They fixed the sound in Raiden Fighters!

This is the main shmups forum. Chat about shmups in here - keep it on-topic please!
User avatar
BPzeBanshee
Posts: 4859
Joined: Sun Feb 08, 2009 3:59 am

They fixed the sound in Raiden Fighters!

Post by BPzeBanshee »

How the hell did I miss this:
- raiden2.c: Solved the remaining issues with Raiden II sprite decryption
- seibuspi.c
. Fixed YMF271 core timer allowing for greater accuracy
. Fixed music in the boss levels of Raiden Fighters and Raiden Fighters 2 play (Mission 1-3, 2-3) slower than the original (ID 02675).
So not only are the speeds finally right for those spots after that MAMEtesters bug was reported several years ago, but the garbage graphics in Raiden II are fixed (although it's still unplayable due to missing logics, so close though).

Game on, Raiden fans! Grab your latest MAME 0.145u6 now! :D
Zeron
Posts: 927
Joined: Thu Jul 29, 2010 2:40 pm

Re: They fixed the sound in Raiden Fighters!

Post by Zeron »

That is amazing but
- raiden2.c: Solved the remaining issues with Raiden II sprite decryption

This means Raiden II and DX soon in playable state?
User avatar
trap15
Posts: 7835
Joined: Mon Aug 31, 2009 4:13 am
Location: 東京都杉並区
Contact:

Re: They fixed the sound in Raiden Fighters!

Post by trap15 »

Not really; that was just for some graphical glitching. There's still a bunch of stuff to do.
@trap0xf | daifukkat.su/blog | scores | FIRE LANCER
<S.Yagawa> I like the challenge of "doing the impossible" with older hardware, and pushing it as far as it can go.
Zeron
Posts: 927
Joined: Thu Jul 29, 2010 2:40 pm

Re: They fixed the sound in Raiden Fighters!

Post by Zeron »

trap15 wrote:Not really; that was just for some graphical glitching. There's still a bunch of stuff to do.

It's in better shape than before though Raiden II,DX and Zero TEAM green status might be a reality sooner or later.


To continue my whining on about Raiden Fighter ACES it seems that the MAME emulation perfection might catch up to the PAL release....
User avatar
BPzeBanshee
Posts: 4859
Joined: Sun Feb 08, 2009 3:59 am

Re: They fixed the sound in Raiden Fighters!

Post by BPzeBanshee »

Zeron wrote:
trap15 wrote:Not really; that was just for some graphical glitching. There's still a bunch of stuff to do.

It's in better shape than before though Raiden II,DX and Zero TEAM green status might be a reality sooner or later.


To continue my whining on about Raiden Fighter ACES it seems that the MAME emulation perfection might catch up to the PAL release....
My thoughts exactly. The only stuff I've seen besides this fix in the GIT repo is basically a whole bunch of developers sidestepping around it and changing comments in the hopes that someone else will get on to it and fix the whole lot. Then there was one guy's rehandling of some stuff in the driver which made no visible difference, and then this. Here's to the remaining unemulated COP-X functions getting done sooner or later. :D

The music in Raiden Fighters still sounds like it's missing some bass or maybe an instrument line somewhere but it does sound A LOT closer to the OSTs now, particularly the Aces OST which has less bass than The Onslaught Raiden Fighters soundtrack.
IseeThings
Posts: 534
Joined: Thu Dec 16, 2010 6:38 pm
Location: California

Re: They fixed the sound in Raiden Fighters!

Post by IseeThings »

BPzeBanshee wrote:
Zeron wrote:
trap15 wrote:Not really; that was just for some graphical glitching. There's still a bunch of stuff to do.

It's in better shape than before though Raiden II,DX and Zero TEAM green status might be a reality sooner or later.


To continue my whining on about Raiden Fighter ACES it seems that the MAME emulation perfection might catch up to the PAL release....
My thoughts exactly. The only stuff I've seen besides this fix in the GIT repo is basically a whole bunch of developers sidestepping around it and changing comments in the hopes that someone else will get on to it and fix the whole lot. Then there was one guy's rehandling of some stuff in the driver which made no visible difference, and then this. Here's to the remaining unemulated COP-X functions getting done sooner or later. :D

The music in Raiden Fighters still sounds like it's missing some bass or maybe an instrument line somewhere but it does sound A LOT closer to the OSTs now, particularly the Aces OST which has less bass than The Onslaught Raiden Fighters soundtrack.
No visible progress is not the same as no progress. I'd hazard a guess that the things you've written off as no visible difference will be a LOT more important to getting Raiden 2 running in the long term than this.

