O&A : Original size mode is needed

[Picsel]

Original size mode is needed by photographers who want to keep the original image size.

Several years ago there was a debat about original size mode. If I remember well, JP Dollangere was a very strong supporter of this option. He was right!

Most of photographers, who make AV shows, want to keep their original image size and do not want to see their work be modified by software. As it is today, even in v750, PTE does not offer this possibility.

If you want to keep original image size, image by image, you have to

1) select a mode (fit to slide or cover slide)

2) set the right size and position in size/ position in pixels.

Doing that, the zoom values displayed by PTE are false and require specific calculation.

That is taking too long time and is a source of errors in case of animations!

It would be very interesting to have a third mode "original size" which will allow to keep the original size of their image at zoom= 100%.

This option will also ease a lot the work for those who want to make accurate 3D animations.

First example.

Screen = 1920x1080

Image original size = 400x400

0&A

Common / Mode

1) Fit to size => image size becomes 1080x1080 / O&A > PTE zoom value =100% (that is false, it should be 270= 1080/400) see att 1 :

2) Cover slide => image size becomes 1920x1920 cropped at 1920x1080 / O&A > PTE zoom value = 100% (that is false, it should be 480= 1920/400) see att 4 :

The fact that zoom value displayed by PTE is false may introduce errors in the choice of image resolution by the AV maker

Then, if I want to recover the original size of my image I have two choices :

1) select Fit to slide

- then size/ position in pixels

- set size 400x400

but in this case PTE Zoom value is no longer 100% but 37,037 although the real image ratio is 1/1

see attachment 2,3

att. 2 : att. 3 :

2) select Cover slide

- then size/ position in pixels

- set size 400x400

but in this case PTE Zoom value is no longer 100% although the real image ratio is still 1/1

see attachment 5,6

att 5 = att 6 =

If you want to make accurate animation using original image size, then you have to modify all zoom value given by PTE depending on the selected mode. That is not logical at all and too much time consuming.

Second example

If you choose "fit to slide" mode for all images whose height > width, then whatever the image definition in pixels, PTE will give them the same number of pixels in height (slide height) and the same zoom value =100% which is totaly wrong and not logical at all. That is very confusing for those who wants to work keeping original image size.

Daniel

[uuderzo]

Daniel,

I well understand your problem. I had a lot of difficulties correctly scaling a lot of images into PTE because they were part of a unique animation and they must all be correctly scaled. And, too bad, all the images were different in size.

Since PTE works (correctly IMO) with a resolution independent approach, the above operation is a pain and i think it has no solution while you work with each single picture.

But maybe adding a kind of "proportional scaling" that works on a group of selected images may help.

Let me explain the idea. Load a bunch of pictures into the tree structure. Select them all. Then apply the "Fit to slide" or "Cover slide" command. PTE should consider the entire group of pictures for the operation and execute the scaling by considering the outer bounding box of all overlapped pictures (or max width and max height) and proportionally scale all pictures within this bounding box. This keeps the "resolution agnostic" approach of the command when used with a single picture.

Greetings! Umberto

[davegee]

Doesn't putting all images in a frame and scaling the frame achieve the same thing?

DG

[uuderzo]

Doesn't putting all images in a frame and scaling the frame achieve the same thing?

DG

No, you are missing the point. I explain.

Suppose that you want to work with a 100x100 picture and a 500x500 picture, but you want the first picture to get a zoom factor 1/5 of the second picture, so you can keep the proportions (pixel size speaking). It's easy to do manually in this case, but think about dealing with 20 pictures, each one with strange sizes (1920x1080, 3260x1560 for example). To get the exact zoom proportions you must start working with a calculator or an excel sheet. And if you need to work with lots of slides in such a way, it's a pain.

My approach is currently to load the biggest picture, set it at 100% zoom, then load other pictures, overlap them on the first picture and work with zoom factor and offset to match them. In my case it was possible because every picture was part of an entire "animated" picture so i had guidelines to act such a way.

Obviously, the suggested proportional scaling doesn't fix offsets, but at least you don't have to deal with zoom factors. Once you correctly scaled all the pictures, you can wrap them into a frame and move it.

