will work, but mind the power of your mini pc and the needs of the planned plugins. Need to keep your system on max of 128sample or even better on 64samples, so first test all things out on your primary computer
Hi, I’m telling you that in our band we use VST Live for backing tracks, and for guitar and bass rigs.
After testing and experimenting for a long time, we found that the most efficient plug-ins for this type of task are Tonex. Neural DSPs sound amazing, but when using multiple instances, they sometimes have small audio dropouts. Amplitube, if you use many instances, tends to crash the RAM.
I’m talking about a workflow, where for each part of the song a dedicated amp is used (for example, clean in the verse, distortion in the chorus, delay in the bridge, etc…). Once you get used to it, putting together a few preset stacks makes putting together a song very fast and efficient, absolutely detailed. For me it was much better than programming the plug-ins with automation. I use Tonex on almost everything, and some Neural DSP on certain songs. I keep all the songs in one big SETLIST, and then from there I build the setlist for the particular show. Per gig a project totals approximately 250 stacks with rig simulators.
The live rig is an X18 console, and a Macbook Pro M1Pro with 32Gb of Ram. For rehearsals and project building we use an RME Fireface and a Mac Studio M1 Max. Each vst Live project has individual inputs and outputs routed for bass, guitar, backing tracks, and Click (we prefer to use the click rendered on an audio track). Live monitoring is handled from the X18.
Exaclty what I figured out from my initial test of VST Live ! The Helix is a great gear but setting up song presets with one snapshot per sound for each part of the song takes a lot of time. And having VST Live automatically change the sound on each part helps focusing on guitar playing.
Moreover with VST Live I will be able to use multiple sound simulation technologies (Genome, Neural, Helix…).
I will try the ToneX plugin as well if it sounds good. Do you use the effects included in the last release of ToneX ?
That is a very large number of stacks !! Are they all prealoaded and activated at the same time ? In my test, I start having some dropouts with only 16 stacks at 256samples. I test on my day-to-day computer (an AMD Ryzen 7 3700X with 8 cores and 32 GB of RAM). Is there a way in VST Live to deactivate the stacks that are not used on a specific song ?
Is the 32Gb of RAM necessary ? One mini-PC candidate for my rig is this one : Asus nuc 14 pro+. Another possibility could be a Mac mini with M4 CPU and 24Gb of RAM.
I need to make some research to see what processors are good for live audio. I guess the M1 Pro is one of them.
I am really happy with the M1 processor for audio, even before the Mabook pro with M1 pro, I used a Macbook Air, M1, and the audio really worked wonders. Super stable, no fan noise, and very efficient. I recommend it. I don’t think that much RAM is necessary, but I preferred to go for an M1 pro with 32, than an M2 with 16. At first, the Tonex effects were limited, so I used delay, octavers and other fx separately, and then I put together different preset-stacks (guitar solo-chorus-etc). Our guitarist plays in D tuning, and for some songs he transposes it to C, for that I usually use Neural DSP Rabea, and Cory Wong for some clean stuff. But always a limited use, because when accumulating several parts with these plug-ins, for some reason, they usually produce an audio dropout when starting a part, after playing about 4 or 5 songs. The structure I use is in parts, two instances playing together simultaneously, one for the bass and one for the guitar. A song can have 10 parts or more. A complete setlist, about 15 songs, 150 parts, 250 stacks or more. The parts where some of the instruments don’t play, I leave free of stacks.
I always preload the entire project (it takes quite a while). We’ve already done three concerts this way, and we’ve had no problems.
As for Tonex, it’s patience and finding the snapshots that work for you, after that, working with each song, effects, and testing. The good thing is that it doesn’t bog down the CPU and runs very smoothly.
Once you have it, you just save the song, and you just dedicate yourself to playing your instrument.
For a metal group 1 bass 2 guitars I made a 18 songs set list only with 3 global stacks and 1 amplitude 5 each global stack (vst2 version because program che fra only works with vst2 for some reason). Just made all the presets I needed and saved inside amplitude, then assigned each preset to a program change slot, then for each song and each part (verse choruses solos etc) made a midi track that send program changes to virtual midi out 1 for bass 2 for guitar 1 and 3 for guitar 2.
Then connected virtual midi in 1 2 and 3 to each amplitude in the global stacks. Works like charm, only 3 plugins loaded, very cpu friendly. If you can have presets saved in the plug-in, why load hundreds of istances and uhundreds of song stacks and parts …. ?!?! Find the optimal cpu route and use it!
