Thanks for the sharemouse tip. I use VNC to control my secondaries setups but I always used a virtual keyboard for the secondary setups, sharemouse will solve that.

Yeah I'm eying an upgrade of my main setup but I'm actually in no rush to do it thanks to this handy solution that I'll keep using even after upgrading. I have 2 other computer as secondary setups right now, older setups that were fine workhorses in their own times. Each is dedicated to its task (one for aquas, one for virtual instrument/samplers with cpu needs), with VEP it's very easy to do a bunch of presets that perfectly fits your needs. The things run in the background and are always ready to go independantly of the load I put on my main setup, which is also a lovely plus. Both main and secondaries are never in the red, you can balance power needs very sweetly. Also since they have very little things installed on them, they require less reinstall and maintenance. The main setup is the one that requires the most time in that domain.

Also why I wanna change screen. I have a dual screen setup right now with one screen dedicated to my secondary setups. I want an ultra wide one where I can perpetuate this behavior.

 

Last edited: Feb 14, 2020

Yes, but I think it's better to try reducing your CPU usage first, by e.g, using less & DAW internal FX instead of 3d pty.

 

I read that servers with multiple cpus were bad for music production because of bad latency performances..So is it wrong?

 

THIS IS a home studio!

 

well we have multicore cpu's which are essentially 2 or more cpu's in 1, so sort of the same as dual cpu's

 

Not at all in this case. You can see he's turned them to Vienna Ensemble "Satellites". Nothing wrong about it. They are used for almost what they were built for, running as servers for VEP. In short, they just run plugins and libraries not DAWs.

Sorry but this is quite ridiculous, something that you would advice a total novice for wasting his comp's resources back in "98. Everyone else today, just can't get enough. Moar powah plz. That's the nature of the beast brah.

Not even sort of the same anymore. Between Intel & AMD, only AMD had this problem with their chiplet design and Ryzen gen 3 has solved it by giving all cores equal access to the memory controller that sits in the middle of the cpu's die. Intel's cpu design is monolithic so far and never had these probs with one cpu, this may change because i hear Intel's planning a chiplet design too.
But with dual cpu mobos you have everything x2. Each cpu has its own chipset, memory controller and ram slots etc etc. The system decides which cpu will run the tasks at hand and audio apps don't handle the latency between the two cpus well...
(EDIT) Whereas rendering and 3d apps, video editing and so forth don't care about this so they scale better with a ridiculous number of threads. But for Windows 10 Home and Pro, the max the os can handle is 64 threads. You need Win Server or special Enterprise version or Win 10 Pro Workstation edition if you want a cpu with more threads.
PS: Win 7 Ultimate x64 supports 256 threads if i recall correctly. Oh dear... so much for the future.
Cheers

 

Last edited: Feb 14, 2020

NetVST (free plugin)
http://netvst.org/blog/
http://netvst.org/wiki/doku.php

 

Last edited: Feb 14, 2020

It only works for FX rendering purposes. When it comes to video FX, it's the equivalent to producing a song with your DAW but the render process takes several hours or even days. No real-time use.

As @taskforce said it's different. I think the important thing to remember is that the weakest link with audio PC slaves is the added latency of the Ethernet communication. It's small for what it was designed for, but for real-time audio 25 ms (random guess) of additional latency can be a problem if you for instance are playing/recording live your guitar/keyboard.

If it's a virtual instrument you basically score/program with the piano roll not that much of a problem or not at all.

 

Last edited: Feb 14, 2020

Xupito
first
you are talking about a hypothetical situation

says nothing about tracking

secondly in toms example he used com over ethernet but an/digi audio for sound

lastly
nowadays thanks to ravenna/aes67/dante, 25ms is nonsens ;)
if one needs AOIP it of course

 

Depends on the pcs but with VEP and two slave pcs it's more like 1- 1.5 ms. Perhaps 2ms roundtrip. A little more when 3 and 4 pcs are added and Hackintosh gets in the mix too. Not 25ms mate. And why on earth would i be recording on a remote pc ? I don't get it lol. Unless you meant recording on the main daw computer but i don't see why would that introduce added latency.

 

Last edited: Feb 15, 2020

Comparison of audio network protocols
https://en.wikipedia.org/wiki/Comparison_of_audio_network_protocols

Sadly there´s not Vienna´s protocol listed, but you´ll get some idea of the latency when moving audio thru network

 

Rather than do that. I think 16-24gb ram and 3.4 ghz is ideal for 16-24 track recording. Have a look at Samplitude and Sequoia which Masterers use and commented to have the best audio and find Greg Wells Start to Finish, he changed my approach he uses 4 plugins on vocals or hardware when he can, entire song on 1 stereobuss and limiter across chorus elements. His methods will save me from some latency headache playing with buffer times the more vsts' added.

 

Wow, hold your horses!

For some reason I'm pretty sure I was talking about something not only the op but two pages later quite some people are interested.

Same thing.

As I stated: random guess to put an example. BTW, I couldn't care less about latest technologies when most people at home have your average joe 1GB ethernet card.

Same thing.

Where on earth did I say that? I was obviously talking about the effects of slave pc's latency when playing live on the master computer. Recording or not.

 

i didn't put a gun to you head ;)
i thought maybe you'd like to know

hmmm........you are doing it again
you can use your existing gear and use adat or aes spdif
or dante or any other stuff like long audio cables. none of it
is latency dependent.
the only thing transmitted over "real" or pure ethernet
is toms video/grafix and mouse, controller signals.
think of it as old-school usb-hub (wich it can have) - monitor switch for peripherals.

that's the whole point of toms rig
super power No latency

 

Just some corrections, i was wrong here, win10 pro supports 128 threads, when a 64c/128t cpu is present though, it divides the processes to two processor groups of 64 threads, which of course doesn't help performance (and even worse for latency) as each processor group has its own priorities. Win10 Enterprise supports 256 threads. It is obvious MS wasn't ready for such a giant leap in consumer technology, as the one AMD offers right now. I'm not sure Apple is ready either, but i read they have been mesmerized by AMD's offerings and have been buying AMD cpus to test on their systems since summer's rls of Ryzen gen3. Don't be surprised if the next gen Apple comp has an AMD cpu hehehe...

That's what Dante,AVB and the rest of Ethernet audio tech leverage. The built-in 1gbit connection found in all comps currently. These aren't latest technologies, they are tried and tested. The majority of live audio works with some form of Ethernet audio as it is the best in terms of connectivity and control. Btw, i wasn't "correcting you" there about the latency, i saw what you wrote, i was just stating my experience, i have also used VEP live (macbook with shuttle mini pc slave) and had no problem performing.
About that recording and remote pc thing, i could swear someone wrote all they 're interested in, is use a remote pc to record live or something like this anyway. But looking back at the threads i don't see it anywhere lol. So no worries, my bad mate

One thing we have to note here about VEP, it's not some proprietary form of Ethernet audio protocol. At least not one that i 'm aware of. It doesn't perform as a dedicated audio card like AVB, Dante etc. have various interfaces, if it did you could have every comp on the network being a slave and a master at the same time and run multiple daws but this is not the case, VEP is just a smart vst/vsti host with a few more perks under the hood.
Cheers

 

Last edited: Feb 16, 2020

Hello, I came across this post while searching for tips on how to control transport between two or three machines, one being the master with the main DAW and the second and third being the VST host machines. I have a mixed setup of:

a) Windows 10 on PC1 with Studio One 6.5, Presonus Studio 1824c (BNC Word Clock Out, only ADAT and SPDIF I/O) and Behringer ADA8200 (BNC Word Clock In, ADAT I/O), and

b) Windows 10 on PC2 with Reaper, NI Kontakt and a huge collection of synths and instruments from a Korg and a Roland subscription, all based on Behringer UMC1820 Audio Interface (no BNC Word Clock, only ADAT and SPDIF I/O) and another Behringer ADA8200 (BNC Word Clock In, ADAT I/O).

c) Windows XP on PC3 with Sonar XI, E-MU Emulator X3 and Proteus X2 and a huge collection of synths and instruments like Proteus X2 and Emulator X3 based on E-MU 1212M Digital Audio System comprising the main card, the I/O daughter card, the daughter sync card and a Micro Dock. This one system is a bomb. BNC Word Clock In and Out, ADAT In and Out, SPDIF In and Out, SMPTE In and Out, MTC Out.

And since one of my interfaces doesn't have BNC, while the other only has BNC Out, I am left only with the EMU 1212M with complete set of I/Os. Therefore no chance of Word clock sync within my system.

And yes, I also have an ART SYNCGEN for external generation of Word clock (four BNC outs) and two SPDIF Outs.

VSL told me they have stopped support for Win Xp and for 32 bit OS.

Therefore, perhaps I am only left with one option - a simple MIDI-Audio connection, e.g. Master sends MIDI commands to slave, slave sends Audio back to Master. But then the question is, how will the transport control work when (a) the VSTs are standalone and (b) when VSTs are hosted inside a host (Reaper, Chainer etc.).

Looking for guidance please.

 

@Sumit Simlai Have you tried audiogridder for offloading? It's network based, so no need for bncs

 

Maybe I misunderstand the scope of this thread; I don't use large scoring templates or whatever. But I run a MacPro 5,1 master and a WinXP laptop together and here's how I do that, if it's of any interest:

The WinXP machine runs on an older audio interface (M-Audio Audiophile USB), syncs via midi time code in the DAW, and both computers send audio i/o back and forth. File folder sharing via ethernet. Works like a charm. I load 32bit plugins in WinXP and trigger them on the Mac over separate midi and/or use it as an FX unit. Latency is no issue.

I wonder if there is anything I'm missing out on with regards to instead using something like VEP? Any advantages it has over my current setup?
EDIT: I see with VEP you don't need any extra interface, for one thing. And I guess it allows you to use just one interface instead of jumping between computers (Neither of which are of much consequence to me, in any case).

 

Last edited: Apr 11, 2024

Honestly I only read about it yesterday and today I pulled up the website to read. Looks interesting. But some posts elsewhere state that it is unstable or something. I will have to try it, but would prefer going for classical solutions like the one @mondomorte has been done.

 

This is exactly the kind of answer I was looking for. VEP is not something I am at all keen to work with, simply because any changes they make to their strategy, for example now they removed XP support etc., is going to severely impact my workflow.

A couple things I would be keen to ask @mondomorte please. You say the M-Audio syncs via MTC in the DAW. Which DAW? The one on your MacPro? or the DAW in the XP Machine? What exactly is meant, would be important for me to understand. Who syncs with whom and why? The idea is to sync a computer (synth) to the main DAW.

You also say that both comps send audio i/o back and forth. I didn't really get this. I thought that normally the Main DAW (Apple in your case) should be sending MIDI commands to the XP machine, while the XP machine in turn send back Audio to the Apple. Please enlighten me.

Thanks in advance.