why isn't this a problem in the other clock out modes?
Other modes are not step dependent, so they are not time division dependent. E.g: 24ppq is always 24 clocks by 1/4th note, 12 clocks by 1/8th note, 6 clocks by 1/16th note, 3 clocks by 1/32th note.
If you change the time division while playing, it does not affect the slaves (or if BSP is slave, it is not affected).
I you have a Master BSP set to "1step", 1/8th time division, at 120BPM, and you connect a slave BSP (slaved to clock din) to it with 1/4th time division, the slaved BSP will think the tempo is 240 BPM.
Why?
120 BPM is 120 Beat Per Minute. A beat is 1/4th note. A minute is 60 seconds. 120BPM = 120 x 1/4th in 60 seconds =
2 x 1/4th note in 1 second.
1/4th note = 2 x 1/8th note
So 120 BPM is
4 x 1/8th note in 1 second.
Your BSP Master will send 4 clocks in 1 second.
Your BSP Slave will receive then 4 clock in 1 second. As it is set to 1/4th, it will go 4 steps forward (1 clock = 1step in "1step" mode), like the Master. But as it is set to 1/4th (the master is set to 1/8th), it will think the tempo is twice as fast. I you change the Master time division in the middle of nowhere, the result will be kind of unpredictable and the two devices will be out of sync (out of phase) because the master cursor will jump to a new position in order to synced with SEQUENCER and SEQUENCER2.
And why does it only occur in the drum section?
1step mode is based on the drum section as the 3 sequencers may have a different time division setting.
Regards,
Y.