@CubeAce Hi sorry for not replying, I've been working but also reading this, & I'll use John CB's words, clear as mud 🤷‍♂️🙃👴😂

@johnebaker

Hi John.

Can you please re-read your post you made at 6:30 please?

I think there is some information missing from it.

Ray.

@browj2, @Former user @CubeAce

Hi

. . . . Clear as mud! . . . +1

I too had trouble getting my head round it too, the initial thoughts I had were - 'wow thats the answer to how to do an effect i wanted to do'. I tried to do the effect before using the old Size/Position and Rotation however is not possibl as there was no anchor point.

After a lot of experimenting things became clearer, however I need to do more experimenting to fully get to grips with it - this may be delayed for a short while due to a change of scenery.

. . . . What is the order? . . . .

The order varies depending on what effect you want to create - in the case of what I call the sling shot effect - ie the image is rotating on it Z axis while moving with and around the anchor point at varying relative distances with the anchor point an approximate curve - hope this makes sense, I know my head was spinning trying to get it into words, I wish I had exported the experimenting I did

Edited:

1. Resize the image
2. Position the anchor point to its location by dragging it in the Preview monitor not using the parameter boxes
3. Position the image to the required offset from the anchor point by dragging it in the Preview monitor
4. Set the rotation
5. Ensuring the keyframe option is set to the top option as shown below (none of the sub sections must be selected) add a keyframe - do not rely on auto keyframing as this appears to have a bug Ray mentioned previously.
   
   
    
6. Repeat 2 - 5 as necessary.

The above sequence is incorrect / not reproducible by other users - see later posts - further work is needed

Using the preview monitor is essential for moving the anchor point and image to maintain the relative positions to each other - usings the parameter value boxes for Position and Anchor give a different result.

Ray the image above is what should have been in the post you mentioned - I will correct it.

HTH

John EB

@johnebaker

Thanks John EB,

I'm redoing my example. I put the anchor point at the upper left, reduced the size, moved to somewhere close to the middle, set total kfs, moved along timeline, z rotation 720, works fine. But I want to start with the full screen image, change size towards upper left, move to the centre, then rotate. Still working on it.

What are all of these kfs?

John CB

@johnebaker

Hi John EB,

I got it.

Now on to resize about centre, move anchor point to upper left, rotate.

John CB

Resize about centre, move anchor point to upper left, rotate.

I succeeded but the anchor point is always in the upper left, even though I managed to get the image to zoom in about its centre. So, don't ask how as I still don't understand.

John CB

@johnebaker

Hi John EB,

1. Resize the image
2. Position the anchor point to its location by dragging it in the Preview monitor not using the parameter boxes
3. Position the image to the required offset from the anchor point by dragging it in the Preview monitor
4. Set the rotation
5. Ensuring the keyframe option is set to the top option as shown below (none of the sub sections must be selected) add a keyframe - do not rely on auto keyframing as this appears to have a bug Ray mentioned previously.

Point 3 - this moves both the image and the anchor point.

Point 4 - I presume this is further along the timeline.

Point 6 - repeat moves the anchor point for the first go-round.

Result is that I can't do one rotation about 1 point, then another about another point.

John CB

No matter what I do, changing the location of the anchor point changes all anchor point kfs to be the same.

John CB

@johnebaker @browj2 @Former user

Hi John EB.

I get it in my mind now. Have yet to experiment with it.

It clicked when I realized in the publicity blurb they referred to it as a 3D effect.

Thinking about it as a 3D effect it makes sense and also how the selection of the xyz order is important.

Say to have made an animation of either 3D text or an object in 3D in blender such as a jet aircraft.

We are looking directly at the screen as if it was a stage in a theater but the red dot is one end of a piece of string that is being held by an actor on stage. Say he is in the center of the stage and swinging the object (jet) around his head in a circle that is attached to the other end of the string.

You could see, if you looked from above, that the object on the other end of the string gets nearer and further away from viewer sitting in the auditorium, but the viewer or person looking at the screen from the front sees the object go from being small on either side of center but grows the closer it gets to center as it goes 'around' the actor. Having a 3D object that turns to show the correct view but whose size remains the same gets the illusion completed by the enlarging and shrinking of the object.

