|
|
Post by the other mark williams on Nov 4, 2024 18:12:21 GMT -6
Man, you know, if the MHLink just had lower latency, I would have considered switching to that from my Thunderbolt setup, and then gotten a MADI Edge Card to bring in MADI that way. That would have been a great solution for my needs. But the latency.... Great sale price though. Damnit! At this price its easier to justify an RME MADI PCIe as the interface to/from a MADI Edge card... Then no need for the MHLink headache... Or the RME PCIe AES card… |
|
|
|
Post by plinker on Nov 4, 2024 21:02:24 GMT -6
It looks like Thunderbolt to PCIe expansion chassis are about $300.
|
|
|
|
Post by gravesnumber9 on Nov 4, 2024 22:31:23 GMT -6
So I've been curious about Metric Halo for a long time and posted such on another thread. I think I was actually part of what got the MH conversation going along with Quint. I was talking about looking for a "forever" interface and felt like this could be it. And now with this pricing discount I could afford to get two.
But... I'm getting a little bit confused about the pros and cons.
On the one hand I really don't like using two apps and that's why I got out of the UAD world to begin with, I hated console.
But on the other hand I like simplicity and if the MIO Console Session thing works smoothly, why not? Also on the other hand I do a ton of hybrid mixing and I feel like the determined latency is more important in that case than the actual latency number. But I'm also getting confused on apparent errors in the MH universe on compensating for that?
Here's my use case. Three things...
1) I do a lot of mixing both for hire and also for my own projects where I was either the producer or the artist. When I produce (myself or others) I track basic tracks in my tracking/rehearsal room through an old A&H console going into a MOTU 24Ao and a MOTU 24Ai. Then I bring that back to my mixing room where I often overdub vocals and ancillary parts. Currently I often use my Cranborne 500ADAT to monitor those overdubs live and my MOTU 828ES and Studio One compensate just fine as they should.
2) Mixing for hire I just mix and I'm doing a lot of hybrid. Not much to explain here.
3) I also do a lot of writing in my mixing room where I'm either playing a keyboard with a VI piano or acoustic guitar and singing. On guitar I monitor direct through the Cranborne. When I do piano I use Studio One's low latency VI monitoring.
I don't track through a ton of plugins except when I'm overdubbing late in the process. And even then I often disable most plugins and have usually printed my instruments. Also Studio One's low latency monitoring works great. I'm assuming it would work great with a MH Lio too but I don't know?
The problem I'm trying to solve is, again, that I'm tired of thinking about this. I like that MH allows you to update hardware and has universally admired human support. I'd like to just be done thinking about interfaces forever.
But other than that I don't really have a problem. My MOTU 828ES / Cranborne 500ADAT / MOTU Ao system in the mixing room is working just fine. It does turn out that I don't need as many line outs as I initially thought I was going to because I'm rarely patching in more than 12 - 14 at the same time having given up on my RND Orbit, but other than the fact that I've got more Line Outs than I need right now, there's no real problem.
And yet... I feel like now is my chance to get into MH affordably with the MH Lio being $1100 off. And I've been flirting with MH for a while.
Am I even a good use case for this device? This thread has confused me.
|
|
|
|
Post by the other mark williams on Nov 5, 2024 0:28:48 GMT -6
It looks like Thunderbolt to PCIe expansion chassis are about $300. Not the good ones that are quiet enough to have in the control room with you. Here's a thread where we were talking about this a couple months ago: realgearonline.com/thread/17592/sonnet-echo-3-pcie-thunderboltActually, I'm not 100% sure if tonycamphd has weighed in on his final fan volume report in real-world usage. Tony, how's it going over there? Can't remember if yours is in a machine room or not? The RME PCIe AES card is 2u in height, so you can't use a 1-slot chassis if you go that route. The Lynx card I think might be 1u high, though? |
|
|
|
Post by the other mark williams on Nov 5, 2024 0:58:05 GMT -6
So I've been curious about Metric Halo for a long time and posted such on another thread. I think I was actually part of what got the MH conversation going along with Quint. I was talking about looking for a "forever" interface and felt like this could be it. And now with this pricing discount I could afford to get two. But... I'm getting a little bit confused about the pros and cons. On the one hand I really don't like using two apps and that's why I got out of the UAD world to begin with, I hated console. But on the other hand I like simplicity and if the MIO Console Session thing works smoothly, why not? Also on the other hand I do a ton of hybrid mixing and I feel like the determined latency is more important in that case than the actual latency number. But I'm also getting confused on apparent errors in the MH universe on compensating for that? Here's my use case. Three things... 1) I do a lot of mixing both for hire and also for my own projects where I was either the producer or the artist. When I produce (myself or others) I track basic tracks in my tracking/rehearsal room through an old A&H console going into a MOTU 24Ao and a MOTU 24Ai. Then I bring that back to my mixing room where I often overdub vocals and ancillary parts. Currently I often use my Cranborne 500ADAT to monitor those overdubs live and my MOTU 828ES and Studio One compensate just fine as they should. 2) Mixing for hire I just mix and I'm doing a lot of hybrid. Not much to explain here. 3) I also do a lot of writing in my mixing room where I'm either playing a keyboard with a VI piano or acoustic guitar and singing. On guitar I monitor direct through the Cranborne. When I do piano I use Studio One's low latency VI monitoring. I don't track through a ton of plugins except when I'm overdubbing late in the process. And even then I often disable most plugins and have usually printed my instruments. Also Studio One's low latency monitoring works great. I'm assuming it would work great with a MH Lio too but I don't know? The problem I'm trying to solve is, again, that I'm tired of thinking about this. I like that MH allows you to update hardware and has universally admired human support. I'd like to just be done thinking about interfaces forever. But other than that I don't really have a problem. My MOTU 828ES / Cranborne 500ADAT / MOTU Ao system in the mixing room is working just fine. It does turn out that I don't need as many line outs as I initially thought I was going to because I'm rarely patching in more than 12 - 14 at the same time having given up on my RND Orbit, but other than the fact that I've got more Line Outs than I need right now, there's no real problem. And yet... I feel like now is my chance to get into MH affordably with the MH Lio being $1100 off. And I've been flirting with MH for a while. Am I even a good use case for this device? This thread has confused me. The disagreement about latency compensation above was in regards to the MIO Console mixer itself. If you're using Studio One's I/O plugin to ping your hardware, everything should work just fine. If you're wanting to keep your Cranborne, you could get an Edge card in the ULN/LIO-8 (there are 2 or 3 different ADAT Edge card options). It doesn't sound to me like you'd experience any issues - you're monitoring live inputs through a hardware mixer at your tracking/rehearsal space. And when you do overdubs at home, you would just either need to monitor through the MIO Console for (essentially) zero latency or use some kind of small analog mixer if you prefer. The ULN/LIO-8 has an analog out mult of the inputs on the back (labeled "Balanced Send 1-8") that are pre-A/D, so it's super easy to send those to your analog mixer. With regards to RTL in the DAW, I'm pretty sensitive to latency, and I'm able to get my RTL low enough with the MHLink (that's the ethernet) connection and driver that I can play VI pianos. If I needed to regularly play guitar through amp plugins and monitor in the DAW late in a production when I have tons of tracks already, it would be dicier. I use an external amp modeler for that reason (if I can't mic a real amp). I'm on a Mac Mini M1 right now, and I usually run at 96kHz, but sometimes 48kHz. Happy to help with any other questions... |
|
|
|
Post by Quint on Nov 5, 2024 7:14:14 GMT -6
Man, you know, if the MHLink just had lower latency, I would have considered switching to that from my Thunderbolt setup, and then gotten a MADI Edge Card to bring in MADI that way. That would have been a great solution for my needs. But the latency.... Great sale price though. Damnit! At this price its easier to justify an RME MADI PCIe as the interface to/from a MADI Edge card... Then no need for the MHLink headache... For me though, the biggest reason I'd be looking at a Lio in the first place is because I would want the Lio to be the actual interface. If I have to use another interface (RME) to get the Lio into the computer, it makes the Lio more or less redundant for me, that is unless I was only looking at doing conversion on a Lio. But I'm not. I'm looking at stuff like Burl, and how to get that into the computer. If MHLink was lower/dependable latency, I would have totally picked up a Lio (especially now that they are on sale) and ran that into the computer using MHLink, and then get a MADI Edge card for the Lio to use for getting the Burl stuff into the computer. The extra conversion on the Lio would have just been bonus for me. |
|
|
|
Post by Quint on Nov 5, 2024 7:18:36 GMT -6
So I've been curious about Metric Halo for a long time and posted such on another thread. I think I was actually part of what got the MH conversation going along with Quint. I was talking about looking for a "forever" interface and felt like this could be it. And now with this pricing discount I could afford to get two. But... I'm getting a little bit confused about the pros and cons. On the one hand I really don't like using two apps and that's why I got out of the UAD world to begin with, I hated console. But on the other hand I like simplicity and if the MIO Console Session thing works smoothly, why not? Also on the other hand I do a ton of hybrid mixing and I feel like the determined latency is more important in that case than the actual latency number. But I'm also getting confused on apparent errors in the MH universe on compensating for that? Here's my use case. Three things... 1) I do a lot of mixing both for hire and also for my own projects where I was either the producer or the artist. When I produce (myself or others) I track basic tracks in my tracking/rehearsal room through an old A&H console going into a MOTU 24Ao and a MOTU 24Ai. Then I bring that back to my mixing room where I often overdub vocals and ancillary parts. Currently I often use my Cranborne 500ADAT to monitor those overdubs live and my MOTU 828ES and Studio One compensate just fine as they should. 