Add transcription for: frames_zips/CGCircuit_RiggingCartoonRealistic_DownloadPirate.com.part3_week05 15 pose reader on character creation_frames.zip

transcriptions/frames_zips/CGCircuit_RiggingCartoonRealistic_DownloadPirate.com.part3_week05 15 pose reader on character creation_frames_transcription.json

@@ -0,0 +1,8 @@
+{
+  "text": " Okay, we are back here with our character. So now we want to set up the post reader for the shoulders, for the auto shoulders, because the moment where we can see what happens is, so far, everything is working as expected. actually we have to turn the auto-shoulders on and what you can see here is what I've done temporarily. I've connected the two, you know, the right side and the left side with a multiply divide node so that we can actually just you know animate one side so that we can compare before and after after we added the post reader here. Let's turn these auto-shoulders on here real quick and then let's see what happens. So so far we can see actually everything seems to be behaving as expected. So we're going down, we're going up, we're going forward, we're going backwards, so that all is fine. The problem however occurs is if you move the pole vector or the elbows here forward, something like that for example, And then you start moving this. Then we can see some popping, for example, so now they're up, now they're down. If we move them forward and backward, play a little bit with this. So we're getting some kind of like undesired results, and really the shoulders seem to go all over the place here a little bit. Even go inside and all that. So the hope is that we can with the post reader avoid all of that, or at least improve it a little bit. So let's set everything back to zero and let's try to create those post readers. Now try to speed this up this time a little bit because we took our time with creating them the first time around with the simple example of the cylinder. Let's try to add that here. So I'll turn off my geo. Let's make sure that everything's set back to zero. Can't stress it enough how important it is. So let's create our locator, our first one, open up the Outliner. And this is going to be our poseReader main locator. And let's group that into itself and call this poseReaderGroup. Now we'll take this poseReaderGroup and we'll parent it under the joint where we want to measure the pose from. And here what we saw, or what we did earlier, is we're actually not measuring the rotations from the IK root join. If you remember, we created a duplicate of that that's outside of the whole shoulder setup. So we have to find that first. And I was, these guys down here, so let's display them again. And let's actually go in and let's go into a different view here and just isolate those so you can see, keep an eye on them. And let's also add our new locator in there. Down here, our post reader group. Let's add that in here as well. Add selected objects. And then let's also say, outload new objects. So now we have our group down there with the post reader inside. Let's take the whole group and let's parent it. Actually, first let's go back to zero. Keep forgetting that. So let's put that back. Then we have back. OK. So now let's take our post reader group and parent it under this first joint here. Parent under the root joint. Let's set everything to zero on the post reader group. So that will guarantee that it's going to be now in the same space as this joint. And we can start working with this. PostReader. So then we want to move the main post reader a little bit back in y here, minus 5, for example. Then we'll duplicate it, and we'll create a target. So we moved it five units forward. So let's call this postReaderTarget locator. And let's create a third one. Actually, we can't really get a main one. And this is going to be our up. Up. OK, there. Maybe moved it to five. And let's make the same adjustments here. I mean, we could leave it as it is. You know, that's the only thing that we need. But just to distinguish them a little bit easier, let's make the sizes a little bit different. So the main one will set that to two. And the up one will set that to 0.5. And I will alter change the color coding on those guys to be yellow. You can see them a little bit easier. Here we go. The last one as well. Okay. And now we create our aim constraint. So we'll go from here and that's where the aim constraint goes on onto the main one. and if we set this to translation, the translation tool, and we make sure that we're on object or local not the world, that's a different one, but object or local will work here and then we want to set the aim constraint exactly how we want it so let's reset everything here just in case and then the aim should be towards that guy, so that would be y, so let's set y as the aim 0 in x, 0 in z, only in y we want to have 1. And then the up vector is going to be in this direction, so that would be our positive set, so set 1, everything else 0. And then we want to have object up, and we have to type in the name here of our post-reader up locator. So post-reader up lock. And then we'd apply. And to double check that we made everything correct, we can see here that we have on the locator, we have created our aim constraint, but there are zero rotations. And on the constraint itself, there should also be zero offsets. So far, so good. Now we want to take this post reader group and parent it to the parent of where we want to measure the post from. So since we want to measure the post from this guy itself, we want to go one step up and that's what we want to have the post reader group go on to. So in our case here this would be one up from the arm, IK shoulder, or the root would be the chest group. So we'll take the whole post