Getting the two in sync to get a convincing effect would be difficult so maybe not a brilliant example but gets me close to understanding the function. XYZ order choice then becomes the choice of which way the orientation of that circle is spinning in. From horizontal to vertical and anywhere between.

This is why any movement appears to be set to a curve and may be the additional points John CB has noticed in the KFE. The anchor points location.

Does any of what I have just typed make any sense to anyone?

As to John EBs point 5.

If I select all of the boxes inside the KFE (Key Frame Editor) holding down Ctrl while left clicking with the mouse it lays down all the key frames at the start of clip length. As soon as you move along the timeline, all but the top box looks as if they are now deselected but do not appear to be disabled in usage thereafter. Only alterations made within the Size Position Rotation effect at each edited point along the timeline get recorded, but they do get recorded automatically as far as I have noticed so far.

Ray.

@CubeAce

Does any of what I have just typed make any sense to anyone?

Dunno about making sense, but this topic sure is hurting my head! 😂 🙂

@CubeAce

Hi Ray,

I think that the anchor point movement/kf effect is separate from the rotation XYZ series. The anchor point is what they will rotate about.

I discussed quite a while back in another thread what the XYZ rotations do.

Try this.

With XYZ, reduce the image way down to about 40%. Move the anchor point to the upper left corner of the image (approx.). Apply X rotation by moving the slider circle back and forth. Set it back to 0. Now, give some rotation, say about 20 degrees. Now do the X rotation routine again and you'll see that the image rotates back and forth about a modified Z axis. With some X, rotate Z around and observe what happens. The image never distorts or changes shape.

Now switch to ZYX and do the same thing. Quite different?

The last one on the list is normal. The first one rotates about the modified axis of the last axis on the list.

This gives you the 3D effects of an object with 0 thickness (a double-sided photo).

The problem that we are trying to resolve is how to kf the anchor point w.r.t. the image. I can't do it.

John CB

@CubeAce, @browj2, @Former user, @AAProds

Hi

First my apologies, the sequence I posted ended up being a mix of 2 separate experiments I tried. I will remove it from my previous post.

Ray analogy of a 3D stage is another way of looking at it, and as Alwyn has commented it does make your head hurt trying to get the effect to work.

@browj2

Is this what you are getting?

. . . . The problem that we are trying to resolve is how to kf the anchor point w.r.t. the image. I can't do it. . . . .

I need to go back to this, hopefully later today. The varying offset between the anchor point and the image I had is not being retained as in the original tests I did.

John EB

@johnebaker @browj2 @Former user @AAProds

If anyone cracks this please post a tutorial 😂😂😉.

Or several.🙏🙏🙏

All I can say so far is it does produce wonderfully smooth action but as to how to get accurate or predictable placement is beyond me. Not all actions are recorded unless manually placed and the anchor point only seems to have one placement within any one given clip at a time. I may of course be completely wrong. I can see it's potential but have no idea how to achieve it. Trying to sort out a 3D space with a 2D view is perplexing to say the least.

Ray.

@johnebaker

Hi John EB,

Yes. The anchor point stays at the same offset throughout.

This is a question for Magix - can the anchor point offset be keyframed?

The other thing that cannot be keyframed is the Sequence of rotations. That, I can understand.

@CubeAce

Hi Ray,

Yes, it is smooth. I found that all actions that need a kf get one, except that one must select the top item (SPR) to get them all if needed.

Predictability likely comes with use and learning, especially for the rotations.

Understanding the right-hand rule for axes helps. When you look at your hand (thumb=x-axis, index=y-axis, middle finger=z-axis) and rotate about 1 axis, look at what will happen when you rotate about the others. As I mentioned, the last one on the list does not get moved from its direction - it is prime and the other axes take on a new direction. That is why the sequence chosen is important. Permutations and combinations.

John CB

Greetings,

I got lost with this thread ages ago and I've come to the conclusion that I'm a normal person and I believe that if it ain't broke don't fix it.

Software engineers are not normal people; they believe that if it ain't broke it doesn't have enough features.

AndyW

@browj2 @johnebaker @Former user @me_again @AAProds

Hi John.

I think the anchor point can be moved to new positions and recorded but it causes crashes.

I was rendering a tryout and I saw something happening in the programs monitor screen I have never seen the program do before. Have a look.

I have never seen the monitor picture update itself during a render. The effect itself seems to be almost totally reliant on the CPU maxing out core 0.

Trying to change the position of the anchor point and manually adding it nearly always ends in the program freezing if you try to play the result back.

Ray.

@browj2, @CubeAce

Hi

There appears to be some intermittent bug(s) with the Size/Position/Rotation.

. . . . changing the location of the anchor point changes all anchor point kfs to be the same . . . .  can the anchor point offset be keyframed . . .

After a lot of retesting, I cannot reproduce what I got the first time.

. . . . . . . .  can the anchor point offset be keyframed . . .

It ought to be able to be key framed as the control can be moved independently.

I am also getting the odd behaviour with the unlabeled - 'ghost' keyframes being generated, however this is not consistently reproducible.

When the ghost KFs are present and KFs are generated automatically, the KF frames are 1 row above where they should be, see the image below, I have not tested this adding the KF manually adding the KF:

Positions 2 - 8 the KFs are displayed one row too high.
Positions  9 & 10 no KF auto generated

The positions indicate which control was adjusted, the Size control width/height were 'Combine input fields ' option was unchecked.

1   Manual addition of all KF to get display active
2 - Width control
3 - Height control
4 - Position horizontal
5 - Position vertical
6 - X-Axis rotation
7 - Y-axis rotation
8 - Z axis rotation
9 - Anchor point horizontal
10 - Anchor point vertical

John EB

Hi all,

Here is a new attempt at trying to explain the XYZ permutations at least partially.

The coordinate system is as shown below with X to the right, Y to the top, Z normal to the screen - coming at you.

In the previous version of MEP/VPX, this system did not change. You could only rotate the image about the fixed axes. Now, the XYZ permutations you to change the direction of the axes (the image follows along).

In each case, the third axis is fixed in position. Giving the third one a rotation, rotates the other two axes about it.

XYZ

In this case, the Z-axis never moves, but it can rotate. Giving Z some rotation causes the other two axes to rotate about the Z-axis (the image rotates as well). In the image below, you can see the new axes locations.

Now, adding X rotation causes the image that is already rotated about the Z-axis, to rotate about the rotated X-axis. Adding Y rotation will cause the already rotated image to rotate about the rotated Y-axis.

Adding X and Y rotations do not change the Z-axis; it remains normal to the page.

ZYX

In this case, the X-axis never moves - it's always to the right. Giving X some rotation causes the other two axes to rotate about the X-axis (the image rotates as well). In the image below, you can see the new axes locations.

Now, adding Z rotation causes the image that is already rotated about the X-axis, to rotate about the rotated Z-axis. Adding Y rotation will cause the already rotated image to rotate about the rotated Y-axis.

Adding Z and Y rotations do not change the X-axis; it remains horizontal to the right.

XZY

In this case, the Y-axis never moves - it's always to the right. Giving Y some rotation causes the other two axes to rotate about the Y-axis (the image rotates as well). In the image below, you can see the new axes locations.

Now, adding X rotation causes the image that is already rotated about the Y-axis, to rotate about the rotated X-axis. Adding Z rotation will cause the already rotated image to rotate about the rotated Z-axis.

Adding X and Z rotations do not change the Y-axis;.it remains vertical towards the top of the screen.

Order of axes - XYz vs YXz

Still working this one out. As far as I can tell, the second one rotates the image about its axis, then the resultant image rotates about the first axis. There is a difference but I'm struggling to find a way to demonstrate this.

Edit: I think that I have a way to demonstrate this, but it will take a couple of days - need time.

So, does the above help clarify the XYZ rotation feature?

John CB

@browj2 @AAProds @johnebaker @me_again @AAProds

Hi John CB.

In my head, no it doesn't help but from a practical viewpoint it may do as it has given me an idea as to how to demonstrate to myself the differences so thank you. 'Every Little Helps' as a well know supermarket here says.

Ray.