Automation Follows Events - "Bypass" Parameters Corrupted

I wonder if anyone else is getting this intermittent issue. “Bypass” automation events are being corrupted when events are moved with Automation Follows Events “on”.

For example:

Screenshot 2023-01-19 at 17.12.49


Screenshot 2023-01-19 at 17.12.53

Bypass should either be “on” or “off”. Sometimes automation is simply removed elsewhere in the track, and on becomes off or vice versa…

Annoying. Bug, or user error?


I think this has been reported by other users.

Ah. Couldn’t find it… will look again. Either way… at least it’s not just me.

A fix would be nice, as I mend a project that’s in a right mess. :wink:

Any word on whether this is a real bug and flagged for a fix? 'tis a nuisance. :wink:


No word on this? Really messes up sessions…

Hello - I still get this. Regularly. It can really mess things up. Any “toggle” parameter (on/off, bypass etc) is fair game for this bug.


So… any investagation of this? Still happening…


Been busy. Remind me this weekend and I’ll check again.

So here’s some more detail and a reproducible example. The basic point is that toggle automation (bypass, eq on/off etc) gets corrupted with illegal “variable” values when copy-pasting. Those values seemingly appear at random. They can really mess up a project.

  1. Make an empty session

  2. Add a track

  3. Drop in a mono audio file

  4. Show panner bypass automation lane (or input filter on/off)

  5. Draw some automation (BYPASS ON) BEFORE the clip but overlapping the front of it slightly

  6. Set “Automation Follows Events” to “on”.

  7. Box-select the clip

  8. Copy

  9. Paste


It’s obviously “wrong”. Not only does this screw up individual copy / paste edits, but it can really make a mess when you are copy-pasting many blocks of data for instance when re-conforming an edit or re-ordering an album master.

Here’s what happened to an eq on-off lane when it was subject to multiple copy-pastes. It took me hours to sort this out (there were multiple automation lanes affected):

Screenshot 2024-06-04 at 00.20.39

And there’s another gotcha to this…

Here I have the cursor at timecode 11:24:10:08 just before the change from an illegal stepped value to “0”. This is automation EQ band 4 on/off. So Band 4 is currently “on”:

I frame advance 1 frame to 11:24:10:09. The automation line is now at 0. Band 4 is now off. Or is it? The band controls are off. But the graphic still shows it as on:

It stays like that if I “play” or if I mouse the cursor around on the time-ruler. Indeed it stays like that until I click the band switch or the display. So is this a graphic issue or what? It’s a mess:

This has been plaguing me for months.



I’ll probably forget to remind you (busy weekend ahead!).

But please see post above.



+1 for getting this fixed asap.
This is a real killer for post people in particular, but I presume it’s also happening in Cubase.


Hello! Any thought on this?

Here’s another example of the chaos this bug can cause. The three blocks on the right were copy-paste rebuilt in a different order to the three blocks on the left. Automation lanes 2-3 are eq band switches. Look at the mess I had to unravel.

Screenshot 2024-06-04 at 12.55.01



Hello - has there been any movement on this?

I should point out that while the most recent instances of this issue have been with Nuendo 13, when I first reported it in January last year we were still on Nuendo 12.



It’s terrible.

Hi Matthias… any thoughts on this, yet?



Yeah I just checked this and you’re absolutely right that this is a bug. Add the “issue” tag and put a repro in your first post.

I don’t have time and brain to post one myself but from what I can tell right now:

  • This seems to appear to the bypass parameter I checked ONLY when drawing automation in, not when using the automation system’s write function.

  • It also seems that we do NOT have to overlap the event itself to create this problem. See the following sets of images:

An event was created and there is one automation point before the event. The automation track shows LC filter bypass. The filter is on throughout:

The next image shows how automation was drawn manually to turn the filter off / bypass it after the event:

This last image shows what happens after the event has been moved just slightly to the left:

EDIT: The middle image shows correct behavior. The cut filter should be off because of where we are located on the timeline.


I’ve already posted a repro. And now you’ve posted another one! :slight_smile:

(I cannot see how to edit my first post - there’s no edit tool; I posted it 18 months ago. And how to I add an “issue” tag…?)

For developers:

Maybe this helps you troubleshoot this behavior - the ‘depth’ of the drop of automation seems to depend on how far away we are from the event when the new automation is drawn in. I have done exactly the same thing as in the above examples but I have done it at different points in the timeline.

In the image in the previous post the drawn automation was right after the event. Here are two images with timeline zoomed out and automation drawn at two further locations:

As you can see there is less drop the further away from the event this happens.

HOWEVER, I’m too tired to go further with this and the distance of BOTH the preceding automation point AND the point written after the event matters. Longer distance before = deeper drop, longer distance after = shallower drop.

Also, in all of the above cases there was no automation point after the event before writing manually and moving the event. With an automation point between the event and the manually written bypass this doesn’t happen (for me).

So two more images for “maximum” and “minimum” drop situations:

EDIT: It looks like I moved the event different distances, but the event is pretty much in the same place. Use the Channel Settings window at the bottom as a visual ‘anchor’ to see the different distances of the automation points.

EDIT #2: Looking at it again it actually seems as if there is an invisible line draw between the two automation points that have been written on the automation lane and that the system then takes the interpolated value on that invisible diagonal line and simply creates a new line at that point. The interpolated ‘invisible’ line is higher under the event in some images and lower in others, and the resulting new line under the event seems to match that.

1 Like

Back to paid work now…

1 Like

Thankyou. And when you do multiple copy-pastes or moves, you end up with the utter mess I illustrate in posts above.

This seems to happen with any “switch” parameter (i.e. any parameter which should only have two states - on/off , bypass etc.).

Please do read all the examples above. I, also, do not have time or brain-spaces to repost it all. :slight_smile: