Fungi Forums
Video Games => Mario Chat => Topic started by: Raccoon Sam on January 22, 2008, 02:44:39 PM
-
The mystery has been solved by our very own Xkeeper. Below is a link to the video and some sweet, sweet copypasta.
Zing (http://uk.youtube.com/watch?v=BOhJk0PAp1A).
This is a tool-assisted recording. That is, it slows down the game and uses save states, mostly to get the timing perfect to demonstrate.
Explanation:
The "glitch" involves the tail wag sound effect and the 1-up sound effect's calls getting merged. The sound code reads both, treats it like a 1-up sound effect, but writes the wrong length, so it plays more notes than it should. We're still doing some research here, but the general consensus among us is that it's just a glitch.
Timing:
Simply wag Mario's tail 5 frames (1/12sec) after the "1UP" symbol appears. Done right, the sound effects will conflict and the longer "version" will play. This is extremely difficult to replicate on purpose, which might explain why it was never caught.
An ongoing discussion of the topic can be found here.
http://jul.rustedlogic.net/thread.php...
Technical explanation:
The tail wag sound effect is written (B0), then the 1-UP (40) is ORed, resulting in F0. It writes the wrong "length" for the 1-up, and as it reads bytes in reverse, the beginning is broken instead of the end.
I never had any idea about this sound, not even while continuously attempting the trick.
Anyone else ever experienced the strange sound?
-
Never done it myself. I was going to try it after he mailed us about it, but then I got wrapped up in SMA4 and forgot.
-
That's a very pleasing sound effect.
-
I recall doing that once or twice during my playthroughs, but never really thought anything of it.
-
In all my years of playing, somehow I never came across the effect. But that's a very cool find and great reverse engineering.
-
But why did they make the sound effect longer than what's used? And how long is the whole thing?
-
They didn't. Read the explanation.
-
I did read the explanation.
-
its a glitch
-
Am I to understand that I'm never going to get a real answer to this question?
-
If you read the explanation, you wouldn't need an answer because the question is incorrect.
-
Do you mean the explanation in the 1st post of this thread? I read it and it doesn't answer the question.
-
That's because your question doesn't make any sense. They didn't program it longer. B0 and 40 get read together and make F0. They are two separate sounds.
-
I already read that. It gets confused about the length and plays twice as much of the 1-up sound sample. The 1-up sample played when the glitch happens is twice as long as the correct one and isn't slowed down or looped. For that to be possible the sample must be at least twice as long as what is supposed to be played. This raises the question of why the programmer/composer wrote a sound effect/musical cue that was longer than it needed to be, and the question of why the whole sample was saved on the hard drive when it was using up space without benefiting the game in any way, and the question of how long the entire sound sample is. I know for a fact that it's possible to decompile the programming of N.E.S. cartridges to view sprites that aren't used in the actual game, so it stands to reason that the same can be done with sound samples.
-
It does not play "twice as much" of the 1UP sound. It plays some of the tail wagging sound slowed down, and then the 1UP sound, one after the other. Also, people write in easter eggs all the time. Like Totaka's song. Not to mention that it's a lot easier to leave random bits of programming in the ROM than clean it before release, since the space is going to be on the chip either way. (That is not the reason for this glitch. However, it is/was frequently done.)
-
Not quite. It combines the lengths of the 1UP and tail wag sound, but starts playing from the 1UP location. Whatever comes before the 1UP data is also played, but that's not necessarily the tail wag data, nor is it "extended" 1UP sound data, just some other random data.
-
That makes sense. You should have said it before.
-
From the first post:
"The 'glitch' involves the tail wag sound effect and the 1-up sound effect's calls getting merged. The sound code reads both, treats it like a 1-up sound effect, but writes the wrong length, so it plays more notes than it should."