|
|
Post by OtisGreying on Dec 2, 2022 20:54:46 GMT -6
I want more flexibility/freedom experimenting with using FX verbs, echoes in my sessions but am running into CPU bottlenecks keeping everything inside one session/computer. I have this idea of buying a separate computer specifically for FX plugins functioning as a send/return system, running a DAW session on this computer with only FX reverbs, echoes receiving signal from my main computer's converter outs and FX signal from this second computer going out back into my main setup.
If I ran this second computer at the lowest possible latency/buffering - could this possibly serve as a realtime ITB FX send/return system? Anyone thought of something similar to this? If its not possible my only other options would be investing into some hardware FX units. Thanks for any thoughts!
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Dec 2, 2022 22:07:24 GMT -6
Just buy a better computer. Get at least an Apple Silicon Mac or an Alder/Raptor Lake Windows desktop. Also it's a reverb. You'll be eqing until it's not obvious anyway unless you're making gross modern pop or modern rap. You can use almost anything if you bandpass and eq it enough. Even the hyper metallic EpicVerb and bx_rooms. You sure as hell can use the D/Renaissance/Oxford Reverbs.
|
|
|
|
Post by drbill on Dec 2, 2022 22:09:15 GMT -6
Sure. Why not. Simple to do. But having 20+ verbs and delays and multiFX haven't even come close to bottlenecking my computer which is OLD!!!! Not sure why any relatively modern computer would run into problems. But you obviously know your system better. So sure, I'd do as you described. I don't think latency will be an issue.
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Dec 2, 2022 22:20:20 GMT -6
Sure. Why not. Simple to do. But having 20+ verbs and delays and multiFX haven't even come close to bottlenecking my computer which is OLD!!!! Not sure why any relatively modern computer would run into problems. But you obviously know your system better. So sure, I'd do as you described. I don't think latency will be an issue. Some of the modern plugs are heavy Bill and adding another fx send with them usually will be the one thing that breaks the sessions. I'm thinking of stuff like Megaverb (my old laptop can only run one instance in a mix), Relab Sonsig, Liquid Sonics, Acon Verberate, and the new Strymon Big Sky. I experience this even on a 12900k Bill on a 5 mic drum kit with just MDWEQ, Molot GE, and U-he Satin on every track with the Strymon until I raised the buffer. I had to switch it out for the ancient Oxford to keep my current buffer. Once most of the cores get utilized, who knows what will freeze or crash the mix session. On a modern session with lots of plugins with oversampling on lots of tracks, who knows what can happen as it gets more and more latent and core cpu usage increases. Once any single core gets overloaded, the daw will freeze up or crash when you hit play and you'll have to freeze tracks.
|
|
|
|
Post by OtisGreying on Dec 2, 2022 23:57:35 GMT -6
I've got a silicon Mac Dan, it just ends up as you describe sometimes - the session gets slower and slower and it becomes really difficult to try out new ideas and virtual instruments/fx down the line as I go. I produce and mix as I write and my sessions can get pretty indulgent from time to time, and if I can offload atleast the FX plug-ins I'd be stoked to still have a smooth session - production, mixing included all the way to the finish line without straining the cpu and getting dropouts or crashes etc. that'd be worth the $$. Other big part of this is I intend to buy the Liquidsonics bundle and I want to really make use of those pretty powerful plug-ins I'm happy you guys think the idea would work - I'm glad you don't think latency will be an issue drbill I assumed that would be tricky. I'm picturing it like: (main computer signal) DA to AD (second computer FX signal fully wet) to DA to AD (main computer) - This checks out right?
|
|
|
|
Post by kcatthedog on Dec 3, 2022 1:45:39 GMT -6
If you are just experimenting , why not print your session with no verbs, import that stereo track into a new session which would start with no plug ins and then experiment with new verbs ?
|
|
|
|
Post by the other mark williams on Dec 3, 2022 2:56:55 GMT -6
I'm happy you guys think the idea would work - I'm glad you don't think latency will be an issue drbill I assumed that would be tricky. I'm picturing it like: (main computer signal) DA to AD (second computer FX signal fully wet) to DA to AD (main computer) - This checks out right? I think you’d be better off keeping the signal digital-to-digital between the two machines if you try this. Whether via AES (preferred) or ADAT. I don’t remember if you’ve mentioned what interfaces you’d be running with the machines. |
|
|
|
Post by seawell on Dec 3, 2022 6:26:42 GMT -6
Another idea would be adding an antelope interface that has DSP or something like a UAD satellite. You could run your FX on those without any further hit on your CPU.
|
|
|
|
Post by thehightenor on Dec 3, 2022 9:05:26 GMT -6
UAD 2 DSP was my answer.
