Questions about PS2 Optimal Timings

[Retrorunner]

Currently I am trying to apply the ps2 optimal timings using junkerhq values:

http://junkerhq.net/xrgb/index.php?title=Optimal_timings

and I run into some question and problems:

1. If I am not mistaken these timings are for 4:3 ratio? So what would be the optimal timings for games which support 16:9?
Using GT3 as an example. The game runs in 512×448. Optimal timings for 4:3 would be:

H.active=512, H.samplerate=686, H.synclen=44, H.backporch=84

For 16:9 its should be:

H.active=582, H.samplerate=686, H.synclen=44, H.backporch=84

I am correct?

2. Using the optimal timings from the wiki the screen ratio is not quite 4:3 (4x Line Bob). Why? Even adjusting v.active to 224. Its because of the border/frame which the OSSC uses to display the image?

3.The image difference between the standard output 4x Line Bob and optimal timings is barely noticeable. I am using the correct values for LFP and sync. Not sure if I miss something…..

Thanks for your help.

[Zacabeb]

1) PS2 games generally run in the same resolution regardless of whether they render to 4:3 or 16:9. The best option is to use the same timings and force the display to stretch the picture to 16:9.

If that can’t be done, you need to use the Generic 4:3 mode and increase all the horizontal values by 4÷3 to generate an oversampled picture that natively displays as 16:9. That would mean the following values:

H.samplerate 1144
H.active 960
H.synclen 83 (approximately)
H.backporch 76 (approxmiately)

The resulting signal may or may not be accepted by the display.

2) Regarding the 4:3 aspect ratio, the standard definition pixel clock of the PS2 was based on a 4x multiple of the ITU-R BT.601 sample rate, resulting in a picture that was a bit narrower than 4:3 given a full 480 line tall picture. Because of its integer scaling, the OSSC also can’t fully compensate the aspect ratio for all resolutions and output modes below 4x.

3) The PS2 had a very high video bandwidth for YPbPr/RGB component across all its revisions (as far as I know). That means you can get a very sharp picture in 4x bob mode with Allow upsample2x enabled, so long as the game itself doesn’t do any softening, reducing the need to use optimized modes unless you need to reduce jitter or other artifacts.

[Retrorunner]

Thanks for your response.

1. It makes from logical perspektive sence but if you take a look on the wii optimal timings for 4:3 and 16:9:

H.active: 640 in 4:3 mode, 720 in 16:9 mode, H.samplerate: 858 for both 4:3 & 16:9 modes
http://junkerhq.net/xrgb/index.php?title=Optimal_timings

its kinda contradictory. The samplerate is the same for both 16:9 and 4:3.

[Zacabeb]

The Wii might have used the wider 720 raster for widescreen games to avoid pillarboxing. As with the PS2, the picture from the Wii is dependent on the display stretching it to fill the screen in 16:9. Games on the PS2 rarely ever used a wider raster than 640 pixels, largely due to the limited VRAM.

[Retrorunner]

thanks for clarification. Just one last question. The vertical resolution is not the original one when using PS2 optimal timings. 448-> 480 so the screen ratio is slidly off. Why it is not recommended to use v.active 224/448 as optimal timings? IMO it does look sharper and the screen ratio is closer to the original.

[Author]

Posts