Metroid Speedrunning Marathon 2014 Aftermath with YOUTUBE VIDEOS

kirbymastah

Edit history:

kirbymastah: 2014-04-29 08:17:23 pm

kirbymastah: 2014-04-29 08:16:52 pm

kirbymastah: 2014-04-22 11:37:12 am

kirbymastah: 2014-04-22 10:42:57 am

kirbymastah: 2014-04-18 12:51:31 am

kirbymastah: 2014-04-17 09:26:04 pm

kirbymastah: 2014-04-17 03:38:10 pm

kirbymastah: 2014-04-17 03:28:37 pm

kirbymastah: 2014-04-17 01:24:20 am

kirbymastah: 2014-04-15 06:37:33 pm

kirbymastah: 2014-04-14 11:56:22 pm

kirbymastah: 2014-04-14 11:51:36 pm

kirbymastah: 2014-04-14 09:18:51 pm

kirbymastah: 2014-04-14 08:49:43 pm

kirbymastah: 2014-04-14 08:27:05 pm

kirbymastah: 2014-04-14 08:17:31 pm

kirbymastah: 2014-04-14 08:17:19 pm

kirbymastah: 2014-04-14 08:16:54 pm

2014-04-14 08:15:18 pm

<(^_^)>

Youtube Videos:

Entire Playlist - https://www.youtube.com/playlist?list=PL3pBMjeS6rYhMwNICQGsj81JUAGx7jKgz

1. Metroid Prime (Any%) -
2. Metroid Prime 2: Echoes (Any%) -
3. Metroid: Other M (Any%) -
4. Metroid (100%) -
5. Metroid: Zero Mission (100% Normal) -
6. Metroid: Zero Mission (Any% Normal) -
7. Metroid: Zero Mission (Low% Hard) -
8. Metroid II Return of Samus (100%) -
9. Another Metroid 2 Remake (Demo 1.3) (100%) -
10. Super Metroid (14% Ice) -
11. Super Metroid (Any%) -
12. Metroid Fusion (1%) -
13. Metroid Fusion (100%) -
14. Metroid Prime: Pinball (Any%) -
15. Metroid Prime (100%) -
16. Metroid Prime: Hunters (100%) -
17. Metroid Prime 2: Echoes (100%) -
18. Metroid Prime 3: Corruption (100%) -
19. Super Metroid (100%) -

Thanks everyone for supporting a very successful marathon! It was really fun hosting it (albeit tiring), all the runners did a great job, and I want to thank everyone for helping make it successful, whether it be helping out in organizing, running a game, commentating, or helping with the layout, and so on.

That being said, while the runs were awesome, there were many things that were unfortunate, so let's go over that.

The Bad

-Internet problems. I've never had so many issues with my own internet, so when I was doing the restreaming for the first half of the marathon, it went down at certain points (especially during the mp2 any% and mzm 100% runs). This really sucked a lot, but thank goodness yashi was willing to spend practically 20 hours of his time in a row to do the rest of the restreaming for us, with a more stable connection.
-Restreaming Set-up. This was most notable in kottpower's run, as the run was unbearably choppy. The setup I used was the same rtmp restreaming set-up that the Mumble Crew Marathon used; the runner streamed to the rtmp, then I/yashi used xsplit to directly take the rtmp stream. Yashi's internet was fine, and kottpower's rtmp stream was fine, so the only explanation was how xsplit was taking the run; it was likely because we streamed at 25ish fps, whereas kottpower's stream was at 60 fps, which likely caused problems. I'll likely have paid xsplit next time we do this, so the fps issue won't be a problem.
-We were generally over 30-60 minutes past our time estimate after the first few runs. This is mostly because of both my unfamiliarity with setting up, and also technical issues. This hopefully won't be a problem next time, since now I've had some experience with organizing an online marathon. Again, with a different restreaming setup, this may be better, or maybe worse. We'll have to see.
-The chat wasn't filled with complete spam, considering it was 2-3k viewers usually. The chat was filled with other m talk literally every 30 f**king seconds. There were times where there were links and such too, so that'll need regulation.
-Essentially I think I tried to do too much on my own, and things got very hectic. I didn't really have much time myself, so next time, I'll have to gather 2-3 other primary organizers to handle things, and be better organized.

The Good
-The runs overall were great. I'm glad I picked the runners that I picked, as they were all top quality. And though some bad things did happen, occasional disasters are bound to happen, and basically every runner (at least from what I've watched)
-We had the AM2R developer on stream, which was amazing! Heck, I think AM2R was the most asked-for VOD overall, and probably had the most positive reception overall
-Again, shoutouts to yashi to being willing to spend several hours in restreaming essentially the second half of the marathon. This took a huge load off my shoulders, and his internet had way fewer issues than me.
-A few people weren't too pleased with mumble, but I was ok with it and I think it worked pretty well for me, besides being required to add a token to chat in grukk's channel. It'd be great if we can make our own custom server or something with more control over it.

Scheduling

The idea behind the schedule was to open with metroid prime 1+2 any%, which were very strong, then the 2D metroid block in chronological order, than all prime games 100%. I squeezed in other m after mp1+2 any% because I didn't really have anything else to put during that shift, and dragondarch wasn't free at the time. I also put super metroid 100% at the end as a finale with kottpower, which theoretically should've worked, but with the framerate issues didn't work well, so that had nothing to do with the actual scheduling.

The main iffy thing was the metroid zero mission block, which had three runs side by side. The good thing was, we kind of had a cool pattern going, from 100% to any% to low%; this was also very convenient for setup since most commentators are the same anyways. The bad is, viewers were kind of confused about why there were three runs in a row, and although the runs are relatively short, it can get rather exhaustive. What do you guys think?

Next Metroid Marathon...

-Dont try to do too much by myself again. Have side organizers with more specialized roles (one for restreaming, one for audio, and a few side-organizers to set up next runs so I can get rest)
-Maybe find a different way to restream that lets commentators watch the run live with less delay
-If using rtmp again, test this THOROUGHLY for EVERY RUNNER.
-Slightly higher estimates; account for setup time in the schedule itself, rather than the run
-Create own mumble server with (hopefully) all issues miles brought up fixed
-Separate layout for GBA games
-Be more strict on regulating commentary, and do audio tests and set some rules for them
-Have rules in chat (i.e. no links, no ascii)
-Do a role call at the start of every run
-Unofficial games to consider: AM2R full version, Super Metroid Impossible

As for new games and categories, the only viable ones not done that I can think of are zero mission any% hard, and super metroid RBO. Hopefully the full AM2R game will be out by then too. People have asked about doing super metroid hacks; I'd be willing to add 1-2 of them, I think lioran said he would've offered to do SM Impossible if he didn't know about the marathon so late. I don't really want more than 2 romhacks, since it kinda gets iffy with non-officla games overall :S

Highlights and Bloopers

I have a list of highlights and bloopers in this spreadsheet - https://goo.gl/i0bnL9

Feel free to suggest cool and/or hilarious things that happened! I'll put together a reel sometime and update the pastebin appropriately.

Also feel free to post any improvements and/or what has gone well! Constructive criticism (besides what's been mentioned already) will be appreciated! Both from the viewers' and runners' perspective!