Personally, I feel like I have more control with preset stacks for each part. I can produce in more detail and faster. I mean, it took me a lot of work to find the right tones that worked best for me, but after that, I have about 14 preset stacks, and it’s just a drag and drop, into the corresponding part. Obviously, everyone finds their own workflow, but I find MIDI automation a bit more cumbersome for my workflow. Also, when I want to add a part, or change something in the song, it’s very simple this way, and I don’t have to automate again.
Understandable, I tried but found that after a certain number of parts and plug-in in each part, crackles happens at part change, while program change inside the same always active plug-in in global stacks feels seamless, anyway, what works works!
Exactly, I started when VST Live came out, and it didn’t have automation. I tried Amplitube (but it overloaded the RAM), Neural doesn’t make crackles, but it produces audio dropouts, I tried others, and then Tonex came out, and I had no problems there, even in setlists of more than 30 songs. Then automation was introduced in VST Live, I tried it, but I had already gotten used to the other workflow, and I had many songs already finished. Luckily, the program improves day by day, the developers listen to us a lot, and I think there is still a lot to grow, and simplify the workflows. Greetings.
May I ask what is your hardware configuration ? While I really prefer the flexibility of Cucuso workflow leveraging parts, I might have to rely eventually on automation and a limited amount of VST if crackles happen.
@Cucuso, do you use Global parts in addition to song parts ?
For everyone, what is your audio samples number : 128 ?
Just to add…. I made midi events clip for each of the sound presets (each clip named after the preset and containing only the relative program change, then it’s a matter of copypaste the events in the songs timelines on the correct guitarist midi track… same as duplicating a part ….
Hi, I replied to the same question about config to your other request in the thread about compliments to developers. Same pc.
I use 64 sample for light projects, 128 to 256 samples if loads of dmx tracks and loads of plugins are used. It’s not about number of plugins, it’s about how “heavy” on cpu is a plug-in. I have a project with more than 50 active plugins, but 2 istances of Antares auto tune artists plus two istances of amplitude 5 makes up 80% of cpu load (single core, can’t use multi core audio because it crackle even with zero plugins I don’t know why, otherwise my cpu load would still be 10% … @musicullum do you know why even with only 1 stack and zero plug-in zero other tracks zero video zero DMX I still can’t use audio multicore?? On a i7 10th gen, win11, 32 gb ram seems strange).
Ciao!
Thank you for your detailed reply on the other thread ! Unless I missed it, you did not mention your computer setup (CPU, RAM, ASIO samples…). It is an i7 10th gen, win11, 32 gb ram as you mention in you previous post ?
Sure! Thats an Asus zenbook duo 13”
16GB / i7 (4core), thunderbolt usb-c’s, HDMI , usb3A.
Wanted to have a laptop that has all the ports I need by default, run the show w.o. port hubs, etc.
Why dual screen? Because there is always a need to have an overview on as much as possible.
How many stacks with effects running simultaneously with this CPU ? Did you have any crackles/dropout issues ?
I just found that when you select “nc” on the output of a stack (i.e. no output), it looks like the stack in somehow unloaded and it frees some CPU. I managed to get rid of crackles by doing this on the stacks I do not use. Doing the same on the input instead of the output does nothing though.
With automation, I guess that it would be possible to shut down the stacks that are not used in a specific song or even in a specific part of the song…
yepp, it looks like VL is suspending to process the hall signalchain when no output is selected.
About crackles, there are different things can cause crackles…
Just telling you a real Windows one… .e.g. when connecting laptop PSU, it does a crackle at my setup…
So my experience and final result for this config is:
I’m using DMX tracks for 800 channels and when they are “moving” my computer seems to struggle with so can’t go below 128samples and even with 128samples I have tiny crackles at 1-2 songs.(still after reducing events). W.o. DMX channels the situation is much better… even can reduce latency to 64samples.
When utilising a massive guitar-rig chain, that’s eating lots of resources… os finally with DMX… I picked some extra horse-power with an UAD audio interface that has on-board DSP… so currenty loading my Marshall amp, the EMT140 reverb and some additional plugins right on the UAD interface. And there are plugins are runned by laptop CPU.
Needless to say, you can’t buy this config anymore… but faster, stronger… as I purchased this laptop in 2020.
But basically the hall show could run from this laptop if dropping the DMX.
Yes… with plugins I took me time to evaulate different plugins and manufacturers, measuring proc-latency values… Yes… that needs some extra attention as the “LIVE” thing must run in the lowest possible latency and very stable.
If you are ok with a one screen setup… I would go grab from the new M4 MacbookPro’s they are definitely mindblowing powerful also OSX - in my experience - does a smoother performance regarding Audio.
If you have any further question, feel free to ask
