|
Post by Johnkenn on Apr 6, 2021 10:38:55 GMT -6
I know popmann had mentioned this...so do I need another interface to do this? Like a little 2x4 or whatever and then record audio in? Is there a way to offload just the vst to a different cpu? Say if I wanted to use my laptop to run Superior but still have the midi being placed...I guess that kind’ve defeats the purpose... I see keyboard guys do this in sessions all the time (with an interface)...just wondered if you could just offload the processing.
|
|
|
Post by avgatzeblouz on Apr 6, 2021 12:31:37 GMT -6
I'm using Audiogridder to run some of my Acustica plugins on a slave computer. Very nice.
|
|
|
Post by popmann on Apr 6, 2021 13:47:03 GMT -6
Yes, you need a second interface. While there ARE solutions like Vienna Ensemble that allow it to all go over ethernet...that makes everything more latent...forces a synchronous clock...and generally makes things more complicated than they need to be.
I think the history one would need is that it's how it's ALWAYS been. The idea of a unified "MIDI sequencer with built in instruments AND DAW" is something that is reactively new...and it always compromise in terms of performance. It's a VERY appealing UI for the musician. Including myself. But, I have not been willing to sacrifice the performance in the past. I AM using a single machine now EXCEPT where I use the Logic Drummer...
I think I would count as "the keyboard guys"...most of the time, I'm using the secondary machine to run say a piano instrument...and I'm record it as AUDIO in the primary DAW. I HAVE also done sequencing with the instruments elsewhere...but, at some point--that's a LOT of midi knowledge that I don't think most people HAVE.
Are you still on that old 2012 Mini? At 48khz? I mean, what wall are you hitting? Are you wanting to leave things MIDI? I have a hard time imagining that box being an issue...unless you're just being non committal across the board.
You use the example of Superior...but, that doesn't use a lot of CPU...so, are you running out of RAM?
|
|
|
Post by Johnkenn on Apr 6, 2021 14:49:50 GMT -6
Yes, you need a second interface. While there ARE solutions like Vienna Ensemble that allow it to all go over ethernet...that makes everything more latent...forces a synchronous clock...and generally makes things more complicated than they need to be. I think the history one would need is that it's how it's ALWAYS been. The idea of a unified "MIDI sequencer with built in instruments AND DAW" is something that is reactively new...and it always compromise in terms of performance. It's a VERY appealing UI for the musician. Including myself. But, I have not been willing to sacrifice the performance in the past. I AM using a single machine now EXCEPT where I use the Logic Drummer... I think I would count as "the keyboard guys"...most of the time, I'm using the secondary machine to run say a piano instrument...and I'm record it as AUDIO in the primary DAW. I HAVE also done sequencing with the instruments elsewhere...but, at some point--that's a LOT of midi knowledge that I don't think most people HAVE. Are you still on that old 2012 Mini? At 48khz? I mean, what wall are you hitting? Are you wanting to leave things MIDI? I have a hard time imagining that box being an issue...unless you're just being non committal across the board. You use the example of Superior...but, that doesn't use a lot of CPU...so, are you running out of RAM? Yeah - the 2012. I’ve only got 16gb ram. Pro Tools and Cubase are pretty good - but if I’m adding B3 or Keyscape to something at the end of a session with SD3, and say a separate ez drummer instance for swells or shaker...or I already have piano or a wurly, et al. I’m usually needing to freeze. It’s not a deal breaker - just kind of a pita. But I’ve been doing more production with Luna - especially now that there are templates. And Alina really doesn’t play well with others.
|
|
|
Post by popmann on Apr 6, 2021 15:53:48 GMT -6
I don't know Alina, FWIW. Is that a VI?
I only have 16gb in my "2018"...the issues are Keyscape and SD3, neither of which has a way to optimize RAM. Kontakt, VSL, Ivory, BFD--I can use nearly none because the storage is so fast. SD3 just has that "SSD" that I think takes it down from 9gb for a kit to 7gb or something...meanwhile, my 45gb Steinway buffers using like 512MB...it's not much of an issue for me, because I print audio. But, I CAN an do regularly have it all going live in an arranging session...but, Keyscape is the one for me that pushes over the edge, so I just rarely use it. It's not better. I kept updating for a while thinking they were going to fix the terrible load times...and crazy RAM usage (compared to comparable in other engines)...I saw the other day there's an update I haven't done, but I literally thought "who cares?" I was going to bring that up in the recent plug in regrets. They never made it performance better...I have better playing everything that ALSO load faster. So, over the years it's jsut been sitting. I keep TRYING it when I need a Wurli or something...but...eh...it just never wins.
I use my Kronos for the Wurli...and even though I've now bought 900 to replace it, the old Scarbee plays best by a long shot. I slap an HD Pultec or something on after it and play/print. You know what the key to that thing is? Turn the noise up to full. I would never have done that ten years ago--who wanst all the noises? But, removal of those noises is what makes it sound less realistic.
Anyway--you have mastering plug ins on your master bus by this point? The reason I ask is because there's no CPU on the planet that will let you do it. The master bus basically has to be processed ON the single core input buffer once you input arm a VI...so, next time you're in that situation, in Cubase if you "alt click" to disable it all together...see if that doesn't solve the issue.
I would GUESS that a secondary machine would annoy you if you're wanting to leave things MIDI. I mean you can do it...store something like Cantabile with a stack per song...but, then you have to manage the VI audio with your hardware (Apollo) mixer...rather than just grabbing the pan in your DAW and moving it...have to know MIDI channel routing and such...but, if in that example you were calling up a Wurli on the laptop...you could really easily jsut play it in as audio-and leave those mastering plugs on the master (assuming you're using Cubase where the compensation engine is accurate)...but, once you put another layer of MIDI into it, it might get to be way more of a "PIA" than freezing.
|
|
|
Post by popmann on Apr 6, 2021 19:25:33 GMT -6
I'm using Audiogridder to run some of my Acustica plugins on a slave computer. Very nice. How does that work with latency? I mean, Acosutica is obviously irrelecant--it's extra DSP power on the other computer...but, does that work for Kenndy's real time VI situation? I feel like Vienna Ensemble doubles your buffer latency. Which for big orchestral projects, fine...but, for SD3 and pianos that won't work (well).
|
|
|
Post by Johnkenn on Apr 6, 2021 19:51:41 GMT -6
Sorry. I was having a margarita and edited your reply instead of hitting reply lol. Here’s what I was trying to say.
Does it say that? It corrected to Alina but I changed it to Luna before I posted. Freaky.
|
|
|
Post by Johnkenn on Apr 6, 2021 19:52:47 GMT -6
I don't know Alina, FWIW. Is that a VI? I only have 16gb in my "2018"...the issues are Keyscape and SD3, neither of which has a way to optimize RAM. Kontakt, VSL, Ivory, BFD--I can use nearly none because the storage is so fast. SD3 just has that "SSD" that I think takes it down from 9gb for a kit to 7gb or something...meanwhile, my 45gb Steinway buffers using like 512MB...it's not much of an issue for me, because I print audio. But, I CAN an do regularly have it all going live in an arranging session...but, Keyscape is the one for me that pushes over the edge, so I just rarely use it. It's not better. I kept updating for a while thinking they were going to fix the terrible load times...and crazy RAM usage (compared to comparable in other engines)...I saw the other day there's an update I haven't done, but I literally thought "who cares?" I was going to bring that up in the recent plug in regrets. They never made it performance better...I have better playing everything that ALSO load faster. So, over the years it's jsut been sitting. I keep TRYING it when I need a Wurli or something...but...eh...it just never wins. I use my Kronos for the Wurli...and even though I've now bought 900 to replace it, the old Scarbee plays best by a long shot. I slap an HD Pultec or something on after it and play/print. You know what the key to that thing is? Turn the noise up to full. I would never have done that ten years ago--who wanst all the noises? But, removal of those noises is what makes it sound less realistic. Anyway--you have mastering plug ins on your master bus by this point? The reason I ask is because there's no CPU on the planet that will let you do it. The master bus basically has to be processed ON the single core input buffer once you input arm a VI...so, next time you're in that situation, in Cubase if you "alt click" to disable it all together...see if that doesn't solve the issue. I would GUESS that a secondary machine would annoy you if you're wanting to leave things MIDI. I mean you can do it...store something like Cantabile with a stack per song...but, then you have to manage the VI audio with your hardware (Apollo) mixer...rather than just grabbing the pan in your DAW and moving it...have to know MIDI channel routing and such...but, if in that example you were calling up a Wurli on the laptop...you could really easily jsut play it in as audio-and leave those mastering plugs on the master (assuming you're using Cubase where the compensation engine is accurate)...but, once you put another layer of MIDI into it, it might get to be way more of a "PIA" than freezing. Oh - and yeah. I disable the mastering stuff.
|
|
|
Post by popmann on Apr 6, 2021 20:06:33 GMT -6
Gotcha...so, it would allow you to host the VIs on a second box and just record them in LUNA? That makes sense. That's really what I did with Mixbus32c...it's VI hosting is challenged...flat buffer and REALLY high CPU use for coming on...so I ran the VIs on the old PC while I ran Mixbus on the Macbook. But, still--that's recording as audio. When I went to do strings, I actually moved the files(tracks) into Cubase on the old PC and did it inside Cubase. There's just so much I end up interacting with the UI for strings. Pianos are more like "dial in sound"--play the part until I get it...strings are constantly sort of adding articulations and modifying keymaps...drawing CCs...
|
|
|
Post by Johnkenn on Apr 6, 2021 23:03:26 GMT -6
Gotcha...so, it would allow you to host the VIs on a second box and just record them in LUNA? That makes sense. That's really what I did with Mixbus32c...it's VI hosting is challenged...flat buffer and REALLY high CPU use for coming on...so I ran the VIs on the old PC while I ran Mixbus on the Macbook. But, still--that's recording as audio. When I went to do strings, I actually moved the files(tracks) into Cubase on the old PC and did it inside Cubase. There's just so much I end up interacting with the UI for strings. Pianos are more like "dial in sound"--play the part until I get it...strings are constantly sort of adding articulations and modifying keymaps...drawing CCs... Yeah - you get what I’m saying. And I remember you saying you felt like Mixbus sounded better - that’s the way I feel about Luna. Maybe “different is the right word with the Neve Summing on the Aux’s and the ability to easily add tape on every channel. Now - you can obviously do that in any other daw...in fact, mixing in Luna has made my process in PTs change. Luna just makes things more intuitive...but It’s also made me rethink some approaches in PTs and I can get similar mixes there now. Honestly - Cubase is the one that kindve does everything right. The comping is better than PTs, hardware compensation, etc. don’t know why I don’t just give in.
|
|
kcatthedog
Temp
Super Helpful Dude
Posts: 15,226
Member is Online
|
Post by kcatthedog on Apr 7, 2021 2:17:48 GMT -6
Ah,why don’t you just change midi to audio, seems a lot simpler?
You could keep the original session with midi for editing or what ever, save it as a new session, convert midi to audio and your ram snd cpu usage should fall a lot ?
|
|
|
Post by mrholmes on Apr 7, 2021 5:45:22 GMT -6
I know popmann had mentioned this...so do I need another interface to do this? Like a little 2x4 or whatever and then record audio in? Is there a way to offload just the vst to a different cpu? Say if I wanted to use my laptop to run Superior but still have the midi being placed...I guess that kind’ve defeats the purpose... I see keyboard guys do this in sessions all the time (with an interface)...just wondered if you could just offload the processing. You use Vienna Server you can tell Vienna to save it session with your daw session and you don’t need a second interface Vienna sends audio midi data through your network connection.
