Jump to content

​Apollo Twin USB 3.0: Creating a Windows Music Production System


Recommended Posts

  • Members

Welcome to another Pro Review, HC’s unique interactive format. We encourage participation from everyone—users of the product, potential users with questions, and the manufacturer. The object is to make this an “open source” review without the limitations on space of print, the potential bias of having a single reviewer, but more importantly, the ability to tap the community’s expertise to dive really deeply into what a product can—and cannot—do, so you can know exactly what to expect. For more information on what Pro Reviews are all about, please check out the FAQ.


This is going to be an interesting Pro Review, to say the least. It’s not “just another” audio interface review, in part because Apollo Twin USB is a high-end, different kind of interface with near real-time processing. However, it’s also going to cover how to create a no-compromise, Windows-based music production system.




Why? Because Windows’s star is rising as an audio platform. Although Thunderbolt adoption has stalled on Windows for several reasons (some practical and some allegedly political), USB 3.0 is common and fast. My first experience with USB 3.0 audio on Windows has been with TASCAM’s US-20x20 USB 3.0 interface which, while obviously targeting a different section of the market than Apollo Twin, shows that the benefits of USB 3.0 compared to USB 2.0 are considerable. With Universal Audio adding hardware-based DSP for near real-time (sub-2 ms) processing on the input and/or output, we have for all practical purposes Thunderbolt-level performance on Windows.


Furthermore, Windows 10 changes the game as well. Microsoft has been paying attention to improving audio performance for their native audio, but remember that Windows 10 is an evolving system. Microsoft’s roadmap includes further audio optimizations that they believe will approach ASIO-level performance. This won’t happen overnight, but the fact that it’s on their “to-do” list is encouraging. Already, having multi-client MIDI support in Windows 10 is extremely helpful to those of us who use MIDI. I’ve been running several applications with Windows 10, particularly Cakewalk SONAR, and I’ve experienced both faster and more stable DAW performance under Windows 10 than Windows 7 or the much-reviled Windows 8.


Now, lest you think I’m a gung-ho Windows fanboi, that’s not the case. I have an older Mac dual-Xeon desktop, an up-to-date MacBook Pro running Logic, and use iPad and iPhone iOS devices. I’m comfortable in both the Mac and Windows worlds, in fact I was-Mac only for a decade, but I use Windows for the heavy lifting with my audio and video work. Part of this is because two of my favorite programs, SONAR and Vegas Pro, are Windows-only but the other is that the cost-effectiveness in terms for power-per-dollar is undeniable. I also appreciate the emphasis Microsoft places on backward compatibility. Although the Mac’s Core Audio is a superior implementation of native audio, ASIO is just fine on Windows, and will likely remain so unless/until Microsoft develops native audio support on the same level as the Mac.


So here’s the plan for the Pro Review, although of course, because it’s interactive and invites participation from all, that’s subject to change.


First, we’ll cover Apollo Twin USB with the existing laptop I use for audio when on the road doing workshops and presentations. That will allow covering the unit itself, the software that’s included with it, and selected UA plug-ins. I’ll also be making sure to stretch the limits of what UA recommends. For example, they test their interfaces thoroughly with only a limited selection of software, so we’ll push that envelope.


Second, Intel has provided (thank you, Intel!) a loaner NUC (Next Unit of Computing).




This is a hip, small, powerful, reasonably priced computer that’s conceptually like a Mac Mini. I’ve always felt that whether you use Mac or Windows, you owe it to yourself to dedicate a computer to music. That hasn’t always been practical due to cost, but something like the NUC not only means you can create a system, but with a compact, high-quality interface like Apollo Twin USB, it’s downright portable. Add a small touch monitor and headphones, and you can fit a top-shelf recording studio in your carry-on luggage.


Finally, although this hasn’t been set in stone yet, I’m trying to borrow a Surface Pro. I know that SONAR runs on it well and even though UA doesn’t recommend using Apollo Twin USB with a tablet, they tend to be extremely cautious and I want to see if that degree of caution is relevant, or whether the concept of a truly portable recording setup based on a Surface Pro is doable. That would be wonderful.


Fortunately, I also have access to all the Gibson Brands products as part of this experiment, which means I’ll be able to borrow the new, and very portable, 4” Les Paul monitors to gauge their usefulness in a mobile-friendly studio (I already use KRK’s KNS-8400 headphones), as well as Neat Microphones for input. And because Apollo Twin USB doesn’t have 5-pin DIN MIDI I/O, I’ll be testing out some DIN-to-USB adapters as part of the review.


Ready? Let’s rock! This is going to be fun...we’ll start with unboxing the Apollo Twin USB, along with some background material on USB 3.0.

Edited by Anderton
  • Like 1
Link to comment
Share on other sites

  • Members

Time to start the unboxing and see what we got...


Here's the box itself. It's not particularly big, so if you want to protect your Apollo while lugging it around, it doesn't take up much space.




Now let's take out the Apollo Twin USB itself. It's pretty obvious UA put some thought into the industrial design...it may be for Windows, but it has the sleek, silver kinda Apple vibe.




