Reading this question I was struck by the fact that the "pointing" of an interferometer is effectively done in the correlator, at least within the common field of view of all the telescopes, which is many thousands of times wider than the imaged area. Could the same data be reprocessed to reveal images of other "tiles" from this common field of view? Could it image the accretion disk, for example, or the base of the jets? Are there any targets that are close enough and bright enough to be worth doing this?

  • $\begingroup$ very cool question!!! $\endgroup$
    – uhoh
    Commented May 26, 2019 at 13:59

1 Answer 1


The question asks about "targets" but doesn't specify black holes. However the other "targets" are related to infall of black holes.

Nonetheless, I think the answer is a solid "yes" either way.

Data is recorded separately at each site. Hours and hours of digitization of one or more 2 GHz wide data channels around 230 GHz, down converted to baseband and synched using a local atomic clock. See this answer and this answer

Any signal reaching the feed horn of each radio telescope is going to be recorded in its data stream. So if there is another source nearby that is covered by the radiation pattern of all of the dishes in the EHT collection, then it certainly could be imaged to some (indeterminate) level of resolution, with (at least) the following caveats:

  1. You have to run the numerical correlations again, with a new range of time-offsets for each telescope. If the source is much brighter, you wouldn't have to use the whole dataset. But if it's just as dim or dimmer, you will have to correlate lots of data.
  2. If the source is far away from the imaged event horizon, there may be several different types of calibrations for each dish and site that have been optimized, and may not work as well for a substantially removed object. The existing data may result in slightly lower resolution far from the target, depending on how calibrations were done. In addition to signals of the black hole, signals from other sources were recorded to assist in calibration of the whole array.

I need to read further on that, I'll update the part about calibration tomorrow.

  • $\begingroup$ Thanks. I was imagining imaging the centre of M87 more widely, rather than something else that just happens to be nearby $\endgroup$ Commented May 26, 2019 at 15:37
  • $\begingroup$ @SteveLinton I thought so, but in addition to that I also wanted to answer for the more general case. Item #1 applies to matter close or far (assuming that the analysis didn't already cover the area in question) but #2 may not be much of an issue at all for other areas associated with M87. I need to do some more reading on the math behind the analysis, I'll come back to this answer again in about 24 hours or less. Thanks! $\endgroup$
    – uhoh
    Commented May 26, 2019 at 15:40
  • $\begingroup$ @SteveLinton I see that I've said I'd do something "in 24 hours or less" and haven't yet done it. I'll get on it this weekend. Perhaps the edit bumping the question back into the active queue will bring in new answers as well. $\endgroup$
    – uhoh
    Commented Mar 26, 2021 at 23:29

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .