|
|
Post by gravesnumber9 on Feb 18, 2023 12:12:35 GMT -6
Also, when I set up singers to overdub with zero latency if they have never overdubbed that way before? I routinely get “wow this microphone is amazing” even though it’s really just the richness that comes from the absence of comb filtering
To me that makes it worth it right there.
(Also. My Signal Arts u48 IS amazing, thanks Chad!)
|
|
|
|
Post by Tbone81 on Feb 18, 2023 12:21:09 GMT -6
Anyone had latency issues from a Hearback Cue system? I had wierd problems I could figure out at a session a few months back. Took sends off the console straight to the hearback and was getting about a 1/16 note of latency no matter what I did. It was terrible.
|
|
|
|
Post by gravesnumber9 on Feb 18, 2023 12:36:13 GMT -6
Anyone had latency issues from a Hearback Cue system? I had wierd problems I could figure out at a session a few months back. Took sends off the console straight to the hearback and was getting about a 1/16 note of latency no matter what I did. It was terrible. Sounds like a loop back somewhere right? But obviously not because I’m sure that’s the first thing you looked for. |
|
|
|
Post by seawell on Feb 18, 2023 12:38:10 GMT -6
Carbon is the only standalone interface that will get you there that I know of. It basically has a HDX card built in. Apollo and Luna are now basically the same as Carbon and Protools with the hybrid engine. So I wouldn't say that Carbon is the only one. Though, to be fair, there are a lot of interfaces that can do monitoring latency, with their own built in digital mixer, at the same level of low level latency as Carbon or Apollo. It's just that Carbon (with Protools), and now the Apollo (with Luna) offer an environment where the built in digital mixer is tightly integrated with the DAW, to the point that they seamlessly work in concert with one another. The auto DSP shuffling, between native and DSP, that Carbon/PT and Apollo/Luna offers is very nice. Carbon is the only one with sub 1ms which is what I was specifically referring to. It's really splitting hairs at this point. It's good to hear Apollo works so seamlessly with Luna because I never quite got used to UA Console with Pro Tools. Lots of great options around and I hope no one thinks I'm trying to sell them on Pro Tools HDX because Lord knows Pro Tools has its headaches as well. It's just that if you do want the lowest latency available and seamless pro tools delay compensation with hardware inserts during the mix then it's still the best set it and forget it option. Some great deals can be found on used systems. |
|
|
|
Post by svart on Feb 18, 2023 13:17:32 GMT -6
Anyone had latency issues from a Hearback Cue system? I had wierd problems I could figure out at a session a few months back. Took sends off the console straight to the hearback and was getting about a 1/16 note of latency no matter what I did. It was terrible. Having recently moved to hearback, I was afraid this might happen but it didn't. Sounds perfect fine. However, I'm using the analog inputs. I had intended to use the adat inputs but my sampling rate was too high. I have a feeling that adat subsystems might have latency depending on his they're encoded. BTW, the hearback system uses adat over Ethernet as it's protocol. |
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Feb 18, 2023 13:22:06 GMT -6
Is there any other solution to get sub 1 ms latency without PT HDX? Lynx and rme cards are even faster if you add plugs to hdx |
|
|
|
Post by seawell on Feb 18, 2023 14:51:31 GMT -6
Is there any other solution to get sub 1 ms latency without PT HDX? Lynx and rme cards are even faster if you add plugs to hdx I'm not sure what you mean here Dan, can you further explain? Thanks. |
|
|
|
Post by Quint on Feb 19, 2023 8:02:30 GMT -6
Apollo and Luna are now basically the same as Carbon and Protools with the hybrid engine. So I wouldn't say that Carbon is the only one. Though, to be fair, there are a lot of interfaces that can do monitoring latency, with their own built in digital mixer, at the same level of low level latency as Carbon or Apollo. It's just that Carbon (with Protools), and now the Apollo (with Luna) offer an environment where the built in digital mixer is tightly integrated with the DAW, to the point that they seamlessly work in concert with one another. The auto DSP shuffling, between native and DSP, that Carbon/PT and Apollo/Luna offers is very nice. Carbon is the only one with sub 1ms which is what I was specifically referring to. It's really splitting hairs at this point. It's good to hear Apollo works