2) Mixing for hire I just mix and I'm doing a lot of hybrid. Not much to explain here. 3) I also do a lot of writing in my mixing room where I'm either playing a keyboard with a VI piano or acoustic guitar and singing. On guitar I monitor direct through the Cranborne. When I do piano I use Studio One's low latency VI monitoring. I don't track through a ton of plugins except when I'm overdubbing late in the process. And even then I often disable most plugins and have usually printed my instruments. Also Studio One's low latency monitoring works great. I'm assuming it would work great with a MH Lio too but I don't know? The problem I'm trying to solve is, again, that I'm tired of thinking about this. I like that MH allows you to update hardware and has universally admired human support. I'd like to just be done thinking about interfaces forever. But other than that I don't really have a problem. My MOTU 828ES / Cranborne 500ADAT / MOTU Ao system in the mixing room is working just fine. It does turn out that I don't need as many line outs as I initially thought I was going to because I'm rarely patching in more than 12 - 14 at the same time having given up on my RND Orbit, but other than the fact that I've got more Line Outs than I need right now, there's no real problem. And yet... I feel like now is my chance to get into MH affordably with the MH Lio being $1100 off. And I've been flirting with MH for a while. Am I even a good use case for this device? This thread has confused me. The disagreement about latency compensation above was in regards to the MIO Console mixer itself. If you're using Studio One's I/O plugin to ping your hardware, everything should work just fine. If you're wanting to keep your Cranborne, you could get an Edge card in the ULN/LIO-8 (there are 2 or 3 different ADAT Edge card options). It doesn't sound to me like you'd experience any issues - you're monitoring live inputs through a hardware mixer at your tracking/rehearsal space. And when you do overdubs at home, you would just either need to monitor through the MIO Console for (essentially) zero latency or use some kind of small analog mixer if you prefer. The ULN/LIO-8 has an analog out mult of the inputs on the back (labeled "Balanced Send 1-8") that are pre-A/D, so it's super easy to send those to your analog mixer. With regards to RTL in the DAW, I'm pretty sensitive to latency, and I'm able to get my RTL low enough with the MHLink (that's the ethernet) connection and driver that I can play VI pianos. If I needed to regularly play guitar through amp plugins and monitor in the DAW late in a production when I have tons of tracks already, it would be dicier. I use an external amp modeler for that reason (if I can't mic a real amp). I'm on a Mac Mini M1 right now, and I usually run at 96kHz, but sometimes 48kHz. Happy to help with any other questions... Bottom line, in your personal experience, when running at 96k, what sort of real world RTL are you able to get over MHLink during tracking, and, when doing so, what system buffer (and MHLink safety buffer) are you running at? Also, for whatever it's worth, I have a M1 Mac Studio with the Gigabit Ethernet port. I hadn't mentioned that previously in this thread. Also, just to be clear, you're using the latest and greatest version of MHLink, correct? |
|
|
|
Post by plinker on Nov 5, 2024 8:12:04 GMT -6
It looks like Thunderbolt to PCIe expansion chassis are about $300. Not the good ones that are quiet enough to have in the control room with you. Here's a thread where we were talking about this a couple months ago: realgearonline.com/thread/17592/sonnet-echo-3-pcie-thunderboltActually, I'm not 100% sure if tonycamphd has weighed in on his final fan volume report in real-world usage. Tony, how's it going over there? Can't remember if yours is in a machine room or not? The RME PCIe AES card is 2u in height, so you can't use a 1-slot chassis if you go that route. The Lynx card I think might be 1u high, though? Wow! For almost the price of a new M5 Mini, I could integrate a PCIe card ?? Sonnet is getting a really good deal on that. I guess they know they have a captive audience. I'm glad this is not a thing I have to deal with.  |
|
|
|
Post by gravesnumber9 on Nov 5, 2024 8:54:37 GMT -6
So I've been curious about Metric Halo for a long time and posted such on another thread. I think I was actually part of what got the MH conversation going along with Quint. I was talking about looking for a "forever" interface and felt like this could be it. And now with this pricing discount I could afford to get two. But... I'm getting a little bit confused about the pros and cons. On the one hand I really don't like using two apps and that's why I got out of the UAD world to begin with, I hated console. But on the other hand I like simplicity and if the MIO Console Session thing works smoothly, why not? Also on the other hand I do a ton of hybrid mixing and I feel like the determined latency is more important in that case than the actual latency number. But I'm also getting confused on apparent errors in the MH universe on compensating for that? Here's my use case. Three things... 1) I do a lot of mixing both for hire and also for my own projects where I was either the producer or the artist. When I produce (myself or others) I track basic tracks in my tracking/rehearsal room through an old A&H console going into a MOTU 24Ao and a MOTU 24Ai. Then I bring that back to my mixing room where I often overdub vocals and ancillary parts. Currently I often use my Cranborne 500ADAT to monitor those overdubs live and my MOTU 828ES and Studio One compensate just fine as they should. 2) Mixing for hire I just mix and I'm doing a lot of hybrid. Not much to explain here. 3) I also do a lot of writing in my mixing room where I'm either playing a keyboard with a VI piano or acoustic guitar and singing. On guitar I monitor direct through the Cranborne. When I do piano I use Studio One's low latency VI monitoring. I don't track through a ton of plugins except when I'm overdubbing late in the process. And even then I often disable most plugins and have usually printed my instruments. Also Studio One's low latency monitoring works great. I'm assuming it would work great with a MH Lio too but I don't know? The problem I'm trying to solve is, again, that I'm tired of thinking about this. I like that MH allows you to update hardware and has universally admired human support. I'd like to just be done thinking about interfaces forever. But other than that I don't really have a problem. My MOTU 828ES / Cranborne 500ADAT / MOTU Ao system in the mixing room is working just fine. It does turn out that I don't