A new test for 32 Mo graphic card

[CorVdK]

I just want to make a call to every member of this forum and every user of PTE to be open minded please.

There are so many clever people here (and also on the French forum) who do the impossible to try to put everybody on the same level.

New ideas are coming very fast for the moment and it is not always possible to absorb everything.

Maybe some tests, or propositions, or lay-outs, or whatever, are not useful for all of us, but they are available if someone need them.

The possibilities of PTE are so enormous that one can take advantage of it for his proper use.

Cor

[LumenLux]

While I never enjoy unpleasantness between members, there are some useful points emerging from this.

The key reminder to me is once again how PTE can be utilized and approached in so many productive ways. Many of us have not the abilities to do what others can. Every thing submitted can be appreciated - as something that will apply to me or something that will not. Thank you all for your contributions. And thank you Lin for the reminder of the "it" in English being "he", "him", "her", etc." in many languages.

[LumenLux]

Now, a question about this test/demo. How many slides and how many images have you used in the project as you have posted it?

[JPD]

Many thanks to everybody.

How many slides and how many images have you used in the project as you have posted it?

There are 32 slides, the 3 first are the same with a title. What is the most difficult for the size of pictures is the zoom function wich need a big picture if we want to have always at least a 1280 x 960 screen definition.

In this test the slides with the zoom at the beginning, used pictures with 1832 x 1374 size wich made 1280 x 960 with a zoom near 70%. It's the maximum limit to run on a 32mb graphic card, it's better if you use a smaller ratio (80%), so that you need only 1600 x 1200 pictures' size.

Of course, for a same final zoom (with the same ratio between start and end) you will need more pictures and slides to do it, but the files will be generally smaller than the same zoom with only one file.

The memory size seems not to be the only reason for which it can be not smooth. I am not absolutely sure, but as there are many calculations to resize, some PC which have enough memory for a slideshow can read it not smoothly because of the calculations for resizing.

I try to see and understand this problem.

Here is a link to the template of the test and there to the template of a zoom 22.7/1 which is smoother because using 0.8 zoom ratio per slide

Sorry Al for the link, it's good now, thanks.

[alrobin]

Here is a link to the template of the test and there to the template of a zoom 22.7/1 which is smoother because using 0.8 zoom ratio per slide

JP,

Apparently the second link cannot be found.

[jevans]

I have made a test using some simples rules in order to run with 32 mb memory on the graphic card:

- Never have more than 27 mb for 3 consecutives slides (Let be images of which the size don't overtake on average 1838 x 1378 (total of the images composing a slide).

- don't have too quick effects (I used 6s for a zoom 100% to 70% on a 1832 x 1374 size's picture).

Jean-Pierre,

Not sure that I understand what all the fuss is about in this post series, but you and a small number of other froum memebrs are providing a major support to this forum and you should be proud of this and ignore any negative comments. I am sure you have strong support from the majority of forum members.

You demonstration ran very well on my ATI Raedon 9200 graphics card, although Igor initially said that this would not be good enough for major PTE PZR effects. The only problem is a small jerk at the beginning of the whole sequence whic I seem to get on all PZR effects. This is a clever idea to use of a number of adjacent images, putting them together in a pan sequence. I would not have thought of doing that. However this will be easier when you are using screen dumps from Google Earth, and not so easy if you are using actual photos, since making a seamless joint between images may be more difficult.

Most of the tests and trials of PTE V5 provide much useful information for the time when a fully working version is available. However the feature that I am missing most, is the absence of synchronisation of sound and images, since it makes creating a show with commentary very difficult.

Jeff

[ronwil]

Jeff

I have created panoramas in Photoshop, using three images as illustrated in a Birthday Greeting which I posted on Beechbrook last year. Since then I have used Serif Panorama Plus software stitching five images together.

Ron [uK]

[CorVdK]

In this test the slides with the zoom at the beginning, used pictures with 1832 x 1374 size wich made 1280 x 960 with a zoom near 70%. It's the maximum limit to run on a 32mb graphic card, it's better if you use a smaller ratio (80%), so that you need only 1600 x 1200 pictures' size.

Hello Jean-Pierre,

My experience is that with 1600 x 1200 JPG-pictures and with my computer Intel Pentium 4 CPU 2,00 GHz, graphics card NVidia GeForce4 MX460 with 64 Mb of graphics memory:

Fading is smooth at all times and also with most effects of PTE, when I go from one picture to an other (fading procedure).

However when I use Panning and Zooming and use more than five 1600 x 1200 JPG-pictures to make an effect, I have sometimes some "hics" (short hop) in the fading to the following picture (after the effect).

