If I understand correctly, the radius of the M87 event horizon is approximately 19 billion km. I used an escape velocity calculator as a sanity check, entering 1.28E40 kg and 1.9E10 km and the calculator yielded an escape velocity of 299,879 km/sec.
I used the same tool to calculate the escape velocity for the ISCO, which is at 3R, which is a distance of 5.7E10 km from the center. The escape velocity from ISCO is 173,135 km/sec. which means it can send us pictures of what it sees, and we can receive them if we adjust for the expected redshift.
Suppose we drop an internally powered spherical lightbulb into the black hole. As it reaches a radius of 1.9E10 km, the gravitational gradient between our satellite at ISCO and the lightbulb will only be (approximately) c-173,135km/sec = approx 127,000 km/sec. Our lightbulb will only disappear from view when the relative difference in the PE of our positions in the gravitational field creates an escape velocity equal to C. And that happens only when the lightbulb is far inside the event horizon as it appears to distant observers. For example, at a radius of 1.2E10 km, the escape velocity is 377,000 km/sec, a meaningless number to a distant observer. But to our satellite at ISCO, the escape velocity difference is only 377,000 km/sec - 173,000 km/sec = 204,000 km/sec. This means we would notice some redshift and some time dilation but it seems as if the lightbulb would be quite visible. Now - what if our lightbulb was a mini sat that was equipped with lights, cameras, and transceivers. And what if we dropped a sequence of them so they could take pictures of what is below as they fell and transmit those images to the mini sat directly above them. As long as the difference in the relative escape velocities remained well below c, redshift and time dilation will be modest, etc. Given that coil d we not daisy chain our way down to the point where spaghettification began destroying our equipment?