Quality Test - Quake III Arena (PC)

2021-01-07 08:55:09 am

Type of quality test: source file
Your operating system: Windows 10
Capture hardware: PC
Capture software: Streamlabs OBS
Capture video resolution: 1920x1080
Name of the game, which console: Quake III Arena, V1.32, PC

Thread title:

Attachments:

q3a_vertex0.mp4

q3a_vertex1.mp4

LotBlind

2021-01-08 08:28:51 am

Nate is the guy confirming A/V quality is good but personally, the option without vertex lighting looks more like I remember the game. These things are very subjective cause everyone had their own settings (and how good the game looks wasn't always the top priority). For an SDA submission I suppose having nicer graphics would be more important. After watching both several times, I think maybe without vertex looks better. I think nate will probably approve of either though.

It always bugged me we didn't have a run for Q3. I think it largely came down to everything thinking that there's always a better player out there and having a low self-esteem (the nature of arena shooters). It seems you're the same guy who played in TDM leagues etc.? I'm really looking forward to whatever runs you come up with for arena shooters that you like. I'm always also asking defraggers why they don't do some speedrunning on the side: seems they could easily take records for a lot of Quake engine shooters if they just wanted to.

Anyway, already very hyped over here if I get to talk about some of these games on the front page Wink

Nate is the guy confirming A/V quality is good but personally, the option without vertex lighting looks more like I remember the game. These things are very subjective cause everyone had their own settings (and how good the game looks wasn't always the top priority). For an SDA submission I suppose having nicer graphics would be more important. After watching both several times, I think maybe without vertex looks better. I think nate will probably approve of either though.

It always bugged me we didn't have a run for Q3. I think it largely came down to everything thinking that there's always a better player out there and having a low self-esteem (the nature of arena shooters). It seems you're the same guy who played in TDM leagues etc.? I'm really looking forward to whatever runs you come up with for arena shooters that you like. I'm always also asking defraggers why they don't do some speedrunning on the side: seems they could easily take records for a lot of Quake engine shooters if they just wanted to.

Anyway, already very hyped over here if I get to talk about some of these games on the front page ;)

fearZZz

2021-01-08 08:57:50 pm

Thanks! Even though I do not hold the IL records atm in I can win (easiest) difficulty, (some due to version of q3 client, which has different bot spawn mechanics in different versions), will try to do my best to submit the fastest runs possible (also, not entirely sure if i should do IL submission and send them separately or club them together in one mp4 file as if segmented run).

And yea, i do believe some defraggers could easily contest the records, but with q3a it all comes down to little limitations of not being able to run any modes on top of original game (like vq3 or osp), and also not entirely being about speed / strafe jumping but rather trying to manipulate and understanding bot spawns / behavior in order to get lucky with fast frags.

Looking at list of games on SDA, Quake 3 sure is missing between Quake II and Quake 4 Cheesy

LotBlind

2021-01-09 09:40:59 am

Yeah, you never know who will actually submit runs ever, so non-WR runs can be accepted as you probably know. Here's the last submission anyone sent, even though Diegowar ended up cancelling it. Also here's the Q3 thread that can be used for further discussion, though there's also probably some Discord out there as well. The post that links to has further links to some old records that you've probably seen already. There was also a bunch of other discussion that could be useful (but I wasn't an admin yet when making my posts there so they don't necessarily represent official SDA views). About changing the sv_fps: it doesn't seem different from allowing changing the framerate in other games that allow that. 200 FPS in Quake 3 in 2021 doesn't seem unreasonable to assume those people who want to be competitive would be able to reach with their computers. The reason such settings are allowed to be used is simply otherwise your run is affected by your CPU speed. This way it at least standardizes it some. The counter-argument would be 125FPS has been a kind of standard for a long time – I wonder if you need a fast computer to get the server framerate to be that high? Is that different from graphics refreshing? The counter to that 125 argument is it doesn't matter what online play uses (you're speedunning a single-player game, not a multiplayer game) and 125FPS doesn't seem like the intended value anyway, so maybe it's a bit arbitrary? Just a compromise between precision and CPU? Wish I understood this a bit better.

Of course the SDA recordings aren't expected to be above 60 FPS but if you're doing ILs, feel free to submit the demos as well and we can host those on the same page. You should send the ILs in separately once you have a complete set for at least one difficulty. You can, however, also provide a second encode where they are all together. The only thing is if someone comes to obsolete some but not all of the runs, then we won't keep updating that mp4.

So the Q3 client... that's just different patch versions of the game? Or how were those distributed? The fastest official patch is the one that should be run.

