What's up with Cubase and automation?

[Functional]

I'm starting to worry that Steinberg is considering that automation in Cubase as it is, is a feature that does not need further improvements. As far as I can recall, automation has not changed for a long while in Cubase. The system is the same as it's always been: you have automation lanes for each parameter / track, where the parameter can be automated by setting up points.

However, to me personally, automation is something I enjoy the least doing. I only do it for mandatory reasons. Here's a list of reasons "why"

Each point is very small.

I have almost crazy DPI settings in my mouse, which means that my mouse is super-sensitive. The reason for this is that in most cases, it's extremely good for workflow and my hand gets less stressed. I also have two screens, so there's that. I spend a lot of time on computer and by previous experience I know that lower DPI settings will stress my hand a lot over extended periods. And the additional benefit is that my friends have hard time using the mouse.

What this results into is that I misclick these points very often. This very fact causes about 90% of my frustration. In fact, because I don't sit next to my screen, these points are also often misclicked just because of that. This was true even before I got 8200 DPI mouse. Even then I did not enjoy working with automation. Especially when you're working with lot of points next to each other, it's very easy to accidentally select points you didn't want to select.

But you need those points for curves!

For reasons completely unknown to me, you can't do bezier curves. They are not implemented. They are always requested, but I've never seen a straight answer as to why they aren't still in. Nobody is telling whenever they are being worked upon or not. No information at all about them. Why do they have to be completely silent on this? I would happily accept an explanation that there are technical restrictions or even something like "We just don't feel that it's worthwhile to invest the resources into implementing it for now; do not expect this in near future."

So why are bezier curves useful though? Well, two main reasons. First is that they require less points. It's not at all about getting perfect curves (which many people seem to think). I just don't want to set up 3-7 of these stupid points to get a curve. Then I want to change the slope a little perhaps? Yeah, off we go to click-hunting the elusive points again. Each one between the first and last points of the curve. Each one. This is anything but fun. It feels frustrating.

The other reason is that by having these curves, more creative possibilities open up. When it doesn't feel tedious to make them, you can do nice things such as automate EQ parameters for some creative sound design or something. Back when I used FL Studio (like, five years ago), I sometimes even didn't use filter envelopes because the automation clips were far more fun, even if it involved more work. And at that time, I didn't even think half of the possibilities I currently do regarding automation. That's depressing too, because you have a vision and you know how to accomplish it, but you don't want to do it because it's far too tedious. All the while you know that it doesn't have to be that tedious.

Why only automation lanes?

Speaking of FL Studio, it had clips. I can understand why some people, especially doing far more simple automation work, would prefer the automation lanes. And I think they're mighty fine and they don't have to replaced. But I do wish that we also had a feature like clips. Like, have an automation folder for each track where you can use automation clips. These clips are much nicer to work with for creative purposes at the very least. The system could work so that by default automation lanes are enabled but you can optionally enable automation clips instead, which would work like midi clips, sort of. It opens a new window where you can set up the automation points in a huge view. Have very big automation points and so on. This would save a lot of headache instead of dragging that automation lane horizontally to make it big.

Frustration deepens

So another issue. In order to lock the horizontal position, I have to press ctrl key. But this means I will also bypass quantization. Okay, this might be sometimes fine. But quite often it's not fine. In fact, for me to copy 4 bars for example, the easiest way is to set up the starting and (possibly) an ending point. But for me to do that, I want the starting point to be quantized tightly AND be locked in horizontal position (so that I do not make any adjustments as to what that parameter is at that point!). But there is no actual way to do this without changing the parameter at least a bit. This doesn't always matter that much (I mean, okay, the parameter changes a little at worst, so what? But sometimes it really matters if that small change turns into big outcomes, depending on what parameter it is and where in terms of signal chain.

And another annoying problem related to the last one. You can't actually make any limitations or "anchor points" (think of it as horizontal quantization). No max value or anything like that. This is sometimes annoying. Let's say I want to automate volume. I want it to be at max -8.25 for whatever reasons. Only way for me to get to that value is to copy it from another point or record automation. "Anchor points" instead would be points where it anchors to certain value. These would obviously require a lot of setup from the user her/himself, but these could be very useful for example to automate dubstep wobbles. This whole practice seems like a mess.

