Oh yeah, that's right, you'd have to do one or the other. You'd either need to lure an enemy in and ice it, or make judicious use of bomb jumping and regaining control of Samus by standing back up in midair.
Probably best to beat both Ridley and Kraid, that way you're also ensured to have enough missiles to beat all the Zeebetites and the MB. Still, with no ice beam, I expect Tourian to be quite impressive when you're finished. I also expect it to be a bit slower since you can't freeze the Rinkas and stand on them.
Well, as expected, Tourian is proving to be impossible. I meant collecting 1 item - the Maru Mari. The missles, I would of course get from the bosses. It can be done, but it's frightfully trying. I can fill up my missles and energy before I get to Tourian from random(ish) enemies, and I only use up 15 on the doors, but the metroids are tough to avoid and energy is scarce. I'll get it, and I hope I don't just break down and grab the ice beam - would be easy as pie if this were a 2 item run.
Yes, please don't, even with a mega-high rerecord count it would be a hella impressive run, to have just the Maru Mari and missiles. It would be a bit more impressive to be able to do it after beating only Kraid (and thus only having 75 missiles), but I'm thinking that would cut you a bit short on missiles in the final battle room.
Actually, I wouldn't imagine it'd be that hard to do with a low rerecord count. Also, at the slower speed, it should be possible to jump between red spheres like crazy to take shortcuts from one side of the map to the other (I know it's possible, I've done it occasionally in realtime), so it should also be quite impressive.
I thought at least 80 missiles were required to finish the game. Ice would come in handy for the Metroids, but might not be required (and would definitely make Tourian more impressive, especially without the ability to freeze the Rinkas in the MB's room to stand on them and fire).
Tool-assisted players are only 'cheating' when compared to realtime players. As long as people continue to try to compare these time attacks with realtime runs, you will have people who complain about it.
But, they are in a different class. Since the tool-assisted scene has set up different rules from realtime speedruns, then I don't consider it 'cheating' because everyone is playing by the same rules. Whether or not savestates and slowdown is cheating depends on the rules that a group is going by, which is why I fear this issue will never be resolved. According to us, it's within the rules, so it's not cheating. According to the realtime speedrunners, it's not within the rules, so it's cheating.
I still feel that we need to have someone work on a way to run these movies on real hardware - that should dispel all notions of them being 'faked'. Especially to the people who see the wall jumps and hat stomps in the SMB vids and scream 'OMG U CAN'T DO THAT FAKE FAKE FAKE'. I don't have the skill to start such a project, but I bet someone out there does =P
Well, every so often, the DVDs could be redone with newer runs. My goal isn't to archive ALL runs on DVD, but merely a series (somewhere between 3 and 5 DVDs) that includes the best of the best at that particular time. Then, like once a year, I could release a single-disc (or possibly even two-disc) 'Best of the last year'. There's no need to try to figure out a way to completely get rid of obsoleted runs, so an 'update' disc should work well, with basically the same information that's already provided on Bisqwit's site for each movie. I'll probably even include the actual emulator movies on the disc as well (since the file sizes are miniscule, this will barely affect the total video length, if at all). I'm not sure whether it'd be a good idea to include the actual emulators - while on the one hand, it would mean that the user would only have to supply the ROM, it also means that people not actually in the scene might possibly get stuck with a buggy version of an emulator after a new version is released.
For the first release, I'd choose the runs to include by popular vote on this forum. I'll be using MPEG-2 VBR at about 8000kbps, and probably AC3 audio (since the runs are done with NTSC ROMs and 60fps Famtasia, the DVDs will be NTSC, so I won't be using MPEG Layer 2 audio), so it should be fairly easy to compute an approximate amount of video to fit on the disc. Some runs might be best to dedicate to their own DVD (or, for example, have a Metroid series DVD), but the shorter runs should give us quite a bit of room.
