Author Topic: CLASSIC COMP - PARANOIMIA 1 (Read 15078 times)

Shockwave · « **on:** April 21, 2007 »

Here is my version of;

THIS CRACKTRO

I have attached a screen shot and exe to this post, you'll need to be logged into your account to download

Many apologies to Roly for the false start, luckilly I managed to get hold of a conversion of this GMC tune, it's not a perfect conversion but it's quite close.

.:] Druid [:. · « **Reply #1 on:** April 21, 2007 »

oh! souvenirs, souvenirs

Ghost^BHT · « **Reply #2 on:** April 21, 2007 »

looks like the real thing here

good work and thanks for the greet

Shockwave · « **Reply #3 on:** April 21, 2007 »

You're welcome mate!

bbear · « **Reply #4 on:** April 21, 2007 »

great one

Shockwave · « **Reply #5 on:** April 21, 2007 »

It was a classic one

Thanks, glad you like it!
I will make one more I think.. Not sure what yet.

benny! · « **Reply #6 on:** April 22, 2007 »

Damn. Outstanding work, shocky !!! Intro conversion roxx. Excellent job!

Emil_halim · « **Reply #7 on:** April 22, 2007 »

you did it shocky, good .

is there any chance to get the source code?

Shockwave · « **Reply #8 on:** April 22, 2007 »

Here it is (no music) attached to the post

Emil_halim · « **Reply #9 on:** April 22, 2007 »

thanks shocky , that is really fast.

now , we have the source code of shocky and Jim.

so ppl will learn and try to make thier own one.

Emil_halim · « **Reply #10 on:** April 22, 2007 »

hi shocky

i have replaced your Asm code with SSE to get more FPS ,so please test it and feed back the results

Code: [Select]

'-------------------------------------------------------------------------------
' CLEAR THE OLD SCREEN BUFFER WITH SOME KIND OF FUCKED UP COPPERLIST :p
'-------------------------------------------------------------------------------

SUB CLEAR_AND_COP()
    DIM AS UINTEGER A
    DIM AS INTEGER TC,SLICE
    DIM AS INTEGER PTR PP 
    
    FOR A=0 TO YRES-1
        SELECT CASE A
          CASE TPBAR-1,TPBAR+21,BTBAR-1,BTBAR+21
             TC=&H000066
          CASE TPBAR TO TPBAR+20
             TC=&H0000AA
          CASE BTBAR TO BTBAR+20
             TC=&H0000AA
          CASE ELSE
             TC=&H000000
        END SELECT
        
    SLICE = XRES
    PP = @BUFFER(A*XRES)   
   
   ' asm
   '     mov eax,dword ptr[TC]
   '     mov ecx, [slice]
   '     mov edi, [PP]
   '     rep stosd
   ' end asm   
    
     asm
         mov eax,dword ptr[TC]
         movd xmm7,eax
         pshufd xmm7,xmm7,0
         mov edi,[PP]
         mov ecx,[slice]
         shr ecx,3
         .lop:
         movdqu [edi],xmm7 
         movdqu [edi+16],xmm7 
         add edi,16*2
         dec ecx
         jnz .lop 
     End Asm

    NEXT
    
END SUB

Shockwave · « **Reply #11 on:** April 22, 2007 »

I tried both, but your version lost about 10fps per frame, I don't know why you'd want to try and optimise something like this anyway?

It runs at well over 400fps on this modest pc as it is....

Emil_halim · « **Reply #12 on:** April 22, 2007 »

It is really wierd.

ok , i tried to optimized this part of asm code to increase the FPS with SSE , but it seems that SEE is not good with Modern computer , 3 years ago or more (with my Pentium 3) i was using MMX and SEE to speed some stuff such as Memory copy or filling memory with some code ad so on......

and it was very effective , so i want to try it with my Pentium 4 , and it seems that new SSE and SSE2 is not good with Modern computer and just wonder why Intel Made it.

Jim · « **Reply #13 on:** April 22, 2007 »

Emil, the problem is you're trying to optimise the wrong bit. SSE can do amazing things when applied to the right problem , but memcpy/memclear isn't the best thing to use it for. rep stosd is going to be within a few % of the fastest way to copy memory under any normal circumstance. There have been a few CPUs where you'd have got some milage from SSE for this, but not really any more. Try SSE for image processing, video decoding, signal processing at it will soar-away beat using standard FPU/CPU code even on a Core2.

Jim

Emil_halim · « **Reply #14 on:** April 22, 2007 »

Yes Jim, you are right.

SSE is good for image processing, I tried it and the results was very good.

But I really did a SEE & MMX MemCpy/Filling and it was very very effective instead of using rep stosd with PIII.

Anyway thanks for your explanation.

Rbz · « **Reply #15 on:** April 22, 2007 »

Excelent work as always

rain_storm · « **Reply #16 on:** April 22, 2007 »

I was wondering which one you would pick to remake great choice Shocky

Shockwave · « **Reply #17 on:** April 23, 2007 »

With luck I may be able to finish my second remake today. Thanks for the nice comments!

Tetra · « **Reply #18 on:** April 23, 2007 »

Nice 1 Shocky

I quite like the Text, reminds me of somthing I tried to do in delhpi so many years ago.

Shockwave · « **Reply #19 on:** April 23, 2007 »

You know, that Paranoimia one is probably one of the most famous cracktros ever made.. Paranoimia released so much stuff all the time, that cracktro was on everything for ages!