Origin points, Copy vs Duplicate, and more...

[WhiteRider]

Ok, it looks like I've been saving up questions, but they are all related, sort of...

First off, I am new enough to ASP that simple things still stall me, but not so new that I missed the 5.5. to 5.6 upgrade. heh

So with that, I've a few questions and searching for advice:

I've a project that uses layers bound to bones exclusively with the only point interpolation being between hand shapes in switch layers. The trick is to keep the various joints lined up so I've been moving origin points to coincide with where the bones pivots will eventually go. I've also made a little target shape to add as a new object to each part (wrist to hand, etc) out of necessity because copying that target from one layer to the next almost always shifts around a bit. Reading the forums here, Vern said something about using Duplicate vs Copy/Paste which will certainly help when the full layer is copied, but is there a way to make copying only one object from one layer to the next behave a bit more?

I've been placing the parts (foot to ankle, etc) by lining up the little targets, and rotating the layer (Key 3) to get the angle right. But is there any advantage to setting the layer rotation to zero and rotating the points instead? Also, is there any way to move the origin point of a selected group of points?

Also, I notice that a vector layer in a switch layer, copied to a new layer and modified (translate points) while still in the switch layer seems to revert to the original point configuration as soon as the timeline reaches frame 1 and beyond. So I presume there is a reason for it happening, but is there a way to turn that off? Sure its only a few extra steps to move it out of the switch layer, edit then drag it back, but I've a lot of points to push around and I'd like to shave time wherever possible.

Thanks in advance...

Edit: Never mind about the third question: Duplicating the vector layer while inside the switch (as opposed to Cut/Paste) gives me a clean layer (no embedded animation in the points) to work with. Thanks Vern!

Edit: I forgot two more questions... Sorry!

What does the little red dot indicate in the layers list, next to the shape icon? Seems like I read it somewhere in the old docs last year, but I cant seem to find it now.

Also, I have a few keyframes in an otherwise empty timeline that simply wont delete? They appear almost at random amongst the layers, and always at frame 1. They represent the unselected points motion and the unselected points curvature.

And why is it that some points that have been angled using the curvature tool cannot later be returned to smooth? The result is usually a gap in the line, especially if the line width is cranked up a bit.

Which reminds me.. Is there a way to *really* weld the endpoints of two separate objects together? I can move and weld the points together and get that nifty little click sound, but they still seem to be two separate objects as evidenced when I click the "new" object with the Select New Object tool and only half of the object turns red.... Just wondering.

[jahnocli]

Let me start you off on the little red dot question. This shows that the layer output is optimised for swf format.

[WhiteRider]

Let me start you off on the little red dot question. This shows that the layer output is optimised for swf format.

Ah yeah. Now I remember. And now I remember where it was in the docs, too. I never out to swf so I didnt think to check there.

Thanks mate...

Cheers!

[mkelley]

I'll answer one question and then provide you with an answer to a question you didn't ask <bg>.

First of all, there is a *big* advantage in rotating your points rather than the layer. Any layer rotations/transforms that aren't part of an animation (that is, things you are doing to construct something rather than animate later on) are going to bite you in the behind if you try and do other constructions in that layer (or even other things -- try using any perspective rotations in a layer and then move the camera to see this).

In essence you are redefining the "base" of that layer. I have, for example, a character I made the mistake of rotating a group layer on to ease construction. I now can not get that character to be sized correctly because the size tool is now skewed by the same rotation (you have to try this to see what I'm talking about -- use scale on a layer after you've rotated it to see). You can, of course, "unrotate" (or return to original) the layer but then you've undone the transform you needed in the first place. My rule of thumb -- in construction *always* work with points.

The answer to the question you didn't ask is that you seem to be doing a lot of unnecessary work by creating "targets" when you could be using onion skin to line things up. When I'm trying to get things lined up I just move to another frame and look at an onion skin for a previous frame and that makes lining things up a snap. True, you will need to copy back your changes to frame 0 if need be, but that's far easier than creating targets you don't need.

Finally, you ought to be able to change a point from sharp to smooth by using the curve change tool regardless -- I've never had a point I couldn't fix this way.

[heyvern]

All the answers look good to me.

I never use layer rotations for characters. I only use layer rotations for "big picture" movements, scrolling backgrounds, change of perspective, big global changes so to speak, not for aligning body parts or animation body parts.

Always use point motion.

