This bit of looneyness during a massive redo of the client’s log database. Basically we found so many problems with the earlier digital database (misassigned  APIs, incomplete log header, incomplete scales [e.g., simply NPHI without apparent matrix unit]) that the client wanted our group to redigitize all of the well logs in the field area. After digitization as part of the QC process we overlaid the new LAS digits onto the raster image and ended up with this garbage. Normally, the generation of the LAS digits was effectively instantaneous (<1 sec) but every now and then it took 2–3 sec, a notable pause and the LAS overlay would show this gobblygook. We found that this error correlated with logs that had “stretches” in the log scan where the scanner infeed struggled with the log folds. The software allowed correction of the measured depth lines but apparently a software bug was triggered when this situation occurred. Oddly the LAS digit issue did not arise at the point of the log stretch but rather randomly elsewhere in the log. We resolved the issue on our end by placing a CD case on top of the log while scanning to push it down better for the scanner infeed. We also advised the company of the bug, which was subsequently corrected. They also added a flap mechanism to weight down the log on their actual scanner. I wonder how many old digital logs encountered this same problem but was not identified. We used version 5.0/5.1 of the software so a lot of LAS files had been generated for years prior without identifying the bug. You’ve got to QC the logs and a great way is to overlay the digits back onto the raster scan. This was a good example of good data in but garbage out!