General Informations

  • Emulator Used: FCE Ultra 0.98.16
  • Aims for fastest time
  • Movie ends as fast as possible
  • Abuses programming errors
  • Takes damage to save time
  • No cheats/passwords
Throughout the most part of the run, I abused the Blade/Item Boost (see below) maily with the ninja blade, and picked up all the shurikens and bombs on my way that were easy to reach. I tried my best to optimize the Blade/Item Boost movements in this run, but I think it could be even better.

Version differences

This game is know as Kage in Japan, and it is also know as Shadow of the Ninja in the USA and Blue Shadow in the Europe. There are (J), (U) and (E) releases, respectively.
Besides the text, the differences among the (U), (J) and (E) versions as far as I know are:
  • The (J) and (U) versions were designed to run in NTSC mode, and the (E) version was designed to run in PAL mode.
    • All the movements in the PAL version were adjusted for lower frame rate when the game was ported, in order to make the PAL version feel as fast as the two NTSC versions. However, they weren't adjusted perfectly due to discrete (integral in this case) mathematical problems. As a result, some movements were made faster (eg. the ROF of the blade weapon is once per 10 frames in the PAL version, while that in the NTSC versions is once per 13 frames), while some were made slower (eg. the walking speed is slightly slower in the PAL version.).
    • For the same reasons as that above, some movements result slightly differences in the PAL version. For example, you can jump so high in the PAL version that the enemy on the box at the end of Stage 1-2 can be jumped over from the ground, as well as two-man higher places can be reached without using the edge-sinking glitch, which are something impossible in the NTSC versions.
  • Level designs were modified during both the (U) and (E) porting. These modifications have changed the difficulty. Overall, the (J) version is the most difficult. Some of the modifications are described below:
    • Some enemies' strengths were adjusted eg. the seamen in Stage 1-1 were made weaker in the (U) and (E) versions, as their hit points were reduced from 2 to 1.
    • Some enemies in the (E) version were replaced eg. in the middle part of Stage 1-1, the monkey in the (J) and (U) versions was replaced by a droid in the (E) version.
    • It is said that the bomb weapons in the (E) version were made not so efficient as in the (J) and (U) versions against bosses. [source requested]
    • Some terrains were modified eg. in Stage 4-2, one conveyer in the first shaft was made longer and higher in the (E) version, while a rising platfrom before the last shaft was added in both the (U) and (E) versions.
  • The (J) version has the fastest intro. It is about 10 seconds shorter because the intro is only played once before you can press the start button to play the game in the (J) version rather than twice in the other two versions.
There could be more differences than I've listed here. I chose to make the movie with the (J) version in favour of the NTSC timing and shorter non-motion periods (mostly the intro, which I think plays twice would be boring).

Some tricks for the NTSC versions

The Odd Rule

In this game, when you are walking on the ground or jumping forward, your horizontal acceleration is 2. When you release the D-Pad, your horizontal acceleration will be -4 on the ground or -1 in the air, which is decelerating your velocity. If your horizontal velocity is going to be below 0, the game will set it to 0. If your horizontal velocity is going to be over 25, the game will set the accelation to 0 and keep it below 26. Therefore, if your current horizontal velocity is even, the game will keep it 24, while if your current horizontal velocity is odd, the game will finally keep it 25. Thus it's better to keep your horizontal velocity odd when you are moving. This is called "the Odd Rule" by me.