Any input on this idea is welcome. All people who contribute to the production of the DVDs will of course be credited. As I mentioned earlier, I will be writing up my method for generating these videos, and publishing it here for others to take advantage of as well. Currently, the software I'm using is VDubMod with LAME CBR DLL, AviSynth, TMPGEnc for encoding and Adobe products (Photoshop, After Effects, Encore DVD) for the DVD interface production.
First of all there is dvd:s (at least here in Sweden) with data you can burn to maximum 4,7 gig, 8 gig, 20 gig, and the current buy-dvd is 40gig so you can put almost every "interesting and popular runs in it". Secondly, you cant just say why dont you just do a .iso file and share it to others because he said he wished to create a dvd movie WITH dvd menues etc and for that you have to create at least 4 SFO-file which is a sort of a link-file which links all the existing file in a dvd to eachother so when you for instance choose a file you can through that SFO file make that choice, and to do that requires alots of time (been there done that and I have a dvd-burner). And the most important thing, you wont be able to never (at least at this stage) share that .iso or .img file because if its at least 4,7gig how do you expect to have any seeds and people willing letting their computers be on for DAYS until you have completed the downloading-process? Because I assume aint alot of people here with decent speed and I certainly wouldnt have the time to download just a single file for days, certainly not worth it.
Well, this is just the early planning stages for such a project, as A) I am waiting for dual-layer burners to come down in price a bit, and B) I need an external hard drive to record the interim AVIs, as it requires roughly twice as much space as the interim AVIs for making Divx/Xvid/whatever videos (because I have to store the AVI as generated by the emulator, and then I also have to store another AVI of similar size to hold the weaved version that will then be fed to an MPEG-2 encoder). As I use either Huffyuv or MJPEG at highest quality and full color resolution, these files do indeed get huge very quickly. I would then have enough space on my laptop's internal drive to build the DVD file structure (VIDEO_TS folder, .IFO, .BUP, .VOB files) and burn it to an external DL writer. As I'm going for best possible quality, I don't want to do it on just 4.7GB media.
These files wouldn't be offered for download by me (although anyone would be free to do so if they choose), it would be offered on DVD. I'd either go free, or media cost only (because I'd imagine DL media is quite expensive at this time - I even saw Circuit City selling an internal IDE DL writer, but I didn't see them selling any DL media). There would definitely be no profit made without the consent of the creators of all runs involved, of course.
Also, anyone would be free to duplicate and redistribute the DVDs - I'm not going to bottleneck their distribution by being the only one to make them. I don't see why I should, anyway, because they're not my runs.
Of course there would be a menu selection to read a bit more information about these speedruns, sort of like Why&Now. I like the idea of making these speedruns more accessible and this would be an attempt to head off yet more screams of 'fake' 'cheating' 'not real' 'game doesn't allow you to do that' 'frauds'. On a side note, this is the same reason I would support a project with the aim of making these movies run on hardware through some sort of controller interface (although I don't have enough experience to initiate such a project). But I'd figure it'd be possible, unless a movie relies on a Famtasia bug.
Well, looks like I'll have to invest in a large hard drive to be able to produce the longer runs for DVD. I use either Huffyuv or PICVideo MJPEG, whichever gives better file sizes, and I have to use PCM audio for at least the initial AVI generation (although during weaving in VDubMod, I can use 256kbps CBR MP3 to save some space - VBR MP3 in AVI is a broken hack anyway, and should not be used IMO). Right now, I'm doing a test encode of the first few minutes of Frenom's SM run, and the initial AVI's filesize is already near 2GB. I can probably do some work on the shorter runs (like I mentioned before, I already have a full test MPEG-2 encode of Bisqwit's SMB2j run), but anything with a substantial length would have to wait.
I also did a test encode of the first act of Gavin's S3&K run, so I can pretty much apply this method to any classic console with an emulator that supports AVI dumping and controller movie recording/playback. I can try to write up my method if anyone with more free hard drive space wants to take a stab at producing a DVD (or at least the video files that would go on such a DVD). The resizing parameters are slightly different depending on whether you're working with an NES/SNES game with a 256 pixel wide frame or a Genesis game with a 320 pixel wide frame. The only snag I might see would be Genesis games that use both 40 and 32 cell mode (and also, the aforementioned issue with interlaced sources, such as S2 2p mode).
