Yeah, looks like it clips it before saving. No idea why the heck they would do that, but there it is:
github.com/Seeed-Studio/DSOQuad … les.c#L385
Doesn’t affect Pawn or custom C programs, though.
Here in the land-down-under, analog TV transmissions ceased over a year ago. Now all terrestrial broadcasting is digital. Unfortunately I am only familiar with PAL SD (standard definition) standards, but not with NTSC, SECAM or HD.
There is still a huge number of gadgets and devices that use analog TV signals (most camcorders can output analog, security system cameras, and of course all the legacy stuff). I am pretty keen on an app that uses the DSO203 to display various things about analog TV waveforms.
I can think of a few…
1 - Display a whole frame : at frame rate (20mS [PAL 50Hz] across the display). Trigger on Vsync.
2 - Display single field : at field rate (10mS [PAL] across the display). Trigger on every second Vsync.
3 - Display a whole frame : at line rate (64uS [PAL] across the display). Trigger on Hsync.
4 - Display selected line : line rate (64uS [PAL] across the display). Count the line number after Vsync and trigger on that Hsync.
5 - Vectorscope type display of colour phase.
6 - Actual image.
Number (3) is what started this thread, on which I have a few oservations and comments…
The maximum display size is 240x320, and that is further cut down by the useable area which varies depending on the firmware (menus, etc).
The original posted image seemed to only have 4 shades: bright, medium, dim and off. Two bits. People are talking about 4 bits (16 shades) being ‘essential’. Why?
At full (DSO) screen resolution (without any menus) two bits only equates to about 19KB.
Custom display drivers would have to be used to expand the 2 bits back to the full range, but this should be doable.
Number (1) and (4) should be possible immediately if only a “TV Vert sync” was available.
Number (2) might need a delay (or field selector) to only display one field.
Number (5) : The bandwidth of the DSO203 is somewhat limited, but the sample rate should be high enough to detect the phase of the colour carrier. This however would probably require FPGA processing to get it fast enough.
Number (6) : For a monochrome image, this would be straight forward. Colour would be somewhat harder and would rely on (5) working properly (at least the sub-structure would need to be correctly decoding colour).
Cheers,
Owen.