Yeah, you never know who will actually submit runs ever, so non-WR runs can be accepted as you probably know. [url=https://forum.speeddemosarchive.com/post/quake_iii_arena__august_2_2016.html]Here's[/url] the last submission anyone sent, even though Diegowar ended up cancelling it. Also here's the [url=https://forum.speeddemosarchive.com/post/quake_3_arena__31.html]Q3 thread[/url] that can be used for further discussion, though there's also probably some Discord out there as well. The post that links to has further links to some old records that you've probably seen already. There was also a bunch of other discussion that could be useful (but I wasn't an admin yet when making my posts there so they don't necessarily represent official SDA views). About changing the sv_fps: it doesn't seem different from allowing changing the framerate in other games that allow that. 200 FPS in Quake 3 in 2021 doesn't seem unreasonable to assume those people who want to be competitive would be able to reach with their computers. The reason such settings are allowed to be used is simply otherwise your run is affected by your CPU speed. This way it at least standardizes it some. The counter-argument would be 125FPS has been a kind of standard for a long time – I wonder if you need a fast computer to get the server framerate to be that high? Is that different from graphics refreshing? The counter to that 125 argument is it doesn't matter what online play uses (you're speedunning a single-player game, not a multiplayer game) and 125FPS doesn't seem like the intended value anyway, so maybe it's a bit arbitrary? Just a compromise between precision and CPU? Wish I understood this a bit better.

Of course the SDA recordings aren't expected to be above 60 FPS but if you're doing ILs, feel free to submit the demos as well and we can host those on the same page. You should send the ILs in separately once you have a complete set for at least one difficulty. You can, however, also provide a second encode where they are all together. The only thing is if someone comes to obsolete some but not all of the runs, then we won't keep updating that mp4.

So the Q3 client... that's just different patch versions of the game? Or how were those distributed? The fastest official patch is the one that should be run.

fearZZz

Edit history:

fearZZz: 2021-01-09 10:30:23 pm

fearZZz: 2021-01-09 10:29:33 pm

fearZZz: 2021-01-09 10:29:02 pm

2021-01-09 10:27:37 pm

I see., I guess I'll be submitting the IL then (once i can be sure of quality, and first of I can win difficulty, then in Nightmare). sv_fps and com_maxfps values are constant always point of discussion to me cause how it affects the gameplay (although as other games utilizing changing fps valujs for fastest time - i dont see why q3a cant as well).

video that might help in understanding the fps role in q3

Personally for single player, I've been running on com_maxfps 333 value that allowed for slightly higher in-game jumps to help with platforms on such levels as q3dm13 (in the video) or q3dm14 (jump pad platform below near BFG).

for q3 client, latest version to date is 1.32 i think (thats the version you get when you buy the game from steam or GOG for example). but i know that certain levels, like q3dm0 for example, had these spawns going on:

, while in 1.32 those close spawns were never triggered by the bot, so they must have modified something that affected spawn behavior in latest version. and lower version of game is probably pain to come by, so, being extremely used to 1.32, i probably won't try to chase older version for sake of 1 level.

For timing the in-game time, i've been using autosplitter that starts as soon as you enter arena/level and stops on last frag., but I guess SDA / validators would be doing something else internal / by frame perhaps cause unfortunatelly the in-game timer does not tell precise picture.

I see., I guess I'll be submitting the IL then (once i can be sure of quality, and first of I can win difficulty, then in Nightmare). sv_fps and com_maxfps values are constant always point of discussion to me cause how it affects the gameplay (although as other games utilizing changing fps valujs for fastest time - i dont see why q3a cant as well).

video that might help in understanding the fps role in q3

https://youtu.be/he02vJvKaRs

Personally for single player, I've been running on com_maxfps 333 value that allowed for slightly higher in-game jumps to help with platforms on such levels as q3dm13 (in the video) or q3dm14 (jump pad platform below near BFG). 
 
for q3 client, latest version to date is 1.32 i think (thats the version you get when you buy the game from steam or GOG for example). but i know that certain levels, like q3dm0 for example, had these spawns going on:

https://youtu.be/FOfRIQQuQPM

, while in 1.32 those close spawns were never triggered by the bot, so they must have modified something that affected spawn behavior in latest version. and lower version of game is probably pain to come by, so, being extremely used to 1.32, i probably won't try to chase older version for sake of 1 level.

For timing the in-game time, i've been using autosplitter that starts as soon as you enter arena/level and stops on last frag., but I guess SDA / validators would be doing something else internal / by frame perhaps cause unfortunatelly the in-game timer does not tell precise picture.