so seamlessly with Luna because I never quite got used to UA Console with Pro Tools. Lots of great options around and I hope no one thinks I'm trying to sell them on Pro Tools HDX because Lord knows Pro Tools has its headaches as well. It's just that if you do want the lowest latency available and seamless pro tools delay compensation with hardware inserts during the mix then it's still the best set it and forget it option. Some great deals can be found on used systems. Oh, got it. Yeah, I didn't realize that you were specifically speaking to options that were sub 1ms. But like you said, is really splitting hairs. Apollo is sub 2ms (1.1ms, to be exact), so I'd say for the purposes of this discussion, it's plenty fast enough, but it's not technically under 1ms. Of note, these RTLs of under 1s and under 2ms, for Carbon and Apollo, respectively, are at 96k. That said, it's worth pointing out that some DSP plugins, can increase that latency on Apollo, and I know that can happen with HDX too, so I assume it's the same with Carbon as well. Console was also something I didn't love having to simultaneously juggle with the DAW too. To be fair, I felt that way with any interface digital monitoring mixer, so it's not like this was a specific issue I had with the Apollo. This is why I got excited when Luna was announced. |
|
|
|
Post by bgrotto on Feb 19, 2023 8:57:50 GMT -6
I've worked with a number of drummers and percussionist who can detect latency on HDX rigs. Have to use the console sends in those cases.
I've also worked with a couple singers who responded better to analog, true-zero latency; pitch improved when using the console sends.
Hell, I myself (drummer) had a soundcheck a few years ago where I noticed a change in monitoring latency when the FOH system was set to 48khz after sound checking in 44.1. I don't know what kind of digital processing was going on that may have added to the latency of their signal path, but I heard the difference right away, much to the FOH engineer's surprise. He actually didn't believe me at first, and we did couple rounds of him switching between rates (or pretending to do so) to see if I could reliably spot the difference. I could, much to his surprise!
In the studio, I'm sensitive to that sort of thing, but I don't know to what extent. I will say that when drumming or percussing, 128 samples at 44.1 starts to become noticeably detrimental. With other instruments, I can usually manage a bit more.
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Feb 19, 2023 11:22:12 GMT -6
Lynx and rme cards are even faster if you add plugs to hdx I'm not sure what you mean here Dan, can you further explain? Thanks. HDX plugins all add latency. HDX no plugin round trip latency through an avid converter is a bit under 2 ms at 44.1 or 48 khz. Avid’s own 0.7ms rtl is at 96khz. Misleading advertising just like most interface manufacturers. HDX plugins latency for brand: Avid usually 9-10 samples. Metric Halo usually 9-10. McDSP and Brainworx usually 33-34. Sonnox Oxford EQ and Oxford Reverb are 7 and Oxford Dynamics 27 and the dynamics has a lookahead! Now consider how junky many of these are and those still good Sonnox plugs are zero latency and 20 samples respectively Native. McDSP is mostly zero latency native. 1000 ms / 44100 samples per second * sample latency = added latency in ms per plugin HDX latency can add up. RME and Lynx PCI-E and thunderbolt interfaces hit under well 3ms RTL at 44.1khz. RME usb hits around 4ms. UAD hits around 4 too but many of the dsp plugins oversample to 192khz and add latency. This is one of the reasons why their non-linear processors are not total junk. Dante with an Audinate card is usually a couple of ms. Dante with Dante Virtual Sound Card is usually 10ms but some computers can do 4ms depending on their Ethernet. Modern Intel, AMD, and Apple Silicon ARM chips on purpose built and configured pcs can run all of the junky old plugins and zero latency modern stuff, eg PSP has some good stuff, at the lowest buffer settings easily. You’re not getting any added latency to the interface’s round trip latency with zero latency native plugins. There are no arrays of dsp chips and I/O protocol chipsets to send data between with at minimum 1 sample latency. You can just keep adding more until the modern cpu craps out and current CPUs are great. Intel 13900k and AMD 7950x can run newer native amp sims and synths at the lowest buffer on higher session sampling rates (no oversampling options that add latency) for under 1ms latency. You’ll need a full atx case and 420mm aio cooler but no dsp add on system can compete with those. the one area dsp chips still out perform native interfaces is with off loading hundreds of simultaneous channels, way more than the 4 to 32 usually seen in rack mount interfaces. |