Thread title:

Youtube Videos:

Entire Playlist - https://www.youtube.com/playlist?list=PL3pBMjeS6rYhMwNICQGsj81JUAGx7jKgz

1. Metroid Prime (Any%) - https://www.youtube.com/watch?v=GfyZyQFmbwc
2. Metroid Prime 2: Echoes (Any%) - https://www.youtube.com/watch?v=cgRkvjZ9Sjo
3. Metroid: Other M (Any%) - https://www.youtube.com/watch?v=oOD7VR8lb_I
4. Metroid (100%) - https://www.youtube.com/watch?v=ERrXTo_fEWE
5. Metroid: Zero Mission (100% Normal) - https://www.youtube.com/watch?v=oJ9qBzW_5-I
6. Metroid: Zero Mission (Any% Normal) - https://www.youtube.com/watch?v=5TTAZjpe2qM
7. Metroid: Zero Mission (Low% Hard) - https://www.youtube.com/watch?v=_MqYHHqxVNA
8. Metroid II Return of Samus (100%) - https://www.youtube.com/watch?v=amMc5k1IqaY
9. Another Metroid 2 Remake (Demo 1.3) (100%) - https://www.youtube.com/watch?v=pvIXW-0vs3k
10. Super Metroid (14% Ice) - https://www.youtube.com/watch?v=YRfKM3BGxwc
11. Super Metroid (Any%) - https://www.youtube.com/watch?v=4SObHvx5juw
12. Metroid Fusion (1%) - https://www.youtube.com/watch?v=HDHjjaQ4C8o
13. Metroid Fusion (100%) - https://www.youtube.com/watch?v=wtI8T2H89Ls
14. Metroid Prime: Pinball (Any%) - https://www.youtube.com/watch?v=SyXupKbKxq8
15. Metroid Prime (100%) - https://www.youtube.com/watch?v=yAwOmPWY4us
16. Metroid Prime: Hunters (100%) - https://www.youtube.com/watch?v=BGhnQ2pbRvI
17. Metroid Prime 2: Echoes (100%) - https://www.youtube.com/watch?v=Emzo1y4HnN8
18. Metroid Prime 3: Corruption (100%) - https://www.youtube.com/watch?v=BGje79meObE
19. Super Metroid (100%) - https://www.youtube.com/watch?v=lAQQqQQhiVQ

Thanks everyone for supporting a very successful marathon! It was really fun hosting it (albeit tiring), all the runners did a great job, and I want to thank everyone for helping make it successful, whether it be helping out in organizing, running a game, commentating, or helping with the layout, and so on.

That being said, while the runs were awesome, there were many things that were unfortunate, so let's go over that.

[b]The Bad[/b]

-Internet problems. I've never had so many issues with my own internet, so when I was doing the restreaming for the first half of the marathon, it went down at certain points (especially during the mp2 any% and mzm 100% runs). This really sucked a lot, but thank goodness yashi was willing to spend practically 20 hours of his time in a row to do the rest of the restreaming for us, with a more stable connection.
-Restreaming Set-up. This was most notable in kottpower's run, as the run was unbearably choppy. The setup I used was the same rtmp restreaming set-up that the Mumble Crew Marathon used; the runner streamed to the rtmp, then I/yashi used xsplit to directly take the rtmp stream. Yashi's internet was fine, and kottpower's rtmp stream was fine, so the only explanation was how xsplit was taking the run; it was likely because we streamed at 25ish fps, whereas kottpower's stream was at 60 fps, which likely caused problems. I'll likely have paid xsplit next time we do this, so the fps issue won't be a problem.
-We were generally over 30-60 minutes past our time estimate after the first few runs. This is mostly because of both my unfamiliarity with setting up, and also technical issues. This hopefully won't be a problem next time, since now I've had some experience with organizing an online marathon. Again, with a different restreaming setup, this may be better, or maybe worse. We'll have to see.
-The chat wasn't filled with complete spam, considering it was 2-3k viewers usually. The chat was filled with other m talk literally every 30 f**king seconds. There were times where there were links and such too, so that'll need regulation. 
-Essentially I think I tried to do too much on my own, and things got very hectic. I didn't really have much time myself, so next time, I'll have to gather 2-3 other primary organizers to handle things, and be better organized.

[b]The Good[/b]
-The runs overall were great. I'm glad I picked the runners that I picked, as they were all top quality. And though some bad things did happen, occasional disasters are bound to happen, and basically every runner (at least from what I've watched) 
-We had the AM2R developer on stream, which was amazing! Heck, I think AM2R was the most asked-for VOD overall, and probably had the most positive reception overall
-Again, shoutouts to yashi to being willing to spend several hours in restreaming essentially the second half of the marathon. This took a huge load off my shoulders, and his internet had way fewer issues than me.
-A few people weren't too pleased with mumble, but I was ok with it and I think it worked pretty well for me, besides being required to add a token to chat in grukk's channel. It'd be great if we can make our own custom server or something with more control over it.

[b]Scheduling[/b]

The idea behind the schedule was to open with metroid prime 1+2 any%, which were very strong, then the 2D metroid block in chronological order, than all prime games 100%. I squeezed in other m after mp1+2 any% because I didn't really have anything else to put during that shift, and dragondarch wasn't free at the time. I also put super metroid 100% at the end as a finale with kottpower, which theoretically should've worked, but with the framerate issues didn't work well, so that had nothing to do with the actual scheduling.

The main iffy thing was the metroid zero mission block, which had three runs side by side. The good thing was, we kind of had a cool pattern going, from 100% to any% to low%; this was also very convenient for setup since most commentators are the same anyways. The bad is, viewers were kind of confused about why there were three runs in a row, and although the runs are relatively short, it can get rather exhaustive. What do you guys think?

[b]Next Metroid Marathon...[/b]

-Dont try to do too much by myself again. Have side organizers with more specialized roles (one for restreaming, one for audio, and a few side-organizers to set up next runs so I can get rest)
-Maybe find a different way to restream that lets commentators watch the run live with less delay
-If using rtmp again, test this THOROUGHLY for EVERY RUNNER.
-Slightly higher estimates; account for setup time in the schedule itself, rather than the run
-Create own mumble server with (hopefully) all issues miles brought up fixed
-Separate layout for GBA games
-Be more strict on regulating commentary, and do audio tests and set some rules for them
-Have rules in chat (i.e. no links, no ascii)
-Do a role call at the start of every run
-Unofficial games to consider: AM2R full version, Super Metroid Impossible

As for new games and categories, the only viable ones not done that I can think of are zero mission any% hard, and super metroid RBO. Hopefully the full AM2R game will be out by then too. People have asked about doing super metroid hacks; I'd be willing to add 1-2 of them, I think lioran said he would've offered to do SM Impossible if he didn't know about the marathon so late. I don't really want more than 2 romhacks, since it kinda gets iffy with non-officla games overall :S

[b]Highlights and Bloopers[/b]

I have a list of highlights and bloopers in this spreadsheet - https://goo.gl/i0bnL9

Feel free to suggest cool and/or hilarious things that happened! I'll put together a reel sometime and update the pastebin appropriately.

Also feel free to post any improvements and/or what has gone well! Constructive criticism (besides what's been mentioned already) will be appreciated! Both from the viewers' and runners' perspective!

kirbymastah

Edit history:

kirbymastah: 2014-04-17 01:24:05 am

kirbymastah: 2014-04-17 01:19:05 am

kirbymastah: 2014-04-17 12:22:13 am

kirbymastah: 2014-04-16 10:23:57 am

kirbymastah: 2014-04-15 03:36:54 pm

kirbymastah: 2014-04-15 03:36:00 pm

kirbymastah: 2014-04-14 08:38:10 pm

kirbymastah: 2014-04-14 08:18:24 pm

2014-04-14 08:15:30 pm

<(^_^)>

Viewer Data

Excel Document with numbers of viewer counts form the graphs and the corresponding games - http://goo.gl/A8N0z2

Key viewer peaks occurred doing MP2 Any%, MZM Low% Hard, the Metroid Fusion block, MP1 100%, and MP2 100%. The viewer count was overall very high, consistently over 3000 viewers during the fusion block and metroid prime 1+2 100% runs. I'm not going to say it's due to popularity of the games (though it very likely is) because there are several potential factors, such as the quality of the stream, the time of day, how far in the marathon we are, and so on.

Twitch VODs

Metroid Prime Any% by MilesSMB
http://www.twitch.tv/ludendi/c/4080968

Metroid Prime 2: Echoes Any% by MilesSMB
Part 1 - http://www.twitch.tv/ludendi/c/4080981
Part 2 - http://www.twitch.tv/ludendi/c/4080993
Part 3 - http://www.twitch.tv/ludendi/c/4081000
Part 4 - http://www.twitch.tv/ludendi/c/4081005

Metroid: Other M Any% by JaggerG
Part 1 - http://www.twitch.tv/ludendi/c/4081022
Part 2 - http://www.twitch.tv/ludendi/c/4081027
Part 3 - http://www.twitch.tv/ludendi/c/4081033
Part 4 - http://www.twitch.tv/ludendi/c/4081039

Metroid 100% by Dragondarch
Part 1 - http://www.twitch.tv/ludendi/c/4081051
Part 2 - http://www.twitch.tv/ludendi/c/4081063

Metroid: Zero Mission 100% Normal by Mobiusman
Part 1 - http://www.twitch.tv/ludendi/c/4081104
Part 2 - http://www.twitch.tv/ludendi/c/4081107
Part 3 - http://www.twitch.tv/ludendi/c/4081118

Metroid: Zero Mission Any% Normal by Dragonfangs
Part 1 - http://www.twitch.tv/ludendi/c/4081158
Part 2 - http://www.twitch.tv/ludendi/c/4081163
Part 3 - http://www.twitch.tv/ludendi/c/4081164

Metroid: Zero Mission Low% Hard by Dragondarch
http://www.twitch.tv/ludendi/c/4081175

Metroid II Return of Samus 100% by EmptyEye
http://www.twitch.tv/ludendi/c/4081184

Another Metroid 2 Remake Demo 1.3 100% (with Developer Commentary!) by Dragondarch
http://www.twitch.tv/ludendi/c/4081197

Super Metroid 14% Ice by Zoast
http://www.twitch.tv/ludendi/c/4081235

Super Metroid Any% by Zoast
http://www.twitch.tv/ludendi/c/4081239

Metroid Fusion 1% by JaggerG
http://www.twitch.tv/ludendi/c/4081250

Metroid Fusion 100% by Kirbymastah
http://www.twitch.tv/ludendi/c/4081267

Metroid Prime Pinball Any% by UchihaSDA
http://www.twitch.tv/ludendi/c/4081281

Metroid Prime 100% by MilesSMB
http://www.twitch.tv/ludendi/c/4081311

Metroid Prime: Hunters 100% by Amhite
http://www.twitch.tv/ludendi/c/4081338

Metroid Prime 2: Echoes 100% by Kirbymastah
Part 1 - http://www.twitch.tv/ludendi/c/4081349
Part 2 - http://www.twitch.tv/ludendi/c/4081368

Metroid Prime 3: Corruption 100% by Kirbymastah
Part 1 - http://www.twitch.tv/ludendi/c/4081377
Part 2 - http://www.twitch.tv/ludendi/c/4081398

Super Metroid 100% by Kottpower
http://www.twitch.tv/ludendi/c/4081418

Credits

We would like to give thanks to the following:

-The staff at twitch for allowing us to be on the front page for the duration of the marathon, and for hosting their website so we can stream this content to you!
-The overall speedrunning community (speeddemosarchive and speedrunslive) for giving their full support in this marathon

Organizers:
-Svenne for initiating and starting with the idea of the metroid marathon. He has been in contact with twitch for getting this on the front page, initiating about the mumble, and allowing this marathon to be streamed on the team ludendi channel.
-Kirbymastah for organizing and putting together the schedule and runners, streaming the marathon, and leading the marathon organization.
-Grukk for helping organize the mumble and providing his server to allow us to use it for this marathon.
-Dickoak and Yashi for allowing us to use their rtmp server for the streaming set-up.
-In particular, Yashi for willing to stay up for the majority of the marathon to set up a restream and restream the second half of the marathon due to my internet issues.

Design:
-Charleon for putting together the streaming layout you see right now, along with helping edit the individual game logos to fit the layout.
-Infinity's End for creating the dark samus sprite and allowing us to use it in the layout.
-Dragondarch for providing several raw game logos.
-Guynzaro for helping edit the individual game logos.
-Heero_fred for putting together a thumbnail for this channel to go on twitch's front page carousel.
-Kari_Fry for putting together a logo for the metroid marathon itself.
-The developers of LiveSplit for their public timer.
-In particular, CryZe for helping with customizing the timer for the marathon.

We would like to thank the developers for releasing these awesome games and Nintendo for this awesome series! We would also like to thank the programmer and the testers for AM2R for allowing dragondarch to run their demo 1.3 at this marathon! In particular, we would like to thank DoctorM64 for joining us on commentary for AM2R while dragondarch ran his demo!

Finally, we would like to thank all of the runners and commentators for putting their time and offering to run and commentate their respective games for this marathon! Without all of you guys, this marathon would not have been possible! And we would like to thank you, the viewers, for watching this marathon and spreading the word about it!

[b]Viewer Data[/b]

[URL=http://s1308.photobucket.com/user/kirbymastah/media/viewers_zps3edd582a.png.html][IMG]http://i1308.photobucket.com/albums/s619/kirbymastah/viewers_zps3edd582a.png[/IMG][/URL]

Excel Document with numbers of viewer counts form the graphs and the corresponding games - http://goo.gl/A8N0z2

Key viewer peaks occurred doing MP2 Any%, MZM Low% Hard, the Metroid Fusion block, MP1 100%, and MP2 100%. The viewer count was overall very high, consistently over 3000 viewers during the fusion block and metroid prime 1+2 100% runs. I'm not going to say it's due to popularity of the games (though it very likely is) because there are several potential factors, such as the quality of the stream, the time of day, how far in the marathon we are, and so on.

[b]Twitch VODs[/b]

Metroid Prime Any% by MilesSMB
http://www.twitch.tv/ludendi/c/4080968

Metroid Prime 2: Echoes Any% by MilesSMB
Part 1 - http://www.twitch.tv/ludendi/c/4080981
Part 2 - http://www.twitch.tv/ludendi/c/4080993
Part 3 - http://www.twitch.tv/ludendi/c/4081000
Part 4 - http://www.twitch.tv/ludendi/c/4081005

Metroid: Other M Any% by JaggerG
Part 1 - http://www.twitch.tv/ludendi/c/4081022
Part 2 - http://www.twitch.tv/ludendi/c/4081027
Part 3 - http://www.twitch.tv/ludendi/c/4081033
Part 4 - http://www.twitch.tv/ludendi/c/4081039

Metroid 100% by Dragondarch
Part 1 - http://www.twitch.tv/ludendi/c/4081051
Part 2 - http://www.twitch.tv/ludendi/c/4081063

Metroid: Zero Mission 100% Normal by Mobiusman
Part 1 - http://www.twitch.tv/ludendi/c/4081104
Part 2 - http://www.twitch.tv/ludendi/c/4081107
Part 3 - http://www.twitch.tv/ludendi/c/4081118

Metroid: Zero Mission Any% Normal by Dragonfangs
Part 1 - http://www.twitch.tv/ludendi/c/4081158
Part 2 - http://www.twitch.tv/ludendi/c/4081163
Part 3 - http://www.twitch.tv/ludendi/c/4081164

Metroid: Zero Mission Low% Hard by Dragondarch
http://www.twitch.tv/ludendi/c/4081175

Metroid II Return of Samus 100% by EmptyEye
http://www.twitch.tv/ludendi/c/4081184

Another Metroid 2 Remake Demo 1.3 100% (with Developer Commentary!) by Dragondarch
http://www.twitch.tv/ludendi/c/4081197

Super Metroid 14% Ice by Zoast
http://www.twitch.tv/ludendi/c/4081235

Super Metroid Any% by Zoast
http://www.twitch.tv/ludendi/c/4081239

Metroid Fusion 1% by JaggerG
http://www.twitch.tv/ludendi/c/4081250

Metroid Fusion 100% by Kirbymastah
http://www.twitch.tv/ludendi/c/4081267

Metroid Prime Pinball Any% by UchihaSDA
http://www.twitch.tv/ludendi/c/4081281

Metroid Prime 100% by MilesSMB
http://www.twitch.tv/ludendi/c/4081311

Metroid Prime: Hunters 100% by Amhite
http://www.twitch.tv/ludendi/c/4081338

Metroid Prime 2: Echoes 100% by Kirbymastah
Part 1 - http://www.twitch.tv/ludendi/c/4081349
Part 2 - http://www.twitch.tv/ludendi/c/4081368

Metroid Prime 3: Corruption 100% by Kirbymastah
Part 1 - http://www.twitch.tv/ludendi/c/4081377
Part 2 - http://www.twitch.tv/ludendi/c/4081398

Super Metroid 100% by Kottpower
http://www.twitch.tv/ludendi/c/4081418

[b]Credits[/b]

We would like to give thanks to the following:

-The staff at twitch for allowing us to be on the front page for the duration of the marathon, and for hosting their website so we can stream this content to you!
-The overall speedrunning community (speeddemosarchive and speedrunslive) for giving their full support in this marathon

Organizers:
-Svenne for initiating and starting with the idea of the metroid marathon. He has been in contact with twitch for getting this on the front page, initiating about the mumble, and allowing this marathon to be streamed on the team ludendi channel.
-Kirbymastah for organizing and putting together the schedule and runners, streaming the marathon, and leading the marathon organization.
-Grukk for helping organize the mumble and providing his server to allow us to use it for this marathon.
-Dickoak and Yashi for allowing us to use their rtmp server for the streaming set-up.
-In particular, Yashi for willing to stay up for the majority of the marathon to set up a restream and restream the second half of the marathon due to my internet issues.

Design:
-Charleon for putting together the streaming layout you see right now, along with helping edit the individual game logos to fit the layout.
-Infinity's End for creating the dark samus sprite and allowing us to use it in the layout.
-Dragondarch for providing several raw game logos.
-Guynzaro for helping edit the individual game logos.
-Heero_fred for putting together a thumbnail for this channel to go on twitch's front page carousel.
-Kari_Fry for putting together a logo for the metroid marathon itself.
-The developers of LiveSplit for their public timer.
-In particular, CryZe for helping with customizing the timer for the marathon.

We would like to thank the developers for releasing these awesome games and Nintendo for this awesome series! We would also like to thank the programmer and the testers for AM2R for allowing dragondarch to run their demo 1.3 at this marathon! In particular, we would like to thank DoctorM64 for joining us on commentary for AM2R while dragondarch ran his demo!

Finally, we would like to thank all of the runners and commentators for putting their time and offering to run and commentate their respective games for this marathon! Without all of you guys, this marathon would not have been possible! And we would like to thank you, the viewers, for watching this marathon and spreading the word about it!

kirbymastah

2014-04-14 08:51:32 pm

<(^_^)>

On the topic of how to restream, I've thought about just having the runner stream directly to a twitch account, and then the restreamer screen capture manually and take the audio. This would work quite well; the main problem would be that the commentators would have the stupid 20-30 second twitch delay. With the rtmp set-up we had, there was only 2-3 second delay for those who watched the rtmp directly which made commentary much easier. Rtmp worked well, but again, as I mentioned, it kinda got ruined for kottpower's 60 fps stream, and the audio had pitch problems for every stream, meaning the rtmp restreaming did something weird to it. If there's a way to compromise this and get the best of both worlds, I'm all ears.

Miles

2014-04-14 08:51:55 pm

I really didn't like Mumble, or at least the way it was used here. Multiple rooms where you can only be in one at a time, and you have to be moved by an admin, meaning you can't freely join in the commentary and when you do do commentary, you're moved out of the main chat... Or was I just missing something? :\

~~Personally I can't wait for your 17% Corruption run next marathon.~~

kirbymastah

Edit history:

kirbymastah: 2014-04-14 09:07:37 pm

2014-04-14 09:06:51 pm

<(^_^)>

Quote from Miles:

I really didn't like Mumble, or at least the way it was used here. Multiple rooms where you can only be in one at a time, and you have to be moved by an admin, meaning you can't freely join in the commentary and when you do do commentary, you're moved out of the main chat... Or was I just missing something? :\

The moving thing was server-specific, many mumble servers people can move around as they wish. This would be convenient for runners and commentators, but at the same time it'd require nobody to do anything dumb. I don't really want people to join in commentary without asking me and/or the runner beforehand (again, it wouldn't happen since 99% of people who I invite to mumble wouldn't do anything like that... I hope) I'll be hoping to make a mumble server with this convenience.

Adding the 'hypethon' token for chatting, I believe, is something specific to grukk's server.

As for not being able to be in multiple rooms, I'm not sure if there's a way around that, and I personally don't think it's that big of a deal IMO, at least for listening to the conversations. I do wish I could see the chats in every channel of the mumble server at once, regardless of the current one I'm in, which is related to what you mentioned Miles. Someone more familiar with mumble, is there a way around this?

Ultimately, I think using mumble again would be fine, with the clear division of the separable groups and such. And ideally, we'll be making a new server (or someone else's more customized one at least) so some of those issues will be addressed. Part of the issues stem from the fact that I sent out the set-up pastebin to all the runners and commentators quite late (I think Thursday) so that was an underlying problem on my part. A lot of this was also my unfamiliarity with using mumble as well, so I'll (and other organizers for next time) spend a lot more time learning about its features and figuring out how to let runners/commentators know about them in the most convenient way possible.

I do want to thank grukk for allowing us to use his server on a last-minute basis Smiley

[quote="Miles"]I really didn't like Mumble, or at least the way it was used here. Multiple rooms where you can only be in one at a time, and you have to be moved by an admin, meaning you can't freely join in the commentary and when you do do commentary, you're moved out of the main chat... Or was I just missing something? :\[/quote]

The moving thing was server-specific, many mumble servers people can move around as they wish. This would be convenient for runners and commentators, but at the same time it'd require nobody to do anything dumb. I don't really want people to join in commentary without asking me and/or the runner beforehand (again, it wouldn't happen since 99% of people who I invite to mumble wouldn't do anything like that... I hope) I'll be hoping to make a mumble server with this convenience.

Adding the 'hypethon' token for chatting, I believe, is something specific to grukk's server.

As for not being able to be in multiple rooms, I'm not sure if there's a way around that, and I personally don't think it's that big of a deal IMO, at least for listening to the conversations. I do wish I could see the chats in every channel of the mumble server at once, regardless of the current one I'm in, which is related to what you mentioned Miles. [b]Someone more familiar with mumble, is there a way around this?[/b]

Ultimately, I think using mumble again would be fine, with the clear division of the separable groups and such. And ideally, we'll be making a new server (or someone else's more customized one at least) so some of those issues will be addressed. Part of the issues stem from the fact that I sent out the set-up pastebin to all the runners and commentators quite late (I think Thursday) so that was an underlying problem on my part. A lot of this was also my unfamiliarity with using mumble as well, so I'll (and other organizers for next time) spend a lot more time learning about its features and figuring out how to let runners/commentators know about them in the most convenient way possible.

I do want to thank grukk for allowing us to use his server on a last-minute basis :)

heero_fred

2014-04-14 09:15:57 pm

one minor nitpick is that I noticed you had a widescreen layout for MP3, yet you used the 4:3 layout for the GBA games and stretched them to fit it, in my opinion it would have been better to either:

a) Use the Widescreen layout without stretching the game feeds
or
b) Have a specific layout for the GBA games

kirbymastah

Edit history:

kirbymastah: 2014-04-14 09:20:35 pm

2014-04-14 09:19:41 pm

<(^_^)>

Quote from heero_fred:

one minor nitpick is that I noticed you had a widescreen layout for MP3, yet you used the 4:3 layout for the GBA games and stretched them to fit it, in my opinion it would have been better to either:

a) Use the Widescreen layout without stretching the game feeds
or
b) Have a specific layout for the GBA games

If we can have someone like charelon make an extra layout, then that would probably work best for the GBA games. I was also hoping widescreen would be used for other m, but jaggerg streams it with 4:3; not a big deal, but would've been nicer.

Which reminds me, I had a lot of trouble cropping and moving the rtmp livestream around the xsplit layout properly (yashi did it fine though), which is another minor point against restreaming using rtmp again Tongue

Overfiendvip

2014-04-14 10:06:11 pm

never a frown

Since I haven't seen anyone giving their input on it, I thought Hunters was a fine run. As someone who hasn't watched Hunters before, I was pleasantly surprised at how much it resembled a classic FPS (the rocket boosting, the weaponry, and the textures come to mind). Even though the chat at that time was abysmal (If there was an "Other M" graph like you had with viewership to show just how much it was mentioned in chat, I'm pretty sure its peak would have been during Hunters, or second only to Other M itself) and the run went over estimate, which may have been due to the runner's inexperience with understanding just how much can go wrong in a marathon run, either that or it was an estimate based around in-game time or something, it was worth the time watching and I definitely think it should have another go, assuming there was another marathon.

Even with the technical issues and lowered frame rate which is understandable, I got a lot of entertainment out of it. I'm sure you know what could have been improved so I'll just end by saying thanks to those who performed. I don't think the runners hear enough about how much it means to people that they took the time to be in the marathon.

Here's a possible blooper, I don't see any listed for Hunters yet even though this and maybe one or two others were candidates: http://www.twitch.tv/ludendi/b/519370712?t=9h19m38s

I'm not sure where in the run but he later comes back to that spot and makes it over first try, which was another bit of humor.

Since I haven't seen anyone giving their input on it, I thought Hunters was a fine run. As someone who hasn't watched Hunters before, I was pleasantly surprised at how much it resembled a classic FPS (the rocket boosting, the weaponry, and the textures come to mind). Even though the chat at that time was abysmal (If there was an "Other M" graph like you had with viewership to show just how much it was mentioned in chat, I'm pretty sure its peak would have been during Hunters, or second only to Other M itself) and the run went over estimate, which may have been due to the runner's inexperience with understanding just how much [i]can[/i] go wrong in a marathon run, either that or it was an estimate based around in-game time or something, it was worth the time watching and I definitely think it should have another go, assuming there was another marathon.

Even with the technical issues and lowered frame rate which is understandable, I got a lot of entertainment out of it. I'm sure you know what could have been improved so I'll just end by saying thanks to those who performed. I don't think the runners hear enough about how much it means to people that they took the time to be in the marathon.

Here's a possible blooper, I don't see any listed for Hunters yet even though this and maybe one or two others were candidates: http://www.twitch.tv/ludendi/b/519370712?t=9h19m38s

I'm not sure where in the run but he later comes back to that spot and makes it over first try, which was another bit of humor.

Mystery

Edit history:

Mystery: 2014-04-14 10:48:58 pm

2014-04-14 10:46:28 pm

OK, so first, awesome marathon. Seeing so many Metroid games run was just great.
The unfortunate downside are, of course, the technical problems, so I'm going to focus on those:

- The bitrate for games were abmyssal. Other M was just a blockfest and was unwatchable. This happened to some other games too, such as the Fusion run towards the end. I understand that not everyone has the greatest internet connection speeds, but the whole point of a marathon is to show it off to other people, and if other people can't watch it, then it just becomes no fun.
- Choppy framerate. Many games had problems with choppy framerates. The fusion runs during the halfpoint or so, I believe was one example. Some games were just choppy always I believe, and some were periodically choppy and sometimes just fine.
- Overpeaked audio was an issue. Much of the commentary had overpeaked audio almost all the time. This usually happens when the mic boost is too high, or one is sitting too close to the mic. This should not be difficult to fix, I believe. The problem is that it detracts from the experience.
- Bad commentary audio quality (technical-wise). Some commentary was just awful, again detracting from the whole experience.
- Noise. It was clear that there was a lot of noise from some people's ends. This could be heard in the very first run. When the connection to the audio commentary would cut out, the noise would disappear. I think this is a typical sign of a bad mic or too high mic boost. The noise was distracting.
- Unwanted sound effects. You could hear keyboard typing (with higher audio frequencies being cut off too) and people talking in the background sometimes. Especially the keyboard is extremely distracting. I think this needs planning in the future. Make sure you aren't disturbed by other people during the commentary or shut off the mic when doing something that causes noise.

Again, great marathon, but these issues almost destroyed the experience pretty much the entire marathon.
Btw, the AM2R run was awesome and absolutely perfect. I don't recall ANY of these issues during that run, which contributed to great overall experience.

Btw, seeing as youtube also has live streaming, I'm curious... has anyone tested it?

As for the estimations, maybe it would be a good idea to take the worst estimate and add 15% to the estimate? Clearly people underestimate what a good estimate should be and an easy way of deriving a good estimate seems like it would help a lot.

OK, so first, awesome marathon. Seeing so many Metroid games run was just great.
The unfortunate downside are, of course, the technical problems, so I'm going to focus on those:

- The bitrate for games were abmyssal. Other M was just a blockfest and was unwatchable. This happened to some other games too, such as the Fusion run towards the end. I understand that not everyone has the greatest internet connection speeds, but the whole point of a marathon is to show it off to other people, and if other people can't watch it, then it just becomes no fun.
- Choppy framerate. Many games had problems with choppy framerates. The fusion runs during the halfpoint or so, I believe was one example. Some games were just choppy always I believe, and some were periodically choppy and sometimes just fine.
- Overpeaked audio was an issue. Much of the commentary had overpeaked audio almost all the time. This usually happens when the mic boost is too high, or one is sitting too close to the mic. This should not be difficult to fix, I believe. The problem is that it detracts from the experience.
- Bad commentary audio quality (technical-wise). Some commentary was just awful, again detracting from the whole experience.
- Noise. It was clear that there was a lot of noise from some people's ends. This could be heard in the very first run. When the connection to the audio commentary would cut out, the noise would disappear. I think this is a typical sign of a bad mic or too high mic boost. The noise was distracting.
- Unwanted sound effects. You could hear keyboard typing (with higher audio frequencies being cut off too) and people talking in the background sometimes. Especially the keyboard is extremely distracting. I think this needs planning in the future. Make sure you aren't disturbed by other people during the commentary or shut off the mic when doing something that causes noise.

Again, great marathon, but these issues almost destroyed the experience pretty much the entire marathon.
Btw, the AM2R run was awesome and absolutely perfect. I don't recall ANY of these issues during that run, which contributed to great overall experience.

Btw, seeing as youtube also has live streaming, I'm curious... has anyone tested it?

As for the estimations, maybe it would be a good idea to take the worst estimate and add 15% to the estimate? Clearly people underestimate what a good estimate should be and an easy way of deriving a good estimate seems like it would help a lot.

TheMG2

2014-04-14 11:18:44 pm

Quote from Mystery:

- The bitrate for games were abmyssal. Other M was just a blockfest and was unwatchable. This happened to some other games too, such as the Fusion run towards the end. I understand that not everyone has the greatest internet connection speeds, but the whole point of a marathon is to show it off to other people, and if other people can't watch it, then it just becomes no fun.

The problem here is that to not have that in order to not have this issue, you would have to reject runners due to low internet quality, and I'm not sure how I feel about this.

Mystery

2014-04-14 11:21:55 pm

Quote from TheMG2:

The problem here is that to not have that in order to not have this issue, you would have to reject runners due to low internet quality, and I'm not sure how I feel about this.

I understand the issue, but for example Other M was unwatchable, no matter how good the runner.
The world is not a perfect place and a marathon kind of has to take a stance on things such as this. What is most important? Quality or runners? The viewers or the streamers? Where does the line go? It's a sad thing, but that line needs to be defined and no matter how one does it, some people are going to be upset.

Miles

2014-04-14 11:23:44 pm

Quote from Mystery:

Other M was unwatchable

I have never agreed more with a statement.

