Regarding using game music On / Off
10 months ago

Hi there, I've been running this game and documenting a lot of stuff regarding the speedrun.

It seems that using the "music Off" option (this mutes the music only inside of missions) saves a considerable amount of time. I calculated the timesave to be around 30 frames (0.5 seconds) per commander turn (that's when the music loads). This results in about a 3:55 to 4 minutes difference between having or not having music on.

I don't really know what we should do about this, I mostly want to open the discussion and get some opinions in the matter. I personally love playing with music and I find really boring to have silence for most of the run, its also not as easy to put random music on top of the run cuz the result screen and overworld music still plays with "music Off", which for me gives a lot of anxiety I get more than 1 song playing at the same time. Should we split categories? accept the silent meta? It also complicates things the fact that the 1st place already has music Off.


This is a very unfortunate discovery that will definitely need further discussion/decision making, but I personally think it will be allowed as it is just personal preference and it'd be weird to force people to play with music on that prefer to listen to their own music and have in game music off. However, even though this would save considerable time, considering the currently limited amount of competition, the time save will likely not have a major effect on the boards. I think most people will just keep it on by default and further discussion needs to happen regarding the rules here.


I just got a 1:24:24 with music on, this time is considerably close to the wr if you discount the 4 minutes of turning the music off. While this game isn't "competitive" in the sense of we don't have a lot of runners, and also most runs are from many years ago, it still has a healthy amount of development (moreso than a lot of other games when I started them), I ran this game just enough to know that both Krazydan and Werster runs are solid in terms of execution and luck grinding, wr even has 1st try on A LOT of luck rolls, which is something you don't just get by doing a couple attempts. I run games that are on the 2 hours+ mark where even just a 1 minute timesave will completely break a category, these are also games where I'm literally the only active runner. 4 minutes is A LOT on this run, is more than retrying all the save and load levels several times, is more than a lot of entire missions.

I'll also emphasize that the 'music off' options doesn't actually put the whole game silent, you'll still hear the overworld map music, the victory and results music, as well as the countries themes and credits music. This makes is particularly hard to put you own preferred music on top of gameplay, or even just listen to the soundtrack of the game on loop.

Sizzyl likes this

I mean you aren't on the victory/result/overworld screens that long so it's not as much of an issue, but it is really concerning how much time music off saves. I'm not an authority figure for this game, nor have I even spedrun it (I just like the base game a lot), so I don't really have any input other than I agree with you that a decision regarding this option should be reached.


Update on Music On/Off research.

I've been doing more research comparing different runs and testing with frame counter on emulator. It seems that music off not only saves time at the beginning of turns (when the music loads) but also saves some loading time when the enemy moves their units. This is barely noticiable on short missions with small maps and few units, but it becomes massive on bigger maps with a lot of enemy units. The big differences start being important around Foul Play, where enemy turns save between 1 and 2 seconds without music, this of course adds up quickly considering the amount of big missions and enemy turns.

With all this the previous 4 minute estimation should go up to 5 or even 6 minutes. I have two takes on what to do about this depending on whether we want to split categories or not.

  1. Split Any% into Music On and Music Off: This is the easiest solution in general, but is always a shame to make splits in a game with a small community, is also sort of detrimental for future competition, since is difficult to compare runs On/Off that are basically doing the same strats (nothing changes apart from loading times, there's no difference in tech or skill on both categories).

  2. Add music loading time to Krazyd4n's run (1:19:58) and my last pb (1:19:56), then ban No Music on the leaderboard for future runs. This solution makes it so all competition on a small community remains healthier, and still keeps the spirit of respecting the solid run that has been record for so long before any discussion was made. Of course we would need to agree on what time we should add, which is definitely complicated.


this discovery is extremely unfortunate lol, I genuinely don't know what the course of action should be and the mods generally only talk on discord, would be nice to get some insight.


This is a really cool discovery, I'm surprised I didn't notice this sooner. My initial reaction is to say that there shouldn't be a category/variant split for this because as you say, this doesn't change anything about how the game actually plays in terms of strats/tech used or route followed. It's the exact same run, just using a toggle that happens to save a substantial amount of time across a full run. The music does add a lot of character to this game so I get why runners would want to still be able to play with it, but my own personal opinion is that in speedrunning we're already ostensibly going for an objective that is seen by a lot of people as "Not fun". Which isn't to say that speedrunning itself isn't fun, but we're running games to see how fast they can be completed. If anyone feels particularly strongly about this I'd like to hear, my own position on the matter is that this is just a meta advancement and we don't need a new category for runs that just weren't aware that this was faster.


It might be seen by a lot of people as "not fun", but for a lot of runners speedrunning a game/category usually means a lot of fun, most top runners really enjoy the categories they grind.

