Edit: I think this was user error. I'd mistakenly assumed that doing an "Auto" would reset all of the options since it resets all the probe options (e.g. reset to 1X DC). Instead, doing an "Auto" does NOT reset the trigger noise reject mode. With noise reject on, the waveform must exceed the hysteresis/noise level before triggering.
I'm continuing to have intermittent triggering problems. It starts out OK but something that I'm doing gets it into a bad state where it won't trigger until the trigger level is well within the waveform. Before it gets into the bad state, I can sweep the trigger level up and down across the whole waveform without losing trigger anywhere on the waveform (of course, it won't trigger when the level is outside the waveform).
I've been looking at a 8MHz canned oscillator and trying to trigger on the negative undershoot. Currently it's in the bad state and is not triggering (the waveform on the display is from a slightly higher trigger level of 320mV). [Normal triggering mode, DC coupling, 10x 150MHz Rigol probe, noise reject off, min holdoff (8ns),
![Rigol DHO800/900 Oscilloscope Bug Reports + Firmware 1942077-0](https://www.eevblog.com/forum/index.php?action=dlattach;topic=395194.0;attach=1942080;image)
I don't know what I did to get it into this state. I've seen it several times now. I'm mostly zooming in and out, while adjusting trigger, and doing some single-shot captures.