After you lift it out, here's what's in the rest of the box: the blue USB 3.0 cable (which is about 6.5 feet) and a cardboard box that holds the AC adapter. Note that you can't bus-power Apollo Twin USB.




Speaking of the power supply, it's one of those "global" types that handles any voltage or frequency, and has snap-in plugs that work pretty much anywhere on earth.




As to the front panel, UA knows how to make me happy...note the hi-Z guitar input right up front, on the opposite side of the headphone jack.




And here's the rear panel. We'll discuss the I/O in more depth later, but it's pretty decent for a compact interface. There's an ADAT in so you have an instant "trap door" for adding eight more mic pres or whatever.




And...there's a getting started document, which is pretty minimal but tells you where to go for the latest info.





Well I guess if UA wants to tell me about "Getting Started"...then that's what Is should do. My next post will be from a hotel room far from home, assuming the wi-fi is functional :)


(P.S. - Don't forget to hit this thread's "Subscribe" button to be advised of when there's a new post.)

Edited by Anderton
fix typoze
Link to comment
Share on other sites

  • Members

Apollo's First Flight


Well, I’m back home. The schedule didn’t quite work out as expected so I never got to set up my impromptu studio :(2. But, I’m not going anywhere for a while, so it’s time to get back into the Apollo groove.


Before doing anything you'll want to download and install the software, which includes the drivers and software plug-ins. If you’re not registered, you need to set up an account with UA. Once that’s done, it’s time to download 1.9 GB of software so if you’re on dial-up, come back in about a week :). Meanwhile, to keep you entertained, there’s a video that describes setup and such. Also, take a look at the knowledgebase article about required Windows settings so you can do a few tweaks that will make Apollo happy, like preventing the USB ports from falling asleep to save power. Note that these tips apply to any USB interface so even if you’re not using an Apollo for now, check them out.


There are four manuals—Apollo Twin USB software, Apollo Twin USB hardware, UA System software, and another one for the Powered Plug-Ins. You want to follow the instructions in the Apollo Software Manual (which advises connecting the hardware before installing the software), as the installation procedure is different from the one described for the UA System software (which advises installing the software before connecting the hardware.)


In terms of plug-ins, the software package includes the UA 1176SE Legacy Compressor, RealVerb-Pro, Pultec EQP-1A Legacy, and the Precision Mix Rack Collection. When you register, you’re offered some deals on plug-ins, however I should point out a very cool aspect of UA’s way of doing business: Once you buy a plug-in, you own the ability to use that plug-in on any UA device. For example I have a Quad card for my desktop with a bunch of plug-ins, and upon authorizing the Apollo Twin USB, all of them showed up as available for my laptop. (This is not like record companies making you buy the same music for every possible format…which kind of undermines the argument that they’re selling intellectual property and not physical media…but I digress.)


So, time to test things out. Given that Windows 10 is not listed as supported, nor is Cakewalk SONAR, I figured my first task was to test the latest version of SONAR on the latest version of Windows 10. And of course, there was no sound—until I had the presence of mind to go into the Console software and select the headphone output. A couple minutes later, I found my way around the console routings, which we’ll get into later. For someone raised on hardware, they make sense; I didn't need to look at the manual for the signal flow to seem logical. Those whose recording career started “in the box” might be momentarily confused, but that’s what documentation is for.




The Console software takes up the center, and is bracketed by SONAR's faders on the right and left. The 1176SE compressor is floating over everything.


Now I gotta say, I was struck immediately by the headphone amp sound quality. The higher price compared to standard interfaces buys you some serious converters. The sound quality is very, very impressive and the level of audio detail is exceptional. It may be a cliché to say it makes listening to music a pleasure, but truly good sound really does massage some kind of pleasure centers in your brain.


I should also discuss why the kind of sound quality is particularly important with UA’s design philosophy. When I told a friend I was excited about checking out the Apollo Twin USB, he said he’d tried an Apollo, but that the sound was too “clinical” and didn’t have “character.” What he clearly didn’t understand is that many of UA’s plug-ins are intended to emulate particular channel strips and processors. As a result, it’s essential that the hardware produce a truly blank, neutral canvas. Any coloration is going to reduce the effectiveness of the emulations, which themselves provide the “character.” To use an analogy, I might like light blue as a color—but if I was doing an oil painting, I wouldn’t want the canvas to be light blue. If I wanted light blue, that’s what I’d paint on a white, neutral canvas.


To wrap things up for the evening, I inserted the 1176 SE and—success. So I think I’ll listen to music for a while, and tomorrow, we’ll discuss latency and why USB 3.0 is a whole lot better than USB 2.0.



Edited by Anderton
  • Like 1
Link to comment
Share on other sites

  • Members

Why USB 3.0?


