This topic is for the purpose of discussing #2101: Rikku & mr_roberts_z's N64 Super Mario 64 in 05:32.17

wat? moar SM64? I'll wait for an encode to watch, but I'd like to express my respect to SM64 TAS'ers for doing a great job on every TAS. I've enjoyed every iteration from the start and it's always just a job well done.

Yay frame war, yay cool authors giving us another good submission. However, I feel the great care that needs to be taken in controlling the camera in this game took a HUGE plunge. What the heck happened? I found myself disagreeing with every camera angle in the entire run. The first half of going into the castle and the first bowser level, the camera was zoomed in too close. In general, having the camera panned out in levels is better because you can get a better picture of mario in relation to his environment. Also, the angle wasn't too good, in that it was far more difficult to see objects/platforms that were approaching. Aesthetically, I always feel it best to take either a behind or diagonal-behind camera angle to get a clear picture of the environment that Mario is approaching. The camera angles inside the castle were mostly fine. First bowser stage, having it zoomed in close and sideways was bad. The change for the BLJ was fine, but then you chose a bad angle for getting to the pipe. The second bowser stage was done zoomed out to start. Good. It was at a sideways angle. Bad. It maintained this angle throughout the entire time. Mega bad! Again, I will repeat, getting some sense of what platform is coming next helps viewers get an orientation of where in the level they are. The pole glitch had a bad angle; you couldn't see anything that was going on. And then comes the biggest camera mistake of the whole run; on the outside of the brick area, just before the pipe, the camera was on the other side of the wall! you can't see mario at all! no one is going to know what the heck he's doing; only someone who watched a previous submission will know what is going on. In a run, going offscreen might be cool to show the author knows what he's doing (like in sonic advance 2 or SDW 120 exit); in a 3D game where you can control the camera, it is essential that you have a "director's sense" of what camera angle to use to best show what is going on. At the top of the spiral staircase, is there really no way after the first jump to increase your speed without bumping into the door? 3rd Bowser stage, again zoomed out sideways angle is still not doing it. You dived behind the gray box at the beginning (the more we see of Mario, the better). at a shortcut, he was hidden by more platforms. At the windy section, having mario come towards the camera didn't look good. man, I haven't had this many camera complaints about super mario 64 submissions in a while. But at this iteration of the 0 star run, I would expect the camera use to be up to snuff, especially because avoiding the lag in bowser boards the sub level is no longer needed.

*Sigh*

Encode Here: http://www.filedropper.com/newmario640star (24mb, 640x480) Very unexpected to see this run :P My thoughts on the run, The new camera angles were, different, but I can't say I really like them any better. Apart from them though, I didnt really notice much difference from the last run though. Still if there was a poll to vote, it would be a yes.

I didn't think the camera angles were bad, aside from just 2 places. Good job on the improvements =o

This never gets old. Hah! Each step get this closer to the Mario Bros TAS record, in fact, if not taking into account the first minute that is spent on cutscenes, it's actually faster. Will watch after it's published, which I've no doubt will be soon.

The camera angles aren't too bad. However, I think it can and should be improved. For the Mario game, it's probably best to pan it out and so that we can always see Mario and where he's heading. Sort of to always see the actions. There are at least 2 (I can only remember two, maybe more) places where I can't see Mario for a short while. Maybe the camera angle should've been switched to "the other side" to begin with to make this more entertaining. I will still vote Yes for the improvements. But I would prefer better camer angles. EDIT: don't see the option to vote any more. I guess this is the new voting system. (judged by comments instead of by votes.) EDIT 2: By the way, I noticed all of the "icons" (post reply, new topic, profile, quote, edit, pm, www, email, ... etc) are blurry. The upper left TASvideos icon is clear -- well, the word TASvideos is clear, the words "Tool-assisted superplay movies" is a bit blurry, but not as bad as the other icons I've mentioned.

I'm sorry to say it, but the camera angles were terrible. This was a yes for just the improvement, but the camera dragged it down. Gonna have to vote no on this, sorry. (mostly the same reasons as Comicalflop)

Ultimately the same as a few of the previous posts... Camera angles were pretty bad in my opinion. I personally prefer the zoomed out camera angle, as it gives the viewer more of a sense of what you're doing/where you're going. The camera angle at the end of BitFS was the absolutely the worst... the tricks shown there are cool-looking and I wish they would have been shown. Optimizing that part of the run would have been very difficult without being able to see where you're going, so I give you guys at least that much. With all of that aside, I am for any SM64 improvements. It's nice to see this game slowly getting beaten faster and faster, as someday sub-5:15 might even be possible at this rate. Overall, I wasn't entertained any more than the last submission, but yay for improvements which seemingly involve intense dedication to find.