I use as many reverbs and delays as I wish and there's no CPU hit.
|
|
|
|
Post by bchurch on Dec 3, 2022 9:08:10 GMT -6
Apple experimented with using additional CPU's as "nodes" for Logic Pro in the late aughts via ethernet cable, though I'm really not sure of what came of it.
The best thing going for offlining right now, imho, is the Antelope and UAD interfaces that include proprietary DSP that offload from your CPU.
Honestly, my brief foray into Apple Silicon had me run right back to the loving embrace of my previous intel i7 (6-core, 3.2gHz, 32GB). Cubase, unfortunately, requires ALL plugins to be m1-native to run native itself. A few daily drivers for me (TC Master X / Brickwall, Naughty Seal Perfect Drums, for example) are not there yet. Cubase 12 runs like a dog on both machines, throwing out wild CPU spikes and glitches. So long story short, I'm back on my 2018 setup and making/mixing music - which is what I have it for.
End of the day, it's not how powerful your shit is, it's how powerful your shit makes you... I prefer to stick with my old Mini, i can drop the sample buffer down to 64 and make a grilled cheese on it without so much as leaving my comfy chair.
|
|
|
|
Post by subspace on Dec 3, 2022 9:14:54 GMT -6
I added a second machine to my set-up for a couple reasons, one being it's nice to have a completely separate interface for musicians using VIs while I'm working in the main DAW. I started farming mix effects out to the second machine as a matter of course and have used a couple plug-ins to tie them together. For sending audio/MIDI between the separate DAWs, Blue Cat's Connector works well: www.bluecataudio.com/Blog/tip-of-the-day/connecting-daws-together-with-connector/To just use the second machine as a DSP farm, AudioGridder opens the GUI of the remote plug-ins on your main machine: audiogridder.com |
|
|
|
Post by kcatthedog on Dec 3, 2022 9:29:00 GMT -6
My m1 mini hardly breaks a sweat: got to say I am curious about how many plug ins you are running ?  |
|
|
|
Post by bchurch on Dec 3, 2022 10:30:41 GMT -6
My m1 mini hardly breaks a sweat: got to say I am curious about how many plug ins you are running ? Barely any. Beyond running a drum kit VI and a few instances of ToneHub (neither of which my i7 balked at), most of my effects are externally routed! I think a lot of it is Steinberg. They've never been a real MacOS-savvy brand and are perpetually a couple of point ones behind the competition. Forces you to get into a "don't tip the apple cart" mindset where you update nothing unless you absolutely must. |
|
|
|
Post by damoongo on Dec 3, 2022 11:50:02 GMT -6
It's called Vienna Ensemble Pro. It does exactly this without any DA/AD conversion and compensates for the latency automatically.
|
|
|
|
Post by drbill on Dec 3, 2022 11:51:07 GMT -6
Weird. My session template has 20 verbs and delays. Doesn't even break a sweat. HDX though. But it's mostly handling mixer duties. All those 20+ plugs are native.
Again, no problem in running a second computer as a reverb farm though. At least in my experience. Maybe there are horrific naive issues that I don't deal with?
|
|
|
|
Post by drbill on Dec 3, 2022 11:52:16 GMT -6
It's called Vienna Ensemble Pro. It does exactly this without any DA/AD conversion and compensates for the latency automatically. Yes. Works great, but slows me down so much I don't use it anymore. Didn't like the workflow, although it seems to be embraced and work great for others I know. |
|
|
|
Post by kcatthedog on Dec 3, 2022 12:01:32 GMT -6
@bc I meant the OP !
|
|
|
|
Post by damoongo on Dec 3, 2022 12:06:07 GMT -6
To just use the second machine as a DSP farm, AudioGridder opens the GUI of the remote plug-ins on your main machine: audiogridder.comLooks cool. Is audiogridder stable? Do the setting on the slave machine get stored in your session on the main DAW? Or is there another step to save and recall sessions on the slave machine? |
|
|
|
Post by bchurch on Dec 3, 2022 12:43:49 GMT -6
This looks really cool (and I also happen to have that same m1 mini still sitting in a box now...)
