So this has been broken since 9.5.
Would love to have some comment by Steinberg team as to whether they view this as broken or not.
In the MIDI CC Automation Merge Modes, when “Replace 2-Last Value Continues” is the selected mode:
Since 9.5 the Automation Lane for any MIDI CC will override the “Last Value” for the same MIDI CC found in a MIDI region once playback has gone past that MIDI region even though no Automation data exists past the MIDI region.
Before 9.5, whatever CC data point was last in the sequence whether it be Automation or MIDI region would remain as the “Last Value”
This should be the correct behavior when “Last Value Continues” is selected
Steps to produce:
(In Preferences-MIDI, ensure that in “Chase Events” “Controller” and “Chase not limited to Part Boundaries” are checked.)
- Create Instrument Track and load a VSTi that audibly responds to CC 1 (modulation)
- At measure 5–Record some notes on the Instrument track while moving the Modulation wheel and make sure modulation data is correctly written into the MIDI part. Have the last value of the Modulation data in the MIDI region be 0.
- Create an Automation Lane for CC 1 and select “Replace 2- Last Value Continues” from the pulldown menu on the lane as the MIDI Automation Merge Mode
- Activate Read on the automation track and add ONE point at measure 2 on the automation lane. Make the value 127.
- The MIDI CC1 data will play fine during the MIDI Region, but when Cubase reaches the end of the recorded MIDI region, the Automation Lane will take over and snap the CC1 (Modulation) value being sent to the VSTi instrument back up to 127 even though the LAST VALUE on the Automation Lane is back at measure 2.
If things were working properly, the LAST value being played back for CC1 should still be at 0 because the actual last point of data is in the MIDI Region.
Previous to 9.5 Cubase would honor the Last Value of CC1=0 even past the recorded MIDI region. In 9.5 once you pass the recorded MIDI region it started reverting back to whatever the automation lane was set at regardless of whether the automation lane data was truly the Last Value or not.
An example of one of the several annoyances caused by this bug would be when overdubbing a spiccato harmony on top of an existing spiccato string line where CC 32 controls the articulations being used by the sample player. I have a default data point in the Automation Lane for my “longs” patch at measure 2. If I play past the existing MIDI region, the automation takes over and changes my spiccato patch back to “longs” while I’m playing.