Let me clarify things. Andreas is the current 'encryption expert' involved with MAME (he did the code decryption on CPS2, CPS3 and a whole bunch of others). He deals with encryption, and only encryption. Raiden 2 sprites had been 99% decrypted before by Olivier, but he was stuck on a few details. Andreas figured those out. It's good, it means the sprites are now clean, but it means absolutely NOTHING when it comes to the protection, which isn't something Andreas deals in.

The protection is what is preventing the game from being playable, not the encryption. Had the protection been figured out you'd see a playable game with some slight corruption in the sprite graphic data. The two things are completely separate.

The protection needs to be understood, at a low level, in order to simulate the behavior of the device with confidence. This will mean lots of small changes, with no obvious visible difference, which often may make things WORSE in an effort to improve them in the long term because quite a lot of what you see now works due to hacking up approximate behaviors for protection commands rather than being able to fully understand them. The encryption is done now, it doesn't need changing.
User avatar
burgerkingdiamond
Posts: 1571
Joined: Wed Oct 06, 2010 9:56 pm
Location: Virginia, USA

Re: They fixed the sound in Raiden Fighters!

Post by burgerkingdiamond »

So is the Raiden II protection particularly tought to crack, or is there just not enough interest from MAME devs?
Let's Ass Kick Together!
1CCs : Donpachi (PCB - 1st loop) Dodonpachi (PCB - 1st loop) Battle Bakraid (PCB) Armed Police Batrider (PCB) Mushihimesama Futari 1.5 (360 - Original) Mushihimesama Futari BL (PCB - Original)
IseeThings
Posts: 534
Joined: Thu Dec 16, 2010 6:38 pm
Location: California

Re: They fixed the sound in Raiden Fighters!

Post by IseeThings »

It's the same chip used on a bunch of 90s Seibu games (or at least a revision of it) and yes, it's nasty.

It has many functions, and even it's own 'macro' style language which the games upload their own variations of the protection routines to but the individual commands can be quite complex, and appear to have lots of subtle behaviors, and flag logic as well as effects on other parts of the protection like the DMA. For the time period we're talking about it's really one of the nastiest entirely custom solutions out there.

Apparently they developed it and had engineers / programmers dedicated to working on the protection for each game individually, not even allowed to talk to the rest of the development team because of the high level of bootlegging with their earlier games, including Raiden.
User avatar
BPzeBanshee
Posts: 4859
Joined: Sun Feb 08, 2009 3:59 am

Re: They fixed the sound in Raiden Fighters!

Post by BPzeBanshee »

Well said Haze, though I still think there's quite a few techheads sidestepping - clearly for the right reasons considering how complex this is. The 'no visible difference' I was referring to was the rewrite of some code that apparently made things easier to look at even though it did exactly the same thing. Optimising code is good and all, I do it all the time on my project, but having seen that and next to nothing but comment changes for the last several years on raiden2.c and co. makes it a bit difficult to spot when something serious is actually going on.

Anyone can write a comment line, but clearly to figure out the 'protection' it takes a real genius who hits his head against the wall quite often. Still a shame that any mention of it seems to be a 'buzzkill' for any of the devs though. I'd like to hope there's progress being made but when MAMEdevs talk like that on their own forum combined with what I'm seeing in the repo for it (and the 'Raiden II Dev Team' looks like they've given up) is it really surprising that I'm left a bit cynical as to whether to distinguish this as any different from say, the impossible to emulate PGM2 hardware we discussed over PM some time ago?

...and having read all that it does make me sound like quite a prick, so I'll say here and now this isn't meant to actually hurt anyone's feelings. It's been a small step in the path to success! I'm just sussing it out like I always do. :oops:
User avatar
Ed Oscuro
Posts: 18654
Joined: Thu Dec 08, 2005 4:13 pm
Location: uoıʇɐɹnƃıɟuoɔ ɯǝʇsʎs

Re: They fixed the sound in Raiden Fighters!

