Moving pivot with grid snapping

[uuderzo]

Hello.

I have the following issue (using v6): i have a sliced picture that i want to "rebuild" into a slide.

The picture is made of several layers (separate images), but not all layers are the same size because they cover different parts of the same picture. But, when correctly assemblied, the whole picture is shown.

I position each layer by looking at each layer coordinates in photoshop, then copy layer coordinates into pte and that's ok. Now, my need is that each layer to have the same pivot point (screen center) because i need this to obtain my animation effect. I turned on the grid but when i try to move the pivot point of each layer (with SHIFT + drag) it doesn't seem that the pivot snaps to grid points. Then it's difficult to correctly align each one to screen center.

How is possible to obtain such grid snapping?

Thanks... Umberto

[thedom]

Hi Umberto,

I think a solution could be to keep the transparent pixels of each layer when you are in photoshop.

This way, all layers will have exactly the same size and you won't have to adjust them in PTE.

[xahu34]

Hi Umberto,

Did you try the Size/Position tool (Original mode tab)?

Regards,

Xaver

[trailertrash]

I position each layer by looking at each layer coordinates in photoshop, then copy layer coordinates into pte and that's ok.

How is that done please?

Andrew

[uuderzo]

@Trailertrash: zoom 100% on photoshop, then CTRL + click on layer to select it and visually find the topmost and rightmost point of the selection with the mouse. Then i write down the px coordinates shown in info panel. Not straightforward but i didnt' find a better solution.

@TheDom: i exclude the transpatent pixels to get the smallest possible images, because some slides work with lots of high-res images (10-12) and this way lets me optimize the video memory occupation. But that's cumbersome to handle...

Greetings! Umberto

[uuderzo]

Or... maybe i missed that...

Is PTE capable of analyze PNG files content and automatically exclude big tranparent areas at runtime while keeping the whole image area when working in O&A editor?

If so, i'd have lots less problems.

[fh1805]

Umberto,

I seem to recall that I saw a post from Igor during v6.0 beta programme that confirmed that transparent pixels (as in a PNG image) add no processor load to the graphics rendering. Obviously they might (will?) add bytes to the memory but the implication seemed to be that there presence would be a static load rather than a dynamic load. It might be worth a quick experiment to try out Dom's suggestion.

regards,

Peter

[Barry Beckham]

Is it just me or are you all making hard work of this. Read the Dom's answer again, he has it.

[xahu34]

Hi Umberto,

In my opinion, it should be possible to program a spreadsheet for the parameters that you need. As depicted in the attached image, you may deal with two coordinate systems, the one of the screen (green), and the one of your image (orange). The positions ( x, y ) of the image and ( a, b ) of the new pivot (both relative to the green coordinate system) are given. The old pivot has the coordinates (100,100) relative to the orange system. What you need to determine is the size of the dotted vector relative to the orange system. This is a task that should be manageable. Good luck

Regards,

Xaver

[uuderzo]

@Xaver: Wow so much effort ... I only asked if it was possible to snap pivot to the grid

Finding the center of the screen is not an issue, i only tought that it was more fast to have a snapping support.

Anyway, i managed to align 6 pictures this evening... now i'm used to it and i feel possible to get the result i have in my mind

@BBDigital: It's a challenge

@Peter: this is good news, anyway i prefer to keep on the different layers size. I need full video card power (read ac-plugged laptop) to get a smooth fading between two rotating parallax pictures... better to stay to cheese side

Anyway, now that finally could begin to assembly all my slicing work into PTE, at first glance looks good... when i have something pretty decent i'll post a video to the forum

Greetings! Umberto

[potwnc]

"Read the Dom's answer again, he has it."

I'm doing something that I think is similar and I'm using the approach of keeping the transparent pixels so all the .png files are the same size as the master .jpg file. I can post this if you like.

Ray

[uuderzo]

I'm doing something that I think is similar and I'm using the approach of keeping the transparent pixels so all the .png files are the same size as the master .jpg file. I can post this if you like.

Ray

Thank you for you offer, Ray. Anyway, i finally get used to center the pivot in nearly no time comparing to the slicing time, so this is no more an issue for me. I still think that minimizing layer sizes is still good for me. I added (only) 14 parallax pictures to my slideshow and i reached 200 mb exe size. Considering that this is equivalent to about 90 seconds of slideshow and i need to reach 20 mins...

Greetings! Umberto