New playerset's frames and animation

Jetryl · Post by **Jetryl** » 05 Feb 2008, 06:28

Doubi, please just try and entertain the possibility that I might know what I'm talking about. I'm not some newbie coming here and spouting off; in fact the playerset you're working on was largely made by me (and talaroc), back before I knew how to sprite. Just listen to me - I'm very experienced at this, and I've already made the mistakes you're making, and learned the hard way. Education is far less bitter a teacher than experience - I'm trying to save you from having to do this the hard way. Please just trust me.

The first and foremost thing you need to do correctly when animating something is make sure the gross physics of it look right. This isn't just me saying this - this is the same principle expounded by all the guys at disney, by any of the better anime studios, and by any of the classic books on animation (like the "Animator's Survival Kit"). If the balance of things, and the counteraction of forces is inconsistent, the whole animation will feel stilted and strange. You can bend a lot of rules of natural law - for example, you can have your characters bound 50 feet into the air; but when they do, their bodies still need to recoil and rotate like any human body would from the force and torsion of their jumping. You can change amounts of motion, but in order to change the direction of it entirely, you need to have some sort of visible cause. In this case, the body is moving into a position that would take it out of balance, and then is moving back into balance without any counteracting force; he looks like he is going to fall on his face. It looks bad, and needs to be fixed.

Getting the motion right is far more important than making it cover the correct pixels ingame; however, I think we're raising a false dilemma here. I think you can have the best of both worlds: I think you can have the sprite both balance his body correctly, AND have the weapon pass through the correct game space.

Suggestions:
- Start by making his back foot stay put.
- Use the same blade motion seen in this image:
http://i6.photobucket.com/albums/y203/M ... k_down.gif
- But make the blade bigger, since that guy's holding something that barely qualifies as a knife. The blade should be at least twice as big as that - in fact, we can even go to 3X the size, based on the flavor of the game. Duran, from Secret of Mana 3, had a sword bigger than he was tall, and cloud was quite the same.
- And most importantly, fix the one thing Modanung failed to do - make the body move! If the swordsman lunges forward with his front foot, that sword will pass over the correct game space

Post by **Crush** » 05 Feb 2008, 12:45

I think that the problem here is that Modanung pictured the AoE of the attack not on breast high of the character but on the ground. Just imagine how it would look when the character attacks to the north and you would still try to cover the area of effect based on the feet. The character would have to lean backwards.

Taking this into account the area of effect that should really be covered looks about like that:

Website · Post by **yosuhara** » 05 Feb 2008, 16:07

Crush wrote:I think that the problem here is that Modanung pictured the AoE of the attack not on breast high of the character but on the ground. Just imagine how it would look when the character attacks to the north and you would still try to cover the area of effect based on the feet. The character would have to lean backwards.

Taking this into account the area of effect that should really be covered looks about like that:

i think this pretty much solve the problem with character leaning too much forward... but now it requires new slash attack animation... maybe some frames of old ones could be re-used

Post by **Dave** » 05 Feb 2008, 17:43

I'm all for increasing the size of the sword, but I'm not sure about making it that much taller than the character.

Maybe we should obtain some pictures of someone actually going through some of these motions.. though I suppose the camera would need to be above them to capture something close to the game's perspective. ..

So the sooner you guys get it right, the less chance that I'll find myself on a ladder, taking pictures of some reluctant volunteer, swinging a wooden sword around.

..but seriously, I guess the original animation looks alright, but his slash could cover a bit more horizontal space.

Whatever the case, you wouldn't hold a sword like this when attempting a slashing motion. That frame looks more like he's about to thrust than slash.

Jetryl · Post by **Jetryl** » 05 Feb 2008, 21:07

dabe wrote:Maybe we should obtain some pictures of someone actually going through some of these motions.. though I suppose the camera would need to be above them to capture something close to the game's perspective. ..

Simple - grab a copy of nearly any fantasy or medieval/ancient combat movie, and watch them swing their swords around. LotR, Braveheart, or 300 would all work fine.