|
|
|
Post by seawell on Feb 19, 2023 13:00:09 GMT -6
I'm not sure what you mean here Dan, can you further explain? Thanks. HDX plugins all add latency. HDX no plugin round trip latency through an avid converter is a bit under 2 ms at 44.1 or 48 khz. Avid’s own 0.7ms rtl is at 96khz. Misleading advertising just like most interface manufacturers. HDX plugins latency for brand: Avid usually 9-10 samples. Metric Halo usually 9-10. McDSP and Brainworx usually 33-34. Sonnox Oxford EQ and Oxford Reverb are 7 and Oxford Dynamics 27 and the dynamics has a lookahead! Now consider how junky many of these are and those still good Sonnox plugs are zero latency and 20 samples respectively Native. McDSP is mostly zero latency native. 1000 ms / 44100 samples per second * sample latency = added latency in ms per plugin HDX latency can add up. RME and Lynx PCI-E and thunderbolt interfaces hit under well 3ms RTL at 44.1khz. RME usb hits around 4ms. UAD hits around 4 too but many of the dsp plugins oversample to 192khz and add latency. This is one of the reasons why their non-linear processors are not total junk. Dante with an Audinate card is usually a couple of ms. Dante with Dante Virtual Sound Card is usually 10ms but some computers can do 4ms depending on their Ethernet. Modern Intel, AMD, and Apple Silicon ARM chips on purpose built and configured pcs can run all of the junky old plugins and zero latency modern stuff, eg PSP has some good stuff, at the lowest buffer settings easily. You’re not getting any added latency to the interface’s round trip latency with zero latency native plugins. There are no arrays of dsp chips and I/O protocol chipsets to send data between with at minimum 1 sample latency. You can just keep adding more until the modern cpu craps out and current CPUs are great. Intel 13900k and AMD 7950x can run newer native amp sims and synths at the lowest buffer on higher session sampling rates (no oversampling options that add latency) for under 1ms latency. You’ll need a full atx case and 420mm aio cooler but no dsp add on system can compete with those. the one area dsp chips still out perform native interfaces is with off loading hundreds of simultaneous channels, way more than the 4 to 32 usually seen in rack mount interfaces. Ok I get what you were saying now, thanks. I have Avid and Lynx interfaces so I thought you meant Lynx was somehow faster if using HDX plugins. I track at 96kHz so I do appreciate having the 0.7ms starting point and then of course you have to consider any plugins you add from there. |
|
|
|
Post by notneeson on Feb 19, 2023 13:53:44 GMT -6
Man, I’ve punched in drums, sans click, to tape while monitoring through the Pro Tools HD10 mixer and I’m still surprised it worked. But that is more down to that particular drummer’s consistency than anything else. And with feel!
|
|
|
|
Post by craigmorris74 on Feb 19, 2023 16:12:21 GMT -6
I might be missing something here, with my converters running into an RME PCI card, there is no noticeable latency. I’ve had hundreds of players here and it’s never been and issue. Of course, the signal going into the DAW is a bit behind after conversion, but the DAW is calibrated so that the recorded tracks are perfectly in time.
Are the latency issues folks are having because of not using some form of direct monitoring? I know USB interfaces typically introduce latency.
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Feb 19, 2023 16:58:11 GMT -6
I might be missing something here, with my converters running into an RME PCI card, there is no noticeable latency. I’ve had hundreds of players here and it’s never been and issue. Of course, the signal going into the DAW is a bit behind after conversion, but the DAW is calibrated so that the recorded tracks are perfectly in time. Are the latency issues folks are having because of not using some form of direct monitoring? I know USB interfaces typically introduce latency. PCI-E cards are the fastest solution and the on interface or card dsp mixes will shave off even more latency but people want to track with plugins and not the onboard dsp mixers where the main latency from the anti-alias filters in the converters. There are also all the Macs reliant on thunderbolt.
If you want to track with plugs and monitor through the daw:
Modern converter chips are pretty quick so it's mostly down to driver quality.