Also whenever working with switch layers I ALWAYS duplicate layers... I don't copy and paste. I don't usually animate points in a switch layer. This can get confusing sometimes. When you animate points in a switch layer you are "doubling" the point animation, you get interpolation movement AND point movement. You have keys for the switch AND keys for the points on multiple layers. I try to keep it simple and only use the switch interpolation.

This is just me... it is not any kind of ground breaking rule or technique. It is just my preference.

-vern

[WhiteRider]

In essence you are redefining the "base" of that layer. I have, for example, a character I made the mistake of rotating a group layer on to ease construction. I now can not get that character to be sized correctly because the size tool is now skewed by the same rotation (you have to try this to see what I'm talking about -- use scale on a layer after you've rotated it to see). You can, of course, "unrotate" (or return to original) the layer but then you've undone the transform you needed in the first place. My rule of thumb -- in construction *always* work with points.

Ok. Fair enough. The logic makes sense, and is pretty much what I expected to hear. Now I can admit that I've been resetting and rotating the points anyway, but at least I know that the "extra" effort is warranted.

The answer to the question you didn't ask is that you seem to be doing a lot of unnecessary work by creating "targets" when you could be using onion skin to line things up. When I'm trying to get things lined up I just move to another frame and look at an onion skin for a previous frame and that makes lining things up a snap. True, you will need to copy back your changes to frame 0 if need be, but that's far easier than creating targets you don't need.

Also fair, but remember that added question about stray keyframes. For whatever reason, there always seem to be leftover keys I cannot delete, and to be perfectly honest, leaving those laying around just violates my overdeveloped sense of order.

Besides, while I agree that creating targets is extra work, there is the added benefit of pre-marking where the bones will start, exactly, when it comes time to rig. And in this particular case, precision is a requirement.
(Now we wait for the end result to see if the methodology is viable. heh)

Finally, you ought to be able to change a point from sharp to smooth by using the curve change tool regardless -- I've never had a point I couldn't fix this way.

Sure, thats logical, and expected. But I still have several points that dont go back to smooth. Testing a few random points just now confirms your statement, on any point that doesnt already have a split in the line.

So perhaps they're just those welded joints between two separate objects I mentioned in the original message. Call it the apparent versus actual closed line.

Thank you for taking the time, mkelly...

[WhiteRider]

All the answers look good to me.

I never use layer rotations for characters. I only use layer rotations for "big picture" movements, scrolling backgrounds, change of perspective, big global changes so to speak, not for aligning body parts or animation body parts.

Always use point motion.

Thats reasonable, Vern. I appreciate the comment. And looking at all of the nifty WIP's in the forum, its more than obvious that point motion excels for organic movement. Just not so much for things like mechanical linkages.

And yes, ASP is designed for organic movement but I've never seen the harm in pushing a tool to the edge of its boundaries. Animating layers seems to work especially well in this regard, and there must be some use for it else why is it possible to bind an entire layer to a bone?

And just so you understand the value of your comments, I should probably mention that it was primarily because of your work demonstrated in the forums about a year ago when I first found ASP that prompted me to make the purchase. Other folks' work as well, but I especially recall that unusual avatar of yours.

And looking at the Liger more recently reaffirms the possibilities. Wow....

Also whenever working with switch layers I ALWAYS duplicate layers... I don't copy and paste.

Yes. It seems such a simple thing but it didnt occur to me that there would be a difference so I never tried it (I prefer the keyboard shortcuts to clicking. Saves time that way.) until I found mention in the forums.
Thank you.

I don't usually animate points in a switch layer. This can get confusing sometimes. When you animate points in a switch layer you are "doubling" the point animation, you get interpolation movement AND point movement. You have keys for the switch AND keys for the points on multiple layers. I try to keep it simple and only use the switch interpolation.

I agree. But I wasnt trying to animate in a switch layer. I just happened to Cut/Paste an existing layer already contained in a switch so I could use it elsewhere. It wasnt until I did a test anim that I discovered that the copied layer also copies all of the animation embedded in the points. Already knew that line width information was copied so I should have realized sooner, I guess...

This is just me... it is not any kind of ground breaking rule or technique. It is just my preference.

-vern

Well, it seems to work well enough.

Anyway, thanks for taking the time, Vern.

[heyvern]

A note on copy/paste for switch layers...

