Performance and Audio Interface

[peter.antoniou]

Does anyone know if there's any correlation between audio interface and Bitwig performance? I've got a computer that should have plenty of horsepower (i7-10875H, 64gb, 2tb ssd) and yet at times even fairly simply projects choke Bitwig - I've got the buffers and latency maxed. I'm using a Komplete Audio 6. I was wondering if a new Thunderbold interface would help matters.

[pikey]

In theory yes it could make a difference to performance, in practice it's one of the last things I would focus on when trying to optimise performance. Thunderbolt interfaces can really only help reduce latency, and only a small amount due to avoiding the USB clock and instead reading directly from memory.

It depends on what the dsp performance graph in bitwig looks like to understand what might be the cause of the performance issues. If it's spiking somewhat erratically even at high buffer sizes but low otherwise then likely other drivers on the PC are causing DPC latency, so you'd want to try updating drivers, or removing hardware and disabling drivers to test if it improves the situation and find the cause.

[zenlinux]

Have you disabled CPU frequency scaling? That's often a culprit. Assuming you're using Windows, this is a good guide to follow to configure your system for low latency audio work: https://help.ableton.com/hc/en-us/artic ... -for-Audio

[drsyncenstein]

the way you work with tracks and fx has a large influence on performance. A single track can overload Bitwig. Yet the same instruments\fx\plugins might use substantially less cpu if you place them in different tracks. I'm not sure what the rules in bitwig are, but in ableton a fx return channel processes all tracks that are send to the fx return. So they easily overload the cpu limits.
So try to expermiment with that and get a feeling how to lower cpu usage.
Even a powerfull 24 core or even 32 cores can push the limits with just one track, if you put too much in it. The rest of the cores can be picking their nose, but you still get crackles and such.

[peter.antoniou]

Quick update - I ended up getting the Antelope Zen Q and there was a huge difference in performance. The same pieces that were maxing out the CPU now sit around 20%.

[JHernandez]

I wonder what it was. Highly doubt it was the thunderbolt that fixed the issue.

I'll just put this here.




In any case, whatever it was it's working now and that's good.

[OdoSendaidokai]

I wonder what it was. Highly doubt it was the thunderbolt that fixed the issue.

If you read the feature list of "Antelope Zen Q" you see that it has an active FX processing unit. So the CPU just can send the data without processing it and the interface is doing a lot of the processing work. Shared power.
This is often the the reason between "cheap tec" and "good tec". In cheap tec you only have some passive DSPs that always need the CPU to process all the data.

[BobDog]

I wonder what it was. Highly doubt it was the thunderbolt that fixed the issue.

If you read the feature list of "Antelope Zen Q" you see that it has an active FX processing unit. So the CPU just can send the data without processing it and the interface is doing a lot of the processing work. Shared power.
This is often the the reason between "cheap tec" and "good tec". In cheap tec you only have some passive DSPs that always need the CPU to process all the data.

Can you explain to me what you are talking about?

[muzicxs]

If you read the feature list of "Antelope Zen Q" you see that it has an active FX processing unit. So the CPU just can send the data without processing it and the interface is doing a lot of the processing work. Shared power.
This is often the the reason between "cheap tec" and "good tec". In cheap tec you only have some passive DSPs that always need the CPU to process all the data.

The offloading only works for their own plugins, just like UAD. There's no way that any of the offloading works on things like Bitwig devices or other plugins.

[drsyncenstein]

I wonder what it was. Highly doubt it was the thunderbolt that fixed the issue.

If you read the feature list of "Antelope Zen Q" you see that it has an active FX processing unit. So the CPU just can send the data without processing it and the interface is doing a lot of the processing work. Shared power.
This is often the the reason between "cheap tec" and "good tec". In cheap tec you only have some passive DSPs that always need the CPU to process all the data.

Are you sure that the active FX processing unit has any function when using normal vst plugins in the Daw?
Yes, there is an FX processor onboard that can run its own Antelope (synergy core) fx plugins. That's basically the same concept as the MOTU 828ES.
But nothing on the website suggests that it does anything for regular plugins (= any other effect or instrument).
The usb or thunderbolt interface always needs to process all data in order to transport it from or to the interface. Native plugins are only processed on the computer.
But the synergy plugins are not native plugins.

Chances are that it's just better drivers for the interface that reduced the load.
Unless that load is only reduced when you run the plugins on the interface itself.
If that is the case, it could even be that the performance for regular plugins has not changed.

edit: i need to type faster. previous post was completed while i wrote this

[OdoSendaidokai]

It's not that the FX Unit is processing your "external" plugins. It only says that there is an active unit doing stuff. And it is logical that this is not the only active unit, when you get that outcome with lower CPU consumption.

Compare it with a wagon you have to push to get it moving, or a wagon where you can switch on a motor that drives it. Here it is the same. AND the wagon has an additional FX "machine" on top that processes some "on wagon" effects.

There are different DSPs what I call "passive" and "active". The passive ones are like the CPU has to tell everything what the DSP have to do: move it here, copy it there and so on. It is like a shelf, where you have to spend all your time to put and sort everything yourself
An active DSP just gets commands like "play", "process", "stop" ..... So you get like an assistant that is already trained and is taking care of putting everything in the shelf, getting it out again and keeping order of everything for you. So you can use your time to do other stuff.

Every other explanation would need to include magic, dragons or alien technology.

[BobDog]

I am pretty sure you may need to educate yourself a little more on this subject.

[backtomusic]

....
Every other explanation would need to include magic, dragons or alien technology.

this cracks me up a bit, because your take on the subject does come of like a fairy tale somehow

[drsyncenstein]

There are different DSPs what I call "passive" and "active". The passive ones are like the CPU has to tell everything what the DSP have to do: move it here, copy it there and so on. It is like a shelf, where you have to spend all your time to put and sort everything yourself
An active DSP just gets commands like "play", "process", "stop" ..... So you get like an assistant that is already trained and is taking care of putting everything in the shelf, getting it out again and keeping order of everything for you. So you can use your time to do other stuff.

Every other explanation would need to include magic, dragons or alien technology.

If dsp's where shelves, they would be called storage.
An efficient driver is no alien technology, but it does make things faster.

Each one has a right to have it's own imagination, but there is no need to imagine computer science and spread that here.

[OdoSendaidokai]

I'm very sorry some still can't understand simple pictures or simple comparisons of two very easy concepts.
The dangerous half-knowledge of some "wonder drivers" that take over CPU tasks, without using the CPU cycles is remarkable esoteric Sci-Fi. Could be a nice fantasy or superhero story.