Im now on Live 12.0 and open a project with 50 tracks. Even if I freeze tracks there is a terrible latency audio + video!!!
How do you guyz working on producing a big project like this (who should ideally growth with more tracks)?

I also wondering how Im going to play a Liveset (midi + voice audio) if just adding auto-tune adds 30 ms latency it's going to be impossible to play a complete hour with interaction....

i5-12400 + B660 asus chipset + ddr4 + AMD Rx5700 xt + MOTU M2

 

bounce & freeze - you should never need 50 tracks armed simultaneously with active VST instruments and shitload of plugins

 

for me Ableton 12 is still unusable with large groups full of audio clips - which is most of my projects because thats how i develop rough ideas...

 

ah yeak ok I got it. You mean stem workflow and you "bounce" per region even if you like... Thx for your quick answer!

 

Yes I know what you say. I also prefer having all options avaible like mixing and producing and arranging in a multi step way of doing... but it's impossible to manage without bouncing on stems like @tzzsmk

And yeah 12.0 is still in beta I prefer think when I use this version alpha who is not bug free!!!

 

exactly, if you bounce (but keep original muted/disabled) you can always go back when needed, but should eat significantly less resources - which should also let you lower total latency without sacrificing stability (not causing dropouts)

 

Yes I have a tuto from a good producer course you can make group of groups and select all your main export groups (consolidate) and render with choosing "selected tracks only".

After doing this you can still drag in some material from the old messy project: (-like instruments, groups, tracks, anyway... ) by going in the browser and put individual files on the current set and continue working.
It's important to label your tracks and stay orginized to free your mind.

Of course you can a make stem out of an existing stem mix to further consolidate the project.

 

Last edited: Mar 15, 2024

hey guys, my 2 cents, altough the last time I used Ableton was in 2009, I did a lot of live actions from 2007-2009, but every DAW has a way to consolidate audio (bounce) and put the plugin to sleep.

Even so, there will be a little latency.

I am not a native english speaker, of course you can see that, but if I understood well, do you sing live and need the autotune?
Well, the way I found to deal with that is using a hardware unit, like roland VT4. I know its not the same, but there will be hardware that could do this for you. . I send only midi to the autotune hardware and It works for me.

cheers

 

i'm not fighting with latency in recordings, but I am trying to conserve CPUs. RAM, I have plenty of.

I commit things to audio. I disable every plugin on the channel, and the channel itself. I put the rendered track below the original, and then I hide the original. If I already have my groups setup, (using Logic Summing Stacks) I collapse them to just be represented by the group channel. So maybe for a 100 channel project, I have something like 10 visible group channel strips.

If I want to go back and change a channel, I have the original right above it inside the project. I like committing stuff to audio, because it helps not get stuck in "forever tweaking" mode. Setup your channel correctly to bounce as a stem, and deal with whatever you printed. If and when I have to go back to an original because the render "isn't working", I consider deleting it altogether. Of course, depending on why it's no longer working.

It's the same thing I would do to reduce latency. But if I was having a latency issue, that M2 interface would be the first thing to go. That is, assuming you already replaced its usb cable. You can find a better usb cable than what they ship that interface with, at your nearest convenience store.

 

Last edited: Mar 15, 2024

Check for any plugin latencies and disable latency causing tracks to narrow down the problem and/or disable latency compensation.

The compensation changes every now and then (and Live has been historically bad for this because it was/is geared towards performance), also plugins can implement latency reporting either differently or start doing it in first place, leading to surprises sometimes. Remember that track latency (+sends) is cumulative, so the highest counting track/return track "wins".

 

I really have this problem, man.
I render my tracks to audio and put plugins to sleep, but I end up waking them every now and then

 

Unfortunately, as the project increases in complexity, it is necessary to raise the buffer size in order to avoid those annoying noises due to buffer underrun. I often find myself dealing with with projects working at 2048 samples only. This is my workaround when I have to play a pattern on a busy project:

• Export a ruff mix of original project
• Open a new project and lower the buffer size in the settings
• import the mix file on the timeline
• Insert vst instrument and play midi pattern (or record audio of the performance from mic, guitar etc...) with low latency
• Save the pattern track (or group track) by drag/drop into user library (browser on the left)
• Open original project, raise the buffersize and dragdrop the performance from browser into a new track
• Done

 

Last edited: Mar 15, 2024