Understanding 3D effects

[KenAdam]

I'm trying to understand how to use 3D effects, starting from the second slide of the WelcomeDemo project.

I don't understand why "Main", which contains "Mask Container" is itself contained in "Frame", nor what the relevance of the different sizes of Main and Frame is to how it all works.

As I understand it, frame applies the Rotate Y of 41 degreees and zoom of .77, but Main then applies a zoom of 1.4 (back where we started?)

The animation of the picture is (solely?) due to the pan of the Mask Continer, so why does perspective get applied?

I'd appreciate if someone could clarify how this works, as I'm obviously missing something here.

Thanks

[fh1805]

You don't give any indication of your previous experience with PTE. As you have just joined the forum, I'm assuming you have very little such experience (if I am wrong, forgive me, I do not wish to offend).

I would suggest that trying to learn how to animate using PTE by "reverse engineering" someone else's work is not the best way to learn. I think you would be better served by downloading and studying some of Lin Evans's excellent tutorials and by experimenting for yourself using some simple "coloured card" images created using your favorite image editor program.

In my experience, you need to understand what the various controls of the animation do in order to understand how someone else's settings achieve their end result.

regards,

Peter

[KenAdam]

Sorry, I could have been clearer.

I've made a few shows in the past, but not for a while due to other committments, and I just installed the new version.

One of the new features is the use of 3D effect on objects, but there doesn't seem to be much information understanding how the parameters work.

I've tried experimenting a bit, and I've managed to use the 3D parameters to create some interesting animation effects (e.g. by nesting a continer in a frame and applying two opposing Y rotations that vary in sync, canelling each other out to create a dynamic moving billboard effect), but I can't get the perspective effect that the example shows, so I tried to understand how the example works (I assumed the example was intended to demonstrate how these effects could be implemented).

Can you point me to a tutorial that covers using the new 3D features? Most the stuff I've found has yet to be updated to V6, but maybe I'm looking in the wrong places...

Other members of my Camera club are hooked on Pro Show Gold, and I want to create some local examples to show them what they are missing, as (to me) the shows they create all have similar failings (which others might call features).

BTW: It seems that Lin Evans's tutorials are not currently available (i.e. no "purchase" option).

[Ronniebootwest]

Can you point me to a tutorial that covers using the new 3D features? Most the stuff I've found has yet to be updated to V6, but maybe I'm looking in the wrong places...

As Peter has already suggested, Lin Evans has produced some tutorials on this subject - go to his web site Here and there maybe something there that will help you.

Ron West

[KenAdam]

As Peter has already suggested, Lin Evans has produced some tutorials on this subject - go to his web site Here and there maybe something there that will help you.

Ron West

If you go to that website (as I noted above) you can't apparently actually get any of the tutorials anymore, however I've now noticed that 3D tutorials have been added to the forum - missed those on first look as I just saw the post was dated 2008 (I'd done theones that were there a year or so ago) and didn't notice that new tutorials had been added there for V6.

Doing them now....

Ken

[Ken Cox]

SEE

MODIFIED LIN EVANS 2 LINKS TO REFLECT DOMAIN CHANGE

http://www.picturestoexe.com/forums/index.php?showtopic=87&st=0

KEN

[Lin Evans]

Hi Ken,

My "PTE for Smarties" tutorial section is continuously updated as PTE evolves, but the tutorials are all still relevant since they show the basics of how PTE works and this has not changed since the version 5 betas.

You have some misconceptions which I think will be easily cleared up if you just follow the logical sequence of the tutorials. In the next beta release of PTE (which will be available to all registered PTE users) there will be some simplifications of 3D transforms which will make that part much easier to understand, but I think your misconceptions stem from an apparent belief that "masking" has something to do with 3D animations. It doesn't really.

First, animations are simply linear changes from one slide or object (an object can be a complete jpg or a PNG with transparency and only the opaque parts visible) to another with a different position. For example, a zoom consists of a starting slide, an ending slide and a number of intermediate slides which smoothly transition between the two. You simply use the automatically placed keyframe for the starting slide, place a second keyframe somewhere (your choice) along the timeline and highlight this second keyframe. You then zoom the image visually while this second keyframe is highlighted and PTE automatically creates all intermediate images necessary to smoothly transition from slide one to the fully zoomed slide. The identical concept is use for pans and rotates or for any combination of pan, zoom and rotate.