Blade/Item Boost

A trick called "Blade Boost" (also by me) features in this movie. When you are jumping forward with high velocity, pressing the B button on the controller to swing your blade or throw your item (a bomb etc.) will set you to a special status, where your horizontal velocity is boosted by about 90% and it can exceed 25, while your horizonal acceleration is limitted to -2 for a few frames, until the game realizes it isn't good and suddenly drops your horizontal velocity heavily and remove the acceleration limit. This procedure can be repeated as much as you want as long as you are in the air. The faster your weapon is, the shorter the cycle can be, and the higher average speed you can get. The ninja blade is the fastest main weapon in this game, whose ROF is one swing per 13 frames. The chain weapon has different ROFs in each grade, which are 1/26f in the firt grade, 1/21f in the secong grade and 1/16f in the last grade. The shurikens and bombs are both as fast as one throw per 4 frames. Since the higher the speed is when you perform this trick, the larger the 90% boost can be, it is best to follow "the Odd Rule" to get the maximum boost.
However, this isn't over yet. As soon as you land, your horizonal velocity will always drop down suddenly. Depending on your moving state before landing, the landing velocity varies.
1) Landing as boosted(still unable to accelerate yet) will halve your velocity.
2) Landing as not boosted with Left/Right Button pressed will temporarily set your velocity to 0 for one frame, then it restores your velocity to how great it should be with acceleration.
3) Landing as not boosted without Left/Right Button pressed will decrease your velocity by 1.
Obviously Case 2 should be avoided when moving forward because Case 3 is always better than it. However, Case 1 or Case 3 may result in an even velocity, and in Case 1 the landing velocity may be decelerated too much. Therefore, you may need to wait for several frames before doing Blade Boost or release Left/Right Button before landing to get an optimal landing velocity. Just like in other games that require constantly jumping for a TAS, you have to adjust the length and maximum height of your jumps according to the terrain everywhere, and it gets harder when you are TASing a game like Kage involved in such complicated accelerate mechanism.

Saving items while making boost

I've found a way to save bombs and shurikens when still get boosted. Just fire them 3 frames before landing. So you may see that I was "just hopping there" after I got bombs in Stage 2-2, but I did got boosted there. However I didn't jump much in narrow passages because of the dropping down of the horizontal velocity when landing.

Sinking into the wall

As far as I know, there are two ways to sink into the wall:
  • Climbing up platforms attaching to the bottom corner of the wall.
Jump to the joint as the "-L" part in "-L___"("-" illustrates the platform and "L___" illustrates the bottom corner of the wall), then climb up as soon as you grab onto the platform while holding down the direction on the D-Pad towards the wall.
  • Falling onto the wall.
The game assumes that you are landing on the ground if you jump and fall to some internal margin. There will be a one-frame window to jump up before you are getting ejected out.
The first way allows you to be ejected upwards very fast through the wall. The second way allows you to reach higher places than expected.

Utilizing invulnerability

When you take damage or pick up an item/power-up, you'll be invulnerable for a while. You can use this to go through enemy without getting hit.

Another way to go through enemy safely

Once an enemy is hit, it will be temporarily invulnerable and unable to hit you for several frames. You can go through it without getting hit, if you are fast enough.

More about items/power-ups

Once you pick up an item/power, bisides temporary invulnerability, it recovers you from some bad status as well eg. being stunned.

More tricks?

There are more tricks in Kage, but I don't want to write a longer discription than this. If you are interested, please feel free to PM me and you'll receive the answer once I've read the messages.

Major improvements

After reading the tricks expalined above, these items should be easy.
  • Faster way of moving.
  • Better strategy of using items for faster movements and boss fights esp. in Stage 2-3, or getting invulnerable to avoid damage (eg. that near the end of Stage 3-1).
  • New zipping tricks for shortcuts in many stages.
  • Reducing more lags by making certain input(hard to find them out during watching, though) and killing certain enemies.
  • Using an unknown glitch to kill some enemy quickly.
  • Reducing waiting time eg. the gap between the gas in Stage 2-1.
  • New shortcuts in Stage 3-1, Stage 4-2 and Stage 5-1.
  • Scrolling the screen better in Stage 4-2 and Stage 5-1.

Drawbacks

  • Having to kill the unskipable monkey in Stage 1-1 rather than the skipable droid there in the (E) version.
  • Having to wait for two moving pads in Stage 1-3. For the first one it was due to weaker jumping ability, and for the second one it was a side-effect of the the first one that I had to wait for the upper pad to show up.
  • Longer boss fights when only the blade weapon is available.
  • For some reasons, some enemy behaved differently and I had to spend more time in dealing with them eg. the mortar soldier one-shot-killed in Stage 4-1 would but shouldn't block the power-wave if I had attacked him on the right, which should have finished off the three of them in two shots.

Potential improvements

  • It can be found that I didn't perform the Blade Boost trick each time I climbed up a platform. I just forgot to do that. Had I done that, it might be either several frames faster now or not, depending on some frame-rule like that in SMB1 and SMB2j.
  • In Stage 4-1, it is possible to sink into the right side of the highest ladder and jump up from there, which can save some frames.