On a side note, using MJPEG for either the initial AVI or the weaved intermediate, combined with the Lanczos resize I do to 256 pixel wide material, gives a slight bit of color artifacting that, on a PC monitor, looks similar to what you might get if you captured SVideo from a real SNES, which I found a bit interesting.
I recently acquired a copy of Adobe Encore DVD, so I'll probably use that to author the menus. I haven't played around with it much, but I understand it has excellent integration with Photoshop and After Effects, which I was planning on using anyway. Gotta use professional-level software to get professional-level results ;)
You know... this might just be me talking but if you wanted and got everybody to OK it, I bet you could pass this around. Personally, having all these on DVD would be pretty rad. I don't know about the legality of everything or how it would work, but I think it would be pretty bitchin to have a collection available.
It'd either be free, or cost of the media only. Due to the amount of work put into these speedruns, I'd think it could possibly be argued that the player owns the copyright over the specific videos (without inferring any control over the characters or music themselves). I'd of course want to distribute it, but it'd be a bit too unwieldy to shuffle around 9.4GB files, thus the consideration of charging the cost of the media (and not a cent more) =P
I also think we might be talking two different things here - I'm talking about actually producing DVD-Video discs that can be played in any standard DVD player. Others I believe are referring to merely a collection of AVIs on a DVD, which is admirable in its own right (since the AVI will be much smaller than the equivalent MPEG-2 for DVD, you could fit many, many more runs in AVI).
Well, I've made a bit of progress on the method I'd use to make video DVDs of the best runs. Using Bisqwit's SMB2j run, I was able to generate a 720x480 MPEG-2 that plays back smooth in DScaler, in several deinterlacing modes, both on the PC and through the nVidia TV-out. Unfortunately, I don't currently own a DVD writer, nor access to a DVD player that can handle miniDVDs, so I can't test it after being actually burned off in DVD format. There was a slight problem with audio sync but this is peculiar to my machine for some reason, it happens with other media and is not related to this process. Now all I need is a dual-layer DVD writer and I'll start work on actually making a series of DVDs. For the record, the total video (spanning from the gray frame before the FDS BIOS screen, to the frame where it goes black after the game is beat and the music finished playing) is just under 400MB (when you consider that DVD DL can hold 9.4GB, that's not a problem at all).
Next is to try the same method with the patched Gens and convert Gavin's S3&K run =P
I think he means games with cheat codes that increase the difficulty. I would think these would be the few cheat codes allowed in these videos, anyway.
i say go for it, I would have already done this if I had a DVD burner. Reduced back to TV resolution these videos probably look very good. A nice menu interface and... oo, i get chills just thinking about it
Actually, it wouldn't be 'reduced back', to get a smooth 60fps playback, you'd actually need to treat the existing NES frames as fields and weave them together, while horizontally resizing to 720 pixels (preferably by computing the correct width to generate a proper NES aspect ratio, then sideboxing to 720 pixels). The weaving would create 480 line frames, which should result in a DVD-ready video once encoded to MPEG-2.
The only possible complication to this method would be the field affinity, although it won't be as much of an issue as it would be with putting Amiga demos on DVD (which is the original purpose for this weaving method), as the NES doesn't generate an interlaced signal. As field affinity really only matters with images sourced from a 480 line source, all we would need to do would be to ensure that whether we denote the first NES frame top or bottom, that we set the resulting MPEG-2 to the same field affinity. Otherwise, you'll have horrendous flickering. Quite a bit of effort, but would produce the best possible quality on DVD. One could also use DVD lowres, but then you'd have to drop half the NES frames.
This would also require redumping the source AVI from the emulator movie, for best possible quality.