Pauan · Post by **Pauan** » 07 Feb 2008, 17:18

Jetryl wrote:Nice huge block of text goes here!

Suggestions:
- Start by making his back foot stay put.
- Use the same blade motion seen in this image:
http://i6.photobucket.com/albums/y203/M ... k_down.gif
- But make the blade bigger, since that guy's holding something that barely qualifies as a knife. The blade should be at least twice as big as that - in fact, we can even go to 3X the size, based on the flavor of the game. Duran, from Secret of Mana 3, had a sword bigger than he was tall, and cloud was quite the same.
- And most importantly, fix the one thing Modanung failed to do - make the body move! If the swordsman lunges forward with his front foot, that sword will pass over the correct game space

Well said! I am glad at least one person here has a basic sense of animation and physics. No matter what Modanung or anybody else says, if the animation does not look/feel right to the player, it will negatively impact the game. Using the excuse "they'll get used to it eventually" is a poor philosophy for a game developer (and people in general). I agree the animation is good, but as has been said, the motion needs to be tweaked.

First, you must realize that swords in this game should not be the same size. As such, ignore the size of the sword. The sole purpose of this is to create a natural looking playerset that we can slap a sword onto after. If the playerset does not look good by itself, what will adding a sword fix? Nothing.

Second, to use such an awkward position in real life may work, but the swing will be very weak. Also, you must realize the Weapons page is just a rough outline displaying the general distinctions between the weapon types. Trying to use such a grid as an end-all "we must fit our animation to this grid" is also quite silly. Fit the grid to the weapon, not the other way around. If the grid is incorrect, fix it. Do not base your animations off of faulty blueprints.

Thus, get rid of the grid. Seriously, delete it. It no longer exists. You have (almost) infinite space to make a good animation. Now, sketch out a rough outline of what you want it to do. Tweak it until it looks natural. Pixel a rough outline, keeping in mind that the grid does not exist. Once you are done, you can delete unnecessary frames, tweak things, and position the sprite properly.

That should fix most of the problems, and if not, then simply tweak more.

P.S. Although this is a fantasy game it is still based in the real-world. The fact you even considered using swords in the first place proves this. As such, if you tried to swing a sword like in your animation in real life, it would simply not work. Far too awkward, no power. If you accept the existence of a real-life object (such as a sword), the concept of gravity, and other physical laws, then you must conclude a sword would function the same within those laws. Let me ask you this: assuming you have a sharp bladed object and the various physical laws are still in place (possibly barring magic), then why would a competent warrior choose to use his blade in an inefficient unnatural stupid way?

P.P.S. I will be greatly surprised if you did not see the unnaturalness in this image. If you saw somebody do frames 4-6, you would immediately expect them to fall over, yet you claim the warriors in TMW would not..? It is a very common problem: working from a single perspective. I struggle with this at times, as do many artists. You work on a piece so long you get locked into one perspective and your brain starts to think that that is how it should be.. even though it is unnatural. I usually fix this by flipping the image horizontally, revealing the obvious flaws. In this case, the best thing to do is look for alternative opinions, as animation is far harder to correct.

Doubi · Post by **Doubi** » 07 Feb 2008, 19:03

Thankyou to everyone for your feedback.

Jetryl, know that I really wasn't dismissing what you were saying, I'm sorry if it came across like that. Lord knows I'm not great at this and I appreciate the clear insight you have. In fact I agree with what you were saying, which is why I ended by suggesting ways Modanung's first animation could be used.

Part of my problem here as Pauan's noticed is that I've come in, looked at what was already done, looked at the info in the wiki, taken it all as gospel and tried to progress within those boundaries.

But clearly game mechanics and all that jazz are secondary to things just looking good. K, got it

I'm working on some sketches of the movement which I'll post up in the next few days. Don't worry, it's completely different

A lot closer to Modanung's first concept.

In the meantime, what are people's thoughts about number of frames of animation? Using blurring of the weapon and going for the look of a nice fast slash rather than someone slowly demonstrating the movement to students of Swordplay 101, I'm thinking 4 frames might be alright. People have told me not to worry about the hypothetical workload for equipment spriters in the future, but, it is in the back of my mind there.