need as many line outs as I initially thought I was going to because I'm rarely patching in more than 12 - 14 at the same time having given up on my RND Orbit, but other than the fact that I've got more Line Outs than I need right now, there's no real problem. And yet... I feel like now is my chance to get into MH affordably with the MH Lio being $1100 off. And I've been flirting with MH for a while. Am I even a good use case for this device? This thread has confused me. The disagreement about latency compensation above was in regards to the MIO Console mixer itself. If you're using Studio One's I/O plugin to ping your hardware, everything should work just fine. If you're wanting to keep your Cranborne, you could get an Edge card in the ULN/LIO-8 (there are 2 or 3 different ADAT Edge card options). It doesn't sound to me like you'd experience any issues - you're monitoring live inputs through a hardware mixer at your tracking/rehearsal space. And when you do overdubs at home, you would just either need to monitor through the MIO Console for (essentially) zero latency or use some kind of small analog mixer if you prefer. The ULN/LIO-8 has an analog out mult of the inputs on the back (labeled "Balanced Send 1-8") that are pre-A/D, so it's super easy to send those to your analog mixer. With regards to RTL in the DAW, I'm pretty sensitive to latency, and I'm able to get my RTL low enough with the MHLink (that's the ethernet) connection and driver that I can play VI pianos. If I needed to regularly play guitar through amp plugins and monitor in the DAW late in a production when I have tons of tracks already, it would be dicier. I use an external amp modeler for that reason (if I can't mic a real amp). I'm on a Mac Mini M1 right now, and I usually run at 96kHz, but sometimes 48kHz. Happy to help with any other questions... This is helpful. I do have a couple of questions. I think I could almost swing it to do one Lio-8 in my mixing room (w/Cranborne) and two in my tracking room paired up with UA-4170 with which which I could take an additional 8 inputs and just toggle between board or UA pre-amp for four of them. I could get pretty close to a lateral price move by selling other gear that is getting underused plus selling the MOTU ecosystem for a pretty conservative price. But I'm not totally sure I could swing it with one Lio-8 in the mixing room because... 1) The monitor outs aren't independent. They appear to be a mult of the standard 1-2 line out as does the headphone out. So this gobbles up two of my line outs and makes an already tightening I/O situation even tighter. Could I use the 1-2 I/O for my two bus chain and then toggle between that and the pre-send signal? Is that how it works? In other words could I switch back and forth between monitoring with and without my analog two bus chain this way? If I could that would make up for the loss of outputs. I would just monitor without two bus when overdubbing or tracking, and with when mixing. Actually that would be a very ideal setup. 2) I have literally zero use for the AES I/O at this point. And I'm struggling to figure out how I could use those. My monitors are Neumann KH310 which I love so I'm not swapping out my monitors for something with digital inputs. Do they make AD/DA boxes for AES similar to like what Ferrofish has for ADAT and MADI? If so I couldn't find any. 3) It does appear that the AES inputs would be useful in the tracking room as the 4-710 can send 8 AES outputs. So am I correct in understanding that the AES inputs on the LIO-8 are in addition to the analog inputs? Dumb question but just checking. |
|
|
|
Post by Quint on Nov 5, 2024 9:54:23 GMT -6
I use MHLink just for the convenience and stability. However, on the MH list-serv they claim that the USB connection is faster. I'm not sure why. The USB connection is faster because MH haven't yet been able to get MHLink to bypass the Apple network interface. They may figure out a way to do that natively, but it also might require a dedicated MH PCI card or Thunderbolt box with an ethernet port on it. Basically, for them to get the RTL latency to an acceptably low level to monitor through your DAW plugins, they're going to have to do something different than they're currently doing. I wish to goodness they would've just gone with a Thunderbolt connector on the ULN-8 3D (or MkIV) to get around these limitations, but the team seems focused on other issues. It drives me slightly crazy, but I love so many other things about the boxes that I just keep on using them. I love their sound and stability. You mentioned "thunderbolt box with an Ethernet port on it". Can you elaborate on this? It's not entirely clear to me where the hurdle lies in getting latency lower over MHLink, but if running network over thunderbolt would eliminate that problem, wouldn't it be simple enough to just get one of those thunderbolt to network adapters that are out there (and not that expensive) and just run MHLink thru that? |
|
|
|
Post by tonycamphd on Nov 5, 2024 10:19:47 GMT -6
It looks like Thunderbolt to PCIe expansion chassis are about $300. Not the good ones that are quiet enough to have in the control room with you. Here's a thread where we were talking about this a couple months ago: realgearonline.com/thread/17592/sonnet-echo-3-pcie-thunderboltActually, I'm not 100% sure if tonycamphd has weighed in on his final fan volume report in real-world usage. Tony, how's it going over there? Can't remember if yours is in a machine room or not? The RME PCIe AES card is 2u in height, so you can't use a 1-slot chassis if you go that route. The Lynx card I think might be 1u high, though? Same noctua fan in mine as the most expensive model, very quiet and can be used in the same room, if you are tracking something very low spl with a sensitive mic, u could put a pillow somewhere directly between it and your mic and it’ll vanish |
|
|
|
Post by Dan on Nov 5, 2024 10:38:05 GMT -6
Windows drivers when for the mk iv?
