Yooo, back in muh days, runners would actually do fights smh smh smh

Joke aside, this looks really dope, gg for finding this.

Hey everyone,

As fhnnhf said, Steam disabled the possibility to pull depots from its console. However, this can be re-enabled through a small patcher. Once you patch your Steam client, you can then proceed to downpatch through the console normally.

I updated the downpatching guide accordingly, adding the link to download the patcher and the few extra steps needed to re-enable the console downpatching method.

Tell me if there is anything unclear in the new version of the guide.

getting a pirated version of a game you actually own might not be illegal, but sharing one definitely is. :/ The problem didn't appear earlier because Transistor is a very niche speedrun which doesn't get much activity, i don't really know how to fix it. Retiming your times might be shady, because the isLoading variable that the autosplitter is tracking doesn't exactly match with the things happening on screen.

lol I was too lazy to do the thread myself, thanks for doing it Strife, and gg to fhnnhf for getting in. Saw the information this morning, glad to see Transistor at SGDQ once again.

I guess this only works because you're not getting locked into the platt area after the fight? If just spamming Turn() at the end of any fight works, that would change a lot of things lol.

that sucks a lot, but I think that even though the load remover works differently between patches, it works consistently across computers: Eskairie's run and mine are probably run on the same patch and seem to get the same stop you get (that would need some more rigorous timing that i can't really do tonight). If that ends up being the case, your old PB is the only one with the disadvantage among top 5 (as savepoints' and Onin's runs don't really have loadless times), so I guess that won't be a problem, because the guide i posted makes you download a very specific version of the game.

We only need to agree on which version we're going to run, that could as well be the version on which you did your WR.

I uploaded a guide to roll back to the previous version of the game through steam console: it's more of a workaround than a proper fix, but it's quite easy and this way people interested in the speedrun can time their run accurately.

Nope, the problem is not solved for the moment and after giving a look to the load remover, it seems to be beyond the scope of my ASL knowledge for the moment. I'm not sure it is hardcoded pointers, or I don't understand what is being done to them in the code.

I'm willing to give it a try if people who got involved can help so I don't need to retro-engineer everything written in the current load remover. Otherwise I'll probably ask the Speedrun Tool Development Discord to help me on that one.

The only thing I can do to help you for now, Cadarev, is to lend you my GoG version of the game if you want to run: btw, most of the time, GoG let's you rollback to previous versions from their Galaxy client.

Yep I did, but to be precise, I had problem with Dxtory while using OBS : while OBS was running, depending on the limit I put, the game was at either 30fps or 66fps. :/

But you are right to mention that my PC might be an exception.

Hey, sorry to come back more than a week later, I've been grinding on the 60fps framerate in order to get ready for my Saturday run, and now that I'm finally comfortable with it, I'm able to reply.

First of all, sorry to have broke that to you savepoints. That sure is bad news, but on a brighter side, the load remover whose design has been initiated with your run is still relevant, because as anyone can verify on the leaderboard, even at 60fps, there are still some differences between loading times from one computer to another (I had to do twice the Port 37 skips in my PB so I can't time it precisely, but there are probably several seconds of difference).

Secondly, the reason I waited so long to reply here was to asssess the stability of the solution I chose to lock my framerate. I tried Dxtory which has been mentioned by Cadarev, but surprisingly, this software can't seem to lock the framerate in Transistor whatsoever. However, I found another software, called RivaTuner Statistics (RTS) which successfully capped my FPS at 60fps, even after hours of playing the game. The only "problem" that I had to face was that OBS lowers the framerate while using RTS, which means when I was capping at 60fps, the game was running at 25fps, but everything was fine after I capped at 120fps.

In the end, my game was running at a stable 59-60fps, and The Empty Set cutscene that I used to check the framerate on the leaderboard lasts a 1m10-11s in my PB, so we are in an acceptable window in my opinion. I don't think it's needed to display the framerate though, because we can always time the Empty Set cutscene and I think that changing maliciously the framerate during cutscenes in order to not get caught would be extremely complicated, given that the game stop as soon as it loses focus.