What, you want me to comment on the actual context? No, I completely disagree on rejecting runners based on their internet quality.

kirbymastah

Edit history:

kirbymastah: 2014-04-14 11:52:10 pm

2014-04-14 11:50:37 pm

<(^_^)>

I feel like the majority of things that Mystery brought up were issues to do with the restreaming method we used. The choppiness and audio desyncing were 90% a result from the rtmp restreams (the actual runner stream was, in the general case, fine, and the restreamer usually wasn't dropping any frames either). None of the commentators actually complained about choppiness either and they were watching the runner's raw stream, which shows it likely has to do more with the rtmp restreaming method. Again, something I'm likely not going to try ever again.

For JaggerG's internet, his streams are actually mostly fine and not that choppy. I've watched countless streams of his, and from what I've watched, the other m run was OKish in terms of choppiness (the resolution was bad but that's not as big of a deal really). I don't recall other m ever becoming completely "unwatchable" so maybe you're exacerbating the issue a bit much? Regardless of that, again it almost definitely came down to the restreaming technique... <_< Honestly I don't think any of the runs were overall completely unwatchable in terms of framerate, aside from the SM 100% run (which I've already addressed, and is basically the same issue... again). I might be remembering wrong though.

For commentary, I kinda started with trying to have more serious, official commentary. Ever since zoasty's super metroid runs, they became more casual and conversational, which isn't a bad thing at all. However, I think I started being rather too lenient over allowing people to commentate (some people left and started commentating away from home, and so on), so I'll have to at least make some restrictions as for making sure the runner can be in a silent room (to the best of their ability) and so on. As for stuff like typing and such, there isn't anything I can directly do about that (heck, I've been guilty of that before as well) besides make sure everyone uses push-to-talk, outside of the runner.

And I think I'd rather have an Other M runner than no game at all, as long as that person has shown to be able to stream properly (which JaggerG has shown). Again, JaggerG's internet is good enough to stream gameplay, despite being at a relatively low resolution for a wii game. I'll just have to find someone to set up a proper restream of good quality (and also figure out bitrates; I'm not exactly stream-tech-savvy either :S)

I feel like the majority of things that Mystery brought up were issues to do with the restreaming method we used. The choppiness and audio desyncing were 90% a result from the rtmp restreams (the actual runner stream was, in the general case, fine, and the restreamer usually wasn't dropping any frames either). None of the commentators actually complained about choppiness either and they were watching the runner's raw stream, which shows it likely has to do more with the rtmp restreaming method. Again, something I'm likely not going to try ever again.

For JaggerG's internet, his streams are actually mostly fine and not that choppy. I've watched countless streams of his, and from what I've watched, the other m run was OKish in terms of choppiness (the resolution was bad but that's not as big of a deal really). I don't recall other m ever becoming completely "unwatchable" so maybe you're exacerbating the issue a bit much? Regardless of that, again it almost definitely came down to the restreaming technique... <_< Honestly I don't think any of the runs were overall completely unwatchable in terms of framerate, aside from the SM 100% run (which I've already addressed, and is basically the same issue... again). I might be remembering wrong though.

For commentary, I kinda started with trying to have more serious, official commentary. Ever since zoasty's super metroid runs, they became more casual and conversational, which isn't a bad thing at all. However, I think I started being rather too lenient over allowing people to commentate (some people left and started commentating away from home, and so on), so I'll have to at least make some restrictions as for making sure the runner can be in a silent room (to the best of their ability) and so on. As for stuff like typing and such, there isn't anything I can directly do about that (heck, I've been guilty of that before as well) besides make sure everyone uses push-to-talk, outside of the runner.

And I think I'd rather have an Other M runner than no game at all, as long as that person has shown to be able to stream properly (which JaggerG has shown). Again, JaggerG's internet is good enough to stream gameplay, despite being at a relatively low resolution for a wii game. I'll just have to find someone to set up a proper restream of good quality (and also figure out bitrates; I'm not exactly stream-tech-savvy either :S)

Mystery

Edit history:

Mystery: 2014-04-15 12:01:20 am

2014-04-14 11:59:15 pm

Quote from kirbymastah:

For JaggerG's internet, his streams are actually mostly fine and not that choppy. I've watched countless streams of his, and from what I've watched, the other m run was OKish in terms of choppiness (the resolution was bad but that's not as big of a deal really). I don't recall other m ever becoming completely "unwatchable" so maybe you're exacerbating the issue a bit much? Regardless of that, again it almost definitely came down to the restreaming technique... <_< Honestly I don't think any of the runs were overall completely unwatchable in terms of framerate, aside from the SM 100% run (which I've already addressed, and is basically the same issue... again). I might be remembering wrong though.

I just watched a little at the beginning in the "beginning area" where Samus is asked to drop bombs, etc. The bombs pretty much blurred the entire screen. That's how bad the quality was.
It might also be me - because I am sensitive to video quality in that way. As long as it's not blocky or blurry, I'm mostly fine. Though usually I'm sensitive to any degradation in quality that I can see. That said, I'm not as picky as others - I typically use CRF 20 for encoding, while others can use down to 16-17.

Quote:

And I think I'd rather have an Other M runner than no game at all, as long as that person has shown to be able to stream properly (which JaggerG has shown). Again, JaggerG's internet is good enough to stream gameplay, despite being at a relatively low resolution for a wii game. I'll just have to find someone to set up a proper restream of good quality (and also figure out bitrates; I'm not exactly stream-tech-savvy either :S)

As for bitrates... well, that depends on the settings used to stream (e.g. what preset for x264), but I think 5 MBps would be the upper-bound for required bitrate.
I can do a few tests to see how it goes, though.

Also, I do think you should be mindful of the quality. It's easy to upset people and it's not easy to please them after that. One look at the stream and seeing poor quality is all it takes for someone to disappear forever. So I do think you should have a minimum required quality for all streams even if it means possibly rejecting runners. That said, the quality does not need to be so high that you have to reject loads of runners. A middle ground.

[quote="kirbymastah"]
For JaggerG's internet, his streams are actually mostly fine and not that choppy. I've watched countless streams of his, and from what I've watched, the other m run was OKish in terms of choppiness (the resolution was bad but that's not as big of a deal really). I don't recall other m ever becoming completely "unwatchable" so maybe you're exacerbating the issue a bit much? Regardless of that, again it almost definitely came down to the restreaming technique... <_< Honestly I don't think any of the runs were overall completely unwatchable in terms of framerate, aside from the SM 100% run (which I've already addressed, and is basically the same issue... again). I might be remembering wrong though.
[/quote]
I just watched a little at the beginning in the "beginning area" where Samus is asked to drop bombs, etc. The bombs pretty much blurred the entire screen. That's how bad the quality was.
It might also be me - because I am sensitive to video quality in that way. As long as it's not blocky or blurry, I'm mostly fine. Though usually I'm sensitive to any degradation in quality that I can see. That said, I'm not as picky as others - I typically use CRF 20 for encoding, while others can use down to 16-17.

[quote]And I think I'd rather have an Other M runner than no game at all, as long as that person has shown to be able to stream properly (which JaggerG has shown). Again, JaggerG's internet is good enough to stream gameplay, despite being at a relatively low resolution for a wii game. I'll just have to find someone to set up a proper restream of good quality (and also figure out bitrates; I'm not exactly stream-tech-savvy either :S)[/quote]
As for bitrates... well, that depends on the settings used to stream (e.g. what preset for x264), but I think 5 MBps would be the upper-bound for required bitrate.
I can do a few tests to see how it goes, though.

Also, I do think you should be mindful of the quality. It's easy to upset people and it's not easy to please them after that. One look at the stream and seeing poor quality is all it takes for someone to disappear forever. So I do think you should have a minimum required quality for all streams even if it means possibly rejecting runners. That said, the quality does not need to be so high that you have to reject loads of runners. A middle ground.

heero_fred

Edit history:

heero_fred: 2014-04-15 12:17:47 am

heero_fred: 2014-04-15 12:17:45 am

2014-04-15 12:16:47 am

While I do agree that the Other M run was pixelated at some points, the quality of the run and the commentary made it worth watching, while yeah, the video quality does affect general viewership to some degree, it was a speedrun marathon, and what matters is having quality runs.

