Patch coordinate/channel alignment

[shawnboltz]

Hi all,

I was playing around with a DAS dataset that has a large gauge length relative to the length of the cable (0.817m/54.72m) and noticed that the cable appeared to be one gauge length shorter than the actual sensing length (dascore said it was only 53.09 m). Digging into it a little bit, I noticed that it has to do with how dascore assigns the coordinates for each channel. dascore puts the coordinate for the channel at the beginning coordinate of the segment of cable whereas the Treble dashboard spans the channel between the starting and ending of the cable (this is difficult to describe, so I've added a graphic).

While this is something that I only really looked into for the Terra15 data, it's due to the design of the Patch/CoordinateManager. As a result, it theoretically affects most, if not all, file formats. Importantly, the spacing between channels is still correct, it's just that the coordinates of the channels are shifted by potentially 1/2 a sample length (in this case, 1/2 the gauge length) from what was intended. In practice, I don't think this makes a huge difference, but there could be some scenarios where it becomes an important distinction.

[d-chambers]

Hey @shawnboltz,

I am in the field this week, but I will try to write more next week. We should come up with definitive meaning/guidance and make sure all the parsers use coordinate consistently. Also, I have had a few other ideas on how we might improve the coordinates, but I havent yet had time to write them down for discussion.