NesVideoAgent: Hi! I am a robot. I took a few screenshots of this movie and placed them here. Oh! I also replaced the ROM name.
  • You indicated Kage (J) [!].nes
  • I updated it to Kage (J).nes

Truncated: Hello klmz, and welcome here.
We already have a movie published for the european version of this game, [856] NES Shadow of the Ninja (Europe) by Xipo in 11:14.56, which regardless of being recorded in the wrong format is faster than this (33728 frames, yours is 35299). In any case you seem to have found a nice glitch, zipping upwards in the wall.
I recommend remaking this movie, on the (U) version, looking at the published movie and incorporating your trick. This one will be rejected.

Truncated: Submission unrejected, see post for details.

adelikat: After discussion we have a new verdict. This movie will be accepted and will obsolete the published blue shadow movie.
Also, any future submissions will be measured from the beginning of the actual gameplay in terms of speed. As a result a theoretical U version submission will not be "penalized" the extra 625 frame it takes at the title screen. All 3 versions (E, U, and J) are "fair game" provided that they are faster using this means of measurement. Also, the E version must run at 50fps.


TASVideoAgent
They/Them
Moderator
Joined: 8/3/2004
Posts: 15586
Location: 127.0.0.1
This topic is for the purpose of discussing #1653: klmz's NES Shadow of the Ninja (Japan) in 09:48.32
emu
Active player (386)
Joined: 1/2/2007
Posts: 188
Location: Germany, Munich
Good game choice and well executed. yes vote edit: I didn't pay enough attention, but didn't you die once? (just for the description text) edit: Nevermind. I didn't know falling into a pit only costs some HP, not a life.
Joined: 2/1/2007
Posts: 245
Location: Israel
This game looks a lot like Blue Shadow, which was completed in less time. Is this a different game?
Skilled player (1605)
Joined: 6/11/2006
Posts: 818
Location: Arboga, Sweden
This is the J-version of Blue Shadow, as far as I can tell.
Warp wrote:
omg lol this is so fake!!!1 the nes cant produce music like this!
Editor, Emulator Coder, Expert player (2157)
Joined: 5/22/2007
Posts: 1134
Location: Glitchvania
Mechuyael wrote:
This game looks a lot like Blue Shadow, which was completed in less time.
Both two of the submissions for Blue Shadow werecorded in NTSC opposite to PAL the proper one. Therefore, their lengths should be multiplied by 60/50=1.2 to be fair, if they are considered as valid records.
<klmz> it reminds me of that people used to keep quoting adelikat's IRC statements in the old good days <adelikat> no doubt <adelikat> klmz, they still do
adelikat
He/Him
Emulator Coder, Site Developer, Site Owner, Expert player (3573)
Joined: 11/3/2004
Posts: 4754
Location: Tennessee
A different version of the game does not constitute a new game (except in some rare circumstances). Comparing to the published , it is 1571 frames slower. And on top of that you say the intro in your submission is 10 seconds faster in your version. This movie is slower, and that is evident when watching the movie. I vote no for failing to improve upon the published movie.
It's hard to look this good. My TAS projects
Editor, Emulator Coder, Expert player (2157)
Joined: 5/22/2007
Posts: 1134
Location: Glitchvania
The Rule wrote:
The PAL/NTSC settings must be correct If you are running a NTSC game (U or J), you must set your emulator to record in NTSC mode. Likewise, if you are running a PAL game (E), you must record your movie in PAL mode. Any other setting will get your movie promptly rejected. For this reason, PAL games are not allowed for Famtasia submissions, because it can only record in NTSC. If you are obsoleting a movie that was recorded with the wrong settings, that movie's completion time will be increased 20% when comparing them to adjust for the fact that PAL games run at 50 Hz and NTSC at 60 Hz. This rule only comes into effect on NES games and possibly Genesis games that have badly programmed region lock-out.
Shouldn't a right one beat a wrong one?
<klmz> it reminds me of that people used to keep quoting adelikat's IRC statements in the old good days <adelikat> no doubt <adelikat> klmz, they still do
Editor, Emulator Coder, Expert player (2157)
Joined: 5/22/2007
Posts: 1134
Location: Glitchvania
Truncated wrote:
Hello klmz, and welcome here. We already have a movie published for the european version of this game, 856M, which regardless of being recorded in the wrong format is faster than this (33728 frames, yours is 35299). In any case you seem to have found a nice glitch, zipping upwards in the wall. I recommend remaking this movie, on the (U) version, looking at the published movie and incorporating your trick. This one will be rejected.
Objection! How can you simply campare a movie for a PAL game with a movie for an NTSC game by their frame counts other than their real lengths? If so, ANY speedruns for the PAL versions are definitely shorter than those for NTSC versions, and them are eternally INVINCIBLE to the NTSC ones simply because them are 50Hz rather than 60Hz and you only consider their frames! Edit: It's even more impossible for a 10 seconds longer run for the USA version to beat the invincible PAL-game-and-recorded-in-NTSC run publised.
<klmz> it reminds me of that people used to keep quoting adelikat's IRC statements in the old good days <adelikat> no doubt <adelikat> klmz, they still do
adelikat
He/Him
Emulator Coder, Site Developer, Site Owner, Expert player (3573)
Joined: 11/3/2004
Posts: 4754
Location: Tennessee
klmz wrote:
The Rule wrote:
The PAL/NTSC settings must be correct If you are running a NTSC game (U or J), you must set your emulator to record in NTSC mode. Likewise, if you are running a PAL game (E), you must record your movie in PAL mode. Any other setting will get your movie promptly rejected. For this reason, PAL games are not allowed for Famtasia submissions, because it can only record in NTSC. If you are obsoleting a movie that was recorded with the wrong settings, that movie's completion time will be increased 20% when comparing them to adjust for the fact that PAL games run at 50 Hz and NTSC at 60 Hz. This rule only comes into effect on NES games and possibly Genesis games that have badly programmed region lock-out.
Shouldn't a right one beat a wrong one?
Only if it is a superior movie.
It's hard to look this good. My TAS projects
Player (88)
Joined: 1/15/2006
Posts: 333
Location: Bangkok, Thailand
adelikat, in xipo's movie description wrote:
This movie uses an European rom but runs it at 60fps. This is unacceptable. But the previous movies were done & published this way too so it is unfair to the author to reject. I am accepting this run. But any future run that is done in this manner (including future Blue shadow submissions) way will be rejected. If using an E rom, it should run in PAL mode.
Had this run also been recorded in this manner, it would have a finishing time of 8:10.27, which clearly beats the existing movie.
print reduce(lambda x,p:p/2*x/p+2*10**1000,range(6643,1,-2))
Editor, Emulator Coder, Expert player (2157)
Joined: 5/22/2007
Posts: 1134
Location: Glitchvania
adelikat wrote:
Only if it is a superior movie.
Just open the Blue Shadow ROM with FCEU 0.98.16 and load the published movie, then look at the "Length: 00:11:15". Isn't it longer than 00:09:48?
<klmz> it reminds me of that people used to keep quoting adelikat's IRC statements in the old good days <adelikat> no doubt <adelikat> klmz, they still do
Joined: 2/1/2007
Posts: 245
Location: Israel
Things like longer opening animations and such are taken into account when looking at new submissions. As long as another version doesn't have some glitch/mechanism to speed the game itself up, it's preferred to use the US version. Edit: Whoops, forgot that if another version is more difficult, then that one is chosen.
Senior Moderator
Joined: 8/4/2005
Posts: 5777
Location: Away
klmz wrote:
Just open the Blue Shadow ROM with FCEU 0.98.16 and load the published movie, then look at the "Length: 00:11:15". Isn't it longer than 00:09:48?
It's longer just because it plays slower. If you discard the difference in playback speed, it's 9:22.
Warp wrote:
Edit: I think I understand now: It's my avatar, isn't it? It makes me look angry.
Player (88)
Joined: 1/15/2006
Posts: 333
Location: Bangkok, Thailand
moozooh wrote:
klmz wrote:
Just open the Blue Shadow ROM with FCEU 0.98.16 and load the published movie, then look at the "Length: 00:11:15". Isn't it longer than 00:09:48?
It's longer just because it plays slower. If you discard the difference in playback speed, it's 9:22.
I think that's exactly the point here. One should not and cannot disregard the difference in playback speed. The current movie is PAL (E) and was recorded in NTSC (J U). Had it been recorded in PAL mode, as it should have been, it would have had a length of 11:15.
print reduce(lambda x,p:p/2*x/p+2*10**1000,range(6643,1,-2))
Editor, Emulator Coder, Expert player (2157)
Joined: 5/22/2007