Greetings! Umberto

[davegee]

It sounds very complicated.

Would an example help your argument?

Another thought - if the images were all sized "correctly" (all the same height or width) what would that do?

DG

[uuderzo]

It sounds very complicated.

Would an example help your argument?

Another thought - if the images were all sized "correctly" (all the same height or width) what would that do?

DG

It's not really complicated, it's more difficult describe it by words.

Answering your first question:

Yes i have a real life example: in my last slideshow i have many "fake 3d" scenes built with several layers stacked in 3D space. Each layer has different location on Z axis to achieve a good 3D rotation effect. Moreover, each layer has different size because of the structure of the image (some layers are small details of the entire scene). My need is to import all layers so that every layer keeps the right proportional size respect of each other.

My situation is a bit more complicated because i obtained all layers from a single picture so placing them at different Z locations requires an additional zoom level adjustment to match them when seen from the front. But this is a different issue. Suppose we don't need this.

Then you see that importing a group of pictures mantaining their respective size is an useful option.

Answering your second question:

If the images are all with the same height and width then they overlap perfectly, hence the bounding box of all images matches with all of them and the images are managed as if they were a single picture. So, Fill frame will fill frame with all images, like if you did it one image at a time.

Greetings! Umberto

P.S. I don't know if initial Daniel requirement is exactly what i'm talking about, please Daniel let me know.

[xahu34]

...

Since PTE works (correctly IMO) with a resolution independent approach ...

That's true, and also in my opinion, the PTE model works well. 100% means that an an image exactly fits or covers the screen (independent from the real, fixed, or virtual size of the screen, and independent from the pixel dimensions of the image. This is a definition, and as a definition it cannot be false. If someone does not like this model, that's ok. On the other hand, PTE works with scene graphs (parent/child model) and relative coordinate systems, and from this point of view, the parameters are well-defined (IMO).

Regards,

Xaver

[Barry Beckham]

If I understand the point correctly, I made it years ago, but not in relation to animation. Not sure it even existed when I raised this. I was saying that it seemed wrong to me that every picture brought into PTE that was smaller than the main screen resolution set, will be expanded up to either:

Reach the edges or the top/bottom of the screen depending on the format of the smaller image. The only option to put this right is to visit evey single image (that is smaller then the set show resolution) in the Objects and Animation screen, select the Common tab > Size position in Pixels > Original Mode and then reset the value to 100%. I said back then that there should be a glabal setting for this.

It would make a nice preferences setting.

It is annoying when you create a small inset image and PTE expands it up to try and fit it to the size of the screen.

[uuderzo]

That's true, and also in my opinion, the PTE model works well. 100% means that an an image exactly fits or covers the screen (independent from the real, fixed, or virtual size of the screen, and independent from the pixel dimensions of the image. This is a definition, and as a definition it cannot be false. If someone does not like this model, that's ok. On the other hand, PTE works with scene graphs (parent/child model) and relative coordinate systems, and from this point of view, the parameters are well-defined (IMO).

Regards,

Xaver

I'm with you on the resolution independent definition. PTE does a good job on this side and it's a big help to avoid the user dealing with pixel resolution sizes. But, on the other hand, sometimes there is this need to keep a set of pictures proportionally sized. I'm not saying that we must start to deal with pixel sizes, i'm just saying that it can be helpful a feature capable to consider the (hidden) pixel size to correctly correlate a set of pictures appearance through zoom level.

Read it as a kind of "macro" that does the dirty work for you. But there will be no more parameters in the interface. Maybe only a right click option on a set of selected pictures. I think it will take only few code lines and make happy many users :)/>

Greetings! Umberto

[Picsel]

P.S. I don't know if initial Daniel requirement is exactly what i'm talking about, please Daniel let me know.

Umberto,BBdigital

Umberto, ...not exactly but the consequences are the same.

The problem was well explained by bbdigital in his last post. Some photographers in our club are complaining that they are enable to put their photography on a slide without PTE recalculation. And as I said and I agree with you if you want to make a slideshow including a lot of photographies that becomes painful with useless work. Obviously it is worst in case of animations.