When we play virtual instruments, we don’t want a delay between the time we hit a note and the time we hear it. The same is true when processing an instrument like guitar through amp sims and other effects. This delay is called latency, for reasons that escape me ("delay" seems fine, but hey, what do I know?). In my experience, drummers are very sensitive to latency issues not just because timing is everything, but because they play in such close proximity to their instrument. Sound travels at approximately 1 foot per millisecond, so with a drummer’s ear being about 3 feet away from a snare drum, that’s 3 ms of latency. On the other hand, guitarists typically play several feet from an amp, so a latency of 6 to 9 ms is not uncommon and we think nothing of it.


fetch?filedataid=117507Computer audio systems have latency because you need to convert your analog signal into digital, process it within the computer, then convert it back to analog again. With a 44.1 kHz sampling rate, these conversions take about 1.2 ms. However, with computers, there’s much more going on. In addition to converting your “analog world” signal to digital data, driver software has the job of taking the data generated by an analog-to-digital converter and inserting it into the computer’s data stream. Furthermore, the computer itself introduces delays because even the most powerful processor can do only so many millions of calculations per second. Besides it’s paying attention to a lot more than your audio, like scanning its keyboard and mouse, checking ports, moving data in and out of RAM, sending out video data, and more.


As a result, the computer places some of the incoming audio from your guitar, voice, keyboard, or other signal source in a buffer, which is like a “savings account” for your input signal. When the computer is so busy elsewhere that it can’t deal with audio, it makes a “withdrawal” from the buffer instead so it can go deal with other matters. The larger the buffer, the less likely the computer will run out of audio data when it needs it. But a larger buffer also means that your instrument’s signal is being diverted for a longer period of time before being processed by the computer, which increases latency. When the computer goes to retrieve some audio and there’s nothing in the buffer, audio performance suffers in a variety of ways: You may hear stuttering, crackling, “dropouts” where there's no audio, or worse case, the program might crash.


One of the bottlenecks is how fast your audio interface can transfer data to and from your computer, and this is why USB 3.0 is important—its so-called SuperSpeed transfer rate can transfer data about 10 times faster than USB 2.0.


On the Mac, Thunderbolt is the high-speed protocol of choice, but it’s expensive and has yet to gain significant traction with Windows. However USB 3.0 comes very close, and USB 3.1 (not in widespread use) has a theoretical maximum speed that’s about the same as the original Thunderbolt protocol.


Like any protocol, USB 3.0 is not without its issues, mostly involving compatibility with (much) older chip sets. If you’ve used FireWire, you know the drill—interface web sites typically have recommended chip sets, but sometimes you’ll run into chip sets that just won’t work. (My desktop computer is in a sort of twilight zone where some USB 3.0 devices work and some don’t, but I added a PCIe USB 3.0 card that seems to work with pretty much everything…we’ll see if it works with Apollo Twin USB.) However this is clearly becoming less of an issue, as my circa 2012 HP off-the-shelf laptop used for the initial testing gets along famously with Apollo. I suspect Intel's NUC will be the same when we start assembling our Windows recording system.


So with respect to latency, how low can Apollo go? Let’s find out, and also compare it to USB 2.0. By the way, note that while some USB 3.0 interfaces will function using USB 2.0 but with degraded performance, Apollo Twin USB requires USB 3.0. I actually consider this a good thing, because it's all too easy to plug a USB 3.0 device into a USB 2.0 port and then you wonder why it's not as fast as you think it should be. That can't happen with Apollo Twin USB.

Edited by Anderton
Link to comment
Share on other sites

Now I gotta say, I was struck immediately by the headphone amp sound quality. The higher price compared to standard interfaces buys you some serious converters. The sound quality is very, very impressive and the level of audio detail is exceptional. It may be a cliché to say it makes listening to music a pleasure, but truly good sound really does massage some kind of pleasure centers in your brain.


I should also discuss why the kind of sound quality is particularly important with UA’s design philosophy. When I told a friend I was excited about checking out the Apollo Twin USB, he said he’d tried an Apollo, but that the sound was too “clinical” and didn’t have “character.” What he clearly didn’t understand is that many of UA’s plug-ins are intended to emulate particular channel strips and processors. As a result, it’s essential that the hardware produce a truly blank, neutral canvas. Any coloration is going to reduce the effectiveness of the emulations, which themselves provide the “character.” To use an analogy, I might like light blue as a color—but if I was doing an oil painting, I wouldn’t want the canvas to be light blue. If I wanted light blue, that’s what I’d paint on a white, neutral canvas.




I think they've made the right choices with respect to that. If you want coloration, you have all kinds of excellent options via he plugins, but you can't take something with coloration and record neutrally with it. Best to start with a clean and neutral foundation that can later be manipulated as desired.


Link to comment
Share on other sites

  • Members

Apollo Twin USB: Latency Tests


Let’s see how Apollo Twin USB performs in the real world.


In the previous post, we mentioned the sample buffer that computer systems use to store audio temporarily. This is usually specs as either a number of samples, a certain number of milliseconds, or both. However while this contributes to latency, it’s usually the least important factor in modern systems. USB “safety buffers” on the way in and out typically contribute more. So, the most realistic latency figure is called “round-trip latency”—the time between sending a signal into the interface, and having it appear at the output. This takes all elements that can contribute to latency into account.


When I first looked at SONAR’s preferences (which provides a readout of round-trip latency, and breaks it down to input, output, and sample buffer latencies), I was surprised that the latency was over 10 ms at 64 samples—I’ve done better than that with other USB 3.0 interfaces. Ah, but then I remembered Apollo’s Console Settings includes an option for setting Input Delay Compensation.




In a nutshell, this allows for audio to line up when using analog and digital processing at the Console input. Most DAWs have a similar feature so that if you’re using something like a limiter with (for example) 2 ms of look-ahead, the other DAW tracks will be delayed by 2 ms so that they line up with the track including the limiter.


Once I turned off Input Delay compensation to level the playing field, the measured latency at 44.1 kHz with a 64 sample buffer was 1.5 ms for the buffer, 4.8 ms for the input, and 3.6 ms for the output, giving a total round trip latency of 8.4 ms.







My dividing line for latency is 10 ms—anything below that doesn’t really matter to me (FWIW with 10 to 15 ms I can still cope, 15 to 20 ms is questionable, and over 20 ms is annoying). 8.4 ms is definitely better than a good USB 2.0 interface (with 64 sample buffers I get 10.2 ms), however there’s a major caveat: you have to convince your USB 2.0 interface to operate reliably at 64 samples, which may be true only for simple projects. As you start adding tracks and plug-ins, you may need to bump up the sample buffers to 128 or 256 sample buffers, for roundtrip latencies of 13.5 ms and 21.5 ms, respectively. So far Apollo Twin USB has been able to handle some pretty complex projects while remaining at 64 sample buffers. I wouldn't be surprised if it could go down to 48 for a lot of projects, but that option isn't available.


However at 96 kHz, things start to get really interesting. Here the roundtrip latency with 64 sample buffers is 3.8 ms.






Think about that for a second…if you’re in a studio and listening to monitor speakers four feet from your ears, listening to Apollo Twin USB through headphones will give less latency than listening through speakers! To put that another way, if you were to split a CD player output, feed one split to your speakers, and then feed the other split into Apollo Twin USB’s input, even after having it pass through the computer and back through Apollo Twin USB’s output the latency would still be less listening on headphones than listening over the speakers.


So the bottom line is…USB 3.0 does indeed make a difference.



  • Like 1
Link to comment
Share on other sites

  • Members

Zero-Latency Monitoring? No, Zero-Latency Processing


You’ve probably heard of zero-latency monitoring as a way to circumvent latency issues. This passes the input signal directly to the audio output for monitoring, thus bypassing the latency added by USB buffering, sample buffering, and the computer itself.


That’s the good news. The bad news: If your sound depends on plug-ins—for example, you’re playing guitar through an amp sim—you won’t hear the processed sound because you’re hearing the signal going into the computer, not the sound coming out.


And this brings us to one aspect of what UA’s Powered Plug-Ins are all about. The reason why Apollo costs more than your average interface (aside from the gorgeous industrial design, of course!) is that it includes on-board DSP built around Analog Devices SHARC processors. Thus, Apollo is not just an interface, but a signal processing sub-system under your computer’s control, via the Console software.


For example, suppose you load up the Softube Marshall amp emulation. You can send your guitar directly into the amp, without going through the computer, and hear all the processing the amp has to offer—while simultaneously recording it into a track in your DAW. There is essentially no latency, other than A/D and D/A conversion, which is under 2 ms (I think the delay caused by conversion alone is around 1.2 ms at 44.1 kHz, but don’t quote me on that—we need a digital guru to confirm). So to put that in real-world terms, that’s the amount of delay you’d have if your ear was about a foot and a half away from your speakers. Would you notice any latency? Of course not, and you don’t here, either. What’s more, the computer latency could be significant, but you’d still not hear any latency because it has nothing to do with the audio sample buffers in your computer.


The screen shot shows what’s going on. SONAR users will see this and think “Wait a minute, he’s recording a guitar in the lower track, but Input Echo isn’t on.” And that’s the whole point: I’m hearing the guitar in near-real-time via the console software. Just to make sure I had things set up correctly I did enable Input Echo at one point, and sure enough, I could hear comb filtering as the delayed signal mixed with the Console signal.




By the way, the Marshall plug-in UI is huge, so you’re only seeing the “channel strip” part that lets you choose your mics and vary their balance. I’m currently running the amp on its 14-day demo, so that should give me enough time to do some recordings so you can hear what it sounds like. Note that this plug-in is not included with Apollo Twin USB; it’s optional at extra cost. However when you download the software, you automatically download trial versions of all UA plug-ins, which you can activate at any time for a one-time, full-function 14-day evaluation.


My understanding is that the latency for “zero-latency processing” doesn’t increase as you add more Powered Plug-ins because again, they’re running in hardware, not software. So you wouldn’t get more latency any more than you would if you had four audio processors with S/PDIF I/O and hooked them in series: You’d have the A/D conversion at the input, and the D/A conversion at the output.


Also note that you can use the UA plug-ins conventionally in your DAW, like any other plug-in. I’ll have more on this later but I will say that despite UA officially supporting only a limited selection of DAWs, so far everything I’ve thrown at SONAR works just fine. I’m going to try some other DAWs shortly to confirm whether the plug-ins work there as well.


Before signing off on this post, a comment. As a computer-addicted guitar player, I’ve gotten used to latency over the years. Having spent a lot of time on stage playing through an amp that was several feet away, I was used to latency anyway but when computers were finally fast enough to go below 20 ms of latency, I felt we were getting somewhere. I don’t complain with 10-15 ms of latency because hey, that’s part of dealing with computers.