Posts: 1134
Location: Glitchvania
moozooh wrote:
klmz wrote:
Just open the Blue Shadow ROM with FCEU 0.98.16 and load the published movie, then look at the "Length: 00:11:15". Isn't it longer than 00:09:48?
It's longer just because it plays slower. If you discard the difference in playback speed, it's 9:22.
Then just play my movie in 6400% speed, isn't it 00:00:09?
<klmz> it reminds me of that people used to keep quoting adelikat's IRC statements in the old good days <adelikat> no doubt <adelikat> klmz, they still do
adelikat
He/Him
Emulator Coder, Site Developer, Site Owner, Expert player (3573)
Joined: 11/3/2004
Posts: 4754
Location: Tennessee
klmz wrote:
moozooh wrote:
klmz wrote:
Just open the Blue Shadow ROM with FCEU 0.98.16 and load the published movie, then look at the "Length: 00:11:15". Isn't it longer than 00:09:48?
It's longer just because it plays slower. If you discard the difference in playback speed, it's 9:22.
Then just play my movie in 6400% speed, isn't it 00:00:09?
Isn't this an argument against YOUR opinion? What we are trying to say is the published movie if faster in terms of number of frames. This difference is a result of more optimized gameplay than your movie. This is more important than the final time real time length. This real time length discrpency is simply from playing the movie faster or slower. You made a worse movie that plays back faster. This does not consitute an improvement. Also this argument is moot because you are failing to understand that there is a problem with the Blue Shadow publication (which could have been avoided had we not made this mistake). While it says it is an E rom, it was played and recorded at 60 fps. Your FCEU gives you a final time of 11:xx because it assumes the movie should be played back at 50fps even though that is an incorrect assumption. An E rom that plays at 60fps should not have been published on the site. A mistake was made. But we can't correct this mistake by putting up a movie that is clearly inferior gameplay.
It's hard to look this good. My TAS projects
Senior Moderator
Joined: 8/4/2005
Posts: 5777
Location: Away
klmz wrote:
Then just play my movie in 6400% speed, isn't it 00:00:09?
And if you play the published movie at 6400%, it's 00:08'47. The point, however, is that you failed to improve upon the published movie. Even if you had just taken all the input from it and applied it to an NTSC game, it would have been 9:22. Yours is 9:48, 26 seconds slower than what's supposed to be optimal. How do you explain the difference?
primorial#soup wrote:
I think that's exactly the point here. One should not and cannot disregard the difference in playback speed. The current movie is PAL (E) and was recorded in NTSC (J U). Had it been recorded in PAL mode, as it should have been, it would have had a length of 11:15.
Movies between NTSC and PAL-unoptimized game versions have to be compared under same conditions. Otherwise, one could have taken the NTSC version of the game, run it in 11:14 and claim it's faster than the current movie, which is actually ridiculous.
Warp wrote:
Edit: I think I understand now: It's my avatar, isn't it? It makes me look angry.
Editor, Reviewer, Experienced player (979)
Joined: 4/17/2004
Posts: 3109
Location: Sweden
I am currently investigating at which speed these games run compared to each other. It shouldn't take too long, so please stay calm until then.
Tub
Joined: 6/25/2005
Posts: 1377
klmz wrote:
Then just play my movie in 6400% speed, isn't it 00:00:09?
now you're just being stupid. NTSC plays at 60 frames per second. PAL plays at 50 frames per second. Therefore, a PAL game will usually run slower than the NTSC version. This is observable on real consoles. On some games, the programmers increased the movement speed in the PAL version to counterbalance the effect, but that doesn't seem to be the case here (if you think otherwise, please share your research). Since the previous movie played a PAL game at NTSC speed, it should run as fast as any other NTSC version. There are differences like the faster intro you mentioned, but the general gameplay should be unaffected. Again, if you think otherwise, please share your research. The point of these movies isn't actually to produce a fast movie, but to produce an entertaining movie. Even if your 9.48 is faster than the published movie had been at PAL speed (11:15), the frame-count suggests that your gameplay is sloppier and thus less entertaining. Since you didn't provide frame comparisons that's a valid assumption. I haven't watched your movie yet, FCEU + wine is giving me trouble as usual.