reader and we will parent it outside into the chest group, just a parent of that joint. And in this joint is where we want to read the post from, so that means that the target has to go under this joint, but not directly we'll group it. And we'll call this pose reader target group. And the pose reader target group, this is what we're parent now under the IK shoulder auto route. Okay, so now if we move this up and down, well, I guess we can't really, but if we take this one in here, show isolate add, select objects, now we move this up and down, then we can see the aim here is actually doing its thing and we are getting values here. So that's good. Let's already create the twist locators as well. So I'll take maybe that one, duplicate it, put it to zero here in terms of the translation to get it into the origin of the post reader, in this case post the target group. And then we want to call this our twist locator. And a twist locator because it's not a target goes into the main post reader group. Okay, I can make this a little bit bigger since it's the main one will set us to two, and then we'll duplicate it twice. One is going to be the target locator and the other one, or sorry, a twist target, what I should call that, twist target. The other one is the twist up. So kind of like twist main, twist target, twist up. So here we can remove the two. And since it's C up, the up we can scale down to 0.5. And the target was one. We're scaling here just again, then we can distinguish them. The target should now move outside a little bit, maybe five number 6 or something like that. And the up, we want to probably move in this direction. Let's see. Let's move in that direction here, maybe minus 3. Let's change the color of the twist locators or the shapes of them so that we can distinguish that from the main ones. Let's make the twist locator is blue. OK, here we go. And let's set up our aim constraint. So I'll select my target twist target, and I select the twist main locator. And I'll go to constraint aim. And now I look at what the orientation is of that main locator. So the aim should be z in this case, z1 positive. And then for the up, we want to have y but negative y. So minus 1 and y, 0 and z. And then we have a new up object. So that would be post reader twist up locator, post reader, twist, up, hit apply. And again, to double check, we can come in here onto the main twist locator and see that we have zero rotations. And also on the aim constraint, we have zero offsets. That all looks good. Then the next step that we need to do is we need to take the target for the twist and parent under the target group. And that those two are together, two targets. And we want to take the remaining two twist attributes, our twist locators, group them. And this is going to be our pose reader twist group. Now, the only thing that's still missing here is that we want to make sure that when we move this up and down, for example, forward, back, that the twist group with those two locators here is actually also moving up. that those blue ones are going to stay aligned, as opposed to like this, they're not aligned here really. Therefore, what we're going to do is we're going to take the result of the rotations here of our main locator, multiply it by 2, because we said that we only have half of the rotation here, multiply it by 2, those outputs, and then we're going to apply that output of the multiplier into the post reader twists rotations so that this rotates kind of the same or similar anyways to the joint. Let's try to set this up. So we come in here, I'll use the hypershape but you can use as always a post the node editor as well. So let's select the post reader twist group and let's select the post reader main locator, create our multiply divide node between and connect the rotation into the input. And we probably only need x and set, I imagine, but let's double check. So that looks correct because x is for bending and z is for side motion. the twisting we don't really need here on the main locator. So we really only need those two. Only x, we need to connect to x, and z, we need to connect to z. And then on the multiplier, we multiply x and z with two. And then the output of those two, x and z, X for bent, Z for side, then goes on to the X and the Z of the post reader twist group. X goes to X, Z goes to Z. And now we can already see that this locator is kind of like moving up here or aligning itself properly. So that all looks good. Set this back to zero. The only thing that's now left to do is we have to create, or we don't have to, but it's nice to be able to create on the object, in this case a joint, where we want to measure our or pose from if we create some custom attributes here. Let's do that. Add a separator in here, 10 underscores. And then we'll create a pose, a bend, pose, twist, and pose, side. And we'll bring that joint in here as well. And now we go from the multiply divide. We already went into the post reader twist group, but now we also go into these two here. So we set x is for bend and z is for side motion. So we'll connect it that way. So we are from output X into BAND, post BAND, and from output Z, that's for our side motion. And then we go select our twist locator, the main one, here in the big one in the center. I will add that one in here as well. Graph, and select it. And that one here would be Y for measuring the twist. So we connect y to twist. Rotate y from the post reader twist locator, rotate y into the post twist. That theoretically should be it. Now we can come in here and test it. So for example, we move this up. We should see here on this drawing some values here showing up. And we can see that. So now the only thing that's left to do is, at the moment we are