I am sorry for late answer, but sometimes it takes me a long time to get to the answers. I have tried the files you have submitted, but without success - I am not sure what is wrong on the image. OK, the interpretation of LRUD according to therion is following (we should probably put this into thbook): If passeges are horizontal (inclination < vthreshold, that can be set in centerline), LR is perpendicular to the shot and UD is vertical. If passages are more or less vertical (inclination > vthreshold), even UD becomes perpendicular to the shot - otherwise passages would not look very good. In the case of vertical (+- 90), UD is interpreted as north-south dimension from the station - to allow tube-like modelling of verticals. In your example - everything seems to me plotted correctly. Do I understand something wrong? Thanks, S. Quoting "Gilbert Fernandes (mailinglist)" <gfe_list@aliceadsl.fr>:

Hello,

I notice a bug in the 3D modeling. In this short exemple you can reproduce the error. The main problem is when you have a pit, ans so a vertical survey ( compas = 90 or -90 ), the up and down value in survey data is used in the 3D model (.lox file) for the with of the pit and not for the hight (or down) of the galery. So pits are always as fat as they are hight. Look the .th file and try to change the up and down value as comment in the exemple, look the effect in the 3D view. Thanks,

Gilbert