Admittedly, I don't speak IT-Nerd so AudioGridder's site lost me quickly.
|
|
|
|
Post by EmRR on Dec 3, 2022 13:02:41 GMT -6
Haven’t seen my new M1’s running over 15% CPU, it’s always the software that’s shitting the bed.
|
|
|
|
Post by the other mark williams on Dec 3, 2022 13:49:07 GMT -6
My M1 Mini has been essentially flawless, and the only music things that have maxed it out have been realtime VIs, so I'm also curious what the OP is using, plugin-wise, that is causing it to max out. I wonder if there are some other settings you have going on, OtisGreying, that are causing slowdowns. Video is a different issue altogether. It's easy to slam my machine when editing video. |
|
|
|
Post by subspace on Dec 3, 2022 18:36:12 GMT -6
To just use the second machine as a DSP farm, AudioGridder opens the GUI of the remote plug-ins on your main machine: audiogridder.comLooks cool. Is audiogridder stable? Do the setting on the slave machine get stored in your session on the main DAW? Or is there another step to save and recall sessions on the slave machine? It’s stable once you get a set-up dialed in with fixed IP, etc, the plug-in settings on the second machine are saved within the AudioGridder plug-in instance on the DAW. |
|
|
|
Post by OtisGreying on Dec 3, 2022 20:31:23 GMT -6
If you are just experimenting , why not print your session with no verbs, import that stereo track into a new session which would start with no plug ins and then experiment with new verbs ? I do this currently, I’m looking for a more convenient workaround. I also use UAD satellites - which are fine but I want to experiment with other reverb plug-ins like liquidsonics without totally handicapping my session that may have a lot of Tokyo dawn plugs or Leapwing plugins - it becomes too much very fast in a dense (and admittledly sometimes unorganized) production. I’m also recording at 96k which doesn’t help my CPU. I basically want a second workstation that can send audio to my main setup in real-time and I can monitor virtual instruments, Reverb’s, FX, on my main monitors, record the audio into my main session, all on another CPU for convenience. The Vienna ensemble thing I’ve looked into and it just seems slow and clunky like drbill said, if it’s slow for him it’s probably not going to work for me - the entire point of doing this is for a workflow gain. I just wonder how delayed the audio will be coming back into my main computer/speakers from the second computer and whether I should try the AES or analog I/O for the least amount of delay hearing the FX in relation to my main sessions playback |
|
|
|
Post by the other mark williams on Dec 3, 2022 21:09:18 GMT -6
If you are just experimenting , why not print your session with no verbs, import that stereo track into a new session which would start with no plug ins and then experiment with new verbs ? I do this currently, I’m looking for a more convenient workaround. I also use UAD satellites - which are fine but I want to experiment with other reverb plug-ins like liquidsonics without totally handicapping my session that may have a lot of Tokyo dawn plugs or Leapwing plugins - it becomes too much very fast in a dense (and admittledly sometimes unorganized) production. I’m also recording at 96k which doesn’t help my CPU. I basically want a second workstation that can send audio to my main setup in real-time and I can monitor virtual instruments Reverb’s FX, on my main monitors, record the audio into my main session, all on another CPU for convenience. The Vienna ensemble thing I’ve looked into and it just seems slow and clunky like drbill said, if it’s slow for him it’s probably not going to work for me - the entire point of doing this is for a workflow gain. I just wonder how delayed the audio will be coming back into my main computer/speakers from the second computer and whether I should try the AES or analog I/O for the least amount of delay hearing the FX in relation to my main sessions playback Two thoughts: First, if you're using Tokyo Dawn (and others like them that have high oversampling modes), I wouldn't leave them on "Insane" mode until you're at the final mixing stage. You want to keep everything as lean as possible while you're still recording and arranging things. You may already be doing that, but I wanted to mention it in case you're not. If you're not freezing things as you go, I'd also recommend that practice - at least given what you've just described above. I use Liquidsonics reverbs a lot, but I don't try to monitor through them while recording. If I really need to hear a big reverb in my cue feed while tracking, I'll use an external unit (I've got a Lexicon MX200 that I got for like $100 or something) or a reverb from my software mixer layer in my Metric Halo ULN-8 ("HaloVerb"). Secondly, if you really, really want to