connecting for our auto shoulder, we are going from the arm, IK shoulder, auto root rotations. So instead of going from the rotations, we have to go from these bent twist side values here. So let's try that. So we select the arm, IK shoulder, auto root that I have in here and show the outgoing connections from there. So you can see one we have the pole vector constraint and the IK handles, so we don't care about those two. We have here the next joint. So we don't care about that one, the child. And we have model panels, so we don't care about that one. So really the only connection we care about is the connection from the joint into this shoulder on off multiply divide node into the first lots here. So this is what we want to reset or reconnect instead of coming from the rotations coming from those custom attributes here. So I'll connect those over. So instead of going from rotate into the first one, y into the second one, set into the third one, we'll now use our custom attributes. So bend goes into x, twist goes into a y, inside goes into set. So now that we've replaced those, let's see what actually changed, or if anything actually changed. So we already can see a difference here between right and left side. So we can see here on this side, we can't even see a difference, while on that side we can. So that looks a little bit better now. Also here, looks a little bit better. This shouldn't be going down or back or whatever it's doing at the moment. But more importantly, if we try to find a good value here, something like that. And now we move to the pole vector here. So here we can, for example, see a problem where this is going down when it shouldn't really, because it's going forward. And that's just because here on the rotations we have all kinds of values. Same thing on the other side here. But on the other side, we can see that it's now actually measuring better rotations than what we're seeing up here. So I can see here on that side where we don't have a post reader. It's going kind of like up and down and doing some crazy stuff. Here this is looking much more reasonable. And it's kind of like trying to stay up. Let's see if I can find a better pose to illustrate that. It all seems to be working fine now. Actually here you can see for example also that the forearm here is moved forward, so that means the shoulders should probably also be moved forward so that is working fine while here on this side it is not really working. I mean this is a pretty extreme pose obviously, but still you can see that it is doing a better job with the post reader than without. Says this a little bit more. So here for example you can see on this side it completely breaks down without the post reader or with the post reader it's doing much better job in like trying to keep what it needs to do while still this being a pretty extreme. Over here, you can see all kinds of popping happening on the other side. So we move the pole vector forward and back while this side behaves a lot more stable and robust.",
+  "segments": [
+    {
+      "text": " Okay, we are back here with our character. So now we want to set up the post reader for the shoulders, for the auto shoulders, because the moment where we can see what happens is, so far, everything is working as expected. actually we have to turn the auto-shoulders on and what you can see here is what I've done temporarily. I've connected the two, you know, the right side and the left side with a multiply divide node so that we can actually just you know animate one side so that we can compare before and after after we added the post reader here. Let's turn these auto-shoulders on here real quick and then let's see what happens. So so far we can see actually everything seems to be behaving as expected. So we're going down, we're going up, we're going forward, we're going backwards, so that all is fine. The problem however occurs is if you move the pole vector or the elbows here forward, something like that for example, And then you start moving this. Then we can see some popping, for example, so now they're up, now they're down. If we move them forward and backward, play a little bit with this. So we're getting some kind of like undesired results, and really the shoulders seem to go all over the place here a little bit. Even go inside and all that. So the hope is that we can with the post reader avoid all of that, or at least improve it a little bit. So let's set everything back to zero and let's try to create those post readers. Now try to speed this up this time a little bit because we took our time with creating them the first time around with the simple example of the cylinder. Let's try to add that here. So I'll turn off my geo. Let's make sure that everything's set back to zero. Can't stress it enough how important it is. So let's create our locator, our first one, open up the Outliner. And this is going to be our poseReader main locator. And let's group that into itself and call this poseReaderGroup. Now we'll take this poseReaderGroup and we'll parent it under the joint where we want to measure the pose from. And here what we saw, or what we did earlier, is we're actually not measuring the rotations from the IK root join. If you remember, we created a duplicate of that that's outside of the whole shoulder setup. So we have to find that first. And I was, these guys down here, so let's display them again. And let's actually go in and let's go into a different view here and just isolate those so you can see, keep an eye on them. And let's also add our new locator in there. Down here, our post reader group. Let's add that in here as well. Add selected objects. And then let's also say, outload new objects. So now we have our group down there with the post reader