Post by Ed Oscuro »

I have a dumb question, and hopefully it's not impolite - aside from the sprite buffering, what's the feasibility of getting MAME-created lag out of drivers so that games respond like they're supposed to?

Don't get me wrong, I love having stuff look and sound as good as it does now, and it's great that this is happening. I have trust in you guys that the new changes have a purpose.

Some random thoughts though, not able to upload these to MAMEtesters at the moment, and not sure they're new issues as a result:
- Blandia sprites are screwed up, don't remember this in the past (some garbage in the intro too)
- Ark Area's graphics smear after about 14 stages in (I think this happened many years ago too).
IseeThings
Posts: 534
Joined: Thu Dec 16, 2010 6:38 pm
Location: California

Re: They fixed the sound in Raiden Fighters!

Post by IseeThings »

Ed Oscuro wrote:I have a dumb question, and hopefully it's not impolite - aside from the sprite buffering, what's the feasibility of getting MAME-created lag out of drivers so that games respond like they're supposed to?
MAME isn't adding any unavoidable lag, unless you start turning on stuff like triple buffer, in which case you're actually asking for it. As I've said countless times before Shmup MAME is hacking drivers so that the emulation is *incorrect* in order to reduce lag. Baseline MAME will not do that, sprite buffers are emulated where they are present. In a handful of cases it's possible emulation issues are causing some lag (communication between CPUs not being quite right) but that's an absolute minoirty of cases, the main one I wonder about being STV, although that could in equal part be original developers not being able to handle the system well.

If your input devices (USB, Wireless, whatever) or monitor (upscaling etc.) are adding lag then MAME can't do a thing about that. Likewise graphics cards requiring images to be buffered / double buffered internally, it becomes outside the control of the emulator.

You're inherently going to get 1 frame of lag with any emulation, because you have to render to buffer (when the original would be rendering to screen) then present buffer in the next refresh, that's simply how current hardware works. You don't have the granularity of control on any modern system or OS to avoid that. I guess in theory you could double the refresh rate output to get that to half a frame, but most modern moderns again don't give you that level of control, nor have the response times.
Ed Oscuro wrote: Don't get me wrong, I love having stuff look and sound as good as it does now, and it's great that this is happening. I have trust in you guys that the new changes have a purpose.


Some random thoughts though, not able to upload these to MAMEtesters at the moment, and not sure they're new issues as a result:
- Blandia sprites are screwed up, don't remember this in the past (some garbage in the intro too)
Yes, I noticed this, possibly a result of the cleanup of the Sprite code, there were 5+ implementations of the same chip before, now there is 1, but the odd game isn't happy with it. I didn't notice it when I was doing the actual cleanup tho, so it could be something else. It's already reported on MT anyway.
Ed Oscuro wrote: - Ark Area's graphics smear after about 14 stages in (I think this happened many years ago too).
This is correct, it happens on real hardware. It's a special effect, not a bug. The hardware sprites are framebuffered, and they specifically turn off the layer clear. A number of other UPL games do the same.
User avatar
BPzeBanshee
Posts: 4859
Joined: Sun Feb 08, 2009 3:59 am

Re: They fixed the sound in Raiden Fighters!

Post by BPzeBanshee »

Yeah, Shmupmame without a doubt is a hack, just as without a doubt there is at least one frame getting removed from input response times at the cost of sprite/background syncing. I'm still not convinced the new builds of MAME are any better in this regard, but trap15's comments about the restructuring so that one change can switch off this buffer for multiple games is quite interesting.

As for refresh rate output, I'm not sure if I'm getting confused between interpolation of frames or outright refresh rate, but the new TVs/monitors aren't exactly consistent at refreshing at say, 120 Hz are they? I certainly avoid them like the plague and I'm hoping the day my 4:3 LCD monitor carks it never happens. I've seen MAME do strange things when you try to force it to sync to specific refresh rates in some cases though as Haze said I asked for it (wasn't literate back then).
captpain
Posts: 1783
Joined: Wed Jul 23, 2008 6:23 am

Re: They fixed the sound in Raiden Fighters!

Post by captpain »

yikes
Last edited by captpain on Sun May 26, 2013 6:43 pm, edited 1 time in total.
IseeThings
Posts: 534
Joined: Thu Dec 16, 2010 6:38 pm
Location: California

Re: They fixed the sound in Raiden Fighters!