Most of these issues, however, wouldn't be as big issues if we had automation clips too. It's not a big problem to adjust volume to -8.25 and copy such points for future. But problem is that you have to do this sort of stuff for a lot of other stuff too. It gets annoying and quick, hence you want to spend less time in doing this stuff.

So am I alone in all of this or does someone actually agree? Because the way it is right now, I'm going to start looking at other DAWs. Once a DAW offers me the same functions that I like in 8.5 but also has improved automation, I'll leave Cubase behind me, because this part of Cubase is just too tedious yet I find that automation can be very powerful tool for creativity.

[xalama qo]

I agree that bezier curves would be really welcome.

However some features you might be missing to make automation editing easier:
- increasing automation lane height increases the size of the points too
- use the "appearing/disappearing" arrow button at bottom left corner of track header to max/minimize
automation lane quickly to avoid headaches
- use the info bar above the arrangement window (ctrl-click on a point for a menu, select Show info) to fine tune point values and horizontal position. Values can be double-clicked and typed in precisely.
- selecting multiple points such as in a curve allows manipulation of those points as a group in various ways by using the controls in the corners, side and bottom edges of the selection box.

Not as quick and easy as curves, but a heck of a lot more doable than aiming for individual tiny points with a hyperactive mouse!

[Functional]

I agree that bezier curves would be really welcome.

However some features you might be missing to make automation editing easier:
- increasing automation lane height increases the size of the points too
- use the "appearing/disappearing" arrow button at bottom left corner of track header to max/minimize
automation lane quickly to avoid headaches
- use the info bar above the arrangement window (ctrl-click on a point for a menu, select Show info) to fine tune point values and horizontal position. Values can be double-clicked and typed in precisely.
- selecting multiple points such as in a curve allows manipulation of those points as a group in various ways by using the controls in the corners, side and bottom edges of the selection box.

Not as quick and easy as curves, but a heck of a lot more doable than aiming for individual tiny points with a hyperactive mouse!

Wow, you're actually right. Lane height does increase the size of the points too! I never actually noticed this and god knows why, because sometimes I do increase the size of the lane. Though I rarely prefer to do so, because again, I only do it for mandatory stuff. Still, it's helpful, gotta do this more. About the arrow I already knew.

The header I also didn't know of, which also migitates some problems!

Regarding the multiple points, that I knew of as well. Unfortunately it rarely is useful (though sometimes is!). Nevertheless, I clearly have to increase the size of the lanes for now, given that no other option exists for now. It really does help because the size they get to (which isn't big) really makes a huge difference. Still kind of small but much, much, much easier to click after some testing. So thanks!

[xbitz]

"Speaking of FL Studio, it had clips" not just clips FL Studio is far ahead of the other DAWs in automation point of view

https://www.youtube.com/watch?v=tV8DspkmD-Y

every DAWs has strength and weakness, automation is definitely one of the strength point of FL

[xalama qo]

I agree that bezier curves would be really welcome.

However some features you might be missing to make automation editing easier:
- increasing automation lane height increases the size of the points too
- use the "appearing/disappearing" arrow button at bottom left corner of track header to max/minimize
automation lane quickly to avoid headaches
- use the info bar above the arrangement window (ctrl-click on a point for a menu, select Show info) to fine tune point values and horizontal position. Values can be double-clicked and typed in precisely.
- selecting multiple points such as in a curve allows manipulation of those points as a group in various ways by using the controls in the corners, side and bottom edges of the selection box.

Not as quick and easy as curves, but a heck of a lot more doable than aiming for individual tiny points with a hyperactive mouse!

Wow, you're actually right. Lane height does increase the size of the points too! I never actually noticed this and god knows why, because sometimes I do increase the size of the lane. Though I rarely prefer to do so, because again, I only do it for mandatory stuff. Still, it's helpful, gotta do this more. About the arrow I already knew.

The header I also didn't know of, which also migitates some problems!

Regarding the multiple points, that I knew of as well. Unfortunately it rarely is useful (though sometimes is!). Nevertheless, I clearly have to increase the size of the lanes for now, given that no other option exists for now. It really does help because the size they get to (which isn't big) really makes a huge difference. Still kind of small but much, much, much easier to click after some testing. So thanks!

[Functional]

Okay, as today is a mixing day for me and I've decided that the vocal track doesn't need de-esser (or more that, I didn't find any sweet spot for the de-esser), I resorted to automation. Kind of, not really in the track but the actual audio clip (gotta have it pre-inserts this time). So I admit, a lot of my headache is gone this time around with having that horizontally as big as possible. Much easier to do work like this at the cost of some aesthetic.

So I'll take some of that back, it's not nearly as frustrating anymore. But I still wish they would improve automation so that it would be less tedious when used especially for creative purposes. And I'm still upset that Steinberg seems to avoid questions regarding the plans on automation.

EDIT: Haha, nope, that was a placebo. The points in audio clips don't actually scale with track horizontal size. I don't even know anymore what to think. Maybe my hands are shaking less now, but doesn't feel like it.

[xalama qo]

If you automate track volume the points in the automation lane will enlarge with increasing auto-lane height. Again, not to hugely massive size, but better than the usual pinheads.

If you're de-essing and resorting to automation, wouldn't it make more sense to insert an EQ with a notch filter at the problem frequency/ies, and either automate the band gain to coincide with the sibilance, OR automate the dry/wet value if you have multiple notched frequencies?

If you're editing in the audio clip because you might want to move stuff around a bit later...say eg in post for AV or TV...then perhaps de-ess the entire vocal, export the de-essed version, and then import the fixed vocal into the project for further edits. That's how I'd do it
I'd probably save the de-essing stage as a separate project so I can go back and tweak the de-essing if needed.

[Functional]

If you automate track volume the points in the automation lane will enlarge with increasing auto-lane height. Again, not to hugely massive size, but better than the usual pinheads.

I'm not automating the volume lane, but rather, the actual audio clip. You can do that by selecting the pencil tool and then clicking the audio file.

If you're de-essing and resorting to automation, wouldn't it make more sense to insert an EQ with a notch filter at the problem frequency/ies, and either automate the band gain to coincide with the sibilance, OR automate the dry/wet value if you have multiple notched frequencies?

Bit OT, but I'm sure it doesnt matter:

In this particular case, it's pretty much simple enough to automate the clip itself (as it comes before any possible inserts). The problem is that the signal chain emphasizes them heavily and they get boosted from the very beginning. So the problem is less in a specific frequency rather than it is in the high-end content in general. As luck would have it, however, I somehow managed to write lyrics where there the problematic consonants are in good spots for automating merely the volume, rather than seeking any specific frequencies with notch filter and take them out.

With that being said, there are couple parts with some issues at 200hz or so, so either I'll use an EQ to automate it and at those frequencies I would consider linear or minimal phase, which by common sense doesn't seem like a smart thing to automate. Or maybe I'll use multiband compressor for that and automate dry/wet instead. Don't know yet, gonna see.

If you're editing in the audio clip because you might want to move stuff around a bit later...say eg in post for AV or TV...then perhaps de-ess the entire vocal, export the de-essed version, and then import the fixed vocal into the project for further edits. That's how I'd do it
I'd probably save the de-essing stage as a separate project so I can go back and tweak the de-essing if needed.

Oh, this is just going to be our first release (a song) and it's gonna be on Soundcloud. All that is left is some sugarcoating (possibly), the mixdown and then mastering. So everything is already arranged here.

[xalama qo]

Good stuff, all the best!

[jancivil]

I use tons of automation. Some of it I interact via a GUI with the mouse, some of it is initially set with controllers (some of it is MIDI CC), a lot of it is host automation. I use Cubase. In the end, there is usually a lot of adjustment to be made, and I'm *writing* automation, as typically via typing as graphically.

I don't know why a 'bezier curve' is in itself a thing. It appears to come from graphics work.
Host Automation is basically where I mix. The decisions I make are not abstract, they all do something actual.

