One of the interesting comments from David was that the seismic...

One of the interesting comments from David was that the seismic processing is taking longer. This is good.

From what I've read and seen video's on, the way the Seismic ambient noise tomography (ANT) works is that an array of sensors are placed. Core used around 100 in a broadly square layout their early work around BP33. How many TG6 are using is unclear. Over a few weeks they collect seismic data.

After collecting all the signals, cleaver maths works out the direction that each processed natural and man-made seismic wave has come from. Knowing the direction allows the sensors to be basically tricked into thinking that the first one to receive the seismic wave sent the signal and all the later one's then receive that signal. The size of this grid goes exponential, particularly if a line of sensors is set down and seismic signals come in aligned to that direction of sensors. Say there were five sensors in the string aligned to the direction the seismic wave came in. You get the pairs of 1:2, 1:3, 1:4, 1:5, 2:3, 2:4, 2:5, 3:4, 3:5 and 4:5. If the string had 10 sensors then there are 45 pairs. If the sensor's are in a grid, each line on the grid has calculations. In addition to this, the number of pair's, each seismic wave has different frequencies which create a velocity image at different depths. For every seismic signal processed, all these calculations repeat. All the calculations are then merged into a 3-d velocity map. There is a huge amount of back-end processing, but thankfully computer processing power is now pretty cheap.

Speculation #1 - If there weren't very many usable seismic signals then the processing would be quicker. If there's lots, then the processing would take longer. The processing is now expected to take a couple of weeks (not the earlier 5 day estimate). Speculation #2 - is the longer processing time due to more seismic waves to process? Speculation #3 - do more seismic waves to process improve the velocity image made?