Post by IseeThings »

I'm telling you, as fact, based on the code, that MAME isn't adding any unavoidable lag.

Maybe the rest of your hardware is, maybe the drivers on your PC are (Windows 7 seems to do some horrific stuff with video) but beyond the unavoidable way modern hardware works, MAME isn't adding anything unless you tell it to, or unless the original hardware had it.
User avatar
D
Posts: 3805
Joined: Tue Feb 01, 2005 3:49 pm
Location: Almere, Netherlands
Contact:

Re: They fixed the sound in Raiden Fighters!

Post by D »

Seibu made one hell of an amazing protection.Those devs must still be very proud to this day
Their games weren´t half bad either :wink:
User avatar
Ed Oscuro
Posts: 18654
Joined: Thu Dec 08, 2005 4:13 pm
Location: uoıʇɐɹnƃıɟuoɔ ɯǝʇsʎs

Re: They fixed the sound in Raiden Fighters!

Post by Ed Oscuro »

IseeThings wrote:I'm telling you, as fact, based on the code, that MAME isn't adding any unavoidable lag.

Maybe the rest of your hardware is, maybe the drivers on your PC are (Windows 7 seems to do some horrific stuff with video) but beyond the unavoidable way modern hardware works, MAME isn't adding anything unless you tell it to, or unless the original hardware had it.
Let me describe what I am seeing so you don't focus on irrelevant things like one-frame lag games and the like.

The standard Shmups Forum test for lag in MAME, which is time-independent (so, as you say, some games may appear to have no lag but still have an unavoidable hit; unavoidable lag is obviously not the issue), consists of pausing, inputting a direction while frame-skipping ahead, and then reversing direction to see how many frames are required to register the movement. It is important to note that this lets us idealize away many of the nuisance factors you mentioned, and this is clear because results are consistent from platform to platform.

Some games appear to have no artificial input deficit - like Terra Force, which registers the input immediately in the next frame.

Others take a very long time and, more to the point, actually feel sloppy when they're being played, like Battle Garegga, which registers the input after something like five frames.

The problem, then, is that MAME appears to take its sweet time (in emulation cycles) to get inputs to some games. How and why this is a problem for accuracy should be obvious. That there is a theoretical limit on how little lag can be presented is obvious (I assume the total formula for lag in getting inputs to games emulated by MAME, when playing the game, would look like "(A) system hardware latency + (B) operating system + (C) MAME interface with the game + (D) game + (E) MAME interface + (F) operating system + (G) output devices (graphics card, monitor,") but that theoretical limit does not (to my understanding) change from game to game. When frameskipping, we have essentially held all the variables constant (and probably rendered those past the emulated game, i.e. F + G and possibly E, completely irrelevant) except for (C), MAME's interface with the emulated game, but there is still a wild divergence. Ergo, MAME is inconsistently laggy.

Of course, there is a hidden assumption here - that the original Battle Garegga board does not have five frames of lag before it registers inputs at the control panel. That doesn't strike me as controversial, especially since (using the frame skip procedure) we can idealize away many nuisance factors.

I could explain more about how I am aware of the subjective vs. objective aspects of MAME lag, and how I account for those - but the point, which I hope is not being missed, is that in the best test so far devised MAME is clearly inconsistent in its lag from game to game, in head-scratching ways.

I assume that the response could be that the contributors of lag before MAME's interface with the game have unusual effects magnified in some games but not others, i.e. lag components A + B + C are somehow going to cause MAME (at D) to not push the user input to the game quickly enough. This argument seems unreal to me, because obviously Terra Cresta has no problem at all. It shouldn't be the data size presented to the game at fault, since my keyboard and operating system don't appear to send different amounts of data to Terra Cresta than to Battle Garegga, and so on, and timing is MAME's job (and we aren't dealing with the situation of MAME being CPU limited in either game); the operating system should be agnostic about whatever game is actually being played in MAME.
User avatar
trap15
Posts: 7835
Joined: Mon Aug 31, 2009 4:13 am
Location: 東京都杉並区
Contact:

Re: They fixed the sound in Raiden Fighters!