This will work sometimes... sometimes it doesn't. For instance if you copy/paste ALL the layers in a switch then it will work. All the points are the same.

What happens is when you say, delete a point then add it back in. AS numbers points internally based on the order of creation. A point added between two points still might be listed as the last point created. when you copy and paste this order can be changed.

I discovered this through trial and error. Like I said, sometimes it works sometimes it doesn't... just to be sure... I don't copy and paste switch layers.

Another really cool trick I like to use when creating switch layers is to use ONE layer as my base to create many switch layers. Create the "switches" as KEY FRAMES on a single base vector layer.

When you are done, duplicate that layer as many times as you have keys. Copy each key to frame zero for each layer and delete all the other keys. Drag these layers into a switch layer. BINGO... all perfect switch layers. Save the original layer and now you can create new ones at any time. You can even add keys between other keys and create inbetween switches much easier.

The reason this works so well is that you have a backup of all your switches if needed later. While creating the different switch "poses" you can copy and paste shapes, add delete points whatever you want with out worrying about breaking the switch interpolation since you are working on one layer.

If you decide to make MASSIVE changes to the switches just go to your base and make them there... much faster and easier than making those changes to multiple switch layers.

-vern

[WhiteRider]

Another really cool trick I like to use when creating switch layers is to use ONE layer as my base to create many switch layers. Create the "switches" as KEY FRAMES on a single base vector layer.

When you are done, duplicate that layer as many times as you have keys. Copy each key to frame zero for each layer and delete all the other keys. Drag these layers into a switch layer. BINGO... all perfect switch layers. Save the original layer and now you can create new ones at any time. You can even add keys between other keys and create inbetween switches much easier.

Hey thats a great idea, Vern! Thank you!

The reason this works so well is that you have a backup of all your switches if needed later. While creating the different switch "poses" you can copy and paste shapes, add delete points whatever you want with out worrying about breaking the switch interpolation since you are working on one layer.

Wait.. I dont understand. I was under the impression that if the number of points mismatch, you couldnt use interpolation so obviously you arent talking about deleting points on an object in a switch layer... So what am I missing here?

Lets say I have a face consisting of separate objects for head shape, mouth , nose and eyes. (No, I wouldnt normally do this, but this is an example for clarity.)

Now lets say I want to droop the eyelids so they close in a slow, languid movement. Id drop the original layer into a switch, duplicate it, and translate the points around the eyes. I'd rely on the switch layer interpolation feature to make the changes as the timeline progressed.

But from your description as I understand it, you are talking about translating the points on the original layer, yes? And if so, what would be the point of using a switch layer at all?

Logic says there would be no point since the timeline keyframes would take care of the point movement, so I have to conclude that I am misunderstanding your suggestion. I apologize.

If you decide to make MASSIVE changes to the switches just go to your base and make them there... much faster and easier than making those changes to multiple switch layers.

-vern

The only real drawback that I can see offhand is that I have yet to get an answer on how to delete those stray keyframes and I run the risk of creating even more if I poke around in the timeline.... Any ideas?

Thanks...

[mkelley]

A note on copy/paste for switch layers...
Another really cool trick I like to use when creating switch layers is to use ONE layer as my base to create many switch layers. Create the "switches" as KEY FRAMES on a single base vector layer.

When you are done, duplicate that layer as many times as you have keys. Copy each key to frame zero for each layer and delete all the other keys. Drag these layers into a switch layer. BINGO... all perfect switch layers. Save the original layer and now you can create new ones at any time. You can even add keys between other keys and create inbetween switches much easier.

The reason this works so well is that you have a backup of all your switches if needed later. While creating the different switch "poses" you can copy and paste shapes, add delete points whatever you want with out worrying about breaking the switch interpolation since you are working on one layer.

If you decide to make MASSIVE changes to the switches just go to your base and make them there... much faster and easier than making those changes to multiple switch layers.

-vern

This should be in the manual (or in the "official" manual :>)

[heyvern]

The reason this works so well is that you have a backup of all your switches if needed later. While creating the different switch "poses" you can copy and paste shapes, add delete points whatever you want with out worrying about breaking the switch interpolation since you are working on one layer.

Wait.. I dont understand. I was under the impression that if the number of points mismatch, you couldnt use interpolation so obviously you arent talking about deleting points on an object in a switch layer... So what am I missing here?