The 3D transform is also done via the keyframes and can be combined with pans, zooms or rotates. 3D transforms allow perspective alterations of any image or object and this includes the ability to completely rotate the image or object in 3D space. Think of a 3D transform as if you were holding a picture printed on a thin sheet of paper at arms length. You can then rotate it horizontally, vertically or any combination of horizontal and vertical using your hands. Let's use a vertical rotation as an example. Hold the paper and rotate the top downward toward you and you will note that the top edge appears wider while the bottom edge appears more narrow in width as you continue to rotate this picture. Eventually, you see only a thin horizontal line and then as you continue to rotate you begin to see the back of the picture in the same perspective. As you continue to rotate you eventually see the bottom of the front of your picture which appears wider than the top. As you continue to rotate through a full 360 degrees the picture eventually looks exactly as it did when you started. This was an example of an "X" axis rotation.

This is how 3D transform works. From the Animations Tab of the Objects and Animations screen click on the 3D parameters button and you can set the X and Y parameters in degrees. This will rotate the image vertically and horizontally. The "Pan Z" is a setting which determines the position in 3D space of the object or image you are working with. This position is relative to how close or far away and whether it is seen directly in front of your line of vision or (in conjunction with Pan) to the right or left. Hold the picture in your hands at arms length. Now bend your elbows and this brings the picture closer to your eyes. That's a "Pan Z" effect. Now imagine that there were two pictures. The first being held at arms length and rotated 45 degrees clockwise while the second was held in the same position it would occupy with your elbows bent and also rotated 45 degrees clockwise. If you had two objects visible on the scree simultaneously seen in these relative positions you would be seeing the effect of 45 positive degrees of Y rotation with one object set to zero Pan Z and the other set to some arbitrary value other than zero. So Pan Z determines how far away or how close the object appears while the animation parameter "Pan" determines the left or right position.

Now you may be wondering about "solid" 3D objects and how they are made to appear to rotate in 3D space. To create a "solid" object with PTE you must place 2D objects in various positions using Pan Z and various X and Y 3D transform positions and "hold" them in the proper perspective as "children" of a controlling object. This controlling object is usually an invisible "frame" When you rotate, pan or zoom the controlling frame, the child objects of that frame are rotated, panned or zoomed in unison.

It is only possible to really do this with PTE using objects which "fit" together from a geometric perspective. PTE is not a 3D modeling software such as Maya or Rhino or Blender, etc., it simply allows three dimensional transforms. For example, if you create some "square" images using your favorite photo editing software, you can place six identically sized image in their proper perspective positions as outlined in my "cube" tutorial and create a 3D "cube" which can be rotated, panned, zoomed, etc., in 3D space via a controlling frame. If you are really ambitious, you could create tetrahedrons, octahedrons, etc., which are much more complex and these also could be manipulated in 3D space. To create and rotate a sphere is a totally different issue and this really isn't done the same way but there are some "tricks" using masks, etc., which can be used to simulate the rotation of a sphere. These are much more complex and you can learn to do this, but first you need a thorough understanding of how PTE works.

My suggestion is to start with the tutorials and go through them in order until you have a thorough understanding of the basics.

Here's a link to my tutorials in the Tutorials section of the forum. If you have problems seeing the AVI's I would suggest downloading and installing the free Kantaris Media Player here:

http://www.kantaris.org/

Tutorials Here:

http://www.picturest...?showtopic=7901

Best regards,

Lin

If you go to that website (as I noted above) you can't apparently actually get any of the tutorials anymore, however I've now noticed that 3D tutorials have been added to the forum - missed those on first look as I just saw the post was dated 2008 (I'd done theones that were there a year or so ago) and didn't notice that new tutorials had been added there for V6.

Doing them now....

Ken

[KenAdam]

Lin,

Thanks for taking the time to write a detailed reply.

I guess I really wasn't making myself clear.

My background is in systems development for aircraft displays, including generating perspective terrain fly-through in 3D, so that wasn't the problem, and I can create the masking effects OK (although I haven't yet understood why the mask over the text in the welcomedemo makes a white blur, regardless of the colour of the text). I only mentioned the mask in the original post to try and keep it clear which elements I was talking about.

I'm doing your 3D tutorials from the forum at the moment, which, so far, confirm my understanding that the "Z position" control is intended to move elements along the z axis.