*edit*
Pauan, if you're still watching and waiting, would it be alright to put me on this issue on mantis? For the sake of keeping things organised, don'tcherknow~

Pauan · Post by **Pauan** » 07 Feb 2008, 19:12

Doubi wrote:Thankyou to everyone for your feedback.

Jetryl, know that I really wasn't dismissing what you were saying, I'm sorry if it came across like that. Lord knows I'm not great at this and I appreciate the clear insight you have. In fact I agree with what you were saying, which is why I ended by suggesting ways Modanung's first animation could be used.

Part of my problem here as Pauan's noticed is that I've come in, looked at what was already done, looked at the info in the wiki, taken it all as gospel and tried to progress within those boundaries.

But clearly game mechanics and all that jazz are secondary to things just looking good. K, got it

I'm working on some sketches of the movement which I'll post up in the next few days. Don't worry, it's completely different A lot closer to Modanung's first concept.

In the meantime, what are people's thoughts about number of frames of animation? Using blurring of the weapon and going for the look of a nice fast slash rather than someone slowly demonstrating the movement to students of Swordplay 101, I'm thinking 4 frames might be alright. People have told me not to worry about the hypothetical workload for equipment spriters in the future, but, it is in the back of my mind there~

Not at all. My point was that since we are working on a physical representation of an attack; it should at least look natural. It is easily possible for it to be both fun (gameplay) and look good (appearances) at the same time. The problem is people forget completely about how it looks. You have to look at it from a player's perspective as well. Consider Seiken Densetsu 3, which is part of the Mana series. It looks fantastic (barring a few ugly 3-D things), but also has the animation looking good too. The goal is to have both good looks and good gameplay. Neither should ultimately have a higher importance. In the beginning stages it is acceptable to use a low-quality image for testing, but the final image should still be so polished it gleams. That was my entire point, roughly speaking.

P.S. It is a wiki so unless otherwise stated you should not take it as the gospel.

Also note that page was a rough mock-up. As I stated earlier, the point was to show the difference between weapon types, not to be the end-all. In other words: swords have a wide attack radius but not that much range. Decent damage, decent delay. Overall average weapon. Copying the weapon grid and forcing the playerset into an unnatural position as a result is not exactly the best of ideas. Instead take the general principles behind the grid and then apply those as best as you can.

P.P.S. 4 frames should be adequate. More may be required though. Also, any thoughts about using the particle engine to assist in the blur? Anybody?

EDIT: I am always watching and waiting.

EDIT: I can't add you as the one assigned.. for whatever reason. So instead I added you as the reporter, and we'll just all agree you're the one doing the work.

Jetryl · Post by **Jetryl** » 07 Feb 2008, 23:45

Doubi: thanks, and I didn't mean to come off as "elitist".

P.P.S. 4 frames should be adequate. More may be required though. Also, any thoughts about using the particle engine to assist in the blur? Anybody?

Uh, 4 frames is nearly a minimum. You probably want to shoot for the range of 4-7 frames.

As for the blur, I would not use the particle engine for that, I would have that drawn-into the weapon sprite (since of course the weapon is going to lie alone on its own sprite-sheet).

However, the particle engine would be very good for providing "splash" effects when the weapons strike the opponent; whether doing damage, or simply glancing off of armor.

Pauan · Post by **Pauan** » 08 Feb 2008, 00:13

Jetryl wrote:Doubi: thanks, and I didn't mean to come off as "elitist".

P.P.S. 4 frames should be adequate. More may be required though. Also, any thoughts about using the particle engine to assist in the blur? Anybody?
Uh, 4 frames is nearly a minimum. You probably want to shoot for the range of 4-7 frames.

As for the blur, I would not use the particle engine for that, I would have that drawn-into the weapon sprite (since of course the weapon is going to lie alone on its own sprite-sheet).

However, the particle engine would be very good for providing "splash" effects when the weapons strike the opponent; whether doing damage, or simply glancing off of armor.