|
|
|
|
Post by veggieryan on Nov 5, 2024 10:56:34 GMT -6
Windows drivers when for the mk iv? Assume never at this point. |
|
|
|
Post by veggieryan on Nov 5, 2024 11:45:34 GMT -6
You mentioned "thunderbolt box with an Ethernet port on it". Can you elaborate on this? It's not entirely clear to me where the hurdle lies in getting latency lower over MHLink, but if running network over thunderbolt would eliminate that problem, wouldn't it be simple enough to just get one of those thunderbolt to network adapters that are out there (and not that expensive) and just run MHLink thru that? You do get slightly better performance by using one of the "known good" ethernet adapters like the apple thunderbolt to ethernet adapter. To get to the next level I think they need to make their own version of this adapter. You see a similar problem with Dante which is another audio over ethernet cable protocol. You can run Dante on a standard ethernet port but there is a reason people buy dedicated PCIe Dante cards. They offer better performance, lower latency and less jitter. RME even offers USB interfaces for Dante which greatly outperform MHLink. It all comes down the skill and knowledge required to build great interface hardware and then to write a great driver for it. I hate to be the "debbie downer" but if Metric Halo had the ability to do this they probably would have done it by now... Or perhaps people just need to be made aware of the problem and complain louder. For me personally I was hoping they would finally fix the I/O plugin in the MIOConsole mixer so that it is latency compensated. That way I could replace my analog mixer with MIOConsole and have all my hardware patched in and able to monitor with decent latency in MIOConsole just like an analog mixer. I got tired of waiting for this to happen and bailed years ago. If they did add latency compensated hardware inserts I would consider switching back and just running through an RME Madi interface from a Madi edge card. I thought the conversion on the mkiii units was mediocre on the AD side especially but the mkiv conversion is reportedly better and most importantly has lower latency in the conversion process itself. For me the built in preamps on the ULN-8 are the standout feature over other interfaces as they are the best preamps I have ever heard in an interface and they have pre converter analog sends. If you don't need those these interfaces make less sense for you as well. If you don't need a DSP mixer then these interfaces make even less sense for you. In your situation I would lean towards an RME Dante interface and an AMS Neve StarNet ADA16 Dante converter or Burl Black 16x16 with Dante since everything seems to be headed towards Dante. For myself in considering replacing my analog mixer instead of MIOConsole there are newer Dante digital mixers like Yamaha DM3 that have inserts at around $2,000 USD. Compare what you get for the price and things are changing... |
|
|
|
Post by Quint on Nov 5, 2024 14:45:13 GMT -6
You mentioned "thunderbolt box with an Ethernet port on it". Can you elaborate on this? It's not entirely clear to me where the hurdle lies in getting latency lower over MHLink, but if running network over thunderbolt would eliminate that problem, wouldn't it be simple enough to just get one of those thunderbolt to network adapters that are out there (and not that expensive) and just run MHLink thru that? You do get slightly better performance by using one of the "known good" ethernet adapters like the apple thunderbolt to ethernet adapter. To get to the next level I think they need to make their own version of this adapter. You see a similar problem with Dante which is another audio over ethernet cable protocol. You can run Dante on a standard ethernet port but there is a reason people buy dedicated PCIe Dante cards. They offer better performance, lower latency and less jitter. RME even offers USB interfaces for Dante which greatly outperform MHLink. It all comes down the skill and knowledge required to build great interface hardware and then to write a great driver for it. I hate to be the "debbie downer" but if Metric Halo had the ability to do this they probably would have done it by now... Or perhaps people just need to be made aware of the problem and complain louder. For me personally I was hoping they would finally fix the I/O plugin in the MIOConsole mixer so that it is latency compensated. That way I could replace my analog mixer with MIOConsole and have all my hardware patched in and able to monitor with decent latency in MIOConsole just like an analog mixer. I got tired of waiting for this to happen and bailed years ago. If they did add latency compensated hardware inserts I would consider switching back and just running through an RME Madi interface from a Madi edge card. I thought the conversion on the mkiii units was mediocre on the AD side especially but the mkiv conversion is reportedly better and most importantly has lower latency in the conversion process itself. For me the built in preamps on the ULN-8 are the standout feature over other interfaces as they are the best preamps I have ever heard in an interface and they have pre converter analog sends. If you don't need those these interfaces make less sense for you as well. If you don't need a DSP mixer then these interfaces make even less sense for you. In your situation I would lean towards an RME Dante interface and an AMS Neve StarNet ADA16 Dante converter or Burl Black 16x16 with Dante since everything seems to be headed towards Dante. For myself in considering replacing my analog mixer instead of MIOConsole there are newer Dante digital mixers like Yamaha DM3 that have inserts at around $2,000 USD. Compare what you get for the price and things are changing... I'm pretty meh on all of the modern AMS Neve stuff, so I'm not terribly interested in their AD/DA. Burl via MADI/Dante is likely the way I'll go. I was just basically looking at the Lio as a way to more or less spend the same amount of money as I would to get a RME card and enclosure, but with the added bonus of getting another 8 channels of conversion and all of the other bells and whistles that MH interfaces are known for. It would have been a win win in that scenario, but the latency seems like it might kill that idea, unfortunately, so the only other option is a RME card in an enclosure. I'm just really pretty committed to the idea of getting a really low latency native solution (no secondary monitoring app) if I'm going to dump my Apollos/Luna. Otherwise, I may begrudgingly stick with the Apollos/Luna if I can't achieve that, despite my feelings about UA. |
