real hardware, but I guess that's only the case if they're optimized to do so.

Emulators can always act faster than the real hardware via coupling speed hacks and tricks with better technology, but I guess on a wide scale the real hardware is obviously far more efficient than the fully-emulated equivalent of it running on different hardware.

Slowdown is generally much more noticeable whenever something is supposed to run at 60FPS. Apparently, it's not noticeable to the human eye, but I've always noticed it to be much more fluid than 30FPS (especially with video, look at film vs live taping).

Sonic is a great example of this. Look at the US/JP version of Sonic 2 followed by the PAL version and finally by something like the Sonic Jam port.. Shows how botched the gameplay can become just by having a little bit of a slowdown here or there.

 

Yeah all emulators of note still aim to make improvements within reason. No matter how accurate their goal is.

A draconian obsession would necessitate recreating supposed "unanimously undesirable" aspects or bugs of the hardware. Say NES having composite video, N64's hideous blurring, Genesis jail bars, PS2's load times, SNES's faint vertical line, FDS's unreliable drive mechanism...

 

When Possible I play with Scanlines I used to use SaiX2 filters or whatever it was called
but they didn't improve the image at all just smoothed them out to the point where some detail was lost.

I have a Physical copy of Slipheed but for whatever reason I can't get it run in either Gens or Fusion (while the rip of it I have will play in gens)

 

Ever try HQ2X/HQ3X? I was always fairly picky about filters, but these actually clean up the image fairly well. They just require a little more power than the other typical filters like Sai or Super Eagle.

I've never liked using scanline filtering because it usually makes it look a lot crappier than it typically would on a real CRT.

 

I've always been curious as to whether or not it would be possible to make a PSX emulator that eliminates polygon bending. It's so bad that I think it's actually better to play in the system's native resolution, because the polygon bending gets really distracting at high resolutions.

 

It depends on the intensity and the Emulator.

 

x

 

PSX Emulator seems to sharpen the graphics up a little than the real PSone hardware. Nintendo 64 runs at a strong 60FPS, and that's really all I notice most the time.

 

I usually use 2xSai or hq2X on KegaFusion. Any of those two combined with 25% scanlines makes it perfect for me. On Zsnes i just go for 25% scanlines (somehow you can't use that with filters like on KegaFusion).

Anyway, in the end i think it's personal to use scanlines or not. I think 25% makes it spot on.

On Mame i think i use the effect Aperture2x4.

 

Eh, I've never really noticed speed differences vs. real hardware, but it is a personal preference of mine to play a game on the hardware it was designed for.

 

25% scanlines is all you need. No need for any other filters. Well, maybe high resolution mode 7 on the SFC but apart from that.