As for quality of the commentary in most runs, most of it was really good, but in some of them the balance between being informative and being entartaining was offset for the worse, thankfully it didn't happen that often.

to finish this comment I will leave these questions, Is it really a re-streaming set up the best option? why not send each runner their layout and have them stream directly to the channel?

(I do know of all the complications the second options has, but in some cases may be a better option, specially with runners that have bad internet connection)

Vulajin

2014-04-15 12:19:42 am

The main problem with that is Twitch frequently has hiccups where it won't allow the next runner to begin streaming. This occurred a lot during the last marathon I saw which did not use restreaming (I think Easter Charity Marathon last year? not 100% sure).

Mobiusman

2014-04-15 02:28:14 am

Having rewatched my run, some of the sound issues were definitely on my end. I can only apologise for this, it's a side effect of the (temporary!) streaming setup I'm using atm.

Quote from kirbymastah:

-The chat wasn't filled with complete spam, considering it was 2-3k viewers usually. The chat was filled with other m talk literally every 30 f**king seconds. We'll probably have to make some rules about the chat...

While obviously we'd prefer a more on-topic chat, I disagree with trying to regulate this. People are going to talk about what the want to talk about, and usually this is other M and "samus is a girl".

Dragonfangs

2014-04-15 02:37:38 am

During the marathon I was exceedingly annoyed that Other M kept getting brought up during every run. But I still agree with Mobius, the chat doesn't exist to please the streamer/organizer/crew, it's there to please the people chatting. And so any topic that stays alive is probably something people genuinely want to talk about.

ap3604

2014-04-15 03:21:08 am

Sorry if it's a stupid question but is there a place that I could see a recording of the AM2R speedrun?

I had to be at work at the time so I couldn't watch the live stream and I'd really like to see this speedrun if it's possible Smiley

Mystery

2014-04-15 03:26:19 am

There are no videos yet, though perhaps it might come.
In the meantime, you could check past broadcasts to see if you can find it: http://www.twitch.tv/ludendi/profile/pastBroadcasts

ap3604

2014-04-15 03:27:42 am

Quote from Mystery:

There are no videos yet, though perhaps it might come.
In the meantime, you could check past broadcasts to see if you can find it: http://www.twitch.tv/ludendi/profile/pastBroadcasts

Thanks for the quick reply Mystery, I'll check it out now and hope I they put up the AM2R video at a later time then.

Mystery

Edit history:

Mystery: 2014-04-16 01:18:40 am

2014-04-15 04:34:37 am

Quote from Mystery:

As for bitrates... well, that depends on the settings used to stream (e.g. what preset for x264), but I think 5 MBps would be the upper-bound for required bitrate.
I can do a few tests to see how it goes, though.

So I have a few tests now. I used Corruption, a 480p game @ 60 fps.
I tried with x264 Very fast preset, CRF 21. I managed to go down to Medium before I started getting under 60 fps encoding (i.e. can't stream at 60 fps). Medium didn't really help bitrate, though. Still waiting for results from Very slow just because. No computer will be able to stream at that preset, though, but it's still interesting if it can push down the bitrate.

Anyway, my results so far are that the average bitrate required for streaming at good quality is 1758 kbps.
Because Twitch is slapping on an incredibly stupid policy that requires all streams to use CBR starting september, I also checked the bitrate curve and found that roughly CBR = 5000 kbps should ensure that 95% or so of the video falls within that range so quality of frames are not sacrificed.

Because Twitch's stupid policy and its evil constant buffering, I also checked youtube. The good thing is that there was absolutely no buffering. The bad quality from my test was that quality was poor even though I was uploading at good settings. The other drawback is that it only supports 30 fps, but that's better than constant buffering.

EDIT: The very slow encode turned out to be the same size as the very fast encode, for anyone remotely interested. Go figure.

[quote="Mystery"]
As for bitrates... well, that depends on the settings used to stream (e.g. what preset for x264), but I think 5 MBps would be the upper-bound for required bitrate.
I can do a few tests to see how it goes, though.
[/quote]
So I have a few tests now. I used Corruption, a 480p game @ 60 fps.
I tried with x264 Very fast preset, CRF 21. I managed to go down to Medium before I started getting under 60 fps encoding (i.e. can't stream at 60 fps). Medium didn't really help bitrate, though. Still waiting for results from Very slow just because. No computer will be able to stream at that preset, though, but it's still interesting if it can push down the bitrate.

Anyway, my results so far are that the average bitrate required for streaming at good quality is 1758 kbps.
Because Twitch is slapping on an incredibly stupid policy that requires all streams to use CBR starting september, I also checked the bitrate curve and found that roughly CBR = 5000 kbps should ensure that 95% or so of the video falls within that range so quality of frames are not sacrificed.

Because Twitch's stupid policy and its evil constant buffering, I also checked youtube. The good thing is that there was absolutely no buffering. The bad quality from my test was that quality was poor even though I was uploading at good settings. The other drawback is that it only supports 30 fps, but that's better than constant buffering.

EDIT: The very slow encode turned out to be the same size as the very fast encode, for anyone remotely interested. Go figure.

JaggerG

2014-04-15 07:04:35 am

Jumping Turtle

Quote from kirbymastah:

-Restreaming Set-up. [...] Yashi's internet was fine, and kottpower's rtmp stream was fine, so the only explanation was how xsplit was taking the run; it was likely because we streamed at 25ish fps, whereas kottpower's stream was at 60 fps, which likely caused problems. I'll likely have paid xsplit next time we do this, so the fps issue won't be a problem.

Oh, I thought you did have the paid version. If that had been the issue, maybe OBS should be looked into. Is there anything you did that literally only XSplit could do?

Quote from kirbymastah:

-We were generally over 30-60 minutes past our time estimate after the first few runs.

...Really? I seem to remember Saturday going pretty smoothly. A lot of the runners have been in marathons before, and understand the importance of very high estimates - because it's not just for their runs.

Quote from kirbymastah:

-The chat wasn't filled with complete spam, considering it was 2-3k viewers usually. The chat was filled with other m talk literally every 30 f**king seconds. We'll probably have to make some rules about the chat...

Again, sorry about accidentally encouraging that at the start of Prime 100%. I realized it after the first room and tried to change the subject back to Prime, then left before I said much more. >.> I'm actually trying to think of some means of getting some kind of public service announcement or something out there so that people understand better. I had an idea for a regular youtube series I could include it in (my pastebin is pretty much a possible script draft for a video), but unfortunately it requires dedicated time I just can't guarantee, not to mention video editing skills I just don't care to learn.

Quote from kirbymastah:

-The runs overall were great. I'm glad I picked the runners that I picked, as they were all top quality. And though some bad things did happen, occasional disasters are bound to happen, and basically every runner (at least from what I've watched)

...that's it? Not "I am a vampire and here's a million dollars?" :StrongBad:

Quote from kirbymastah:

Scheduling

I felt that starting at 7am had been a mistake. Although European audience is definitely out there and very significant, the vast majority of activity is generally in the late afternoons and evenings in the US timezones. Then again, I could be mistaken, and perhaps ludendi hosting ESA makes a considerable difference in Primetime hours.

And again, I feel in general it's bad to put the same game more than once consecutively, unless it's something like LADX glitched% for a donation incentive or something. But as you said, this had probably been at least the easiest solution to the schedules of each runner.

Quote from kirbymastah:

-Slightly higher estimates, and "practice setup" would be good.

Alternatively, for a long time I've thought it silly that Speedrun marathons, focusing on fast runs, have had estimates that are like 15 minutes, or in some cases half an hour, of buffer included for set-up and potential error costs. I understand these costs are VERY necessary to consider, but I personally don't think it makes sense to include set-up time if you're timing the run on-stream. I think in the future, it may be better to put the runner's estimate, then have set-up time after each run that includes those considerations. I think the potential excitement of the audience comparing the run to the estimate is great for a series that has been pushed this far.