The first time I tried a Thunderbolt interface (the Focusrite Clarett on my MacBook Pro) and started getting sub-4 ms round-trip latency, I could definitely ”feel” the difference. It wasn’t enough to make me jettison Windows, because fortunately USB 3.0 came along just in time. Frankly (which may not be something UA’s marketing department would like to hear), even when used conventionally Apollo Twin USB’s latency doesn’t interfere with my enjoyment of playing guitar or keyboards through a computer—anything below 10 ms is fast enough for me. But using the Powered Plug-Ins’ talent of giving you no latency is something else, as now you really do have a real-time playing experience. (And note that unlike UA, which calls using Powered Plug-Ins a near real time playing experience, I just called it real time. One or two ms of latency is near real time, so technically UA is correct; but in reality, I don’t know anyone who puts their head one to two feet away from a speaker, so you are getting a real-time experience.)

Link to comment
Share on other sites

  • Members

Cool info Craig. I'll be watching the progress. I've been half glancing at the Twin for a while. Even though a tiny bit off-topic, when you get to the Marshall testing, I'd like to know how well Softube have nailed old jmp heads. If they even were trying to get that happening.


Being from the old days, I'm thinking I'd often use the Apollo with stuff like Studer or Neve plugs dialed in at the front end so that my incoming signal is committed and printed to that sound right away. Purposely ...so I wouldn't try guessing at a zillion possibilities later.





Link to comment
Share on other sites

  • Members
Cool info Craig. I'll be watching the progress. I've been half glancing at the Twin for a while. Even though a tiny bit off-topic' date=' when you get to the Marshall testing, I'd like to know how well Softube have nailed old jmp heads. If they even were trying to get that happening.[/quote']


I'll do some audio examples, and you can judge for yourself :)


Being from the old days, I'm thinking I'd often use the Apollo with stuff like Studer or Neve plugs dialed in at the front end so that my incoming signal is committed and printed to that sound right away. Purposely ...so I wouldn't try guessing at a zillion possibilities later.


Well you may or may not be able to resist temptation...one of the cool aspects of Apollo Twin USB is you don't have to print with the effects, although of course you can. You can simply hear what they would sound like, with AFAIC zero latency, but apply the effects later.


Link to comment
Share on other sites

  • Members

My First Apollo Twin USB "Gotcha"


Pilot error is always fun...I was getting intermittent operation with the AC adapter today. I thought maybe it was defective, or my barrier strip had a dead outlet, or something. The moral of the story is to remember that the AC adapter plug that goes into Apollo Twin USB is a locking type, you don't just plug it in randomly - there are little flanges that require plugging it in properly, and giving it a half-turn to lock it into place.


I believe this may be a CCPF (Curious Cat Prevention Feature).

Link to comment
Share on other sites

  • Members

The Console


Now that we've taken a first look at the performance and found out that yes, USB 3.0 makes a difference, let's take a look at the Console software.




The above screen shot show the basic configuration: two analog inputs (mic, line, or instrument), S/PDIF digital input, and two “virtual” inputs (more on these later). The top of the Analog channel strips duplicates what's on the front panel: mic/line switch, highpass filter, +48V phantom power, pad, phase (polarity) flip, and link, which gangs channel pairs (e.g. Analog 1 and Analog 2) in stereo instead of treating them as two mono channels. The communications between Apollo Twin USB and the Console app is bi-directional. For example, if you enable the front panel pad switch, the Console pad button follows and vice-versa.


Incidentally when you do switch, you’ll hear a click. This implies that the switching is done by relays, which guarantees no signal degradation because you’re dealing with what’s essentially an electronically-controlled mechanical switch, not electronic switching via FETs or CMOS integrated. Don’t get me wrong; the quality of semiconductor switching can be excellent…but there’s nothing better than a relay.


If you choose ADAT optical input instead of S/PDIF, the mixer grows 8 more inputs at 44.1 or 48 kHz (choosing 88.2 or 96 kHz restricts ADAT to four channels, while 176.4 or 192 kHz yields two channels of ADAT). While eight more channels starts to get unwieldy, UA gives you plenty of view options. You can show/hide individual inputs, hide aux returns, choose an overview that shortens the fader throw somewhat to see both inputs and inserts, or choose just inputs, inserts, or sends to have the longest possible throw on the faders. These views also have keyboard shortcuts.


Furthermore, there are flyouts. For example, suppose you want to see Analog 1’s sends without choosing the view that shows all sends at once, click on the overview’s Sends thumbnail for Analog 1 and voila—a flyout with a “close-up” on the sends:




You can also do the same for inserts, where you can insert up to four of UA’s Powered Plug-Ins. Each slot has an enable/bypass button and “replace” button for choosing a different plug-in:




You can also call up presets that populate multiple inserts at once, including several "signature sounds" from various audio luminaries.




The input stage also lets you open up the “Unison” option for calling up “Unison-enabled” plug-ins. Now if you’re a synth player…no, this doesn’t stack all the voices on one key! Instead, it provides control from the plug-in over the input stage’s analog characteristics. It’s getting late so we won’t take this any further for now, but because we’re still in “input-land,” tomorrow let’s look at what Unison is all about—it’s very clever.