try this "second computer as an FX host" idea, I would absolutely go digital from your main machine to the "FX host" machine. If you're using something like Metric Halo or I believe MOTU, you can literally hook up another interface with an Ethernet cable and have very minimal latency between the two boxes/machines. If you're using more traditional audio interfaces, I would use a multichannel digital connection to get from your host machine to your FX machine. If it's just reverbs, the added predelay you'd get from going D --> A --> A --> D --> A --> D might not be a big deal -- it's just predelay. But other FX you might like to try could matter. Certainly any delay FX would matter. You still haven't mentioned what your DAW is or what audio interfaces you're using. That could definitely matter as to how feasible this plan is. I certainly don't want to try and dissuade you from attempting this, just trying to help you think through the possible snags (or even showstoppers) before they happen. |
|
|
|
Post by OtisGreying on Dec 4, 2022 0:29:37 GMT -6
I do this currently, I’m looking for a more convenient workaround. I also use UAD satellites - which are fine but I want to experiment with other reverb plug-ins like liquidsonics without totally handicapping my session that may have a lot of Tokyo dawn plugs or Leapwing plugins - it becomes too much very fast in a dense (and admittledly sometimes unorganized) production. I’m also recording at 96k which doesn’t help my CPU. I basically want a second workstation that can send audio to my main setup in real-time and I can monitor virtual instruments Reverb’s FX, on my main monitors, record the audio into my main session, all on another CPU for convenience. The Vienna ensemble thing I’ve looked into and it just seems slow and clunky like drbill said, if it’s slow for him it’s probably not going to work for me - the entire point of doing this is for a workflow gain. I just wonder how delayed the audio will be coming back into my main computer/speakers from the second computer and whether I should try the AES or analog I/O for the least amount of delay hearing the FX in relation to my main sessions playback Two thoughts: First, if you're using Tokyo Dawn (and others like them that have high oversampling modes), I wouldn't leave them on "Insane" mode until you're at the final mixing stage. You want to keep everything as lean as possible while you're still recording and arranging things. You may already be doing that, but I wanted to mention it in case you're not. If you're not freezing things as you go, I'd also recommend that practice - at least given what you've just described above. I use Liquidsonics reverbs a lot, but I don't try to monitor through them while recording. If I really need to hear a big reverb in my cue feed while tracking, I'll use an external unit (I've got a Lexicon MX200 that I got for like $100 or something) or a reverb from my software mixer layer in my Metric Halo ULN-8 ("HaloVerb"). Secondly, if you really, really want to try this "second computer as an FX host" idea, I would absolutely go digital from your main machine to the "FX host" machine. If you're using something like Metric Halo or I believe MOTU, you can literally hook up another interface with an Ethernet cable and have very minimal latency between the two boxes/machines. If you're using more traditional audio interfaces, I would use a multichannel digital connection to get from your host machine to your FX machine. If it's just reverbs, the added predelay you'd get from going D --> A --> A --> D --> A --> D might not be a big deal -- it's just predelay. But other FX you might like to try could matter. Certainly any delay FX would matter. You still haven't mentioned what your DAW is or what audio interfaces you're using. That could definitely matter as to how feasible this plan is. I certainly don't want to try and dissuade you from attempting this, just trying to help you think through the possible snags (or even showstoppers) before they happen. I've got a Lynx for my main interface and a Audient iD44 that I would use as my second mark. My DAW is ableton. Would a "multichannel digital connection" mean AES? I wonder if I'd be able to send signal from my main session computer 1 to computer 2 (fully wet FX) and receiving the wet signal back into computer 1 with little delay this way? I definitely wouldn't mind not using up channels on my converter if its possible
Also to clarify mark I didn't mean monitoring FX while tracking like Apollo Unison i just meant using the FX computer 2 as a send/return system that would be playing back the FX in time with the main session on my main computer. Basically sending dry signal from computer 1 to computer 2 then sending wet signal back into computer 1. Is all of that possible through AES?
|
|
