Proper Way to Manage PNGs within Slides

[BBorges]

Hi guys,

After a zillion tutorials and videos I few a lot more comfortable with the software, though I am still very far away from what I actually need to create.

I have some fundamental questions that I would like to clarify before I attempt to work on this project and then realize I am headed in the wrong direction.

Background: I would like to create a slideshow where certain components of each image "detach" from it and move around the frame. My understanding is I need to create PNG of these cutous to achieve this effect. With that in mind, I went to the photo editor, made selections and created the PNG cutouts. So now coming back to PTE I have the following issues:

1) do I use one slide and put all images/PNGs as objects in this single slide or use multiple slides?

2) Is it recommended to utilize frames for something like this?

3) Should the PNGs be linked to its main photo or not?

4) what is the best structure this presentation?

I did attempt a slideshow and here is the structure I used:

- (1) slide with (4) pictures as objects inside of it (no frames)

- Pictures were all rooted (not child relationship)

- The structure looked like this:

Objects:

- picture1

- picture1DNG

- picture2

- picture2DNG

Here is the result slideshow:

https://swa.sharefile.com/d/sfe063b701044bdda

Trying to create the presentation this way seemed very counter intuitive, because I will need a LOT more than two photos (an two correspondent PNGs) and it was very hard to keep track of the KeyFrame Points. Very confusing, and I had to keep on trying to match the PNG size/location to the original JPG so it looked aligned, til the moment the PNG would make its move.

I suspect I should be using either frame or child-mother connections with this type of slideshow, but I am not sure.

Any help with this regard would be highly appreciated.

Thanks in advance,

BB

[Lin Evans]

Hi BB,

There are a number of issues you need to consider. First, for the brief time your show ran, the memory size was pretty large so you may want to use smaller images in terms of initial size. For example, if you are going to show the slides on an HD TV, the only reason for them to be larger than 2 megapixels is if you are doing deep zooms, etc. So first resample the originals and save in no larger than 1600x1200 pixels.

The second thing is that "if" precise position is essential, then you want to keep all the excess "transparency" surrounding the image for the PNG. That way the alignment will be automatically perfect. Otherwise you will have the issue of trying to resize the png's and precisely position them. That takes lots of time.

The next issue is, if you want the image left behind the png cutout to look normal, then you will want to remove the subject from the original after creating the PNG. This way, you can clone over the area left from the cutout and use a cutout with much smaller overall file size. You can then remove the excess transparency thus making the memory load "much" smaller for your show.

If you have Photoshop or a similar program, it should be fairly easy to clone background trees, grass, whatever, over the area where the subject originally was. Then your cutout can be sized and placed in the approximate original position and when it moves, what is left behind will look normal.

The only reason to use frames for what you are doing is if you want to move two or more png cutouts simultaneously. Otherwise there is no real advantage.

If you could verbally explain exactly what you want to accomplish with the png cutouts, perhaps I could suggest the best way to proceed.

Best regards,

Lin

[BBorges]

Thanks for the detailed response Lin.

1- I will re-size the photos accordingly the next time. I had received guidance on this before but I didnt realize i needed for what I thought it'd be a small slideshow. I guess it emphasizes how important this is.

2- If I understand what you call "excess border" I think I did have that, because once I insert them as objects, they line up correctly (and they have the same frame-box as the original picture). The way I created then was very simple: open original JPG, double the layer, make the selection, invert selection and delete everything, then delete background and save-as PNG. I am attaching a paired JPG/PNG so if you could tell me I am creating the files accordingly would be great as well. The position issue I was mentioning is after I move the JPG around, I have to find the keypoint for the PNG and match the timeframe and then match Zoom/Position. It seems like every time I move the JPG, I have to make sure the PNG is following the same trace.

In this case, making the PNG a child of the original would solve the issue of coordinating the movements with the PNG

Your method of creating the PNG is fine.

3- Not totally clear here: you are saying that I should remove the cut-outs from the original JPGs as well? I could do that, but which pixels will make up the image then? Black background? I can almost see what you are trying to say but I am not clear how this would work or help me. Could you clarify this a bit?

In the case where the PNG remains and obscures the original image, there is no need. What I was alluding to might be seen in a scenario like this: Let's say you resized the cut-out much smaller. If so, then the original jpg would hold a "duplicate" of your cut-out allowing both to be visible simultaneously. So in the event that you needed to make it appear that the cut-out was "leaving" the original image, you could clone something over the subject in the original jpg so it would appear that when the cut-out zoomed out and/or moved it's position, there would be no duplication.

4- Thanks for clarifying about the frames. I wasnt sure if I should be using it or not.

5- Lin, let me try to explain the goal. For clarity, let's say "JPG" is a regular photo with elements in it, and "PNG" is a cut-out of one element from the same "JPG". With that in mind, lets say I have a JPG_1 slowly moving (or zooming) in the frame, and at some point towards the end of the movement, the correspondent PNG_1 moves faster and takes over the frame. Right at that time, a second JPG_2 fades in across the frame (or zooms) and towards the end of that movement, the second PNG_2 takes over the frame. And this goes on for a list of JPG/PNG.

I think so (clear). In such a case you probably won't need to clone over the original and you are presently on the right track. Just make the PNG a child of the jpg (no need for a frame) and the motion will correspond until such time as you decide to move the PNG independently. A "child" can have it's own zoom, pan, rotate, etc., independent of the parent, but will inherit the parent's motion, zoom, etc., as well.

Is that clear?

I understand the PTE file I created was very limited, but it does illustrate the goal, though it needs a lot of refinement.

Any other suggestions?

I think you are proceeding correctly with the exception of making the PNG a child of the jpg.

Thanks again,

BB

[Lin Evans]

see above,

Lin

[BBorges]

In such a case you probably won't need to clone over the original and you are presently on the right track. Just make the PNG a child of the jpg (no need for a frame) and the motion will correspond until such time as you decide to move the PNG independently. A "child" can have it's own zoom, pan, rotate, etc., independent of the parent, but will inherit the parent's motion, zoom, etc., as well.

Nice!

That will definitely save me half of the trouble!!

Thanks for the tip! I knew there had to be a simpler way of doing it!

If anyone has any other ideas, it is very welcome!