JPD

As I promised, here is the template "Screensaver.pte" of the photo album (56 photos), it's now abble to run on 32 Mb graphic cards. It's a completly new one, lighter than "Elle le dit..." it use less than 25 Mo of memory. You can use it to do screensaver with your own photos (family for instance) or put it (or only a part) in your slideshow. The explanation about how to use it are inside. If an english spoken people could correct the english explanations and send them to me, I could put in the zip a better english explanation, my english is really poor. You can also find this template on my page dedicated to PTE V5 were you can find many tests and exemples with V5. Most of them are in french, there is a good tool to translate here and Google has also a tool for that.

At least I succeeded to download it, this version is better, without tourists and less panoramics which were not really necessary. It's interesting to use PAN and Zoom, but not too much, that's OK like that. Good job.

Sorry, I have forget the good explanation. Look at the third slide of the exemple. The blue rectangle is this which give the number of ellipses. If you start at 0 to go at 3600 you will have ten times the ellipse. The red rectangle must have twice the negative value of the blue one so you will start from 0 and go to -7200 If you want to start at the opposite, the blue one will from 0+180=3780 and its end at 3600+180=3780 and the red one will start 180 x -2 = -360 and finih at 3780 x - 2 = -7560 If you want to start at 90° replace 180 by 90, if you want to start at 330° replace 180 by 330 (or -30, it's the same thing). If you do which is above, the ellipse will be horizontal. If you want to have it vertical (-90°), add 90 x 2 = 180 at the values (beginning and end) of the red rectangle, if you want it start at -45° add 45 x 2 =90 at the 2 values. If you want to have several objets on the same orbit, they must start at differents degres so you will add for instance 0 on the blue values of the first one, 72 for the second, 144 for the third one, 216 for the fourth one and 288 for the last one, and on the red one, 0, -144, -288, - 432 and -576, they will be at the same distance each from other. If you don't give any value to the electron, it will have always the same part looking to the center If you want it is always in the same direction, put a value equal at the negatine sum of the red and blue rectangle For the first exemple above it would be -3780. You can add at this value a constant value if you want to change the direction were is lookng the electron. I hope it's clear for you, because it's very difficult for me to explain in english.

Thanks Al for your comment, thats's right, I use the vertical zoom during the page turn in order it stay inside the frame. To day it's impossible to have the trapeze we need for that on a picture. The frame is made with PNG file and is above the photo. Thanks Bernard, Danielle will be happy that you like her photos. I was making a template for a screensaver in which people can put there own photo (child, family or flowers, all what they want), when she saw that Danielle ask me to put her photos, what I did, and it's the result. For the screensaver I have just a little problem, to day if you move the mouse, the screensaver don't stop, we have to wait a little for that, we are obliged to use escape. When I finish it and write explanations to use this template you will have it. You will have possibility to copy several slides of it with the Hawk's method to put them in your own slideshow, no problems. It will be possible to move, zoom, rotate the book when the pages turn (but this need a good PC). With this system, it's possible to open (or close) a window.

Here is a little slideshow made as a photo album, there is absolutely no calculation to do it. I hope you will enjoy.

I am just a man, as everybody. I am happy that previous demos is greatly appreciated. Thanks.

You just have to change the value of start and finish angles on the hight level object (for instance substract 30°) and to change with the same inverted value on the lowest level object, that's all, the object will run on the same orbit but at another place. Look at the slide 7 of the exemple "Ellipse" 21/08/06 on my page, there are two objects on each ellipses. If you add 180° on the first level (at beginning and end) and substract 180° at the lowest object's level, the new object will be exactly in opposition with the same without modification.

Lin, I hope you will soon recover your possibilities to see as well as before, and you will be abble to continue to do so nice photos you do.

You made a vey nice work with many good ideas and also found nice photos from Teide, the choice of the music is also very good. I enjoy it. Sorry, I don't speak spanish and my english is bad :-(

Normally yes, even sometimes it seems to me to have best result with a hight resolution for some slideshows, the tests confirm it's quicker with low resolutions.

When I read what Al wrote, I understand that I didn't understood what Igor wrote, I agree that the screen definition is important, but I also found that the size of the pictures is important too. I made in november hundreds tests with different sizes of pictures at different screen definition and I found with a method we call in french "les moindres carrés" an approximative formulas with a little fixed part of time for each picture plus a time depending of the size in pixels of the picture plus a time depending of the size of the picture on the screen. Here is a little part of the results I found In order to have a simplest rule I search the values without the fixed part, the results where at +/-15% with the first rule they are at +/- 18% with the second one, so it's enough good. On my PC the values I gave are about 4ms/ Mpx on the screen, 3.2 ms/Mpx for the picture. Notice that a rectangle with the same size than a picture is quicker than the picture : it's because it has only the time of calculation for the screen, not for the picture itself (of course, it's a rectangle). That mean there is a time of calculation for the picture itself. The values I found are for my PC, but I think it must be the same for all PC with others values : a part for the picture, a part for it's image on the screen. I am sorry to have too much difficulties to well explain what I found.