My puzzle is in understanding the compound elements and transforms which create the effect of an image sliding into the view as shown in the "welcome demo", as this doesn't seem to use the Z axis. One problem I find with PTE (probably because I want to understand the underlying transforms) is that it is not easy to track how compound objects are interacting to produce the resultant matrix transform effect - e.g. as you look at one element, you can only see the transform sequence that is attached directly to it, not what other key frames there are on containing elements that are themselves impacting the effect on that object. When you are developing the effect, you know what you are trying to achieve (such as the build up of rotation and skew effects in the helicopter example), but if you try to understand how something was achieved (even when going back to an old project!) it is not clear what is going on. AFAIK there is no equivalent of writing comments in an animation sequence.

I was trying to make sense of the combination of elements each with their own transforms that make the image move correctly (I'd like to create related effect where the image enters the display something like that, rotates to face the viewer and then uses the reverse effect to leave the display on the other side, probably as the next image enters).

As a technolator, I probably focus too much on the technology before getting a good understanding at a higher level, but I guess it is too late for me to change my ways!

I usually learn faster by trying and experimenting than by only reading, so I tend to take things apart to see how they work, which is fine if everything you find is part of how it works, but sometimes you trip over things that don't matter at all, and then waste time worrying about them.

Thank you for your time here and for the time you have put into creating the tutorials.

Regards,

Ken

[Lin Evans]

Hi Ken,

Here's a quick sample of what I think you describe. Just unzip and load the PTE file and you can see how is was done by looking at the 3D transforms.

http://www.learntomakeslideshows.net/sample/sampletest.zip

Best regards,

Lin

Lin,

Thanks for taking the time to write a detailed reply.

I guess I really wasn't making myself clear.

My background is in systems development for aircraft displays, including generating perspective terrain fly-through in 3D, so that wasn't the problem, and I can create the masking effects OK (although I haven't yet understood why the mask over the text in the welcomedemo makes a white blur, regardless of the colour of the text). I only mentioned the mask in the original post to try and keep it clear which elements I was talking about.

I'm doing your 3D tutorials from the forum at the moment, which, so far, confirm my understanding that the "Z position" control is intended to move elements along the z axis.

My puzzle is in understanding the compound elements and transforms which create the effect of an image sliding into the view as shown in the "welcome demo", as this doesn't seem to use the Z axis. One problem I find with PTE (probably because I want to understand the underlying transforms) is that it is not easy to track how compound objects are interacting to produce the resultant matrix transform effect - e.g. as you look at one element, you can only see the transform sequence that is attached directly to it, not what other key frames there are on containing elements that are themselves impacting the effect on that object. When you are developing the effect, you know what you are trying to achieve (such as the build up of rotation and skew effects in the helicopter example), but if you try to understand how something was achieved (even when going back to an old project!) it is not clear what is going on. AFAIK there is no equivalent of writing comments in an animation sequence.

I was trying to make sense of the combination of elements each with their own transforms that make the image move correctly (I'd like to create related effect where the image enters the display something like that, rotates to face the viewer and then uses the reverse effect to leave the display on the other side, probably as the next image enters).

As a technolator, I probably focus too much on the technology before getting a good understanding at a higher level, but I guess it is too late for me to change my ways!

I usually learn faster by trying and experimenting than by only reading, so I tend to take things apart to see how they work, which is fine if everything you find is part of how it works, but sometimes you trip over things that don't matter at all, and then waste time worrying about them.

Thank you for your time here and for the time you have put into creating the tutorials.

Regards,

Ken

[KenAdam]

Lin,

Thanks for that.

At a quick look before I quit for the night,I think you're getting the same effect I was. The image comes in sideways, then rotates into position.

In welcomedemo, the image moves into the z axis of the screen (a more realistic pseudo 3D slide, sort of "star wars sideways"), which is what I'm not getting yet.

I'll spend more time on this tomorrow, when maybe it will suddenly click...

Thanks agin,

Ken

[Lin Evans]

Hi Ken,

Is it the red flower with the yellow floral background and reflection which appears just after the first screen and moves in from the left side that you are referring to?

If so, that's just done by changing the Y axis rotation and using the pan feature on a slide like this:

http://www.learntomakeslideshows.net/sample/slidein.zip

Best regards,

Lin

Lin,

Thanks for that.

At a quick look before I quit for the night,I think you're getting the same effect I was. The image comes in sideways, then rotates into position.

In welcomedemo, the image moves into the z axis of the screen (a more realistic pseudo 3D slide, sort of "star wars sideways"), which is what I'm not getting yet.

I'll spend more time on this tomorrow, when maybe it will suddenly click...

Thanks agin,

Ken

[KenAdam]

Lin,

Yes that is the bit I'm referring to.