inside. Let's take the whole group and let's parent it. Actually, first let's go back to zero. Keep forgetting that. So let's put that back. Then we have back. OK. So now let's take our post reader group and parent it under this first joint here. Parent under the root joint. Let's set everything to zero on the post reader group. So that will guarantee that it's going to be now in the same space as this joint. And we can start working with this. PostReader. So then we want to move the main post reader a little bit back in y here, minus 5, for example. Then we'll duplicate it, and we'll create a target. So we moved it five units forward. So let's call this postReaderTarget locator. And let's create a third one. Actually, we can't really get a main one. And this is going to be our up. Up. OK, there. Maybe moved it to five. And let's make the same adjustments here. I mean, we could leave it as it is. You know, that's the only thing that we need. But just to distinguish them a little bit easier, let's make the sizes a little bit different. So the main one will set that to two. And the up one will set that to 0.5. And I will alter change the color coding on those guys to be yellow. You can see them a little bit easier. Here we go. The last one as well. Okay. And now we create our aim constraint. So we'll go from here and that's where the aim constraint goes on onto the main one. and if we set this to translation, the translation tool, and we make sure that we're on object or local not the world, that's a different one, but object or local will work here and then we want to set the aim constraint exactly how we want it so let's reset everything here just in case and then the aim should be towards that guy, so that would be y, so let's set y as the aim 0 in x, 0 in z, only in y we want to have 1. And then the up vector is going to be in this direction, so that would be our positive set, so set 1, everything else 0. And then we want to have object up, and we have to type in the name here of our post-reader up locator. So post-reader up lock. And then we'd apply. And to double check that we made everything correct, we can see here that we have on the locator, we have created our aim constraint, but there are zero rotations. And on the constraint itself, there should also be zero offsets. So far, so good. Now we want to take this post reader group and parent it to the parent of where we want to measure the post from. So since we want to measure the post from this guy itself, we want to go one step up and that's what we want to have the post reader group go on to. So in our case here this would be one up from the arm, IK shoulder, or the root would be the chest group. So we'll take the whole post reader and we will parent it outside into the chest group, just a parent of that joint. And in this joint is where we want to read the post from, so that means that the target has to go under this joint, but not directly we'll group it. And we'll call this pose reader target group. And the pose reader target group, this is what we're parent now under the IK shoulder auto route. Okay, so now if we move this up and down, well, I guess we can't really, but if we take this one in here, show isolate add, select objects, now we move this up and down, then we can see the aim here is actually doing its thing and we are getting values here. So that's good. Let's already create the twist locators as well. So I'll take maybe that one, duplicate it, put it to zero here in terms of the translation to get it into the origin of the post reader, in this case post the target group. And then we want to call this our twist locator. And a twist locator because it's not a target goes into the main post reader group. Okay, I can make this a little bit bigger since it's the main one will set us to two, and then we'll duplicate it twice. One is going to be the target locator and the other one, or sorry, a twist target, what I should call that, twist target. The other one is the twist up. So kind of like twist main, twist target, twist up. So here we can remove the two. And since it's C up, the up we can scale down to 0.5. And the target was one. We're scaling here just again, then we can distinguish them. The target should now move outside a little bit, maybe five number 6 or something like that. And the up, we want to probably move in this direction. Let's see. Let's move in that direction here, maybe minus 3. Let's change the color of the twist locators or the shapes of them so that we can distinguish that from the main ones. Let's make the twist locator is blue. OK, here we go. And let's set up our aim constraint. So I'll select my target twist target, and I select the twist main locator. And I'll go to constraint aim. And now I look at what the orientation is of that main locator. So the aim should be z in this case, z1 positive. And then for the up, we want to have y but negative y. So minus 1 and y, 0 and z. And then we have a new up object. So that would be post reader twist up locator, post reader, twist, up, hit apply. And again, to double check, we can come in here onto the main twist locator and see that we have zero rotations. And also on the aim constraint, we have zero offsets. That all looks good. Then the next step that we need to do is we need to take the target for the twist and parent under the target group. And that those two are together, two targets. And we want to take the remaining two twist attributes, our twist locators, group them. And this is going to be our pose reader twist group. Now, the only thing that's still missing here is that we want to make sure that when we move this up and down, for example, forward, back, that the twist group with those two locators here is actually also moving