So to conclude, being too constraining to potential new runners of the game is not necessary imo, we probably won't get any massive number of submits, so mods should have plenty of time to verify new ones if they do look fishy.

After what Onin pointed out in his thread and our discussion, I investigated a little bit about the framerate issue and here are some interesting things that I discovered.

So, for those who didn't read it, Onin rejected one of my run at the beginning of the week, explaining that my game was running faster than other's and that it was probably due to my framerate which is supposed to be at 60fps thanks to the Vsync being activated. After checking, he was right and my framerate was higher than 60fps, at 66fps more precisely, and that even though the Vsync was activated. I was then surprised to be the first one to have to face a faulty Vsync feature among the runners.

In fact, I am not the only one who didn't run at 60 fps. I did the experiment on several videos on the leaderboard and I tried to time the cutscene just before Sybil's fight : the starting point was the apparition of the mention "The Last Night" and the end point was the spark in which Red appears. And here's what I found :

• Cadarev, Onin and Eskairie's cutscenes lasted 1m11s.
• savepoints' cutscene lasted 1m07s.
• Mine and Vvenzg lasted 1m04s.

Thanks to a very sophisticated cross multiplication, given that it's highly likely that Cadarev, Onin and Eskairie were running at 60fps, I found that savepoints was probably running with a framerate of 63 fps and that Vvenzg, like me, had a 66fps framerate. So there is a little discrepancy in the way the game handles the Vsync framerate from one computer to another and that is quite a pity.

Funnily enough, savepoints' run was the starting point of the development of the load remover because their run saved time on the loading times: actually, their shorter loading times were probably due to their higher framerate, because as I experienced both, I have significantly longer loading times now that I run in 60fps. ^^

To conclude, I'm fully aware that the game is almost not run anymore, and I don't really ask for anything. I just wanted to sort this out a little bit and exposed what I found. There's nothing much to do anyway, the rules are now clear, and I gave a point of comparison in case the moderator verifying has a doubt about the framerate of a run.

Now don't forget to watch Transistor at Bourg-la-Run, December 9th at 2:30 UTC ? (Onin and Strife, if you're reading this and are still interested in commenting on the english restream, don't forget to send me your info as stated in the other thread)

Oh and getting my game to run at 60fps while recording with OBS was a nightmare, I might write a little tutorial for those facing the same problem as me and wanting to submit a PB.

Nope, 200fps is faster, and significantly. I can't see why the speed would be more accurate at 200 fps than at 66fps. x)

It's as if my game didn't want to be capped at 60fps. I tried using Dxtory, but it will either cap at 30fps (which is awful to play but does get me longer segments) or at 66fps. I'm completely at loss right now. I'm reinstalling the game right now, will try some compatibility mode (I'm on W10) and hope for a miracle

EDIT : Ok, I may have something. The only way that I found is to deactivate completely the Vsync thanks to "nofixedstep" and to externally cap the FPS of the game thanks to DXtory. It seems to do the trick, even though I'm really surprised to be the only one having to go to such lengths in order to have my game running at 60FPS. Anyway, it might be better to clarify the rules as "run at 60 fps capped" as I'm deeply convinced that the "Vsync feature" of the game is making it run at 66fps on my computer.

Ok, forget my previous comment. In short, I was saying that my game was faster and that the -SwapInterval=1 fixed that but it actually didn't.

So my game IS faster than 60FPS: the FPS displaying tool in Steam indicates a framerate of 66FPS. What is funny is that the ratio between 66 and 60 is approximately the ratio between my first segment and Cadarev's one, so it is probably the explanation behind my time.

However, all the things I tried to activate the Vsync didn't work, and my framerate is still at 66FPS. The only thing that I managed to do is having my framerate at 200FPS thanks to the /nofixedstep launching option. But using either -1, 0 or 1 on SwapInterval doesn't seem to change anything. Not sure how to continue from there, any idea ?