The interface manufacturers can also select minimum phase filters on the da ouputs that are ineffective at aliasing prevention but have lower latency or a shorter linear phase filter followed by a minimum phase filter that is a little better at aliasing prevention but still sounds bad. This is one way that RME has kept latency all the way down in addition to very efficient and stable drivers. The solution for accurate monitoring with an interface like this is to get a separate DAC for your monitoring and use the on interface outs for the headphones mixes.
Or a cleaner external interface with quick drivers like Lynx Aurora N or Apogee Symphony over thunderbolt but that might be impossible on many computers.
|
|
|
|
Post by EmRR on Feb 19, 2023 17:12:42 GMT -6
I might be missing something here, with my converters running into an RME PCI card, there is no noticeable latency. I’ve had hundreds of players here and it’s never been and issue. Of course, the signal going into the DAW is a bit behind after conversion, but the DAW is calibrated so that the recorded tracks are perfectly in time. Are the latency issues folks are having because of not using some form of direct monitoring? I know USB interfaces typically introduce latency. Apparently correct. I've never tried recording 'through the DAW' with plugs in monitoring, DP definitely will not do it, not even with no plugs. I keep my buffer at no less than 1024 to stay out of processing overload territory. Something that's a few tracks might work at a much lower setting, but not once plugs are involved. I tried using the MOTU AVB mixer for awhile, and latency wasn't a bother, but driving the damn thing really requires a dedicated assistant to handle that portion, or you really can't actually pay attention to the music and capture side properly. I punted, added another patch bay mult so my inputs hit the analog mixer for headphone sends in parallel with the DAW, vastly easier to manage. I should try reducing buffer size on the M1 Studio and see where the cut-off is now, it'll surely still be related to DP and not the M1. Was just mixing a 10 piece classical guitar orchestra with vocals and sound efx, not a lot of processing other than reverbs, and that was borderline banging on the real time processing part of the CPU meter in DP11 at 1024. Cut it in half, it'd be warning city. |
|
|
|
Post by gravesnumber9 on Feb 19, 2023 21:50:36 GMT -6
I might be missing something here, with my converters running into an RME PCI card, there is no noticeable latency. I’ve had hundreds of players here and it’s never been and issue. Of course, the signal going into the DAW is a bit behind after conversion, but the DAW is calibrated so that the recorded tracks are perfectly in time. Are the latency issues folks are having because of not using some form of direct monitoring? I know USB interfaces typically introduce latency. Apparently correct. I've never tried recording 'through the DAW' with plugs in monitoring, DP definitely will not do it, not even with no plugs. I keep my buffer at no less than 1024 to stay out of processing overload territory. Something that's a few tracks might work at a much lower setting, but not once plugs are involved. I tried using the MOTU AVB mixer for awhile, and latency wasn't a bother, but driving the damn thing really requires a dedicated assistant to handle that portion, or you really can't actually pay attention to the music and capture side properly. I punted, added another patch bay mult so my inputs hit the analog mixer for headphone sends in parallel with the DAW, vastly easier to manage. I should try reducing buffer size on the M1 Studio and see where the cut-off is now, it'll surely still be related to DP and not the M1. Was just mixing a 10 piece classical guitar orchestra with vocals and sound efx, not a lot of processing other than reverbs, and that was borderline banging on the real time processing part of the CPU meter in DP11 at 1024. Cut it in half, it'd be warning city. How is the MOTU mixer made by the same company as Digital Performer? Is this where they put the interns? That thing is so hard to use! |
|
|
|
Post by EmRR on Feb 20, 2023 7:21:29 GMT -6
Apparently correct. I've never tried recording 'through the DAW' with plugs in monitoring, DP definitely will not do it, not even with no plugs. I keep my buffer at no less than 1024 to stay out of processing overload territory. Something that's a few tracks might work at a much lower setting, but not once plugs are involved. I tried using the MOTU AVB mixer for awhile, and latency wasn't a bother, but driving the damn thing really requires a dedicated assistant to handle that portion, or you really can't actually pay attention to the music and capture side properly. I punted, added another patch bay mult so my inputs hit the analog mixer for headphone sends in parallel with the DAW, vastly easier to manage. I should try reducing buffer size on the M1 Studio and see where the cut-off is now, it'll surely