i am currently using mmf standard could i use hwa beta to increase the speed of mmf?
do i also need to use an extension to activate the hwa beta?
i am currently using mmf standard could i use hwa beta to increase the speed of mmf?
do i also need to use an extension to activate the hwa beta?
You don't need an extension, just install the HWA patch, both versions will co-exist on the machine, you can use either the normal version or the HWA version.
About the speed, that depends on your applications. If you use non optimised graphical extension it won't be faster, it could even be slower. Just install it and test your apps.
If your application has slowdown resulting from having too many objects being displayed, or graphical effects, HWA will greatly "speed it up", that is, decreasing the rendering time. It also allows you to apply graphical effects that are not possible in MMF2 without it- pixel shaders that can make modifications to images at runtime. It will not, however, speed up projects that are slow because of heavy processing obstacles; it would not help a chess program compute a move any faster.
As Yves said, simply download the HWA version and install it, and you will be able to open either HWA or MMF2-normal by their .exe's, and both will operate on the same folder of extensions/projects/resources.
Hey, on my old computer, with a Geforce 6600 from 2004 or 2005 (ancient!), I had 200 FPS in HWA and 2 FPS in standard MMF in my test application. That is an extreme case though, lots of active objects scaling and rotating. Still, as you can see, HWA might greatly improve the FPS. It also does, in my opinion, smoothen your gameplay. Even if your standard MMF game runs at 60 FPS, it's not perfectly smooth like in HWA to me.
thanks how many active animations can you have before slowdown?
is the size of frame related to any slow down?
would you reccomend controling the anamations to a specific area? i.e player enters area activate animation
what suggestions or options are there?
there are no specifics, as it is completely depends on the computer it is being run on. A computer will have both a CPU (processor), and a GPU (graphics processor). They can do virtually the same things, however a GPU, usually on your graphics card, is optimized for doing graphics manipulations- matrix functions.
In normal MMF2, all processing for an application is done on your central processor. When you add HWA, graphics are instead handled on the GPU. Because the GPU and CPU can work at the same time, this is much like multithreading.
Lets say it takes your game 100 time units to compute your event editor list per frame of gameplay, and 300 time units to render your graphics. Now, if your processor rendered all of that, it would take 400 time units to render a single frame. Now, lets say the GPU does it instead. Now because they're side by side, it only takes 300 units, the greater of the two. However, a GPU might be more optimized for graphics, and instead of 300 time units it takes 3 time units, because it works 100x as fast. Now it only takes 100 time units to render a frame.
The larger your window size, the more graphics are being rendered. MMF2 does not render the graphics of objects that are offscreen. It does not matter what animations are being used on offscreen objects because they are not drawn. So having a larger game window means that more graphics are being rendered at a time- HWA can help reduce the load on the CPU from this
thanks for the information.
my plan was to use a single frame area of 6000 x 6000 9.53mb and thats just the area with no events.
my main chacter 5.67mb
would you reccommend that i separate the frame into areas?
hi pixeltheif i have looked over your kind explantion if mmf does not render the graphics of objects that are off screen. So i could use a vertual frame size of 640 x 480 and the frame window of 6000 x 6000 as only the graphics that are in the current virtual window will be rendered tech the other graphics are off screen.
Am i correct?
I realy need to know before changing the current graphics which took 3 months to place in the frame
That is correct- graphics are only rendered within the virtual frame area, and off-screen graphics are not rendered until they are in view. However, it can still cause performance degradation.
Offscreen objects still require the same amount of memory, and there is a tradeoff between memory and processing speed. If you have too many objects offscreen, the game can slow down just the same. Fortunately, this restriction is, in general, a lot more lax than graphical rendering- you can usually have insane amounts of things in the collision map and with stored variables before you'll notice any slowdown. For example, my current project uses over 50 megabytes of runtime RAM via large arrays and an expansive collision mask, and it runs with no slowdown.
But do remember- large graphics can take up large memory very quickly. A full screen-sized object in 16 million colors mode with no compression can easily take up megabytes on its own.
ok thanks i suppose i must decide personaly i belive it makes more sense to create the game area over a number of frames and use global values to move between the frames
example: player reaches edge of frame leave frame load new frame and continue each new frame 1000 x 1000 which i could map out using my orignal as a template.
im curious could the window object help me?
I want to have a full screen display and i dont understand what you mean when you say
"A full screen-sized object in 16 million colors mode with no compression can easily take up megabytes on its own."
How do i compress a full screen?
Posting Permissions
Reply With Quote