|
|
|
Post by plinker on Nov 5, 2024 14:53:20 GMT -6
But I'm not totally sure I could swing it with one Lio-8 in the mixing room because... 1) The monitor outs aren't independent. They appear to be a mult of the standard 1-2 line out as does the headphone out. So this gobbles up two of my line outs and makes an already tightening I/O situation even tighter. Could I use the 1-2 I/O for my two bus chain and then toggle between that and the pre-send signal? Is that how it works? In other words could I switch back and forth between monitoring with and without my analog two bus chain this way? If I could that would make up for the loss of outputs. I would just monitor without two bus when overdubbing or tracking, and with when mixing. Actually that would be a very ideal setup. 2) I have literally zero use for the AES I/O at this point. And I'm struggling to figure out how I could use those. My monitors are Neumann KH310 which I love so I'm not swapping out my monitors for something with digital inputs. Do they make AD/DA boxes for AES similar to like what Ferrofish has for ADAT and MADI? If so I couldn't find any. 3) It does appear that the AES inputs would be useful in the tracking room as the 4-710 can send 8 AES outputs. So am I correct in understanding that the AES inputs on the LIO-8 are in addition to the analog inputs? Dumb question but just checking. I'll chime in with what I know: 1a. Correct; the monitor outs aren't independent -- they are a mult of the 1-2 line outs. 1b. re: monitoring -- If I understand your question... I'd recommend that you keep 1-2 hooked to your monitors (via the monitor out jacks). Then insert an I/O plugin in the input strip that comes from your DAW -- it's called "Host 1/2". You'll configure that I/O plug to route to your outboard via the remaining output channels (e.g., 3 through 8), and return through whatever analog inputs are unused. Then, when monitoring, you can simply turn the I/O plug on (or off) to route to the outboard. This is what I do with individual outboard processors. I'll try to upload a picture of the MIOConsole configured like this later. I have yet to have a routing challenge that couldn't be addressed by the MIOConsole. Anything is possible. And don't forget the preamp sends on the back of that box -- you can route a mic into the excellent ULN preamps (via Mic inputs) to outboard (via Balanced Send) then back in (via Line inputs). Note: you still get only 8 channels of AD conversion. 2. I can't answer your question but will admonish the audio tool-making world for not migrating to AES digital as a standard. ADAT is so old and clunky that it hurts my brain to consider it an option. 3. Not a dumb question and, yes, the AES inputs are in addition to the analog inputs. Good luck! |
|
|
|
Post by veggieryan on Nov 5, 2024 15:48:42 GMT -6
I'm pretty meh on all of the modern AMS Neve stuff, so I'm not terribly interested in their AD/DA. Burl via MADI/Dante is likely the way I'll go. I was just basically looking at the Lio as a way to more or less spend the same amount of money as I would to get a RME card and enclosure, but with the added bonus of getting another 8 channels of conversion and all of the other bells and whistles that MH interfaces are known for. It would have been a win win in that scenario, but the latency seems like it will kill that idea, unfortunately, so the only other option is a RME card in an enclosure. I'm just really pretty committed to the idea of getting a really low latency native solution (no secondary monitoring app) if I'm going to dump my Apollos/Luna. Otherwise, I may begrudgingly stick with the Apollos/Luna if I can't achieve that, despite my feelings about UA. We are in a very similar situation. Running the numbers it ends up not being worth it to ditch the Apollo ecosystem right now. Mainly because you have to buy a separate Dante or Madi interface before you even think about conversion. I am also hearing positive reports that the Gen 2 Apollo's actually did greatly improved the headphone outputs which has been my main beef with my Apollo X6 outside the limited routing and DSP. If the headphone outputs are really improved then I would consider ditching my analog monitor controller for a Gen 2 Apollo. Getting usable DI inputs and improved conversion would be icing on the cake at that point for me. Would simplify things for me. I am sure some people would really like the room correction in DSP. Per our discussions I am now happily monitoring through Logic with my Apollo X6 instead of Console and its working well at 96 kHz with 64 sample buffer size on my M1 Pro MacBook Pro. I use the Console like an FX rack for now using the virtual channels like bus sends in Logic for reverbs and delays. For context, I will post my RTL with input compensation disabled in Console and no plugins in console using Oblique's RTL Utility: Apollo X6 Line out -> Apollo X6 Line In 3.281 ms or 315 samples RTL at 64 sample buffer at 96 khz Apollo X6 Line out -> JCF AD8 Line in -> RME ADI-8 DD AES In -> Apollo X6 ADAT In 2.986 ms or 278 samples RTL at 64 sample buffer at 96 khz Note: The group delay of the AD conversion in the JCF AD8 is very low at 11 samples at 96 khz with PEP disabled. This means the AD conversion on the X6 is 37 samples longer than the AD8 minus how ever many samples the ADAT input adds. As you can see the Apollo X is really no slouch when it comes to real world *usable* RTL including the actual conversion latency... How much better would RME really be? How close to this real world usable RTL can MHLink get? |
|
|
|
Post by Quint on Nov 5, 2024 16:24:34 GMT -6