Meanwhile, here’s one more observation before signing off: I was pleasantly surprised how effortless it was to change sample rates. With SONAR, some interfaces require closing SONAR, or closing and re-opening a control panel, or some other gymnastics. I thought that was just “part of the deal” but with Apollo, when I changed sample rates on SONAR, the Console followed right along without a second thought. It balked only when I wanted to see what would happen if I selected 384 kHz, which Apollo Twin USB doesn’t support. SONAR just said “It’s not supported” and reverted to the Console’s current sample rate. This is quite slick, and since I assume this is a property of the drivers, then I assume you’ll get similar results with any DAW.

Link to comment
Share on other sites

  • Members

Cool. The Unison feature on input is something I want to be sure to get a handle on. Even though the focus here is the overall interface, I hope UAD pops the new Fender Tweed Deluxe over for you to incorporate into even just a tiny part of the review :)

Link to comment
Share on other sites

  • Members

Unison Mode


So here's the deal. UA accesses parameters within the hardware preamps that control gain and impedance so that even if you're using digital control, you're not changing the digital data stream, but rather, the analog hardware itself.


Clearly some people at UA are guitar players :) because an amp's impedance makes a big difference on the sound of a guitar plugged into it. Most "DI" instrument inputs are high impedance, so any effects caused by an impedance change either need to be handled externally to the input (e.g., MOTU's zBox DI, the "Drag" control on Radial Engineering preamps) or emulated digitally. However, the interaction between input impedance and a passive magnetic pickup is quite complex. The pickup has resistance, reactance, and capacitance, all of which interact with the impedance. While digital emulation can approximate how impedance will effect a pickup, unless they're modeling is based on your pickup, it won't be the same.


But remember that the cable you use will also be a factor, and that's not emulated. So if you want the same sound you get on stage going through one of the amps that UA models, then you need to use the same cable you use on stage if you want to be a stickler for accuracy.


Ditto gain staging, although the effects tend not to be so dramatic. If you have two gain controls in a preamp, turning up the first one while turning down the second - or turning down the first one while turning up the second - may produce the same level, but the character may not be the same.


Note that UA has also built some user interface goodies into the Apollo Twin USB when working in Unison mode so you don't get too lost, and are aware when you are controlling these physical parameters. Rather than describe them all, UA has made a video which explains the topic pretty well.




However, I was confused when the video said that the 500 ohm impedance setting gave a brighter, louder sound with a mic than the default 2 kOhm setting because there was less mic loading. That certainly goes against my experience, where it takes a higher impedance to produce less loading. Maybe the setting is 500K instead of 500 Ohms? Or maybe the labels are reversed...or maybe the star system where I was raised has different laws of physics. Anyway, that doesn't detract from the feature itself.


So is this some game-changing mega-feature that will cause other interface companies to fold up their tents and say "We don't control analog with digital, might as well give up"? I don't think so, because the differences are subtle. However UA is all about subtlety, and this kind of attention to detail is one more reason why UA is considered a master of its art. UA seems to believe something that I believe as well: every little dB adds up, and a chain is only as good as its weakest link. While I don't think anyone considers UA's preamps a "weak" link anyway, the Unison mode makes it a stronger link.

Link to comment
Share on other sites

  • Members
Cool. The Unison feature on input is something I want to be sure to get a handle on. Even though the focus here is the overall interface' date=' I hope UAD pops the new Fender Tweed Deluxe over for you to incorporate into even just a tiny part of the review :)[/quote']


I'm downloading the 8.6 software now, which has the Fender Tweed...I was on 8.5.2, which doesn't have it. I don't know if it's a Unison-compatible amp (not all of them are), but I'll find out. I'll fire up the demo once I have everything installed.

Link to comment
Share on other sites

  • Members

Virtual Console Output Section


I’ve mentioned previously that the Big Deal with these inserts is that the plug-ins are in series with the input path, so you can route the input through the plug-ins in real time. Although the obvious use is recording with effects, or auditioning inputs with effects, don’t forget that this has a lot of potential for live performance. For example, if you use something like an Ableton Live with a laptop for live performance, you can do what you normally do with Live as well as feed in mics or other instruments. With the continuing trend of combining conventional instruments with live EDM setups, real-time effects is a big deal.


While we’re on the subject of inserts, although I typically open up an effect’s interface as needed, there’s also a “Channel Strip View.” This stacks all the effects in one scrollable window so you can move from one to the other easily for editing, with an additional option of picking the insert effect you want to edit from a list and seeing only that particular effect.


Let’s transition from the inputs to the outputs via the two Aux bus returns (outlined in red), which also have inserts for effects. Below the inserts are monitor (cue) controls for the headphone and Line 3/4 outputs. Note that this also gives you a physical aux bus of sorts - you can send signal to a headphone jack, pull out of that to go into something like a classic hardware reverb, then return that into the analog inputs.