|
|
|
Post by avgatzeblouz on Apr 7, 2021 8:03:56 GMT -6
I'm using Audiogridder to run some of my Acustica plugins on a slave computer. Very nice. How does that work with latency? I mean, Acosutica is obviously irrelecant--it's extra DSP power on the other computer...but, does that work for Kenndy's real time VI situation? I feel like Vienna Ensemble doubles your buffer latency. Which for big orchestral projects, fine...but, for SD3 and pianos that won't work (well). I had not understood the specific need for VIs.
|
|
|
Post by timcampbell on Apr 7, 2021 9:26:05 GMT -6
Well I never really thought about slaving 2 computers together but if you did it opens up all kinds of possibilities. Besides unlimited processing power you could run 2 different, incompatible platforms at the same time without needing to transfer files.
|
|
|
Post by popmann on Apr 7, 2021 12:37:37 GMT -6
There was a time when there were three here...a G3 running Digital Perfomer (MIDI sequencer)...a hardware Akai DAW (handling audio)...and a Windows box running Gigastudio and various other VIs.
Prior to that, it was a reel to reel analog machine SMPTE sync'd to a MIDI sequencer with external ROMpler boards.
I remember getting Logic 5 on Windows 2000 or something with their software instruments and going "THIS is gonna be cool when they get the sampler close to Giga"...I was like THIS is what I want--the consolidated everything.
...and then every time in the past I've tried to do it, it's SUCH a PIA vs just having a second box running an asynchronous clock.
I mean I do it on my recent (2018) build...but, I have to watch my step so to speak, in a way I never had to with the elderly PC tower+Macbook(the last dual machine set up I had here)...
|
|
|
Post by OtisGreying on Nov 20, 2021 9:43:58 GMT -6
So is there any way to incorporate a second computer without dealing with latency changes, clocking issues, etc.?
Should I just set my second computer up with a second interface and second pair of speakers and just use it as a second seperate setup, hit play on my main computer in the same room with the main speakers and record MIDI on my second smaller set-up while listening to the session on my main set-up? And bounce the audio down of the VI's and MIDI and transfer to the main setup?
|
|
|
Post by popmann on Nov 20, 2021 11:26:33 GMT -6
The thing is….usign a second computer with a secondary interface and monitoring at the same place in time you do for other analog audio inputs IS the set and forget low latency way to work. I did all that HD work on the macbook with a buffer of 2048. Hardware level VI latency. How? It ran on the PC tower at a few ms….monitoring on the Macbook interface hardware mixer. You could save a couple more on an analog mixer, and just make sure your compensation engine is set for that to be the reference.
Running it inside your DAW (if its not cubase) means futzing with buffers up and down to balance VI latency and audio DSP.
Using VEPRo adds at least an extra buffer of latency. For the entire local AND remote machine.
|
|
|
Post by OtisGreying on Nov 20, 2021 16:58:12 GMT -6
The thing is….usign a second computer with a secondary interface and monitoring at the same place in time you do for other analog audio inputs IS the set and forget low latency way to work. I did all that HD work on the macbook with a buffer of 2048. Hardware level VI latency. How? It ran on the PC tower at a few ms….monitoring on the Macbook interface hardware mixer. You could save a couple more on an analog mixer, and just make sure your compensation engine is set for that to be the reference. Running it inside your DAW (if its not cubase) means futzing with buffers up and down to balance VI latency and audio DSP. Using VEPRo adds at least an extra buffer of latency. For the entire local AND remote machine. Sorry popmann but all I understood was perhaps what I suggested IS in fact the only set and forget low latency way to work? Right? Everything else went over my head, do you suggest I work this way? - with two separate setups? Is it easier than going the Vienna ensemble route? I’m just trying to get more VIs going quickly with less latency so I can be more creatively trying different sounds during productions.
|
|
|
Post by nobtwiddler on Nov 21, 2021 13:31:56 GMT -6
I run 3 x mac's when I mix.
1 x Mac mini 2020 runs Luna, and or DP, all the multitrack files. 1 x Mac mini 2016 Runs, the Neve Recall software for all my AMS NEVE STUFF. 1 x MacBook Pro 2020 acts as my Mixdown machine. Which is either setup to record a stereo mixes at 192KHZ, and or 16 x stems, and a stereo Mix of those stems, also @ 192Khz.
