cde wrote on 2020-08-21, 21:41:
The Philips 252B9 looks really good! Looking forward to your review 😀 BTW I thought the VGA output for 320x200 games was 720x400, not 640x400. Is this correct? In either case line doubling would work the same I suppose.
Looking forward to receiving and reviewing the 252B9 . I hope it will end up being a good gamble .
As for the VGA output matter
a) VGA mode 13h (320x200@70Hz in 256 colors), the output coming of analogue VGA card was line-doubled to 640x400 by the card itself .
b) VGA text mode is composed of 80 columns and 25 lines of character cells each of which is composed of 9x16 pixels, which gives 720x400 actual output resolution .
The timings for a) and b) (horizontal/vertical front porch, back porch, sync pulse length) look essentially identical to a digitizer (unless you start to analyze the actual pixel data, which no digitize does, AFAIK), inside an LCD monitor's VGA input . In doubt, most LCD monitors assume the actual resolution is always 720x400 and sample accordingly . This gives un-distorted text mode but craps out when the actual signal is 640x400 (uneven pixel sub-sampling on the horizontal axis) .
To avoid this issue, some higher-end/better scalers/digitizers allow manually configuring the number of total, active, front porch, back porch and sync lines on the horizontal axis in order to accurately sample 640x400@70Hz . Two examples of those are the Extron DVI-300 variants and the OSSC . Of the two, only the OSSC preserves 70Hz for output .
Over DVI/HDMI/Displayport, this is never an issue as the number of actual pixels is always known to the scaler (whether it be the one in the video card or the one in the monitor), so no guesses compromises need to be made .
EDIT : On a CRT, there is no issue displaying either a) or b) , as the monitor does not care and directly displays however many active pixels there are on a horizontal scanline .