Bin copied to track with weird processing

[DonGateley]

When I copy a bin to a track some weird processing is applied to it. I've uploaded to a bin a sin sweep with a rectangular timing pulse preceeding and another trailing the sweep. Then I copy the bin to a track. When I download the track I just created (over the wifi with the embedded web server) I find that the track has had a 30 Hz brick wall high pass applied to it in the process of copying it.

The strangest part though is that the leading rectangular pulse has been converted to a positive going ramp, i.e. it is integrated to the level of the pulse and then drops within a sample to zero. But the trailing pulse is converted to a ramp that is the time reversal of the leading one. i.e. it jumps to the level of the pulse within a sample and than ramps down to zero. There is no linear process that can do that. What on earth is going on?

Downloading the bin shows exactly what I uploaded; no change.

I want to be able to copy the bin to the track, play it and record the response to it without it being modified in any way at all other than by the system it is passing through externally. Is there a way to do that?

My purpose here is to measure the impulse response of linear systems and this is defeating that effort.

[DonGateley]

My apologies. The high pass filter is in my data and is not an artifact of the MTD application. But then, you knew that.

I still don't understand how or why the rectangular pulses become ramps when copying a bin to a track but that's not of consequence to what I'm doing.