EDIT : Apparently the game is locked at 66 FPS in "VSync", so my game is actually not faster ( https://steamcommunity.com/app/237930/discussions/1/540743128801326102/ )? I'm started to get lost in all that...

I looked it up a little bit on Google, and the culprit might be the Nvidia Control Panel which can bypass Steam and deactivate the Vsync on its own. But I won't be sure before getting back on my computer tonight.

And yeah, after rewatching it, there's now way I can do a shorter first segment than Cadarev by more than 10s.

I'll have to check on that this evening, but I played it in online mode (and as a proof, you can see my steam page: https://steamcommunity.com/profiles/76561198012270276 ). I didn't make any modification to the game, except activating the "disable display scaling on high DPI settings" option because the game has problem with full screen on my computer, but this has normally nothing to do with vsync. If there is eventually a problem with vsync, that might have something to do with an update of the game (it may have been made an option in the menu and I would have unchecked it).

This may also be a video bug, because it does feel faster than when I am in game. I'll try to get better/smoother video (it may looks faster because of the 30fps so I'll switch to 60fps), and I will check if the video is not simply running faster by comparing my real time and the length of the video.

Anyway, nothing malicious on my part, and sorry for the trouble. I'll investigate that matter and submit a proper PB.

Onin : well, indeed I asked for one caster, but I think there won't be any problem if you and Strife end up commenting together. From my experience, it's easier to comment with someone, and then the staff will be assured to have at least one caster for the run, because Strife still hasn't confirmed he'll be available yet. If you're definitely ok, DM me the following info via Twitter or twitch (it goes for both you and Strife) :

• Username as it will be displayed on the stream (with spaces, capital letters and stuff)
• twitch channel (if you want it displayed somewhere)
• Discord username
• Concerned games (so in our case, Transistor oc)

Then I'll forward that to the staff and you will be invited to their Discord server. ^^

My bad, I wrote this in the middle of the night and it clearly lacked precision. x)

So yeah, the run is scheduled Saturday December 9th at 3:30am in France (2:30am UTC). The full schedule is also available here : https://horaro.org/bourglarun/2017 .

Strife, that would be an honour given that I learnt the run basics through your tutorial. I didn't improve the route in any way (I think), so you're pretty much up to date, I'm sure you'll be perfect, even though I'm not really the one deciding: it's the restream staff (feasel is in charge, I think) that'll have the last word, I'm just trying to help them by asking around, but I can't see why they would refuse. Just tell me if you're available Saturday morning and I'll give you the instructions.

And thanks a lot Cadarev! I PBed last night with an amazing last segment, finishing 20s behind you. My SOB is now under 37 minutes, I'm definitely coming for your WR, hope you don't mind. :P

Hello everyone,

As some may have noticed (probably not), I'm training the game at the moment (5th in the leaderboard rn, and I have an unsubmitted PB that would rank 3rd but still not satisfied with it) and I submitted the Any% run to a French speedrun charity marathon (Bourg-la-Run) held between the 8th and the 10th of December.

Luckily enough, I was selected and will subsequently run Transistor at about 2:30AM UTC (3:30AM in France) EDIT : Saturday December 9th, but I'm not here to brag: an english restream of that event is supposed to happen on SpeedGaming, and they are looking for casters to comment the runs. That's why I logically post here in order to know if anyone would be interested in assuming that role for my run. Tell me if you're up to it (by DM on twitter or here) and I'll let you know what info the restream staff needs.

I also seize this opportunity to thank everyone who's been part of the community and who's participated to the speedrun one way or another: Transistor is actually the first speedrun I've ever completed, and it's all the more meaningful as this game's always been very important to me. I don't really know if there are a lot strats that could be added to the current route (Cadarev's one) but I'd love to discuss the topic if some of you are interested. Take care !