I'm pretty meh on all of the modern AMS Neve stuff, so I'm not terribly interested in their AD/DA. Burl via MADI/Dante is likely the way I'll go. I was just basically looking at the Lio as a way to more or less spend the same amount of money as I would to get a RME card and enclosure, but with the added bonus of getting another 8 channels of conversion and all of the other bells and whistles that MH interfaces are known for. It would have been a win win in that scenario, but the latency seems like it will kill that idea, unfortunately, so the only other option is a RME card in an enclosure. I'm just really pretty committed to the idea of getting a really low latency native solution (no secondary monitoring app) if I'm going to dump my Apollos/Luna. Otherwise, I may begrudgingly stick with the Apollos/Luna if I can't achieve that, despite my feelings about UA. We are in a very similar situation. Running the numbers it ends up not being worth it to ditch the Apollo ecosystem right now. Mainly because you have to buy a separate Dante or Madi interface before you even think about conversion. I am also hearing positive reports that the Gen 2 Apollo's actually did greatly improved the headphone outputs which has been my main beef with my Apollo X6 outside the limited routing and DSP. If the headphone outputs are really improved then I would consider ditching my analog monitor controller for a Gen 2 Apollo. Getting usable DI inputs and improved conversion would be icing on the cake at that point for me. Would simplify things for me. I am sure some people would really like the room correction in DSP. Per our discussions I am now happily monitoring through Logic with my Apollo X6 instead of Console and its working well at 96 kHz with 64 sample buffer size on my M1 Pro MacBook Pro. I use the Console like an FX rack for now using the virtual channels like bus sends in Logic for reverbs and delays. For context, I will post my RTL with input compensation disabled in Console and no plugins in console using Oblique's RTL Utility: Apollo X6 Line out -> Apollo X6 Line In 3.281 ms or 315 samples RTL at 64 sample buffer at 96 khz Apollo X6 Line out -> JCF AD8 Line in -> RME ADI-8 DD AES In -> Apollo X6 ADAT In 2.986 ms or 278 samples RTL at 64 sample buffer at 96 khz Note: The group delay of the AD conversion in the JCF AD8 is very low at 11 samples at 96 khz with PEP disabled. This means the AD conversion on the X6 is 37 samples longer than the AD8 minus how ever many samples the ADAT input adds. As you can see the Apollo X is really no slouch when it comes to real world *usable* RTL including the actual conversion latency... How much better would RME really be? How close to this real world usable RTL can MHLink get? Its even better in Luna, where DSP is used for record-armed channels, in which case it's 1.1 ms at 96k. But I get what you're saying about native monitoring RTL thru the DAW when using an Apollo. It's still not bad, though other TB interfaces fare even better on latency when monitoring natively thru the DAW. I think it just generally speaks to the positives associated with using TB as the connection protocol between a computer and an interface. It's certainly why I generally prefer TB over the other options. Speaking of connection types, do you suppose there is any latency advantage to having a Gigabit Ethernet connection to connect MHLink to? My Mac Studio has a Gigabit Ethernet port, for whatever that's worth. |
|
|
|
Post by veggieryan on Nov 5, 2024 17:52:20 GMT -6
Its even better in Luna, where DSP is used for record-armed channels, in which case it's 1.1 ms at 96k. But I get what you're saying about native monitoring RTL thru the DAW when using an Apollo. It's still not bad, though other TB interfaces fare even better on latency when monitoring natively thru the DAW. I think it just generally speaks to the positives associated with using TB as the connection protocol between a computer and an interface. It's certainly why I generally prefer TB over the other options. Speaking of connection types, do you suppose there is any latency advantage to having a Gigabit Ethernet connection to connect MHLink to? My Mac Studio has a Gigabit Ethernet port, for whatever that's worth. Gigabit Ethernet port won't really make difference over the recommended ethernet adapter for MHLink AFAIK. As for the 1.1ms at 96k in Apollo Console. That is only with no plugins or with only plugins that don't add latency. Once you start adding plugins that have latency like all the desirable mk2 ones you actually want to use the latency goes up... You have to check the latencies for UAD plugins chart here: help.uaudio.com/hc/en-us/articles/15247166304148-Additional-UAD-Plug-In-LatencyLet's take an example vocal chain in Apollo Console since you might want to use a few of those fancy plugins you paid full price for a few years ago: Latency in Samples - Plugin 55 - V76 Preamp 55 - Pultec EQP-1A 55 - UA 1176AE 56 - Teletronix LA-2A 221 Samples total - 2.3 ms at 96 khz or 4.6 ms at 48 kHz If you have more than one mic you probably will need to enable input delay compensation and set to medium-long for that example chain in Console and thats 300 samples. Now your RTL is 300 samples higher which is huge... quite a different story than the Apollo "no latency" DSP marketing... Its crap like this along with the limited routing and DSP that led me to just ditch it all and monitor in the DAW with plugins like MBSI, Mixwave VLC or CA-70S that sound better anyways and add 0 samples of latency at 1x oversampling. Note that UADx plugins have suspiciously high latency ex: usually 118 samples for preamp/eq/channelstrip UADx plugins... I hope it's not to prop up the idea that DSP is still relevant. So UADx are basically useless for DAW monitoring. Very convenient... People often forget about the group delay of the converter chips so you might find that the actual RTL of most PCIe and thunderbolt interfaces are basically the same as the Apollo or worse. Maybe you gain 0.5 ms at 64 sample buffer size. Probably not. The bigger question is if a better driver like RME makes a 32 sample buffer usable and if the real RTL is actually lower. Look at this chart from the blog of the user "ramses" on the RME forum and you will see that with the JCF AD8's low input delay I am beating a lot of RME scores even at 32 sample buffer size. Who knows if the 32 sample buffer is even usable with a reasonable DAW load? RME latency chart: www.tonstudio-forum.de/attachment/3028-rme-rtl-comparison-v2-jpg/from this blog: www.tonstudio-forum.de/blog/entry/105-rtl-comparison-of-different-rme-products-en-de/ |
|
|
|
Post by Quint on Nov 5, 2024 18:03:56 GMT -6
