gap in recorded stimuli (block RVG06_B06)

[jihyunbak]

Another stim parsing problem for a tone150 block:

ValueError: ToneTokenizer: Incorrect number of stimulus onsets found in block RVG06_B06. Expected 4500, found 4496.

Originally posted by @JesseLivezey in #114 (comment)

[jihyunbak]

This block actually has a glitch period in the middle of the recording.

logs from my local debugging branch...

DEBUG:nsds_lab_to_nwb.components.stimulus.tokenizers.base_tokenizer:found 4496 onsets by thresholding mark track
DEBUG:nsds_lab_to_nwb.components.stimulus.tokenizers.base_tokenizer:Tokenizing tone150 stimulus.
DEBUG:nsds_lab_to_nwb.components.stimulus.tokenizers.base_tokenizer:audio file start time: 9.19559904
DEBUG:nsds_lab_to_nwb.components.stimulus.tokenizers.base_tokenizer:stim onset: 14.32059904
DEBUG:nsds_lab_to_nwb.components.stimulus.tokenizers.base_tokenizer:last marker: 1365.0673664
DEBUG:nsds_lab_to_nwb.components.stimulus.tokenizers.base_tokenizer:audio file end time: 1369.19559904 
DEBUG:nsds_lab_to_nwb.components.stimulus.tokenizers.base_tokenizer:recording end time: 1390.91509248
...
ValueError: ToneTokenizer: Incorrect number of stimulus onsets found in block RVG06_B06. Expected 4500, found 4496.

Upon inspection of TDT-recorded aud1 and mrk1 streams, all these values above are correct. But there was this mysterious period in the middle of the recording (see plot). Found this while checking if all mark-to-mark intervals are correct.

@jthermiz - any idea how this happened?