LotBlind

2021-01-11 07:19:46 am

By SDA rules, you're not just allowed to change the FPS not only for each level, but in theory even during each level as well (I think... at least last time we talked about segmented runs, I didn't hear any objections to allowing changing it whenever). Also, you can and should bind those commands on separate keys if you make use of them during the running. I think the broad reasoning why it's allowed is if we don't allow for setting precise frame values, it makes the ground uneven depending on the speed of your CPU. However, if a particularly high FPS is advantageous, there have also been community rulings about capping it at something that most players should be able to reach without massive investments. This is the case for The Elder Scrolls: Oblivion at least, which they capped at 60 FPS. The FPS meter can be showing for such games of course. I don't know what would be reasonable for Q3 in 2021... You could also argue that values that don't have any rounding in the code that's run frame-by-frame are more like what was intended? Maybe? The argument Berry makes here is a pretty convincing one for how 200 (and I think other multiples of 40?) would represent how the game was supposed to work. I don't know if you can make the jump onto megahealth on Lost World though. So aside from 333, 200 would also probably make sense as a cap. If we think any value up to 333 isn't difficult to achieve nowadays, I guess we can also just leave it to the runner's discretion.

Based on that video, 125 FPS has something to do with the package send rate, and is thus in some way arbitrarily tied to online play. I wouldn't think of that as making more or less sense than 333.

Yeah, if you use map_restart, as you can see in the discussion in that other thread, such a run will be allowed to be submitted with the "with resets" tag, whether or not any runs are ever submitted that don't use it. It's more to futureproof and to be consistent. And the timing will be precise frame-counting by the designated staff member.

I see GOG at least doesn't have any alternative installs that don't use 1.32. So yeah, not something we would require to find an older client.

By SDA rules, you're not just allowed to change the FPS not only for each level, but in theory even during each level as well (I think... at least last time we talked about segmented runs, I didn't hear any objections to allowing changing it whenever). Also, you can and should bind those commands on separate keys if you make use of them during the running. I think the broad reasoning why it's allowed is if we don't allow for setting precise frame values, it makes the ground uneven depending on the speed of your CPU. However, if a particularly high FPS is advantageous, there have also been community rulings about capping it at something that most players should be able to reach without massive investments. This is the case for The Elder Scrolls: Oblivion at least, which they capped at 60 FPS. The FPS meter can be showing for such games of course. I don't know what would be reasonable for Q3 in 2021... You could also argue that values that don't have any rounding in the code that's run frame-by-frame are more like what was intended? Maybe? The argument Berry makes [url=https://forum.speeddemosarchive.com/post/quake_3_arena__61.html]here[/url] is a pretty convincing one for how 200 (and I think other multiples of 40?) would represent how the game was supposed to work. I don't know if you can make the jump onto megahealth on Lost World though. So aside from 333, 200 would also probably make sense as a cap. If we think any value up to 333 isn't difficult to achieve nowadays, I guess we can also just leave it to the runner's discretion.

Based on that video, 125 FPS has something to do with the package send rate, and is thus in some way arbitrarily tied to online play. I wouldn't think of that as making more or less sense than 333.

Yeah, if you use map_restart, as you can see in the discussion in that other thread, such a run will be allowed to be submitted with the "with resets" tag, whether or not any runs are ever submitted that don't use it. It's more to futureproof and to be consistent. And the timing will be precise frame-counting by the designated staff member.

I see GOG at least doesn't have any alternative installs that don't use 1.32. So yeah, not something we would require to find an older client.

LotBlind

2021-01-15 05:56:47 am

Okay, well I thought nate would come here as he usually does but he thought I "had it covered" so just to reiterate: either one of those videos would be fine but IMO without vertex looks best. Go ahead with recording runs! Did you see anything useful in the thread I linked? It seems easier for the community to decide whether 333 FPS is too high for an reasonably powerful PC. Let's hope no-one complains about it later. 200 FPS does seem to make the game behave as intended, even though that's usually a flimsy argument. There are other cases where first-person shooter runners have used framerates that definitely don't make things happen "as intended".

nate

2021-01-15 09:32:44 am

just chiming in here to say that i think either of those videos looks good to me. i'm not as familiar with this game so i can't say which would look better to people who are.

fearZZz

2021-01-15 05:37:29 pm

Ok; noted. Thanks for the feedfback!

fearZZz

Edit history:

fearZZz: 2021-03-14 01:58:33 am

2021-03-13 09:41:51 pm

hmm took longer than i expected hehe. not sure if size will be a problem. im 90% done and mp4 take 4 GB in total already across 20 segments

fearZZz

2021-03-14 01:57:49 am

Attachment:

rdy2go.png

LotBlind

2021-03-14 06:27:39 am

No problem at all! Imagine some long JRPG run :/

fearZZz

2021-03-24 01:24:37 am

haha; ok! submission form sent ^^