up. that those blue ones are going to stay aligned, as opposed to like this, they're not aligned here really. Therefore, what we're going to do is we're going to take the result of the rotations here of our main locator, multiply it by 2, because we said that we only have half of the rotation here, multiply it by 2, those outputs, and then we're going to apply that output of the multiplier into the post reader twists rotations so that this rotates kind of the same or similar anyways to the joint. Let's try to set this up. So we come in here, I'll use the hypershape but you can use as always a post the node editor as well. So let's select the post reader twist group and let's select the post reader main locator, create our multiply divide node between and connect the rotation into the input. And we probably only need x and set, I imagine, but let's double check. So that looks correct because x is for bending and z is for side motion. the twisting we don't really need here on the main locator. So we really only need those two. Only x, we need to connect to x, and z, we need to connect to z. And then on the multiplier, we multiply x and z with two. And then the output of those two, x and z, X for bent, Z for side, then goes on to the X and the Z of the post reader twist group. X goes to X, Z goes to Z. And now we can already see that this locator is kind of like moving up here or aligning itself properly. So that all looks good. Set this back to zero. The only thing that's now left to do is we have to create, or we don't have to, but it's nice to be able to create on the object, in this case a joint, where we want to measure our or pose from if we create some custom attributes here. Let's do that. Add a separator in here, 10 underscores. And then we'll create a pose, a bend, pose, twist, and pose, side. And we'll bring that joint in here as well. And now we go from the multiply divide. We already went into the post reader twist group, but now we also go into these two here. So we set x is for bend and z is for side motion. So we'll connect it that way. So we are from output X into BAND, post BAND, and from output Z, that's for our side motion. And then we go select our twist locator, the main one, here in the big one in the center. I will add that one in here as well. Graph, and select it. And that one here would be Y for measuring the twist. So we connect y to twist. Rotate y from the post reader twist locator, rotate y into the post twist. That theoretically should be it. Now we can come in here and test it. So for example, we move this up. We should see here on this drawing some values here showing up. And we can see that. So now the only thing that's left to do is, at the moment we are connecting for our auto shoulder, we are going from the arm, IK shoulder, auto root rotations. So instead of going from the rotations, we have to go from these bent twist side values here. So let's try that. So we select the arm, IK shoulder, auto root that I have in here and show the outgoing connections from there. So you can see one we have the pole vector constraint and the IK handles, so we don't care about those two. We have here the next joint. So we don't care about that one, the child. And we have model panels, so we don't care about that one. So really the only connection we care about is the connection from the joint into this shoulder on off multiply divide node into the first lots here. So this is what we want to reset or reconnect instead of coming from the rotations coming from those custom attributes here. So I'll connect those over. So instead of going from rotate into the first one, y into the second one, set into the third one, we'll now use our custom attributes. So bend goes into x, twist goes into a y, inside goes into set. So now that we've replaced those, let's see what actually changed, or if anything actually changed. So we already can see a difference here between right and left side. So we can see here on this side, we can't even see a difference, while on that side we can. So that looks a little bit better now. Also here, looks a little bit better. This shouldn't be going down or back or whatever it's doing at the moment. But more importantly, if we try to find a good value here, something like that. And now we move to the pole vector here. So here we can, for example, see a problem where this is going down when it shouldn't really, because it's going forward. And that's just because here on the rotations we have all kinds of values. Same thing on the other side here. But on the other side, we can see that it's now actually measuring better rotations than what we're seeing up here. So I can see here on that side where we don't have a post reader. It's going kind of like up and down and doing some crazy stuff. Here this is looking much more reasonable. And it's kind of like trying to stay up. Let's see if I can find a better pose to illustrate that. It all seems to be working fine now. Actually here you can see for example also that the forearm here is moved forward, so that means the shoulders should probably also be moved forward so that is working fine while here on this side it is not really working. I mean this is a pretty extreme pose obviously, but still you can see that it is doing a better job with the post reader than without. Says this a little bit more. So here for example you can see on this side it completely breaks down without the post reader or with the post reader it's doing much better job in like trying to keep what it needs to do while still this being a pretty extreme. Over here, you can see all kinds of popping happening on the other side. So we move the pole vector forward and back while this side behaves a lot more stable and robust."
+    }
+  ]
+}