I particularly enjoy playing this game with music, and I'll keep grinding 'music any%' cuz I can't enjoy 'no music'. I only got the record as an statement to show how much time 'no music' saves and to make the discussion more visible, but I think playing without music was such a dull grind, I can't imagine expending a competitive amount of time in such a boring endeavor.


As the consensus among people who have logged runs of the game appears to be that Music Off being the standard would discourage them from continuing to play, I think we can move forward with tracking which runs are playing with Music On/Off, and decide what to do regarding a timing offset. I'm unsure about how exactly this should be handled as it's not a simple case of "Music Off saves X amount of time across a run", as everyone's runs are likely to take a different # of turns to complete. If we can get some objective data regarding how many frames per turn are saved by having music disabled and whether these frame values ever change (Frames saved on turn start + frames saved on unit moved), that would be a good place to begin. Assuming this frame data doesn't change, it wouldn't be very difficult to track in a spreadsheet, just time-consuming during verification, if we decided to go that route.

For now I've added Music On/Off as a variable with each run on the board tracked appropriately. If the above suggestions turn out to be impractical, worst case scenario is that Music Off can be banned for future submissions, it's not an elegant solution but I don't see a good alternative.


Loading time with music on at the beginning of each turn is 30 frames average, but varies randomly between 24 and 40 frames. Then long turns time difference due to enemy unit movement seems extremely inconsistent, on Foul play for instance turns can randomly go from 50 frames up to 110+ frames. This makes so it's unreliable to fairly compare runs with and without music.

This is why I suggested banning no music after adding a certain amount of time to the runs that already have no music, just so we have them on the same leaderboard for historical reasons, the time added can range from 4 to 5 minutes I think. I could calculate the exact difference but that would take some time, cuz it means analysing every single turn in the run.


I made a short video by TASing Mission 1 with both Music Off and Music On to show the frame data involved

Music Off: Andy Turn 1 End: 1817 Flak Turn 1 End: 1979 Andy Turn 2 End: 2142 Flak Turn 2 End: 2283 Andy Turn 3 End: 2494 Flak Turn 3 End: 2638 Mission Complete: 2787

Music On: Andy Turn 1 End: 1817 Flak Turn 1 End: 2013 Andy Turn 2 End: 2208 Flak Turn 2 End: 2383 Andy Turn 3 End: 2626 Flak Turn 3 End: 2803 Mission Complete: 2984

The Enemy Unit timesave is obviously negligible on Mission 1, but the Turn Start timesave is pretty clear. With Music On, Flak's turns took 34/34/33 additional frames respectively, that 1 fewer frame on Turn 3 is supposedly accounted for by him only having 1 unit on that turn. I'd love to see if this Turn Start timesave ever deviates from this value, 30~ does seem like an accurate ballpark figure.

Edited by the author 10 months ago

I got my values comparing Krazyd4n's run with one of my music pb's. There's more details to get the estimate I got, the very first turn in the mission doesn't add time, as well as the last turn when there's a country theme at the end (Liberation for example plays the orange star theme no matter your music option).

Here are some of the results comparing runs (on the music tab):


I don’t know why I didn’t get notified of these posts…

Music off is nothing new, I’m fairly certain all AW games are faster with music off and all of Frozer’s and Gippy’s WRs in the series play with music on knowing that it’s slower, purely because music off is incredibly boring both to play through and to watch. Also, I’m pretty sure there are hundreds of games where turning off music would save time. I don’t think it’s any more complicated than: CPU has fewer instructions to run, hence game runs faster, hence saves time.

I don’t think music off should be allowed, it’s the most ridiculous time save there could be. You can literally just look at a music on run and say “oh music off would be ≈ 5 minutes faster” and there’s nothing more of interest to say about music off. It doesn’t really change anything as far as I know.

Music off was first discovered in AW Dual Strike by either me or other top players (STL/fidelcastro/idk) in I think around 2009 but the only proof with a quick search I can find is from a 2010 TAS STL and I made:

In Dual Strike Time Survival records music off is actually abused because they’re all like 10 seconds long at most, it’s nothing like grinding 1h+ of silence. lol


Me and Krazyd4n are currently running with music in order to don't have splitted categories, and reading all the opinions I think the most healthy decision for the leaderboard is to ban no music.

Game stats
Latest news
A new look and resources

Hi there, I took the freedom of upgrading the look of the leaderboard a little bit.

Also posted all the resources we use for the speedrun (emulators, splits, etc).

The AW speedrun discord is now linked on the leadeboards.

7 months ago
Recent runs
Latest threads
Posted 7 months ago
13 replies
Posted 8 months ago
2 replies
Posted 2 years ago
6 replies
Posted 2 years ago
4 replies