Post by trap15 »

Ed Oscuro wrote:retarded conclusions
The games originally have that much 'input lag' (sprite buffering). MAME does not add any additional frames that aren't present on original hardware (unless the system is improperly emulated, but that's not a wide many).
Basically, listen to IseeThings, he knows what's up.
@trap0xf | daifukkat.su/blog | scores | FIRE LANCER
<S.Yagawa> I like the challenge of "doing the impossible" with older hardware, and pushing it as far as it can go.
User avatar
Ed Oscuro
Posts: 18654
Joined: Thu Dec 08, 2005 4:13 pm
Location: uoıʇɐɹnƃıɟuoɔ ɯǝʇsʎs

Re: They fixed the sound in Raiden Fighters!

Post by Ed Oscuro »

trap15 wrote:
Ed Oscuro wrote:retarded conclusions
The games originally have that much 'input lag' (sprite buffering). MAME does not add any additional frames that aren't present on original hardware (unless the system is improperly emulated, but that's not a wide many).
Basically, listen to IseeThings, he knows what's up.
Ed Oscuro wrote:Of course, there is a hidden assumption here - that the original Battle Garegga board does not have five frames of lag before it registers inputs at the control panel. That doesn't strike me as controversial, especially since (using the frame skip procedure) we can idealize away many nuisance factors.
If I was being retarded, I wouldn't have considered the right conclusion, would I?

You can go fuck yourself, by the way.
User avatar
trap15
Posts: 7835
Joined: Mon Aug 31, 2009 4:13 am
Location: 東京都杉並区
Contact:

Re: They fixed the sound in Raiden Fighters!

Post by trap15 »

If you weren't being retarded, you would have listened to Haze and not come up with any other conclusion.
@trap0xf | daifukkat.su/blog | scores | FIRE LANCER
<S.Yagawa> I like the challenge of "doing the impossible" with older hardware, and pushing it as far as it can go.
User avatar
AntiFritz
Posts: 1630
Joined: Sun Apr 03, 2011 2:34 am
Location: Australia

Re: They fixed the sound in Raiden Fighters!

Post by AntiFritz »

Ed Oscuro wrote:
trap15 wrote:
Ed Oscuro wrote:retarded conclusions
The games originally have that much 'input lag' (sprite buffering). MAME does not add any additional frames that aren't present on original hardware (unless the system is improperly emulated, but that's not a wide many).
Basically, listen to IseeThings, he knows what's up.
Ed Oscuro wrote:Of course, there is a hidden assumption here - that the original Battle Garegga board does not have five frames of lag before it registers inputs at the control panel. That doesn't strike me as controversial, especially since (using the frame skip procedure) we can idealize away many nuisance factors.
If I was being retarded, I wouldn't have considered the right conclusion, would I?

You can go fuck yourself, by the way.
You shouldn't disregard things that people like iseethings have to say. It's not making you look very intelligent.
RegalSin wrote:Rape is very shakey subject. It falls into the catergory of Womens right, Homosexaul rights, and Black rights.
IseeThings
Posts: 534
Joined: Thu Dec 16, 2010 6:38 pm
Location: California

Re: They fixed the sound in Raiden Fighters!

Post by IseeThings »

MAME does the same things at a core level for the games where you have the expected lag, your example being Terra Cresta, as it does for others like Battle Garegga.

In the case of the Raizing games I believe the ShmupMAME guys actually found where, in the *original 68k game code*, they buffer the *inputs* by several frames. There is also some kind of sprite buffer logic for a *single* additional frame as evidenced by it being required to keep the sprites and backgrounds in sync. It all adds up.

Now WHY they would want to do that is another question, but they do. It could be some method to filter out erroneous inputs or just devise evil ways to kill you by knowing your movements in advance of them happening ;-)

The frequency at which inputs are polled on modern systems could be a problem tho, yes. On the original boards you basically have a pin relating to each input and the main cpu could get the state of that pin at any time*. On a modern PC you have all sorts of logic between input and the value given by the hardware to the Windows driver. Again this isn't MAME adding lag tho, it's modern hardware.

* later hardware like STV and many Namco titles are messy tho, driving inputs through a secondary processor which has the potentially to add lag on real hw too because the main cpu is not provided with direct access to the hardware state.
User avatar
Ed Oscuro
Posts: 18654
Joined: Thu Dec 08, 2005 4:13 pm
Location: uoıʇɐɹnƃıɟuoɔ ɯǝʇsʎs

Re: They fixed the sound in Raiden Fighters!