At this point I forgotten how I locked the MacBook, and the main Mac Mini together, but I did. (I think it was thru Midi?) Somehow it all works..
|
|
|
Post by popmann on Nov 21, 2021 14:54:15 GMT -6
The thing is….usign a second computer with a secondary interface and monitoring at the same place in time you do for other analog audio inputs IS the set and forget low latency way to work. I did all that HD work on the macbook with a buffer of 2048. Hardware level VI latency. How? It ran on the PC tower at a few ms….monitoring on the Macbook interface hardware mixer. You could save a couple more on an analog mixer, and just make sure your compensation engine is set for that to be the reference. Running it inside your DAW (if its not cubase) means futzing with buffers up and down to balance VI latency and audio DSP. Using VEPRo adds at least an extra buffer of latency. For the entire local AND remote machine. Sorry popmann but all I understood was perhaps what I suggested IS in fact the only set and forget low latency way to work? Right? Everything else went over my head, do you suggest I work this way? - with two separate setups? Is it easier than going the Vienna ensemble route? I’m just trying to get more VIs going quickly with less latency so I can be more creatively trying different sounds during productions. Don't fully understand the disconnect. Vienna Ensemble increases latency. It IS "more like what you're used to"---because all your VIs remotely load--initiated from the VEP plug in inside your main DAW project. The audio comes back into audio channels in your DAW. You don't have to interact with a second mixer UI. It is CLOSER to the philosophical idea of "offloading them to another machine". If you use a secondary interface on a second computer it's lower latency. It's a completely different workflow than having the VI hosted in your DAW. You need to use a hardware mixer (like nearly all interfaces have) for monitoring. You set the main DAW to 2048 (or whatever makes sense to be big enough for mixdown)...and you set the VI machine at 128 or whatever makes for a snappy response. You shouldn't have to adjust them really ever. Certainly not on a regular basis. But, also--you won't get your VI presets loaded when you open a project. So, "easier"....isn't a single thing. To ME....VEP has some very specific use cases, but seems like way more trouble to set up than it's worth--but, I sequenced external MIDI devices for 10 years before software. It is ABOLSUTELY a learning curve that might be like learning a new language for someone.
|
|
|
Post by Vincent R. on Nov 25, 2021 15:40:55 GMT -6
I use Vienna and have no issues. I also work in a studio with 4 computers all through Vienna. No issues at all. Full orchestral mixes for my music, film, and TV. humblecomposer…
|
|
|
Post by Quint on Nov 25, 2021 16:03:34 GMT -6
Is the need to offload processing at all related to a need for a low buffer size for tracking VIs?
If so, have you considered using Blue Cat's Patchwork to host your VIs?
|
|
|
Post by OtisGreying on Dec 9, 2021 20:08:59 GMT -6
A second computer with midi keyboard running a session that has a few VST/soft synths, lowest latency as its only running a select few of these plugs, that going into DA converter, routed into my main computer/set ups AD converter as a stereo signal, would work I think?
Then tracking the VSTs/soft synths as if they were analog synths effectively with as little latency as possible?
|
|
|
Post by OtisGreying on Dec 10, 2021 1:18:28 GMT -6
Tried routing a VST on my second computer into the line output of my second interface and plugging that line output into the INPUT of my Lynx converter in my main setup - no sound. Not sure why it wouldn’t work but it isn’t currently - are you not able to route sound out of a daw as an analog in?
Edit: got it working. Are any of you guys doing things this way? It doesn’t retain midi information but it seems to accomplish having all VSTs/VI’s at lowest possible buffer settings - I guess the point of Vienna ensemble pro is that it retains your midi info on the master computer?
|
|
|
Post by popmann on Dec 10, 2021 8:58:44 GMT -6
Thats how i geenrally do it.
If/when you need MIDI on the primary DAW, you plug the keyboard into the primary’s midi and the primary midi out TO the secondary. Audio connections remain the same. Youuse a midi track in the primary. You can sequence midi just like everyone did for the first 20 year of midi.
|
|