And in one from the Willy Stages its looks like that you wait a bit and lose some Frames. Was this for luck manipulation or lag avoiding? Or for something else. Or was it lag, that it only looks like waiting?

Where do you exactly mean?

I watch again the movie. Yes I think I mean this part. 1. Wily Stage after climb the 2 ladders and equip the weapon from Elecman (near Frame 39600) Looks like waiting for me and shoting in the air much. But why? Near nothing get hit and there seems no use for it.

There are bricks megaman has to get rid of either by shooting them with Elec's weapon or throwing them off with Super Arm. Usually there is a major lag, but it's avoided by waiting a bit. I guess.

Okay, who voted no?

1. Wily Stage after climb the 2 ladders and equip the weapon from Elecman (near Frame 39600) Looks like waiting for me and shoting in the air much. But why? Near nothing get hit and there seems no use for it.

To reduce lag as you suspected earlier, also note that there are some big rocks ahead that you probably did not notice when Bisqwit fired at them. And the very minor delay (just a few frames) getting the refills near the ladder where for luck manipulation.

Ok, I test it now with a Save State infront of it and saw these blocks too. Now everything is clear for me. Great job.

Wow, the dance with Elecman is amazing! This is a (sort of unnecessary) yes. Bombman: "Why can't you just use the door like everybody else?"

I kind of liked the v6's Elecman battle more than this, but it's not a bad idea to make a new movie look different.

MM1 movies keep on getting better. This should grab MM2's star. Yes vote.

Haha... so gooood..... would it be possible to vote no ???? AngerFist, Bisqwit, you guys did great work together... YES !!!

Wily over. Move = Very Yes. I'm speechless, really. Great job.

great job guys

tool23, I was going to congratulate you for actually hacking that in, until I realized the text was misaligned and figured out it was fake :(

Okay, who voted no?

Uh... oops? Sorry, misvoted.

... This rom should file for abuse and tool-assisted assault. :[ (Yes vote.)

Why would you watch a movie of a game you don't like? I'm asking because I'm curious.

I was intrigued by the comments. Besides, Bisqwit said that "the next Megaman movie will be MAD", or something like that. Well, it is mad enough. :D

Regarding entertainment, each person is different.

Absolutely.

Unforunetly, it seems others don't notice it, Bisqwit himself told me he didn't notice the faster movment, he only knows it was faster movement because of the tools he made :(

Well I'm not surprised then. I don't know Megaman good enough to say whether he's moving at a normal speed, that kinda reduces the resulting effect. :| On the other hand, I don't like the game quite enough to educate myself further for that kind of issues. P.S. By the way, this one seems to be the most appreciated movie on the site for now. 52 (actually, 53) yes votes out of 53 total is awesome. Is there a game which could, if properly TASed, gather in the similar harvest of votes?

Well, I wouldn't say Meggy's "moving faster", but rather just moving along a slightly better path.

Well, I wouldn't say Meggy's "moving faster", but rather just moving along a slightly better path.

He does also move faster, because I ensured that every time*Rockman starts moving horizontally from a standstill (like coming ahead an obstacle or clearing a ladder), the subpixel part of his horizontal position is optimal. For example, if there is a wall at positions 64...128, and Rockman is descending down from the left side of the wall, by the time he clears the wall, his X position is 63.95, rather than 63.10 for example. You can't observe the difference from what you see within the first few frames, but you'll know that you reach some place a frame faster than normally. Similarly, when going left, it's better to wait at 129.00 than at 129.35, when the direction is going to be to the left. If you hold right while facing a wall, Rockman's subpixel X position will fluctuate, but the integer part will be kept at constant value. (Acceleration works even when facing an obstacle; the position is fixed separately.) For example, assume Rockman's X speed is 1.3 pixels / frame. When you hold right while facing a wall, the coordinate may go through a sequence like this: 　50.0 (50.0 if there was no obstacle) 　50.3 (51.3 -"-) 　50.6 (52.6 -"-) 　50.9 (53.9 -"-) 　50.2 (55.2 -"-) 　50.5 (56.5 -"-) 　50.8 　50.1 　50.4 　50.7 　50.0 Knowing this effect, you can choose to stop movement when the coordinate is 50.9, and then start moving when it's possible for the coordinate to reach 52.2 at once. (Or rather, because of the acceleration delay, time it so that it reaches 50.9 just before the edge and immediately 52.2 thereafter.) Of course, you need to be able to read (and interpret) the RAM to do this, because you can not see the subpixel position from the screen (pixel is the smallest display unit). (* Actually, not every time. I realized this opportunity about half-way to the movie.) Ps: This same thing applies to Rockman 2 as well with exactly the same numbers.

