Index
> Main > object oriented animation file. |
| Author |
|
|
vid 27 Nov 2008, 00:21
Sooo, what is advantage of this format against existing FLV (true and tested), for example? And then, as we all know, FLV is much inferior to DivX and modern stuff, what is advantage against those? Does your format allow for HW acceleration (present on virtually all GFX cards today)?
|
|||
27 Nov 2008, 00:21 |
|
|
Borsuc 27 Nov 2008, 02:31
Xvid is way better than DivX
also, who cares about HW acceleration that much? It's not like movies are hard to decode with a CPU or too slow...  _________________ Previously known as The_Grey_Beast |
|||
|
27 Nov 2008, 02:31 |
|
|
revolution 27 Nov 2008, 03:39
Borsuc wrote: also, who cares about HW acceleration that much? It's not like movies are hard to decode with a CPU or too slow... |
|||
|
27 Nov 2008, 03:39 |
|
|
Borsuc 27 Nov 2008, 03:48
oh yeah you need so many apps when watching a movie
and if you need that use a multi-core CPU _________________ Previously known as The_Grey_Beast |
|||
|
27 Nov 2008, 03:48 |
|
|
revolution 27 Nov 2008, 03:54
Not all apps need an active user in front of them. My background tasks run 24/7 at 100% CPU, so any program wasting my CPU cycles will quickly get booted off and more efficient replacements found.
|
|||
|
27 Nov 2008, 03:54 |
|
|
edfed 27 Nov 2008, 17:19
the advantage of this format is that it is fool compatible.
then, it is a format for my hobby. after, idon't know the flv format, but one thing is sure, this format is very modular because it can display everythings graphic. assuming an evolution of the structure, like addition of current item's size instead of a fixed value ( add [ptr],item's size ) and then, your animation file can be composed of a string of many elements of different nature. box,text,oscilloscope,image,line,effect, etc. for the moment, i don't care about HW acceleration. i focus only on functionality. this is the test animation, a pulsed light ball. Code: tstanim dd f.anim,100,50,20,20,0,0,@f-$-4 dd f.put,0,0,18,18,white-1,bille,1 dd f.put,0,0,20,20,white-6,bille,2 dd f.put,0,0,18,18,white-7,bille,3 dd f.put,0,0,16,16,white-5,bille,4 dd f.put,0,0,18,18,white-3,bille,5 dd f.put,0,0,20,20,white-2,bille,2 dd f.put,0,0,18,18,white-1,bille,3 dd f.put,0,0,16,16,white-5,bille,4 dd f.put,0,0,18,18,white-3,bille,5 dd f.put,0,0,20,20,white-2,bille,2 dd f.put,0,0,18,18,white-1,bille,3 dd f.put,0,0,16,16,white-2,bille,4 dd f.put,0,0,18,18,white-3,bille,5 dd f.put,0,0,20,20,white-2,bille,6 @@: it works in the last shell update. this is the function to include Code: anim: .call=0 .x=4 .y=8 .xl=12 .yl=16 .c=20 .bmp=24 .t=28 .ptr=24 .s=28 .img=32 push esi edi mov edi,esi mov eax,[esi+.ptr] lea esi,[esi+eax+.img] mov eax,[esi+.call] or eax,eax je @f mov eax,[eax] or eax,eax je @f call eax @@: pop edi esi ;here is the plce where to put the timing routine. add dword[esi+.ptr],.img mov eax,[esi+.s] cmp [esi+.ptr],eax jl @f mov dword[esi+.ptr],0 @@: ret this is the line to add to flist.inc Code: f: ... .anim dd anim ,0 ... db 'anim.inc' align 16 include 'anim.inc' align 16 |
|||
|
27 Nov 2008, 17:19 |
|
|
baldr 27 Nov 2008, 18:57
edfed,
How do you think, which byte value will be omnipresent in your .OAF/.FAF/whatever file? Methink, it's 00.
vid and Borsuc, DivX, XviD… Let him play with self-written tools, may be it will get some result. "Tread lightly because you tread on my dreams." — William Butler Yeats. |
|||
|
27 Nov 2008, 18:57 |
|
|
edfed 27 Nov 2008, 22:51
it is not always zero that is omnipresent.
sometimes, it is -1, sometimes it is 10 etc etc. sometimes, it is random numbers, or a count. but at least, it will be possible to have a pointer instead of a sub object inside the parent. Code: time1=10 time2=10 time3=0edfedh anim dd f.anim dd obj1,x,y,color,time1 dd obj2,x,y,color,time2 dd obj3,x,y,color,time3 obj1 dd f.box,x,y,xl,yl,color obj2 dd f.txt,x,y,xl,yl,color,text,font obj3 dd f.button,x,y,xl,yl,color,bmp1,bmp2,variable??? or pointer??? with 0edfedh as time value, the current selection will never change, for example, if you have finixhed the reading of a video stream, you have an endless menu that needs to click to activate something. one interresting way where to work is the parameter computation with ascii formulas as parameter. then to just move an item, or play a loop, you just leave a formula, and it will compute parameters each times it changes frame. maybe for the 00 redondancy problem, it is possible to crete a little code to compress the images and everithing on the 00 value if one byte = 00 then, it will be like this: 00 01 if two bytes = 00 00 02 if 329 bytes = 00 OO FF 00 4A to use transparentlly in the file system routine? maybe later, it will be interresting to seek a mathematical way. the file would be composed of vectors, counters, matrix, arythmetics, etc etc... BTW, it means that programer and user should have an arythmetics omnipresent tool, like ctrl+C ctrl+V, or prtscr. |
|||
|
27 Nov 2008, 22:51 |
|
< Last Thread | Next Thread > |
Forum Rules:
|