One cool feature at the inputs is if you right-click on a fader or panpot to copy the respective mix to the headphone, 3/4 outs, or either or both of the two aux mixes. This is really convenient if, for example, you're monitoring while overdubbing; you can copy the mix to the headphone out, then tweak the headphone cue levels to bring individual inputs up or down.


Moving along to the console's output section itself, metering is important and all Console meters have global adjustable peak and hold times of 1, 3, 5, or 10 seconds (clip hold can also be set to infinite, so you can tell clipping occurred even if you look away when the clip occurs). The Control Room section (outlined in pink) has the expected options...“dim” to reduce levels (e.g., when you need to take a call but the client wants to keep hearing what's happening) and monitor source.


As to the Monitor strip to the right, you have the main meters at the top. The Insert Effect switches don’t relate to inserting plug-ins on the master output, because you can’t—which would be helpful for a live situation if you wanted to process an entire mix. Instead, these switches determine whether inserts are routed to your DAW (UAD Rec) so you can record the signal with effects, or recorded dry but you can monitor the effects (UAD Mon). If both are selected, whether an effect is recorded on monitored depends on the related switch for a given effect.


Below that, the Show buttons choose show/hide for the Aux returns and Control Room. Next up are the Cue options (clicking on it opens the dialog box outlined in yellow). You can also sum to mono and/or mute the output; the big output knob's ring turns red if you select mute. This simple idea should be applied to all software mixers in all DAWs to avoid those “why am I not hearing anything?" situations. At the bottom, you can save and load “sessions,” which are basically console presets.

Link to comment
Share on other sites

  • Members

Console Recall


This is another slick feature: a VST/AAX plug-in for Windows systems (the Apollo interfaces that support Mac also have an AU version) that lets you store console settings within a DAW’s project. You can stick the plug-in pretty much anywhere; the master bus is a logical place but you’re not limited to that. It basically syncs the DAW to the Console settings.




An analogy (although it doesn’t work this way) is how you can store MIDI sys ex in most projects, and play that back when you open a project to restore settings of MIDI gear.


When I first installed the UA software, SONAR couldn't "see" the Console Recall plug-in. It turned out that Apollo installed the 64-bit version in C:\Program Files\Steinberg\VSTPlugins\Powered Plugins, which is also where the other powered plug-ins install (the ones you can use as insert effects in your DAW). Once I included this folder in SONAR’s VST scan path, everything showed up as it should. I didn’t move the plug-ins to my usual plug-in folder just in case that screwed up some kind of file path association that Apollo wanted to see, but I assume you could do that.


If for some reason your program doesn't recognize the Console Recall plug-in, you can search on Recall.dll using File Explorer to find it.


The Console software has lots of other little features and options, but this is a review, not a manual! We’ve covered the high points, so it’s time to move on to our next topic.

Edited by Anderton
Link to comment
Share on other sites

  • Members

Powered Plug-Ins: The Backstory


Apollo Twin USB includes several basic plug-ins, but these aren’t VST; they’re in a proprietary format Universal Audio calls “Powered Plug-Ins" that are designed to run on the Universal Audio's hardware DSP, whether on a card or built into Apollo Twin USB. You might think the concept of dedicated hardware may seem quaint in this age of super-fast native computers, but there’s a good reason to supplement a computer with DSP.


Although some people assume powered plug-ins will make your computer run faster, that’s not strictly true. However, it may seem snappier if you use external DSP because you’ve offloaded processing tasks from your CPU. An analogy would be if someone is carrying two heavy suitcases and you offer to carry one. The other person won’t go faster because of suddenly growing stronger muscles, but because of not being loaded down by the other suitcase.


Back in the days of Windows 98 and Pentium I processors, no computer had the power to do decent native processing. A German company called Creamware packed a bunch of SHARC DSP chips (made by Analog Devices) on a card to handle audio, virtual instruments, and effects; and of course Digidesign (now Avid) specialized in supplementing hardware with software via Sound Tools and later, Pro Tools.


When Universal Audio introduced their UAD-1 DSP card, computers had become more powerful but still lagged behind the needs of audio professionals. By having a dedicated DSP card, Universal Audio could throw processing cycles at it and not have to compromise their plug-ins, which would otherwise be drawing from the same CPU as the DAW software that was running tracks, following automation, feeding soft synths, and also doing housekeeping like scanning the keyboard and mouse, communicating with hard drives, etc. etc.


A more subtle advantage was that the amount of DSP in a UAD-1 was fixed and you could count on it. If you were running it at 99% capacity, it would work and wouldn't crash. With native processing on your computer, CPU power is always a moving target. When you first start a project, you may be able to get very low latency with several plug-ins. As you add more virtual instruments and effects, the computer has to work harder, but the load on the computer fluctuates. You may be using 60% of your CPU power on average, but it could spike up to over 100% occasionally when too many functions happen at the same time - and your computer may very well freeze or even crash. There are many workarounds: freeze or bounce tracks, remove plug-ins temporarily, turn off plug-in oversampling if possible, etc. Although companies try to produce plug-ins that are at least somewhat CPU-friendly, the compromise is that they have to scale back on how far they can pursue the sound quality.


