Posted: 1/18/2015 11:57 PM
(Edited: 1/19/2015 12:11 AM)
Post subject: Obsoleting a Snes9x 1.43 TAS with a Bizhawk 1.9.1 TAS
Quote
Hetfield90
Skilled player (1219)
Joined: 8/29/2014
Posts: 301
GBA TAS of 2018 TASer of 2016 SNES TASer of 2016 PSX TAS of 2016 TASer of 2015 Rookie of 2015 SNES TASer of 2015
I just finished making a Mega Man X 100% TAS that beats the currently fastest published run by 400 frames and am working on my submission text, but I'm curious to know how much more optimal my TAS is since Mega Man X has a substantial amount of lag which isn't accurately emulated on Snes9x 1.43. So I have a couple of questions regarding this: Is something like counting lag frames and subtracting them from total frame counts something that would be done during the adjudication process, done only if my TAS was slower due to emulator differences and this was necessary to prove I had more optimized gameplay, something I would have to do myself before submitting, or something I can do if I want but isn't necessary for submission? Also, what would be the most efficient way to go about doing this? Is there any alternative to frame advancing all the way through both TASes and identifying and counting each lag frame individually? The lag frame counter doesn't seem to be very reliable in either emulator.
Posted: 1/19/2015 12:04 AM
Quote
Nach
Emulator Coder
Joined: 3/9/2004
Posts: 4588
Location: In his lab studying psychology to find new ways to torture TASers and forumers
Since Snex9x 1.43 is known to under emulate lag, I think we'd accept anything that was very close to the existing TAS in time if it's clear there's additional lag anywhere. Since you went one step further and beat the current published run in actual frames too, there's not much more that needs to be done. I was sick of how inaccurate the existing run is, and would welcome the improvement.
Warning: Opinions expressed by Nach or others in this post do not necessarily reflect the views, opinions, or position of Nach himself on the matter(s) being discussed therein.
Posted: 1/19/2015 12:18 AM
Quote
Hetfield90
Skilled player (1219)
Joined: 8/29/2014
Posts: 301
GBA TAS of 2018 TASer of 2016 SNES TASer of 2016 PSX TAS of 2016 TASer of 2015 Rookie of 2015 SNES TASer of 2015
Ah I see, thank you. I think I will still want to find out how much more optimized my run is for my own curiosity if nothing else though. Is there a still accurate but less tedious way to do this than the method I had in mind?
Posted: 1/19/2015 12:22 AM
Quote
Nach
Emulator Coder
Joined: 3/9/2004
Posts: 4588
Location: In his lab studying psychology to find new ways to torture TASers and forumers
Completely accurate? Probably not. The best way to measure improvements though IMO, is to calculate the level start time (where X can first begin moving) and level end time (where X delivers the final blow to the boss), and subtract one from the other in each level. Then provide a table in your submission which says how many frames the old run used for a level, how many frames the new run uses, and the difference. There may be differences elsewhere, in terms of how long lag is between cutscenes and level loading times and so on, but we ignore such things when looking at a run anyway. We primarily focus on the changes in gameplay.
Warning: Opinions expressed by Nach or others in this post do not necessarily reflect the views, opinions, or position of Nach himself on the matter(s) being discussed therein.