still be related to DP and not the M1. Was just mixing a 10 piece classical guitar orchestra with vocals and sound efx, not a lot of processing other than reverbs, and that was borderline banging on the real time processing part of the CPU meter in DP11 at 1024. Cut it in half, it'd be warning city. How is the MOTU mixer made by the same company as Digital Performer? Is this where they put the interns? That thing is so hard to use! Ugh i know. But better than the similar Presonus I’ve had the displeasure to drive on a live broadcast gig with PA! None built for professional action, esp when i think how much easier it would be with a mackie 1604 and a couple graphic EQ! All the smaller companies probably suffer from the same problem: contract coders who come and go, no one after the initial design who really knows enough to truly fix anything. |
|
|
|
Post by gravesnumber9 on Feb 20, 2023 10:00:28 GMT -6
How is the MOTU mixer made by the same company as Digital Performer? Is this where they put the interns? That thing is so hard to use! Ugh i know. But better than the similar Presonus I’ve had the displeasure to drive on a live broadcast gig with PA! None built for professional action, esp when i think how much easier it would be with a mackie 1604 and a couple graphic EQ! All the smaller companies probably suffer from the same problem: contract coders who come and go, no one after the initial design who really knows enough to truly fix anything. I always got the impression that MOTU was a bunch of nerds and MIT type kids. Maybe they can only afford to keep so many full time nerds on the software side. Everything else they do is like "how did you do this for this price?" But then that online mixer... wow it is bad. It makes UA's Console look easy to use and that thing is also a mess. |
|
|
|
Post by EmRR on Feb 20, 2023 11:18:10 GMT -6
Ugh i know. But better than the similar Presonus I’ve had the displeasure to drive on a live broadcast gig with PA! None built for professional action, esp when i think how much easier it would be with a mackie 1604 and a couple graphic EQ! All the smaller companies probably suffer from the same problem: contract coders who come and go, no one after the initial design who really knows enough to truly fix anything. I always got the impression that MOTU was a bunch of nerds and MIT type kids. Maybe they can only afford to keep so many full time nerds on the software side. Everything else they do is like "how did you do this for this price?" But then that online mixer... wow it is bad. It makes UA's Console look easy to use and that thing is also a mess. The meters in it have always been buggy, they just stop displaying at times. Like weeks on end. Then suddenly work again. Lately for me channels go mute until a reboot, or the mute toggle is required to get a channel back, that's new. I don't know who wants an LA-2A emulation as a master bus limiter. Really? You can't use the more full featured channel compressors on a bus? Really? The aux mixer windows are totally confusing. Then there's the alt aux (touch console) window that's for a different version of personal monitoring? But they all show up on desktop too? Which am I supposed to use for what? Here's some options, none of them right. I don't know how the boot time of those interfaces is as slow as it is, feels like 1996. A few digital console manufactures have virtual offline versions in which you can build a show, then install it when you're with the hardware. I don't get the majority that don't. It's all too complicated to throw into any time constrained live situation, at best it's still a huge sidetrack. I don't think we'll ever be happy with any of this until we have large GENERAL PURPOSE touch screens that are truly responsive, no perceived lag. They (anyone designing these new things) might call themselves DISRUPTORS, but I don't want DISRUPTION in my work environment. Rather, they are ignoramuses who haven't bothered to learn the craft before devising additional ways to do things. No one asked for reinvention of the wheel, yet it continues to happen. |
|
|
|
Post by gravesnumber9 on Feb 20, 2023 12:01:42 GMT -6
I always got the impression that MOTU was a bunch of nerds and MIT type kids. Maybe they can only afford to keep so many full time nerds on the software side. Everything else they do is like "how did you do this for this price?" But then that online mixer... wow it is bad. It makes UA's Console look easy to use and that thing is also a mess. They (anyone designing these new things) might call themselves DISRUPTORS, but I don't want DISRUPTION in my work environment. Rather, they are ignoramuses who haven't bothered to learn the craft before devising additional ways to do things. No one asked for reinvention of the wheel, yet it continues to happen. THIS!!! Disrupt things that don't work. Don't disrupt things that are already running smoothly. I suppose that's the modern tech equivalent of "don't fix it if it ain't broke." |
|