Well, Igor there is something that I don't understand, I have made hundreds tests to find the values I give above, one of these tests is here. I put 2 slides, the first one in which there is nothing (just a very little file 1 x 1 pixel, and the second one in which there are several times the same picture (to test on my PC I put 100 times the same file, for a more power PC it will be necessary to put more times this file to be abble to count the number of images which are calculated). between the 2 slides there is a rectangle transition of 99 seconds. I can count the number of images that PTE calculate during these 99 seconds. On my PC, this exemple give 99 images/99 seconds for 1600 x 1200 screen definition and 153 images for 99 seconds with a screen definition of 1152 x 864, and all the others test I did last month say the same thing. With my Radeon 7500, it take more time to calculate the same picture with 1600 x 1200 than for 1152 x 864. It seems to me, before, when looking a slideshow that it was smoother with hight screen definition, but the test above say the contrary and all the other tests I did have the same result +/- 18 % I did also a test which is on Diapositif which permit to know how many of Ms/s are between the memory and the memory of the graphic card. As I use bmp files and that there are preloaded at the begining of the test there is only the time between memories. On my PC, I found 17 Mb/sec (2.4 ghz), one of our members found 24 Mb/s, that mean that if we want to be good on the most part of the PC today we have to calculate the time of loading between memories with a rate between 10 and 12 mb/sec. (+ read from hard disk + calculation for jpg or png). I am sorry, it's in french, but what is important is the result. The explanations are inside the template, you can translate with Google if you want to use it.

--> (Mr B @ Jan 1 2007, 11:36 AM) ←I to have a NVIDIA GeForce FX5200 and get jerky transitions from the Forum members shows in PTE5. I understand the need for more processing power but how does this help when producing a show that is played on a lower spec machine? Please try this 2 slideshows and tell me if there are jerk : Test PZR 1280 Fantaisies florales Thanks I think so.

Which is important is the size in the memory which is Width x heigth x 4 in Bytes and for calculation the size in pixels of the part of the picture which is in the visual zone, the size of the jpg has no matter. The calculation are on all pictures which aren't at 0% for opacity and only for the part which is in the visual zone, even there is another picture above. For the part which is out of the visual zone, there are some calculations but very smaller (about 15% on my PC). Always on my PC, it need about 3.2 ms/Mpx of the size of the pictures which are inside the visual zone and about 4 ms/Mpx for the calculation of the picture in the screen definition. For the rectangle, there is only calculation of what is in the screen definition even if opacity is at 0%. If I have two pictures 1280 x 960 inside the visual zone and a screen definition of 1600 x 1200, my PC need : 1280 x 960 x2 x 3.2 ms/Mpx = 7.5 ms + 1600 x 1200 x 2 x 4ms/Mpx = 14.6 ms, total = 22.1 so the frequency is 1000/22.1 = about 45 hz, which is good, this number must be above 25 hz to have smooth effect. In this exemple, I just have the calculation for a rectangle transition. 1 Mpx = 1024 x 1024 pixels Note : during the transition, the calculation are on two views and if there are several times the same object, there will be the same number of calculations than the number of object, but it will be only one time in memory.

Hi Nicolas, Best wishes for the new year.

It happens some times, but I haven't find to day a good exemple to give to Igor. What also happen on old graphic card is a non linearity in the effect when the volume of calculation change. The problem (start late and jump) is not only on zoom, but also on rotate and horizontal Pan (I never see that on vertical pan, I don't know why). The reason seems to be too much calculations at the same time.

Merci Maureen, Thanks Robert and Lin. I hadn't enough time to correct some defaults, next time I will ask to put Christmas on the 27th of december :-) Merry Christmas and Happy new year for you and your family.

Very nice show and very nice photo. Many thanks Bernard.

Thanks Al, thanks Ken for your wishes. Ken, it's a good idea :-)

I have made a little show to wish you a Merry Christmas and Happy New Year. My best wishes to Igor and his team.

I agree with you that it will be better for everybody, even for me, when there will be the function, but use the future, as I read, about the result we want isn't right, it will be only easier and quicher to do. Many people want this function for opacity to have better 3rd image for instance, it will be a good thing to do that quicker than to day.

It's possible today with PTE. You just have to use the keypoints for that.

I made a little tutorial (in french only) at this adress which explain how to do that enough easily and give a very good result, for Pan, zoom, rotate or opacity or several of them. It's PTE which make itself the most part of the calculations.

I don't want that my files about test, demo and my PTE files needs to be log to download them, there are free on my page You knew my position about that, so please move them from your page. Thanks.

Thanks Igor, I must wait to have a LCD projector and hope to obtain the same result, it must be astonishing to see.