Post by Ed Oscuro »

Well, the first part about the original games' input buffers being the major cause is just the answer I was looking for!
IseeThings wrote:The frequency at which inputs are polled on modern systems could be a problem tho, yes. On the original boards you basically have a pin relating to each input and the main cpu could get the state of that pin at any time*. On a modern PC you have all sorts of logic between input and the value given by the hardware to the Windows driver. Again this isn't MAME adding lag tho, it's modern hardware.

* later hardware like STV and many Namco titles are messy tho, driving inputs through a secondary processor which has the potentially to add lag on real hw too because the main cpu is not provided with direct access to the hardware state.
I am confused why you still feel the need to point this out for me, when I already discussed it above; I'm pretty well aware of the various slowing factors in a PC's operation. I hate USB 3 and its drivers, but I hated USB 2 and its drivers too. This constellation of issues almost certainly cannot be a (major) factor in frameskip operation, where there is ample time for commands to filter through the buffers and into MAME to be passed onto the game. Otherwise, lag frames could be notably shorter during frameskip than otherwise. Theoretically, I suppose one could hide some of the impact of too-slow propagation of commands through system hardware and the OS to MAME during frameskip operation, even if it wouldn't help in normal operation, although I view that as a tangent to our main topic.

Earlier, I did mention Ootake's author discussing using more CPU time to more quickly poll inputs so as to shorten this time, rather than tying it explicitly but arbitrarily to the beginning of a frame; I am guessing what he did would be analogous to going into the original game code and removing input buffers, as you mention. I am completely willing to agree that it is contrary to the purpose of MAME to do that; there's no need to even have a discussion on that point. So it's an interesting wrinkle, but not the major focus of the discussion. Given what I've heard here and elsewhere, I doubt one frame or more could be won back in frameskip (due to the commands having reached the emulated game well before the next frame), though even fractional parts of frames would be welcome. The main point I wanted to clear up was where the discrepancy in frame lag was coming from, and I suppose that has been at least mostly accounted for.

_____________

About the stupid crap up lousing up the thread:

I don't want to spend too long fighting about this, so I'll explain it briefly. Is anybody familiar with the phrase "If you understood what I said, I must have misspoken?" That was from the good old days the Federal Reserve, with Alan Greenspan as Chairman. Clearly, the Fed did not expect that people had a right to understand what was going on, so long as "somebody qualified" knew. Yet that sort of attitude caused bewilderment and great sadness, because the people (many of whom were academics or bankers who actually COULD understand what was going on) felt that it did not respect their ability to comprehend what was going on. Of course, the real story is that the Fed felt (and from time to time people have said this again, even like Paul Krugman) that they could best attain their objective by misleading the public.

Clearly that isn't the objective of a project like MAME, whose stated focus is to document the games. Some of us are not so clever or well-educated (or can spend the time) that we can correctly read *and interpret* the code for ourselves, so the layman-appropriate explanations are genuinely helpful. However this does not mean that we do not understand the logical implications of things (hence my little A + B + C + etc. above).

To put it bluntly, I don't know how the hell anybody can sift through the years' worth of confusion over the many pages of the Shmupmame thread without noticing that we only had partial answers about what was going on. When you have things like people stating that Shmupmame's sound is better than newer MAMEs, it should be clear there is a problem with the level of popular understanding.

At the same time, the standard, pat answers like "blame your computer" (i.e. "the lag is in the system and the OS") were clearly known and also clearly not sufficient to resolve the problem. (I hope I made that clear, anyway.)
trap15 wrote:If you weren't being retarded, you would have listened to Haze and not come up with any other conclusion.
There is a difference between being told "MAME isn't causing the lag, but you don't warrant a better explanation" and being given an explanation. I can't reach a conclusion without being able to explain it. The very fact that we have to go over all sorts of confusion factors repeatedly should have made it clear to you that an explanation would be necessary to reach a conclusion. And, for what it's worth, you were either unwilling or unable to give me the last piece of the puzzle, so like I said, you need to take a hike.
AntiFritz wrote:You shouldn't disregard things that people like iseethings have to say. It's not making you look very intelligent.
On top of what I just wrote (I wasn't disregarding things, I just didn't find them compelling reasons), you might not want to mix up Haze and trap15, there - or assume that just because trap15 states that Haze originally gave an adequate explanation that it actually was adequate.