Beyond the abstract shapes (which I don't have much use for in CCs, let alone Host Automation), the idea of extremely fine resolution is def. not a Cubase thing, as it resolves it right away to the simplest form.

So I don't really relate to or 'get' a lot of the thrust of that concern, as to a musical decision.

I also do a lot of selecting ranges in the audio editors rather than automate, owing to the Offline Process History (or Cubase's way of basically clip or object editing).

[sadriabi]

I wish there was a seperate volume trim automation lane like in Pro Tools.

[xbitz]

additive and multiplicative curve based automation modulation also missing from Cubase, but it's a "just" a post-processing DAW ...

https://www.youtube.com/watch?v=CggQXSA_LuA

not a creative type one as Ableton Live, FL, Bitwig

[Functional]

I don't know why a 'bezier curve' is in itself a thing. It appears to come from graphics work.
Host Automation is basically where I mix. The decisions I make are not abstract, they all do something actual.

In the end, I think the issue here, so to speak of, is indeed that Cubase seems to be targeting audience such as you, who I guess would not benefit from the bezier curves in most cases. I still find it odd though, because it just doesn't seem like a feature that is truly that hard to implement. If you draw automation, Cubase will already make an estimation of whatever you draw and delete the points it considers unnecessary. This is very much how bezier curves would be implemented.

Beyond the abstract shapes (which I don't have much use for in CCs, let alone Host Automation), the idea of extremely fine resolution is def. not a Cubase thing, as it resolves it right away to the simplest form.

Oh, I think here is a possible misunderstanding. To elaborate, most of us fanatics who are in love with bezier curves have used FL Studio in the past. Once we switch to another DAW, we notice that automation works a lil' it different than it did in FL Studio. A lot, in fact. So we google about it and we hear the word "bezier curve". So essentially, we start calling that very feature as a "bezier curve".

What the feature is? Well, I'm not sure how high the resolution is in FL Studio, but it's not at all about the resolution. Human ears become rather insensitive to super fine changes as you know, so there's no point in having that super resolution. The actual feature works like this: between each two points in automation "clips" (as they are clips in FL Studio), another dot is automatically created which you can drag up and down. With this point you define the curve. So you want a downwards slope or upwards slope? No need to click at least 2 points or even 5 at worst (in my experience), you just click this one handy point and set it as you like.

What actually is going on the DAW itself is another thing. The bezier curve is just a representation, I highly doubt that FL Studio actually could keep it as accurate within the engine. Could be wrong, but still, that's not what I'm looking for.

So essentially this is the feature we are talking about when we mention "bezier curves". It's very addictive and actually fun enough to use to the point where you might start using it instead of doing things internally in a synth, for example. Not necessarily everyone would find this feature useful, but I think it would work well if we had automation clips as well (as an optional feature) where these curves would be included. I find this more as a creative tool rather than necessary for traditional automation, unless one is into the EDM-oriented stuff, which love to use curves on their buildups and what not.

I also do a lot of selecting ranges in the audio editors rather than automate, owing to the Offline Process History (or Cubase's way of basically clip or object editing).

To be honest, I've not tried this feature yet properly despite knowing of its existence. It actually does seem super useful in some ways, but also, it only does dry/wet, which kind of limits its functionality in replacing automation completely. Perhaps I should give it a closer look and try it out on whatever I will be working after this.

[LawrenceF]

I don't know why a 'bezier curve' is in itself a thing.

Me either. I mean, I don't know why it's apparently such a "big" thing. Don't get me wrong, it is handy to be able to draw a curve a little faster, which is handy, but there's nothing bezier curves give me (mmv as usual) that I couldn't easily enough do before I had them.

I know for certain that there are people who have legitimate practical reasons for wanting them, no doubt about that. I also know some of the asks for them are from some who only want them because they look kinda cool and some other DAW has them.

If I made a personal list about the practical / useful things that most modern DAWs are missing, bezier curves for automation wouldn't even make the first page of that list.

[chk071]

Is it safe to say that bezier curves is to Cubase what is different time signatures to FL Studio?