Post subject: Is this normal for frame advance?
Joined: 2/14/2007
Posts: 127
When using frame advance, I've noticed a two-frame delay from the frame I press a button and the frame the game actually responds to it. Is this the normal way to use frame advance? If so, it seems a lot more awkward than having the game register the button press the same frame it is pressed. If this is not normal, do you have any suggestions on how to fix it? None of the FAQs mention anything about this.
JXQ
Experienced player (750)
Joined: 5/6/2005
Posts: 3132
This was a visual inconsistency in older versions of SNES9x. V9 corrects this problem. http://tasvideos.org/EmulatorHomepages.html#snes9x_ Edit: Hmm I think I misunderstood what you were saying. Buttons register in the game two frames later than they are pressed. This is true whether or not you are using frame advance. After a while, you'll probably get used to it. (I'm so helpful)
<Swordless> Go hug a tree, you vegetarian (I bet you really are one)
Post subject: Re: Is this normal for frame advance?
upthorn
He/Him
Emulator Coder, Active player (388)
Joined: 3/24/2006
Posts: 1802
Floogal wrote:
When using frame advance, I've noticed a two-frame delay from the frame I press a button and the frame the game actually responds to it.
This is normal. The game has to wait until it actually reads a button press to respond to it. So when you hold a button and advance a frame, you don't see it respond on that frame, because it didn't read the button being pressed until then. Instead, you see it update the next frame after that. For Sega Genesis, there is a delay of one more frame, because the Sega Genesis prepares one frame in advance. This is even more the case with the Sega 32X attachment, which prepares an additional two frames in advance. You get used to it, but having RAM watches (which do update instantly) on important values (like speed, and position) can help.
How fleeting are all human passions compared with the massive continuity of ducks.
Former player
Joined: 10/1/2006
Posts: 1102
Location: boot_camp
I find that using a frame counter to time the button presses is extremely helpful, and it also saves the trouble of repeatedly redoing an action because you can't remember on which frame you pressed a button.
Borg Collective wrote:
Negotiation is irrelevant. Self-determination is irrelevant. You will be assimilated.
Emulator Coder, Skilled player (1301)
Joined: 12/21/2004
Posts: 2687
It could actually be fixed by making the emulator display 2 frames ahead, which would make TASing slightly less tedious by always giving more immediate feedback, but Gens and VBA are better-equipped for that change than Snes9x.
upthorn
He/Him
Emulator Coder, Active player (388)
Joined: 3/24/2006
Posts: 1802
nitsuja wrote:
It could actually be fixed by making the emulator display 2 frames ahead, which would make TASing slightly less tedious by always giving more immediate feedback, but that's currently harder to do in Snes9x than in Gens or VBA.
Which reminds me, I managed to get rid of the staticy sound issue that Gens had when the S3K camhack used a similar method.
How fleeting are all human passions compared with the massive continuity of ducks.
Joined: 2/14/2007
Posts: 127
Thank you all for your prompt replies. Nuts. This slightly discourages me from ever trying to make my own TAS. Perhaps this should be mentioned in the site's FAQ?
Joined: 6/14/2004
Posts: 646
This may simplify, or it may complicate, but you can think of it this way: [current view]-->[detects input]-->[acts based on input]
I like my "thank you"s in monetary form.
Editor, Skilled player (1941)
Joined: 6/15/2005
Posts: 3247
I never knew there was a two-frame delay despite making several Megaman X series TASes. So it is transparent matter to me.
NrgSpoon wrote:
This may simplify, or it may complicate, but you can think of it this way: [current view]-->[detects input]-->[acts based on input]
That's a nice way of thinking about it, though I never think about it.