The camera angles didn't bother me at all, the fact that they where so different from the last submission made the run look more different than it actually was, which I liked. The improvements where naturally pretty subtle, but at this point, considering how many revisions the 0 star run has had already (and the 1 star run too, which is mostly the same), it is still pretty impressive to see almost another 2 seconds saved over the previous version. And BitDW was noticeable faster. So to sum up. I liked it just about as much as the last run, where impressed by the improvements and think it should be published.

I dislike the idea of making the gameplay different than the last submission just because. I think each submission should strive for the perfect movie, and that means including the best solutions present on other movies, if no better solution is found (on the aesthetics point of view). I'm not sure if this was the case here, or if it was just 100% focus on frame shaving. All in all I have to say it's impressive to keep improving movies. I hope it won't stop. Good job.

Another Mario 64 run, this is what I like about this TAS, you guys are always striving to make this near as perfect as possible. Who knows? It could be improved once again or this could be it. It should be published. EDIT: That reminds me, since some improvements were found in this submission, no doubt could it be used in the 100% TAS.

Camera didn't bother me that much, though I'll agree that I don't like the constant mario close-ups and on the fire bowser level, mario's moves got entirely hidden for a certain period of time, which is silly. I was still thoroughly entertained and see this as worthy of publication, though I would reccomend the authors to fix the camera angles if the process only requires hexing the inputs and changing the c-button inputs.

Yeah, quite unexpected, but looks pretty decent overal. Some of the camera angles were a bit "different," but it still made a unique watch, though I would have prefered different camera angles. Good job guys.

Last thing I was expecting was for there to be an improvement in the final battle. I expected that to be picked over so much that it would have an unbeatable time already. The new camera angles made the game look a lot different, especially with the action hidden in a couple spots, but easier to see in others.

I would reccomend the authors to fix the camera angles if the process only requires hexing the inputs and changing the c-button inputs.

As I understand it, there are two significant problems with this suggestion:

1. In Mario 64, joystick movement is relative to the camera angle. So if you change the camera angle, you also have to adjust the joystick input for all frames afterwards, until the camera angle is reset to what it was during the initial recording.
2. There can be slowdown if too many things are on the screen at once, so there could be additional lag to compensate for as well.

BitFS could have had a better camera angle at the end (unless that reduces lag), but otherwise, it seems good. sub 5:30, here we come!

i independently came to the same conclusion as several others i have never cared at all about the camera angles previous to this mario 64 tas and when a couple people each time whined about the camera angles i thought they were being ridiculous but this one is where even i have to draw the line: these camera angles were bad the majority of the time and a few times even made me literally cringe also falling into the fire against bowser #2 might be something that other people do not care about but it looked ugly to me i feel that given the frequency of submitted improvements to mario 64 and how much time it would take to clean up the angles why not pass on this one in favor of the next? the published tas i thought was overall more entertaining also props to the encoder/uploader; that was good quality and file hosting services are more and more becoming more convenient, reliable, and faster in general than bt

also falling into the fire against bowser #2 might be something that other people do not care about but it looked ugly to me

This was required. [22:12] <mr_roberts_z> The position after he comes up makes Mario grab Bowser's tail closer to the bomb, so there's less distance Bowser has to fly.

Woah, I never knew that!

also falling into the fire against bowser #2 might be something that other people do not care about but it looked ugly to me

This was required. [22:12] <mr_roberts_z> The position after he comes up makes Mario grab Bowser's tail closer to the bomb, so there's less distance Bowser has to fly.

It's obviously not neccesary to fall into the lava to get into the right position for that sideflip dive, and it's possible to grab Bowser equally close to the bomb with other methods for diving into the tail as well. As for the run, poor camera angle choices. However, there were some nice improvements to make it a somewhat worthy submission.

It's obviously not neccesary to fall into the lava to get into the right position for that sideflip dive, and it's possible to grab Bowser equally close to the bomb with other methods for diving into the tail as well.

exactly if doing that had saved time then go right ahead but on your list of saved frames it says nothing about gaining frames during the second bowser battle

also falling into the fire against bowser #2 might be something that other people do not care about but it looked ugly to me

This was required. [22:12] <mr_roberts_z> The position after he comes up makes Mario grab Bowser's tail closer to the bomb, so there's less distance Bowser has to fly.

It's obviously not neccesary to fall into the lava to get into the right position for that sideflip dive, and it's possible to grab Bowser equally close to the bomb with other methods for diving into the tail as well. As for the run, poor camera angle choices. However, there were some nice improvements to make it a somewhat worthy submission.

It doesn't seem appropriate to include this word here.