I'm guessing it's 2V/div, 5V supply, and the ramps are open bus (with weak pull-ups, but nothing reading during so it doesn't matter). But this would be a good place to have multiple channels showing the strobe signals.
Another tweak that may be diagnostic: loading the signal (particularly the suspicious pin) with a resistor of say 1k, down to even a couple hundred ohms, pulling to VDD or GND, or with a capacitor of some 10s of pF, and see how much is necessary to cause data corruption. Most likely, all the other signals will be well-behaved (tolerant down to, say, 200 ohm or so, or up to 30pF or more), and the suspect signal fails with much less (or in the pull-up case, potentially improves). You will see the effect of these loads on the scope trace.
Tim