Your latest example still doesn't make it move into the screen (in the z-axis sense), but simply changing the z-position setting to about 200 at the second keypoint does get the right effect.

I can start from here to get what I wanted to do in the first place (I still can't see where the WelcomeDemo adjusts Z, but I'll go back to trying to undestand that once I've worked through setting up this effect)

[fh1805]

Ken,

In the current implementation of 3D control in PTE v6 there is no direct control over the Z axis other than Pan along the Z axis.

What you cannot do in PTE is, for example, build a cube and then work on the assumption that the three axes of that cube (which we as people would probably refer to as the cube's X, Y and Z axes) remain the X, Y and Z axes as the cube changes its orientation. The Rotate X control does behave in this manner. If you apply some "static rotate" on the Y axis and then rotate on the X axis, the effect is as you would expect. Similarly, the Rotate Y control behaves entirely as you would expect. But the only Z axis is that which is perpendicular to the 2D plane of the monitor; and the only controls are Rotate about Z (from the Animation tab) and Pan Z from the 3D parameter window.

I found this very confusing and non-intuitive when I first tried to get to grips with the 3D animation capabilities.

regards,

Peter

[KenAdam]

Thanks for all the help.

I think I've got the sequence I was trying to work out in the first place, although there may be a better way of doing it.

You can see my results in this small project

http://www.standpretty.co.uk/pte/SlideAround.zip

featuring one of our chickens.

I wasted some time, as I set the movement on the image, then realised that I needed it on a frame containing the image.

I managed to fix that by cutting and pasting inside the project file, as I couldn't spot how to copy the animation sequence from one type of object to another in PTE (is there a way?)

Now I just need to decide if I actually like what I was trying to achieve for the show I'm making, but I need to chain a few slides together for that.

[Guest Yachtsman1]

Jerky on my machine, have you ticked the mipmapping option in O&A?

Yachtsman1

[KenAdam]

I have checked it (I'm sure I had already in the pte file that is in the zip).

It is a little jerky on 1 PC here, but not on another. (Both with similar Windows performance scores).

I've tried various settings (low quality, edge anti-aliasing etc) but get the same result.

The image is not very big (much smaller than the one that Lin had in the sample).

I'd appreciate suggestions for anything else that might be causing this.

Edit: I just went back to the "slidein.zip" file that Lin posted above, and it shows the same jerky result on this PC.

I'll start a new project and build up the effect in steps...

[Lin Evans]

Hi Ken,

The larger (in dimensions and file size) the file used for this type of animation, the more likely there will be jerky movement. The "anti-shimmer" or mipmapping really has nothing to do with this issue. There is lots of complexity in this type of animation and only with top-end video cards will certain motions be smooth.

In my development system, I have what was at one time a top-end video card (the nVidia 8800GT) but which now isn't even close to the really good ones. If I use the large eleven megapixel image used in my demo and try to zoom on the "Z" axis while panning a 3d transformed image rotated on the Y axis, the motion is jerky even on my system which generally will play "almost" any PTE animation smoothly. If I resize this image to around 1084x768 I can do it smoothly. It's just asking too much of most systems to create on-the-fly 3D transformed images of large sizes with "Z" axis zooms simultaneously with pans.

The answer is to use as small an image as useful for your viewing audience. I wouldn't recommend doing this type of animation even with a two megapixel (1920x1080) image unless you are certain that the system your show will be played on has a top-end video card and plenty of available RAM. Also be certain that no anti-virus software is running in the background and usurping resources.

Best regards,

Lin

I have checked it (I'm sure I had already in the pte file that is in the zip).

It is a little jerky on 1 PC here, but not on another. (Both with similar Windows performance scores).

I've tried various settings (low quality, edge anti-aliasing etc) but get the same result.

The image is not very big (much smaller than the one that Lin had in the sample).

I'd appreciate suggestions for anything else that might be causing this.

Edit: I just went back to the "slidein.zip" file that Lin posted above, and it shows the same jerky result on this PC.

I'll start a new project and build up the effect in steps...

[KenAdam]

Lin,

My desktop which runs the demos fine is an 8800GTX.

The laptop I'm using at the moment ia an ATI Radeon HD4330, which is jerky.

Curiously, my digital map software (panning, zooming, rotating, fetching and decompressing tiles to give worldwide coverage, with complex threat and terrain overlays), which normally kills any low performance graphics card, runs fine on both.

Perhaps the ATI chip does better for OpenGL than it does for DirectX.