Also the "hic" problem seems to be frequent in a Pan-Zoom effect when I use one or more PNG-pictures.

I think that the "hic-problem" is also a question of my videocard, only 64 Mb, and the NVidia GeForce4 MX460 is not what you say a brilliant quality videocard, also aged.

Recently I made a show about Rajasthan (India). It's a show of 34,5 minutes and about 233 Mb "big" with music, narration and 334 pictures with a size of 1600 x 1200. To run it (completely) smooth on my computer I had to bring the pictures back in size to 1024 x 768 (JPG high quality). The file is now 141,5 Mb and runs very smooth.

Cor

[JPD]

Many thanks, Lin, I think the same for you and several members who works to help people.

However this will be easier when you are using screen dumps from Google Earth

That's right, but I hadn't pictures enough large, my Minolta A1 is only 2560 x 1920, I would happy to do a zoom with important zoom ratio if somebody can send me a very large picture.

But even with Google earth, I was obliged to resize many pictures at the same scale as the picture at the end of the slide before, but it's enough easy to do, and it's even possible to resize the pictures if you use a large screen definition (ie 2048 x 1536) with PTE itself and make a screen copy, so that it's the same algorythme which resize and you have a very little difference between 2 slides.

I'll try to explain the different way I used to do when I'll finish the 1440 and the 1024 size test.

[JPD]

Also the "hic" problem seems to be frequent in a Pan-Zoom effect when I use one or more PNG-pictures.

The size of the picture in the graphic card memory is sometimes the reason, but it seems that some graphic cards which are old of 2 or 3 years haven't enough power to do the calculations of the picture, but to day, I am not absolutely sure of that, I have to make other tests later to know.

It's necessary also there is enough time to load the pictures in the memory, with my PC it's about 12mb/s, it's not very quick and may be sometimes a problem.

[Tomuk]

Jean-Pierre Hi!,

I have just downloaded your test sequence and it ran very smoothly on my system,but I am realy interested in the images you used.

You say you borrowed them from Google Earth at a high resolution, would you be so kind as to explain exactly how you do that as all the images I have seen on Google Earth when zoomed in are very poor quality indeed.

Tom

P.S

Please keep your tests and demos comming, I don't know what us lesser mortals would do without you.

[alrobin]

... all the images I have seen on Google Earth when zoomed in are very poor quality indeed.

Tom,

JP is probably busy right now, but if I may be so bold as to reply to your post, if you zero in on the Eiffel Tower in Google Earth, you should find that it is just as detailed as in JP's slideshow. Some places on the earth warrant more attention to detail than others!

[JPD]

Tom, Al is right, it's depend the place you see. Paris and the place where I live (Champs sur Marne in the east suburb of Paris) are at the same definition as New-York, as I made each of my pictures with 9 screen copy at a definition of 1600 x 1200 so I haven't the distorsion made by Google Earth's tool and I didn't use the option "Relief" which doesn't permit to paste exactly differents views.

I have made several pictures at different height (8) with the same method. As you see, there is nothing special in these pictures.

[Tomuk]

Tom, Al is right, it's depend the place you see. Paris and the place where I live (Champs sur Marne in the east suburb of Paris) are at the same definition as New-York, as I made each of my pictures with 9 screen copy at a definition of 1600 x 1200 so I haven't the distorsion made by Google Earth's tool and I didn't use the option "Relief" which doesn't permit to paste exactly differents views.

I have made several pictures at different height (8) with the same method. As you see, there is nothing special in these pictures.

Al and Jean-Pierre,

Thank you both for you very prompt replies, I shall now return to Google Earth and try and emulate (to a far lesser degree) Jean-Pierre's sequence with images I have snatched.

I shall post a link to my effort if I consider it has any merit, (but don't hold your breath).

Tom

[Ken Cox]

Tom

see

http://bbs.keyhole.com/ubb/z0302a1700/sep06.htm

New Imagery & Data Updates etc

ken

[Tomuk]

Thanks Ken the link was very helpful,lots of useful info about Earth and how to get the best from it.

Tom.

[JPD]

I continue to test PZR for 32 mb graphic cards, after the 1280 pictures' size, here's the 1024 picture's size.

In this project, there must have never more than 20 mb in the memory of the graphic card and we can have more effects that we need really. You can download the test and the album. here

[Ken Cox]

smooth as silk on my system

ken

[canteau]

Hello JP,

Very very smooth as silk on my laptop with 32 mb graphic card

Bravo JP very good work...

Bernard