There are a few highly eccentric, low morph & long period targets which will not fit without intervention. It looks like the key is to give JKTEBOP input values for ecosw & esinw as close to the final values as possible. The preds for ebop maven are pretty good, especially for ecosw, but esinw is more likely to be inaccurate when the eclipses are very narrow.
We can calculate values from ecosw and esinw from the eclipse data coming from TESS-ebs (see Hilditch page 238 for eqn). These data will already be confirmed as quite accurate by successful flattening.
Based on detecting a potentially challenging system (morph, period and/or phiS criteria) we could calculate these values and override the ebop maven predictions.
I've already done this by hand for TICs 0279741942 & 0350298314 which require this intervention to get anything other than a nonsense fit.
There are a few highly eccentric, low morph & long period targets which will not fit without intervention. It looks like the key is to give JKTEBOP input values for ecosw & esinw as close to the final values as possible. The preds for ebop maven are pretty good, especially for ecosw, but esinw is more likely to be inaccurate when the eclipses are very narrow.
We can calculate values from ecosw and esinw from the eclipse data coming from TESS-ebs (see Hilditch page 238 for eqn). These data will already be confirmed as quite accurate by successful flattening.
Based on detecting a potentially challenging system (morph, period and/or phiS criteria) we could calculate these values and override the ebop maven predictions.
I've already done this by hand for TICs 0279741942 & 0350298314 which require this intervention to get anything other than a nonsense fit.