I checked the code for RT130, actually it already used check_reftek_header() to exclude the files that not belong to time range. So we don't have to do constraint for reftek. If any suggestion, because read_start and read_end will be the exact day and each data file will be in that day, so I want to read just the first few bytes to decide the day of the package to decide to exclude it or not. I will see if it is any faster than the current code.
MseedReader.check_time() is used to exclude data that is out of plotting range, so the data read are time constraint for each record. However, with non-multiplexed datasets, we can consider exclude the files.