That's true, and also in my opinion, the PTE model works well. 100% means that an an image exactly fits or covers the screen (independent from the real, fixed, or virtual size of the screen, and independent from the pixel dimensions of the image. This is a definition, and as a definition it cannot be false. If someone does not like this model, that's ok. On the other hand, PTE works with scene graphs (parent/child model) and relative coordinate systems, and from this point of view, the parameters are well-defined (IMO).

Regards,

Xaver

Xaver,

perhaps I was not accurate enough! There were 3 modes

- fit to slide

- cover slide

- original size

I agree with you, PTE works perfectly well in case of fit and cover modes but that is not a convenient way for everybody!

(I disagree with what you said about "definition". When I said PTE gives false zoom values, I was also refering to zoom value given when you use size/position in pixels function. You were speaking about zoom value in case of using only fit and cover modes that is not my point here...although even in fit and cover modes the quality of the slideshow is depending on the zoom ratio which is applied to the images as it changes the image definition! If whatever your original size image the information given by zoom value is 100% you may be lost!)

As I try to explain above, some photographers would like to use their photography in original size without PTE recalculation and without loosing their time specifying image by image the dimensions of their original photography that PTE should be able to get via the exif for instance.

Regards

Daniel

PS : IMO???

[Picsel]

....it's a big help to avoid the user dealing with pixel resolution sizes. But, on the other hand, sometimes there is this need to keep a set of pictures proportionally sized. I'm not saying that we must start to deal with pixel sizes,....

Greetings! Umberto

Umberto,

I cant follow you, it is very important for an AV maker to know exactly what are the sizes in pixels of their images.

Daniel.

[xahu34]

...

PS : IMO???

IMO = In My Opinion (Internet slang expression)

Regards,

Xaver

[xahu34]

... I disagree with what you said about "definition" ...

Sorry, I am a mathematician, and people of that kind do not discuss on the correctness of a definition (as long as it does not include a contradiction). A definition is correct by definition, and I just gave the definition of 100% as it is in the present mathematical model used by PTE. You do not seem to like this model, but it is well-defined and free from contradictions. It is up to you to propose another model, or to extend the present one by re-introducing some kind of original mode. Why not? Keep in mind that an original mode typically refers to a screen resolution, and the machine on which you program the show must not have the same resolution as the monitor on which the show will be presented. The present size/position tool in PTE takes care for this aspect.

Regards,

Xaver

[uuderzo]

Umberto,

I cant follow you, it is very important for an AV maker to know exactly what are the sizes in pixels of their images.

Daniel.

Daniel, perhaps we have different requirements.

In my opinion it's not important to know the exact size in pixel of a picture, and i appreciate the way PTE works.

But sometimes i need to put on screen a set of pictures that must keep their relative size. I mean, if i need to work with, say, a 500x500px and a 100x100px image, sometimes i wish that they appear on screen the first 5 times bigger than the second.

I don't really need how many physical screen pixels they will take, just because PTE is resolution agnostic and it does a good job rescaling all pictures at the current work resolution.

Obviously, i must be aware that a picture needs as many pixels as the surface it will cover on the slide, just because if not i fall into interpolation losing picture sharpness. But really the physical size is not important IMO.

By working resolution agnostic, you can get your project and apply at different resolutions.

In my last project, i rendered it three times, one at 1440x1080 at 60FPS, one at 1024x768 at 60FPS and one at 1024x768 at 30FPS with just few clicks.

If i had my project pixel-resolution tied, i'd have a lot of annoyances during this conversion.

Greetings! Umberto

[Picsel]

....You do not seem to like this model,...

Xaver,

among other people who ask me to report this question there is a mathematician, member of our club and photographer, very funny indeed!

As I said in my previous post, the model you are speakink about is perfect for me within the both options

- fit to slide

- cover slide

but you should consider that zoom 100% does not mean the same thing in both cases and so has different "definitions"....

let assume that

- slide dimensions are W (width in pixels) and H (height in pixels)

- image dimensions are w (width in pixels) and h (height in pixels)

- z= real zoom value applied by PTE to the image

then :

in case of "fit to slide"

if (w/W)<(h/H) then z= w/W displayed as "100%"

if (h/H)<(w/W) then z= h/H displayed as "100%"

in case of "cover slide"

if (w/W)<(h/H) then z= h/H displayed as "100%"

if (h/H)<(w/W) then z= w/W displayed as "100%"

So you must admit as a mathematecian there is not only one definition but 4 different definitions for zoom=100%

As I also said previously : a lot of people, mainly photographers do not want their photographies to be recalculated by PTE and want to keep the original size of their image.

To day, as it is, they have to select size/positions in pixels then specify or click on size, and they have to do that image by image. The zoom value displayed in that case is wrong, it should be 100% for original size (that is just another definition of zoom = 100% for z=1 in such a particular case !)

So to be clear enough, I just would like a third option which could ease the way to do that. All the "ingredients" already exist inside PTE.

.... Keep in mind that an original mode typically refers to a screen resolution,...

I disagree, the original mode refer to image dimensions, the screen resolution has nothing to see with that but obviously we must assume that image dimensions are compatible with the slide format which has been specified.

Regards

Daniel

[Picsel]

Daniel, perhaps we have different requirements.

In my opinion it's not important to know the exact size in pixel of a picture, and i appreciate the way PTE works.

But sometimes i need to put on screen a set of pictures that must keep their relative size. I mean, if i need to work with, say, a 500x500px and a 100x100px image, sometimes i wish that they appear on screen the first 5 times bigger than the second.

I don't really need how many physical screen pixels they will take, just because PTE is resolution agnostic and it does a good job rescaling all pictures at the current work resolution.

Obviously, i must be aware that a picture needs as many pixels as the surface it will cover on the slide, just because if not i fall into interpolation losing picture sharpness. But really the physical size is not important IMO.

By working resolution agnostic, you can get your project and apply at different resolutions.

In my last project, i rendered it three times, one at 1440x1080 at 60FPS, one at 1024x768 at 60FPS and one at 1024x768 at 30FPS with just few clicks.

If i had my project pixel-resolution tied, i'd have a lot of annoyances during this conversion.

Greetings! Umberto

Umberto

I understand your point.

That is your way of working, i do not want to say it is right or it is wrong if it is convenient for you.

But all AV makers who want to keep a control on the quality of their slideshow must know perfectly well the dimensions of their image.

I think, that in your post, you are confusing image resolution (which also depends on display dimensions) and image dimensions ( or definition); I am speaking about image dimensions!

Just another little comment about your FPS : as an European guy ,I am surprised to see that you are using american standards!

Regards

Daniel

[uuderzo]

Just another little comment about your FPS : as an European guy ,I am surprised to see that you are using american standards!

Regards

Daniel

You are right, i'm using american standards. But my work is designed for use with a digital projector directly attached to the PC or iPad and those FPS play well with this configuration, letting me squeeze a bit more smoothness (you can notice the difference from 25 and 30 FPS).

Anyway, i also understand your point. Your workflow is aimed to the maximum image quality and you don't want any PTE intervention on interpolation, that may slightly blur your sharpening. This requires you to deal with screen resolution and image resolution and to work well to match them. I simply never approached this way just because i use a lot of slight pans and zooms, and when you involve zooms the sharpening is blurred. So... i tend to avoid thinking about this issue. But, i repeat, i understand what you mean.

Greetings! Umberto

[Picsel]

...This requires you to deal with screen resolution and image resolution and to work well to match them. ...

Umberto,

to be clear, an image has dimensions (definition in pixels x pixels) but cannot have a resolution (pixels/inch)without speaking about the dimensions of the support on which it is displayed. If your slide is 1920x1080 and the image on the slide is 400x400 it will keek this definition whatever the dimensions of the support on which it is displayed except if you are zooming in. So that is the reason why the dimensions are so important. Depending on the "support/display/screen" dimensions, the resolution will change (but the dimensions of the image remain the same) and you have to take care of that only for the distance between the "support/display/screen" and the "spectator"

That is also the reason why some photographers would like to keep their work without any uneeded "manipulation" which is not under their control.

Daniel

[ComputerUser]

Maybe only a right click option on a set of selected pictures. I think it will take only few code lines and make happy many users :)/>

Yes, I'd very much like to Ctrl-select several slides, right-click and "Set to Original size"