Okay... you can delete and add points shapes whatever to a vector layer that is NOT a switch layer. It is just a plain old vector layer with point motion keys. Those keys represent the switch layers... at some point in the future.

When you are "done" with this base layer and want to create the switch layers, obviously you can't delete points and add points. If LATER ON you decide to make big changes to your switch layers created this way you GO BACK to your base layer with the key frames and make those changes. The animation keys are still there, there isn't nearly as much work involved.

THEN you repeat the steps to create the switch layers from before. If you name the layers the same and replace them in the original switch layer folder the switch keys will still work.

The "base layer" is never "used" in the animation. It isn't seen or rendered. It is there as a back up. The biggest problem with switch layers is making changes to them. It is a NIGHTMARE to delete or add points and/or shapes. You have to do the same thing for each layer and it is very difficult to find the points when they are moved around.

The only real drawback that I can see offhand is that I have yet to get an answer on how to delete those stray keyframes and I run the risk of creating even more if I poke around in the timeline.... Any ideas?

Do you have the time line settings for keyed points and selected keyed points turned on? You should be able to look at the time line for the key frames for the points, select them and press delete.

When you use this technique (using a base layer for creating switch layers) you should key ALL the points on each frame to insure that only the points you want to move are moved.

Another tip... check out Actions... they are really cool and many people like them better than switch layers. When using Actions you should only key the points you want to move NOT key all the points.

-vern

[slowtiger]

This is an opportunity to remind people of the concept of duplicating files and templates.

Don't just think inside of AS inside one file only. Get yourself used to plan a production, which includes to invest some thinking about file organisation.

Of course there is no "one true way" of doing this. I found the following workflow helpful to me:

In a project with a fixed cast of characters who will appear in many scenes or episodes, it is a good idea to keep each character within its own file, complete with all turnaround views. These "character master" files get to be imported into the different scenes. If a character needs improvement, this is done in the master file, so the next new scene will have those improvements as well. Old scenes/episodes are not touched again.

[WhiteRider]

Okay... you can delete and add points shapes whatever to a vector layer that is NOT a switch layer. It is just a plain old vector layer with point motion keys. Those keys represent the switch layers... at some point in the future.

Ah. Now I get it. Sorry for being obtuse.

The "base layer" is never "used" in the animation. It isn't seen or rendered. It is there as a back up.

Yes. What a grand idea.

The only real drawback that I can see offhand is that I have yet to get an answer on how to delete those stray keyframes and I run the risk of creating even more if I poke around in the timeline.... Any ideas?

Do you have the time line settings for keyed points and selected keyed points turned on? You should be able to look at the time line for the key frames for the points, select them and press delete.

Yep. Settings, All. At zero every layer has a few keys as expected. BUt on a few layers there are also one or two keys at frame one. (Usually at the point motion and point curvature positions)
Highlighting these stray keys so they turn red and either pressing the delete key or clicking the delete button does nothing except turn them grey again. Moving the scrubber back and forth from frame zero to one shows no obvious changes to the layer, however, so I dont know what is actually been keyed. shrug

When you use this technique (using a base layer for creating switch layers) you should key ALL the points on each frame to insure that only the points you want to move are moved.

Ah... How do I ensure all the points are keyed? By highlighting them all before adding a keyframe? Or something else...

Another tip... check out Actions... they are really cool and many people like them better than switch layers. When using Actions you should only key the points you want to move NOT key all the points.

-vern

Yes. Another good idea. Thanks...

[WhiteRider]

In a project with a fixed cast of characters who will appear in many scenes or episodes, it is a good idea to keep each character within its own file, complete with all turnaround views. These "character master" files get to be imported into the different scenes. If a character needs improvement, this is done in the master file, so the next new scene will have those improvements as well. Old scenes/episodes are not touched again.

Thank you SlowTiger....

File organization is one thing at least I dont need to ask for help.
I come from the very early days of 3D apps where you literally never knew if a Save or Save As would work, and if it failed the original scene file was trashed so pure habit keeps me making backups every so often, plus branches at every major change.

Probably too much for anyone else, but it works for me. Plus as you mentioned, original characters are untouched. I am always fascinated by how much a character evolves over time and this way I have records, as it were.

Cheers!

[synthsin75]

Sometimes you can create keyframes that have no 'base' key on frame zero. These become the 'base' and can't be deleted normally. All you need to do is copy these to frame zero. With a 'base' key at zero, you can now delete this key.