Footleg I think the answer is 'yes'. Hopefully you can view the images embedded in the image rather than attached, otherwise this won't make much sense. In this example there is a single centreline (comprising both cave and surface survey) and 4 scraps. A short section of cave at the left hand end does not have any scrap drawn. Compiled with Therion 5.3.11 One interesting point is that regardless of the 'walls' setting, there is never a tube generated there once we have scraps drawn anywhere in this centreline. My deduction is that Loch uses only one type of tube generation per centreline, because in almost all my larger projects that transition between traditional survey and paperless survey, I have a combination of each type of tube generation within the same Loch model. This fits with the description in the Therion Book "walls <auto/on/off> turn on/off passage shape generation from LRUD data for subsequent shots. If set auto, passage is generated only if there is no scrap referencing given centreline." In these examples I have not used any 'point passage-height' or 'point dimensions' in the scraps. Paperless survey (no LRUD), walls auto, off or on (or not specified), No plan scraps. [Loch always guesses tube dimensions] -same if LRUD present and walls off Paperless survey (no LRUD), walls auto, off or on (or not specified), plan scraps drawn [Loch uses scraps for tube width and guesses tube height] Now, I added a few LRUD to the same centreline, including some with only UD data dimensions station left right up down 1.4 - - 20 20 1.5 - - 20 20 1.6 2 2 20 20 3.13 10 10 20 20 3.14 10 10 20 20 Tube generation, where it occurs extends one station beyond those specified. Paperless survey (some LRUD), walls on or auto (or not specified), No plan scraps. [Loch uses LRUD for tube dimensions and only for stations where partial data is provided, it guesses the missing data ie if a station has no data, no tube is generated] And now adding some scraps and things get interesting. Paperless survey (some LRUD), walls off (or not specified), Plan scraps drawn. [Loch uses scraps, not LR, for tube plan dimensions, and UD for height where specified, otherwise height is guessed. Where there are no scraps, LRUD is not used.] That is perhaps contrary to what one might expect from the Therion Book 'walls' entry. It also produces perhaps the best model. Paperless survey (some LRUD), walls on or auto, Plan scraps drawn. [Loch uses scraps AND LR for tube plan dimensions, and UD for height where specified, otherwise height is guessed. Where there is no LR, then it is guessed. Where there are no scraps, LRUD is used.] If I change the LRUD definition like this. data dimensions station up down 1.4 20 20 1.5 20 20 1.6 20 20 data dimensions station left right up down 3.13 10 10 20 20 3.14 10 10 20 20 ..then the guessed width troubles continue. Paperless survey (some LRUD), walls on or auto, Plan scraps drawn. [same as above] So my final conclusion, for paperless survey, no need to include walls statements in each survey, or if you do, walls off is better than on or auto. If you happen to provide some UD data, then your model will be improved (more realistic). I surmise that with these same settings, scraps with point passage-height or point dimensions will have a similar effect, but I have not tested this. Bruce