m00
Editor, Reviewer, Experienced player (979)
Joined: 4/17/2004
Posts: 3109
Location: Sweden
Okay, here's the deal: >On some games, the programmers increased the movement speed in the PAL version to counterbalance the effect, but that doesn't seem to be the case here (if you think otherwise, please share your research). According to my tests, they DID increase the movement speed when porting game, but not enough. In real time, the character in (J) on NTSC still runs faster than he does on (E) on PAL, but not by a lot. In (E) on NTSC, he then naturally runs a LOT faster than (J) NTSC. This makes it highly questionable wether we should apply the 20% time bonus for the current run (see the Rules on this), which would actually make the published run's correct completion time around 11 minutes. After discussing this a bit with adelikat, we decided that: 1) for this game (and others with adjusted movement speeds), time is compared in seconds rather than frames. 2) for this game ONLY, use whichever version you like, J U or E. We still favor U versions, and choosing the J version just because of a faster intro is not a good idea. Submission will be unrejected for now.
adelikat
He/Him
Emulator Coder, Site Developer, Site Owner, Expert player (3573)
Joined: 11/3/2004
Posts: 4754
Location: Tennessee
I'd like to be extra clear: You can use J, U , or E roms. But the E rom must be run at 50fps or else be rejected. The better movie will be decided by its final time in seconds.
It's hard to look this good. My TAS projects
Active player (411)
Joined: 3/16/2004
Posts: 2623
Location: America, Québec
Imo, using E is not the good way. I didn't watched this run nor xipo's one. I can't tell which one is better. What I am sure is that movies using PAL version of any games did always confused decisions and such. If it's proven that this run is really faster than xipo's run, even if there is some flaws, I suggest it to accept it. But I have a question. Which one does have the fastest intro? J or E?
Skilled player (1605)
Joined: 6/11/2006
Posts: 818
Location: Arboga, Sweden
J. E "waits" at the title screen for a good amount of time before the player seems to be allowed to press start, while J allows the player to press start almost instantly, at least judging from the movies (I haven't tested yet, but I assume that the movies are optimal during the title screens)
Warp wrote:
omg lol this is so fake!!!1 the nes cant produce music like this!
Editor, Emulator Coder, Expert player (2157)
Joined: 5/22/2007
Posts: 1134
Location: Glitchvania
Thanks to Tub and Truncated, the truth is known. I thought you judges should have known that all the movements in the (E) version of this game has been adjusted when porting. I was wrong, so it was I who was being stupi*(just to stay civilized here) in arguing with you without sharing my research as Tub said. It was all my fault, right? So I was taught a good lesson: never to overestimate others too much, especially those who make decisions. Well I'll add some stuff about the differences among the three versions to the Submission Description later. Proceed with your work. Edit: Sorry for being rude. I apologize.
<klmz> it reminds me of that people used to keep quoting adelikat's IRC statements in the old good days <adelikat> no doubt <adelikat> klmz, they still do
Player (88)
Joined: 1/15/2006
Posts: 333
Location: Bangkok, Thailand
Understand that the judges can't possibly be familiar with every game in existance. This is particularly true when dealing with the differences between PAL and NTSC versions of a game. The fact that you had initially made no reference to the differences or even to the publication you were trying to obsolete didn't help matters either. I think this could have avoided a lot of the confusion. Also, it is not a given that PAL versions run faster than NTSC versions, many are adjusted, but many are not. With absolutely no information to work with, it was judged solely by frame count, and was found to be slower. Truncated should not have had to do the research he did; upon submitting a movie, the author should be well-informed and be able to state clearly why their submission is an improvement to an existing movie. Anyway, it seems that everything has been cleared up. I'll leave some comments about the actual submission after I have time to review both.
print reduce(lambda x,p:p/2*x/p+2*10**1000,range(6643,1,-2))