Quote from kirbymastah:

-Be more strict on regulating commentary, and do audio tests and set some rules for them

Push-to-talk for commentators makes sense, in retrospect. I should have thought of that, being a regular Raidcall user myself. I did try to keep out of commentary that already had informed commentators. As much as I've learned about the Prime series and Super Metroid, I'm far less qualified to talk about the most important subjects for speedrunning each category as those you already had. Near the end, I did try to fill in for some silliness to liven it up a little, but the priority for commentary for these marathons should definitely be Runner > Designated Commentator (this includes special guests of any kind) > anyone else. It actually kind of looks odd to place the runner above potentially a developer, but I think most of the time the runner is quiet often enough, or is just as interested (if not moreso) in listening to the dev as anyone else.

Quote from kirbymastah:

As for new games and categories, the only viable ones not done that I can think of are zero mission any% hard, and super metroid RBO. Hopefully the full AM2R game will be out by then too. People have asked about doing super metroid hacks; I'd be willing to add 1-2 of them, I think lioran said he would've offered to do SM Impossible if he didn't know about the marathon so late. I don't really want more than 2 romhacks, since it kinda gets iffy with non-officla games overall :S

I'm inclined to hope for any% runs of EVERY game next time, in order to make it more like a relay run through the series. This is gonna be tough to schedule, but cool if it pans out. Other categories and fangames/hacks could come after.

As for hacks, if AM2R is in, I think some hacks are just as valid. Are there any Zero Fission hacks runners around? But yeah, a limit on those is definitely best.

Quote from kirbymastah:

Highlights and Bloopers

ZM 10% Hard: The fails and success of Iron Ted?
You missed Hunters and Pinball
Pinball: A brief introduction to Meteor Prime Pinball.

About Mumble, one thing you maybe could do is mute everyone not commentating. I personally tried to stay out of Commentary I'm not involved in, so I ended up PMing you a lot. I recall you PMing others a lot, too.

Also, man, it's hard explaining a game I've never played before. Seems like the chat thought I did alright, so that's cool at least. The delay made it even more odd, though.

[quote="kirbymastah"]-Restreaming Set-up. [...] Yashi's internet was fine, and kottpower's rtmp stream was fine, so the only explanation was how xsplit was taking the run; it was likely because we streamed at 25ish fps, whereas kottpower's stream was at 60 fps, which likely caused problems. I'll likely have paid xsplit next time we do this, so the fps issue won't be a problem.[/quote]
Oh, I thought you did have the paid version. If that had been the issue, maybe OBS should be looked into. Is there anything you did that literally only XSplit could do?

[quote="kirbymastah"]-We were generally over 30-60 minutes past our time estimate after the first few runs.[/quote]
...Really? I seem to remember Saturday going pretty smoothly. A lot of the runners have been in marathons before, and understand the importance of very high estimates - because it's not just for their runs.

[quote="kirbymastah"]-The chat wasn't filled with complete spam, considering it was 2-3k viewers usually. The chat was filled with other m talk literally every 30 f**king seconds. We'll probably have to make some rules about the chat...[/quote]
Again, sorry about accidentally encouraging that at the start of Prime 100%. I realized it after the first room and tried to change the subject back to Prime, then left before I said much more. >.> I'm actually trying to think of some means of getting some kind of public service announcement or something out there so that people understand better. I had an idea for a regular youtube series I could include it in (my pastebin is pretty much a possible script draft for a video), but unfortunately it requires dedicated time I just can't guarantee, not to mention video editing skills I just don't care to learn.

[quote="kirbymastah"]-The runs overall were great. I'm glad I picked the runners that I picked, as they were all top quality. And though some bad things did happen, occasional disasters are bound to happen, and basically every runner (at least from what I've watched) [/quote]
...that's it? Not "I am a vampire and here's a million dollars?" :StrongBad:

[quote="kirbymastah"][b]Scheduling[/b][/quote]
I felt that starting at 7am had been a mistake. Although European audience is definitely out there and very significant, the vast majority of activity is generally in the late afternoons and evenings in the US timezones. Then again, I could be mistaken, and perhaps ludendi hosting ESA makes a considerable difference in Primetime hours.

And again, I feel in general it's bad to put the same game more than once consecutively, unless it's something like LADX glitched% for a donation incentive or something. But as you said, this had probably been at least the easiest solution to the schedules of each runner.

[quote="kirbymastah"]-Slightly higher estimates, and "practice setup" would be good.[/quote]
Alternatively, for a long time I've thought it silly that Speedrun marathons, focusing on fast runs, have had estimates that are like 15 minutes, or in some cases half an hour, of buffer included for set-up and potential error costs. I understand these costs are VERY necessary to consider, but I personally don't think it makes sense to include set-up time if you're timing the run on-stream. I think in the future, it may be better to put the runner's estimate, then have set-up time after each run that includes those considerations. I think the potential excitement of the audience comparing the run to the estimate is great for a series that has been pushed this far.

[quote="kirbymastah"]-Be more strict on regulating commentary, and do audio tests and set some rules for them[/quote]
Push-to-talk for commentators makes sense, in retrospect. I should have thought of that, being a regular Raidcall user myself. I did try to keep out of commentary that already had informed commentators. As much as I've learned about the Prime series and Super Metroid, I'm far less qualified to talk about the most important subjects for speedrunning each category as those you already had. Near the end, I did try to fill in for some silliness to liven it up a little, but the priority for commentary for these marathons should definitely be Runner > Designated Commentator (this includes special guests of any kind) > anyone else. It actually kind of looks odd to place the runner above potentially a developer, but I think most of the time the runner is quiet often enough, or is just as interested (if not moreso) in listening to the dev as anyone else.

[quote="kirbymastah"]As for new games and categories, the only viable ones not done that I can think of are zero mission any% hard, and super metroid RBO. Hopefully the full AM2R game will be out by then too. People have asked about doing super metroid hacks; I'd be willing to add 1-2 of them, I think lioran said he would've offered to do SM Impossible if he didn't know about the marathon so late. I don't really want more than 2 romhacks, since it kinda gets iffy with non-officla games overall :S[/quote]
I'm inclined to hope for any% runs of EVERY game next time, in order to make it more like a relay run through the series. This is gonna be tough to schedule, but cool if it pans out. Other categories and fangames/hacks could come after.

As for hacks, if AM2R is in, I think some hacks are just as valid. Are there any Zero Fission hacks runners around? But yeah, a limit on those is definitely best.

[quote="kirbymastah"][b]Highlights and Bloopers[/b][/quote]
ZM 10% Hard: The fails and success of Iron Ted?
You missed Hunters and Pinball
Pinball: A brief introduction to Meteor Prime Pinball.

About Mumble, one thing you maybe could do is mute everyone not commentating. I personally tried to stay out of Commentary I'm not involved in, so I ended up PMing you a lot. I recall you PMing others a lot, too.

Also, man, it's hard explaining a game I've never played before. Seems like the chat thought I did alright, so that's cool at least.  The delay made it even more odd, though.

CBenni

Edit history:

CBenni: 2014-04-15 08:18:35 am

2014-04-15 08:17:51 am

About chat: I wasnt able to read it all day, there were 3 events going on simultaneously and uni was giving me alot of work, but a lack of rules made it hard to decide what you guys wanted and what you didnt, I generally just made sure to look out for spam and I am sorry other unwanted chat went unchecked. Deactivating all spam filters in moobot was a bad idea however imho, we didnt set that up for no reason.
There should be some discussion about this before ESA kicks off.

Svenne

2014-04-15 08:35:34 am

Smörgåsbord

When the marathon started I had to un-activate the moobot because it banned people for bs reasons. I think this was the best chat I have seen in a online marathon, the only times people spammed was on when Kirby wanted to see FrankerZ's etc! Pretty sure only like 5-10 people got timed out altogether.