If you assume we use sprites for the blurring, then yes, you are right.

Hmm... any particular reason? Other than the aesthetics and the possible computer/server load? I mean, I do agree, but perhaps it would be nice to use the particle engine to spruce up the sprite blur? Just throwing out suggestions.

On another note.. I tried a small experiment. I noticed with the animation the primary problem was where the player leans forward very heavily. I also noticed that once the blur became added the leaning-forward frame became unnecessary. Keeping this in mind, I redid the animation, but removed the 4th frame. Tell me what you think:

Yes yes I know it isn't perfect, and yes I know it still needs a LOT of tweaking before it's usable... but don't you agree it looks a lot better than the original? The added bonus is that it still covers the same basic area as the original, thus it fits into the Weapons grid. Comments?

Post by **Crush** » 08 Feb 2008, 00:26

Looks pretty nice, pauan. But I think that at least one of the three "preparation" frames could be dropped. For gameplay reasons the attack animation should in fact have as least "foreplay" as possible.

Also remember that we also need the sideways and the back animation.

Pauan · Post by **Pauan** » 08 Feb 2008, 00:28

Crush wrote:Looks pretty nice, pauan. But I think that at least one of the three "preparation" frames could be dropped.

Also remember that we also need the sideways and the back animation.

Considering I only modified an existing image and the side/top frames are not completed (nor are being done by me), I suggest you ask Modanung or somebody else to make them.

Also, I considered dropping one. I would need to test it myself to be sure, but I think it would make the animation look far less smooth and powerful. After running several animation tests I found that the reason it looks like a good swing is because of the initial backstroke. When I made this too short the main strike appeared weak. As such by removing a frame, it could very easily make the attack look.. well.. weak. As said I will have to test it personally to be certain, but if I am right, then at least it is only one frame more, ne?

Pauan · Post by **Pauan** » 08 Feb 2008, 01:02

Alright.. I decided to drop the first frame. Yes it isn't quite as smooth, but it still looks good and it's one less frame! Which in the long run could mean a ginormous amount less work. Yes it will require a ton of tweaks to get it to the point where I say it's "complete," but it's still a good start, in my honest opinion.

Anyways, here it is. I included the standing frame at the beginning for reference.

leeor_net · Post by **leeor_net** » 08 Feb 2008, 01:20

I think the animation is pretty good as it is now. I would suggest, however, that the front foot come forward instead of the rear foot sliding back. That seems unnatrual for some reason.

Also, I think the idea of 'effects' being separate animations would 1) complete the animation by filling in the details not necessary from the base and 2) would make the lives of the artists a hell of a lot easier. A game that very successfully pulls this technique off is MapleStory (cringe, I hate to mention the game but hey, the artwork is truelly fantastic). I can post a few animation pieces to demonstrate how they used effect animations to fill in the gaps of the primary character's animation.

Just thought I'd put in my $0.02...

Pauan · Post by **Pauan** » 08 Feb 2008, 01:31

leeor_net wrote:I think the animation is pretty good as it is now. I would suggest, however, that the front foot come forward instead of the rear foot sliding back. That seems unnatrual for some reason.

Also, I think the idea of 'effects' being separate animations would 1) complete the animation by filling in the details not necessary from the base and 2) would make the lives of the artists a hell of a lot easier. A game that very successfully pulls this technique off is MapleStory (cringe, I hate to mention the game but hey, the artwork is truelly fantastic). I can post a few animation pieces to demonstrate how they used effect animations to fill in the gaps of the primary character's animation.

Just thought I'd put in my $0.02...

*Steals the two cents...!*

I agree with you completely. All I did was remove some obselete frames, though I highly considered fixing the foot thing.

That is already planned. Actually, right now the weapons are listed separately from the playerset. This is just a quick mockup: the weapons (and possibly blurs as well) will be separated.

P.S. It seems unnatural because normally somebody would lunge forward to attack, but in this animation the player is reaching backwards with his foot.

P.P.S. I used to play MapleStory, but for the benefit of others, please do post some images.