Quote:
Originally posted by corysama
Even if the new PS2's DVD playback is progressive scan it will have no effect on games. Enabling progressive scan in game requires specific code and restrictions on texture usage --full-sized frame buffers eat extra video memory that could have been used for more textures. The PS2 hardware was actually never designed to do progressive scan. The developers figured out a software hack to make it work after the hardware launched.
|
That's not true at all. There was nothing about the design that hindered the usage of progressive scan. In fact I have technical documents about the GS given to developers with their development kits from when the PS2 was released where it specifically states all the modes the GS can handle which includes 480p and 1080i. Enabling progressive scan is not a software hack as you have direct access to this resolution on the GS.
See the problem is elsewhere. First, the way they designed the PS2 was in such a fashion that was so different than what developers were used to. Developers were used to the concept of VRAM. You have memory that stored your texture data which you would access as well as an area for the frame buffer. The PS2 architecture was designed to have a huge bus between memory and the GS. The embedded DRAM on the GS was not meant to be use as VRAM in the way people were used to. The bus was so big that it was designed to stream textures into the embedded DRAM which would act like a frame buffer. Now since the bus was so fast, you could stream numerous textures into the frame buffer before it would need to draw that single frame to the screen.
Now a misconception of even this among people who don't develop games is an example like a fighter. If that was the case that you constantly streamed texture data from memory, well then when you have a fighter, you have everything in view at once. See the 4 MB of DRAM, doesn't mean you need to cram all your texturse in there at once. You only need the textures in memory for the object your are drawing at that time. Again, the key here is that the bus is so fast that you can do this.
So that was the first problem, the fact that nobody knew exactly how the architecture was different. As a result, some people were only drawing half a frame broken up into even and odd scanlines in order to save memory. In fact the reason they were doing this is because it is a supported mode in the GS that was designed in the hardware. That is why many first generation games had bad jaggies because the fields would get out of sync often. Even then, you never had a full frame drawn in memory to allow a progressive scan image either. It was a hack job for developers to fix it because they were using it wrong to begin with.
Now getting to the second problem is Sony's lack of way of educating the developers on how they should be thinking of developing their games. This is sort of the developers fault too, but mostly Sony's. This goes back further to when the original PS was released. Back then they gave everyone libraries which made it easy to develop for, but developers were so used to coding to the metal on systems like the SNES and Genesis, that they wanted more freedom with the PS hardware. They constantly asked for access to the actual hardware/registers itself, but Sony refused for awhile which made the developers having to stick to libraries yet still complain about the lack of access to the hardware to increase performance. Eventually that would change in the PS life, but as a result would have a negative effect on the PS2. Since everyone complained about the restrictedness of having only libraries, this time Sony just gave them the pure technical documents of the hardware and pretty much said, have at it. This of course was overwhelming to the developers, and to some degree Sony was only giving them what they claimed to have wanted. But with such complex hardware, it became a chore to get anything done without libraries. That is why the PS2 became known difficult to program for.
So that leads to the third problem; Sony didn't provide libraries making the development environment easier to work with. This is all Sony's fault of course as I'm not trying to take the blame off them, but it also is partly the fault of the developers too. All of these factors is why the reputation exists today and the misconceptions that clearly still are passed around today about what the PS2 can and can't do. It's clearly evident when you look at the PSP environment which is completely developer friendly with an extended usage of developed libraries for the system.
So the bottomline is the PS2 was not nearly ill concieved in hardware design, as the concept of streaming textures into a dedicated frame buffer is pretty standard today, but it was ill concieved in how the developers were taught to take advantage of the system. The PS2 is very well capable of handling progressive scan and even HDTV resolutions. Memory is not the bottleneck that many of you believe it is. It took forever for gamers on serious game forums to overcome the concept of vram vs frame buffer. It's not an easy concept to grasp either if you're not tech savvy. Some games out there actually use a full frame buffer but don't have an option to enable progressive scan. Other games take advantage of the memory saved from using the interlaced mode as interlaced TVs still make up a majority of the TVs used to play games on. But one needs not to look any further than games like Burnout 3 and SSX3 to see that the system is very well capable of doing nice looking progressive scan games. It's all in the hands of the developers to enable it.