I really should have replied here earlier. I just watched the movie, and what can I say? It exceeds my expectations, even though I watched pretty much every movie in the Mega Man discussion topic. Rockman v6 was the best example of a tool-assisted speedrun, and this version is nothing less. There is really nothing to criticize!

Absolutely a beautiful run. Yes even though my vote hardly matters at this point.

Awesome run! Especially the way Bisqwit entered Bombman's room. The funny thing is that I cursed that I don't have a working FCEU, but then I realised that Bisqwit gave me the sources some weeks ago :) Now I'm just waiting for an encode to be archived. Hopefully the encode is good, unlike some encodes lately.

Hopefully the encode is good, unlike some encodes lately.

Although it is off-topic, which encodes have not been good lately?

This is the source code of the emulator extension I used in the making of this AVI (not in the making of the movie). It makes the magnet beams look much more solid.

void RockmanBeams()
{
    unsigned a;
    int scrollbegin = RAM[0x1B]*256 + RAM[0x1A];
    //for(a=0; a<256*240; ++a) XBuf[a] = (XBuf[a] & 15) | 192;
    if(RAM[0x1E] != 0) return; // vscroll
    for(a=0; a<16; ++a)
    {
        unsigned power  = RAM[0x5B0+a];
        int      xmid   = RAM[0x5D0+a]*256 + RAM[0x5C0+a];
        int      ypos   = RAM[0x5F0+a];
        unsigned length = RAM[0x5A0+a];
        unsigned units = 1+length/4;
        if(!power) continue;
        
        int screenbegin = (xmid - scrollbegin) - units*8;
        int screenend   = (xmid - scrollbegin) + units*8;
        
        if(screenbegin < 0) screenbegin = 0;
        if(RAM[0x1A] != 0 && screenend   > 255) screenend = 255;
        if(RAM[0x1A] != 0 && screenbegin > 255) continue;       
        if(screenend   < 0) continue;
        
        if(RAM[0x1E] != 0) ypos -= RAM[0x1E]-256;
        
        while(screenbegin <= screenend)
        {
            unsigned x=screenbegin&255;
            int y, ycount=0;
            for(y=ypos-2; y<ypos+4; ++y,++ycount)
            {
                int color=12;
                if(screenbegin > 255)
                {
                    continue;//if((x|y)&1)continue;
                }
                if(y < 0 || y > 239) continue;
                switch(ycount)
                {
                    case 0: color=13; break;
                    case 5: color=13; break;
                    default:                
                    {
                        int offs = y+screenbegin+RAM[0x1A];
                        int hi=0x2C, lo=0x11;
                        //if(power==156 && (RAM[0x23]&1)) hi=0x3C;  
                        //  -RAM[0x23]/8; // frame counter for visfx
                        color = (offs&2) ? hi:lo;                   
                    }
                }
                //if(((x^y)&1) != (RAM[0x23]&1)) continue;
                XBuf[x + y*256] = color | 128;            
            }
            ++screenbegin;
        }
    }
}

This function is placed in video.c and called from FCEU_PutImage() before any other line. (Actually, this is not precisely the same function; the actual code uses a copy of the RAM that is updated on every NMI (to prevent it from predicting the magnet beam locations on next frame), but the difference is small and this one was easier to post.) It has a visual quirk: the decorations in the real magnet beams go \\\\\\\\, but the decorations of this filter go ////////. It renders the beams more properly than the game, so during some zippings you may see the game's rendered beams appearing erroneously in the next screen, due to wrapping.

Hopefully the encode is good, unlike some encodes lately.

Although it is off-topic, which encodes have not been good lately?

Sorry, I think I said it too harshly, but if I remember right there were one or two lower quality encodes this year (not the DKC ones). I'll investigate, I may have already deleted the videos or I may have been really drunk while watching them. I'll tell you in PM or IRC or something, if I find out.

P.S. By the way, this one seems to be the most appreciated movie on the site for now. 52 (actually, 53) yes votes out of 53 total is awesome. Is there a game which could, if properly TASed, gather in the similar harvest of votes?

The Mario 64 20-star TAS was getting just as many votes when it first came out.

I actually think there should be an entirely glitch-free version of this game available too. I don't care much about this one, but I'm sure it will be accepted and obsolete the old one (which also had way too many glitches to be interesting for me). I'll refrain from voting.