OK. Are we all good now?
IseeThings
Posts: 534
Joined: Thu Dec 16, 2010 6:38 pm
Location: California

Re: They fixed the sound in Raiden Fighters!

Post by IseeThings »

I'm simply explaining everything for the benefit of everybody without them having to read significant chunks of the thread.
User avatar
Ed Oscuro
Posts: 18654
Joined: Thu Dec 08, 2005 4:13 pm
Location: uoıʇɐɹnƃıɟuoɔ ɯǝʇsʎs

Re: They fixed the sound in Raiden Fighters!

Post by Ed Oscuro »

Fair enough, it's good to have it all in one place finally!
xMetalliCx
Posts: 71
Joined: Fri Jun 17, 2011 4:08 pm
Location: UA

Re: They fixed the sound in Raiden Fighters!

Post by xMetalliCx »

about all this "lag" stuff - it's common feature of the principles of emulators, and in most cases not related to input inself, but to graphics.

lets take some simple framebufer-based system like ZX-Spectrum for example:
how original hardware works - at the start of every frame videologic starts to fetch data from memory, output it to videoDAC and doing this until frame ends.

but in emulator we can't just take from RAM all videodata and show it on PC screen, because we want have raster (multicolor) effects, famous border lines during game load and so on, right ?
so we must take framebuffer byte by byte (and border color) and copy to buffer, with the same speed as hardware does (parallel with CPU and other stuff emulation), and show that buffer to user at the start of NEXT frame.
so, as you can see under emulation users see the picture 1 frame later than users of original hardware.

and as was said - all this is common to almost all emulators, they just works this way.
IseeThings
Posts: 534
Joined: Thu Dec 16, 2010 6:38 pm
Location: California

Re: They fixed the sound in Raiden Fighters!

Post by IseeThings »

Yeah, that's basically what I was referring to as unavoidable lag.

It's not a 'MAME' problem, it's just inherent to the concept of emulation, because you can't render to a modern screen directly in the same way, in perfect sync with the timings and quirks of the hardware you're emulating. If you run at double refresh rate (which most current screens probably won't allow anyway) you reduce that to half a frame because that present frame arrives faster, but it's still always going to be an issue.
xMetalliCx
Posts: 71
Joined: Fri Jun 17, 2011 4:08 pm
Location: UA

Re: They fixed the sound in Raiden Fighters!

Post by xMetalliCx »

I can't say its totally unavoidable.
In the case of simple systems we can throttle emulation at max speed until we got screen for that frame (from one vblank to another in ticks of emulated system), immediatelly show it and then sleep for rest of frame time.
so if emulation didnt take much time we can show current frame picture almost fast as real hardware.
IseeThings
Posts: 534
Joined: Thu Dec 16, 2010 6:38 pm
Location: California

Re: They fixed the sound in Raiden Fighters!

Post by IseeThings »

yeah, but most simple systems aren't THAT simple if you drill down to pixel / sub-instruction accuracy level anyway, and you'd only have a tiny tiny window to run the emulation in, so for all practical purposes it's unavoidable.
User avatar
casualcoder
Posts: 347
Joined: Sat Apr 21, 2012 4:35 am
Location: West Coast, Canada

Re: They fixed the sound in Raiden Fighters!

Post by casualcoder »

'fixed' the sound huh? So you mean someone completely rewrote the soundtrack? :lol:

Well, seriously, I know what you are referring to but as a music producer it always struck me just how bad the sequencing is for the Fighters series. It's like whoever composed it had done so with off-time samples and then decided at the end to just double the tempo and call it 'drum and bass".

With that said, that part never bothered me as much as the slow progression, especially in the later levels. Is it just me or does it take forever to finish the game, 1cc or otherwise?
User avatar
trap15
Posts: 7835
Joined: Mon Aug 31, 2009 4:13 am
Location: 東京都杉並区
Contact:

Re: They fixed the sound in Raiden Fighters!

Post by trap15 »

It never seemed off-beat or anything to me. They're very complex though, lots of weird rhythming. But it works well, and I for one love the RF soundtracks :D
@trap0xf | daifukkat.su/blog | scores | FIRE LANCER
<S.Yagawa> I like the challenge of "doing the impossible" with older hardware, and pushing it as far as it can go.
Post Reply