Its even better in Luna, where DSP is used for record-armed channels, in which case it's 1.1 ms at 96k. But I get what you're saying about native monitoring RTL thru the DAW when using an Apollo. It's still not bad, though other TB interfaces fare even better on latency when monitoring natively thru the DAW. I think it just generally speaks to the positives associated with using TB as the connection protocol between a computer and an interface. It's certainly why I generally prefer TB over the other options. Speaking of connection types, do you suppose there is any latency advantage to having a Gigabit Ethernet connection to connect MHLink to? My Mac Studio has a Gigabit Ethernet port, for whatever that's worth. Gigabit Ethernet port won't really make difference over the recommended ethernet adapter for MHLink AFAIK. As for the 1.1ms at 96k in Apollo Console. That is only with no plugins or with only plugins that don't add latency. Once you start adding plugins that have latency like all the desirable mk2 ones you actually want to use the latency goes up... You have to check the latencies for UAD plugins chart here: help.uaudio.com/hc/en-us/articles/15247166304148-Additional-UAD-Plug-In-LatencyLet's take an example vocal chain in Apollo Console since you might want to use a few of those fancy plugins you paid full price for a few years ago: Latency in Samples - Plugin 55 - V76 Preamp 55 - Pultec EQP-1A 55 - UA 1176AE 56 - Teletronix LA-2A 221 Samples total - 2.3 ms at 96 khz or 4.6 ms at 48 kHz If you have more than one mic you probably will need to enable input delay compensation and set to medium-long for that example chain in Console and thats 300 samples. Now your RTL is 300 samples higher which is huge... quite a different story than the Apollo "no latency" DSP marketing... Its crap like this along with the limited routing and DSP that led me to just ditch it all and monitor in the DAW with plugins like MBSI, Mixwave VLC or CA-70S that sound better anyways and add 0 samples of latency at 1x oversampling. Note that UADx plugins have suspiciously high latency ex: usually 118 samples for preamp/eq/channelstrip UADx plugins... I hope it's not to prop up the idea that DSP is still relevant. So UADx are basically useless for DAW monitoring. Very convenient... People often forget about the group delay of the converter chips so you might find that the actual RTL of most PCIe and thunderbolt interfaces are basically the same as the Apollo or worse. Maybe you gain 0.5 ms at 64 sample buffer size. Probably not. The bigger question is if a better driver like RME makes a 32 sample buffer usable and if the real RTL is actually lower. Look at this chart from the blog of the user "ramses" on the RME forum and you will see that with the JCF AD8's low input delay I am beating a lot of RME scores even at 32 sample buffer size. Who knows if the 32 sample buffer is even usable with a reasonable DAW load? RME latency chart: www.tonstudio-forum.de/attachment/3028-rme-rtl-comparison-v2-jpg/from this blog: www.tonstudio-forum.de/blog/entry/105-rtl-comparison-of-different-rme-products-en-de/Oh, I'm totally aware that the plugins add latency in Console. But the native version of those same plugins add latency on the native side too. I was simply pointing out an apples to apples comparison in a scenario where no plugins have been added. I'm not sure which UAD plugins you are talking about that are 118 samples though? Most of the UAD plugins that have latency are 55 samples. |
|
|
|
Post by veggieryan on Nov 5, 2024 18:15:49 GMT -6
You are aware but most people are not. Also the native UADx plugins usually add more latency than the DSP versions which I find suspicious.
|
|
|
|
Post by Quint on Nov 5, 2024 18:25:23 GMT -6
You are aware but most people are not. Also the native UADx plugins usually add more latency than the DSP versions which I find suspicious. I remember that there was some confusion back in the day about plugins adding latency in Console, and that had to do with UA's marketing and vague pronouncements. I was confused by it myself many years ago. I think Matt Hepworth's latency chart, and UA getting their feet held to the fire on this issue, ultimately resulted in not being as much of a problem as it used to be, because the info started to get out there in ways that it previously was not. But, yeah, I'm sure there are still people out there that don't know that the DSP plugins add latency. I generally hate how UA does shit like that. Just be honest and upfront. It's one of the things that appeals to me about MH. The MH company stance doesn't appear to be one where they treat their customers like they are dumb. I need to go and look, but as far as I was aware, I thought the native and DSP plugins had the same latency. That would be interesting if they did not. |
|
|
|
Post by the other mark williams on Nov 5, 2024 19:03:01 GMT -6
Speaking of connection types, do you suppose there is any latency advantage to having a Gigabit Ethernet connection to connect MHLink to? My Mac Studio has a Gigabit Ethernet port, for whatever that's worth. [...] Gigabit Ethernet port won't really make difference over the recommended ethernet adapter for MHLink AFAIK. [...] Actually, on Quint 's machine, it would make a difference. On the M1 Mac Studio, there's an issue with the ethernet adapter that causes it to drop packets unexpectedly. The MHLink driver can handle that, but it's the reason people have needed a larger safety buffer on that machine in particular if they're using the native ethernet port. MH recommends using the Sonnet TB3-to-Ethernet adapter for the M1 Mac Studio. It costs a bit more than various el cheepos, but it's guaranteed to have the right (fastest, most stable) chipset. I think it's this one: --TB3 Sonnet adapter on Amazon--
The issue is not present on the M2 Mac Studios, nor any of the Mac Minis, I don't believe. |
|
|
|
Post by veggieryan on Nov 5, 2024 19:03:38 GMT -6
Most of the UADx plugins are 118 samples compared to the usual 55-ish for the DSP versions. 429 samples for UADx Capitol Chambers probably since it's a reverb. UADx Ampex is 326 samples which is interestingly far less than the DSP version.
|
|
|
|
Post by veggieryan on Nov 5, 2024 19:11:06 GMT -6
Gigabit Ethernet port won't really make difference over the recommended ethernet adapter for MHLink AFAIK. [...] Actually, on Quint 's machine, it would make a difference. On the M1 Mac Studio, there's an issue with the ethernet adapter that causes it to drop packets unexpectedly. The MHLink driver can handle that, but it's the reason people have needed a larger safety buffer on that machine in particular if they're using the native ethernet port. MH recommends using the Sonnet TB3-to-Ethernet adapter for the M1 Mac Studio. It costs a bit more than various el cheepos, but it's guaranteed to have the right (fastest, most stable) chipset. I think it's this one: --TB3 Sonnet adapter on Amazon--
The issue is not present on the M2 Mac Studios, nor any of the Mac Minis, I don't believe. I didn't word that very well but yes I agree nothing beats the current recommended ethernet adapter for MHLink which is currently the Sonnet unit. Think of all the money Metric Halo could make by selling their own dedicated low latency thunderbolt to MHLink adapter instead of giving 200 bucks to Sonnet... are you reading this BJ? |
|