As processors became more powerful, many people (however, not including me!) predicted that products like UA's DSP cards would become obsolete because eventually, it would all be done inside the computer. And often you can do it all inside the computer; yet UA is doing better than ever.


That may seem counter-intuitive, but as processors became more powerful, designers couldn’t resist taking advantage of that power with more detailed and realistic algorithms. So not only has the idea of powered plug-ins not become outdated, but now, being able to monitor through plug-ins at near-zero latency adds yet another element that native systems can’t duplicate. Maybe they will someday…but for now, there are some real and practical advantages to supplementing software with hardware. Let's look at the plug-ins that come with Apollo Twin USB.

Link to comment
Share on other sites

  • 2 weeks later...
  • Members

The LA-2A Legacy Compressor/Limiter


Unfortunately taking a road trip always puts a bump in the road for a pro review, but let’s get back into the swing of things by checking out the bundled LA-2A Legacy compressor/limiter. This is of course one of the best-known, and probably the most emulated, of all the classic compressors (it also does limiting, and while useful, I think most would not consider that the LA-2A’s main claim to fame).


Note that the legacy version has limitations that don’t exist with UA’s later LA-2A emulations. The Emphasis setscrew, which determines whether highs are compressed more when turned clockwise, is locked to the default position where all frequencies are compressed equally. Also, the plug-in doesn’t oversample. However, when used as a VST plug-in within SONAR as opposed to an insert effect in the console, you can use SONAR’s upsampling feature—I tested it, and the plug-in was more than happy to be oversampled. Finally, the Legacy version doesn’t emulate the transformer and input/output distortion. Whether this is a positive or negative is up to you, but one upside is the plug-in uses less DSP, so you can load up more of them.


The controls are about as basic as you can get: Peak Reduction sets the threshold that determines the compression amount, while Gain makes up for the reduced output level that’s inherent in compression.






However, despite this simplicity—or perhaps because of it—those who were raised on “modern” compressors with their multiplicity of controls sometimes have a hard time getting a natural sound out the LA-2A, and they just assume this lack of subtlety is part of its “vintage” quality. So rather than wax eloquent about the “smooth, analog-style compression (which is indeed an accurate assessment), here’s the LA-2A technique I’ve been using since the days of tape for gentle compression that gives program material a lift, or provides what some might call that elusive “glue” quality.

  1. Set Peak Reduction to 0.
  2. Turn up Gain until the output reaches the maximum desired level.
  3. With the meter switch (upper right) set for Gain Reduction, watch the meter until you start seeing about 2 dB of gain reduction (and certainly no more than 3 dB).
  4. Turn up the Gain to match the level attained in step 2.

That's it. You probably won't hear an obvious effect, however, if you turn off the “power” switch to bypass it the uncompressed will probably sound just a bit more “distant.” However, note that due to the attack time, you might go "over" from time to time. A tight limiter following the output that doesn't limit, but rather traps transients, will let you get away with more compression and gain with the LA-2A.


That said, the LA-2A also works well with large amounts (and hey, sometimes obscene amounts!) of compression, especially (at least to my ears) on vocals and bass. I’ve yet to hear a vocal of mine that didn’t sound better after being processed by an LA-2A-style compressor, and despite this being the “legacy” version, when you want smooth compression it will do the job...and then some.

Link to comment
Share on other sites

  • 3 weeks later...
  • Members

UA 610-B Tube Preamp and EQ


First, major apologies for the delay in doing additional posts. This time it wasn’t just my travel schedule, but I also moved into a new home and the last of the move happened yesterday. So I’m set for doing more pro review posts, at least until GearFest kicks in…


The 610-B is another plug-in that’s included with Apollo Twin USB, and provides the sound quality and behavior of a tube preamp. This includes the transformer component as well as various phase shift and non-linearities. The 610-B features UA’s “Unison” mic preamp technology when used with the appropriate Apollo Twin USB input, so that it incorporates the subtle alterations that occur when using the mic, line, and Hi-Z inputs; much of this due to matching impedances properly to the original hardware. The plug-in also upsamples internally, which increases latency more than some of the other plug-ins.


Although UA points out today’s users typically use the 610-B for vocals, bass, horn, or strings, and it is useful in those contexts, I’d advise going further. For muchof my testing, I tried the 610-B on complex program material to hear what it would do…more on this later.


Here's the 610-B inserted in an input, using Unison mode, which was my starting point for testing with guitar.






The top control sets the tube input stage level, in five switch-selectable positions that range from 10 dB of attenuation to 10 dB of gain. Below that, the EQ section provides +/- 9 dB boost/cut in 1.5 dB stepped increments for low and high frequencies, with those frequencies set by the toggle switch above the respective knobs. This is a gentle, “character” EQ as opposed to having the surgical qualities of something like a parametric.


Going further down brings us to additional preamp controls. The left toggle switch chooses a 500 ohm or 2 kilohm input impedance when using the mic input, or the line input. The middle toggle provides a pad that’s active when using the mic input, and the right toggle flips polarity.


The “big knob” level control affects the output stage gain, so you can stage gain at both the input and output. Below that, the smaller output control trims the output level but without affecting the “character.”




Generally, “character” preamps and channel strips have a subtle effect. However with the 610-B, the gain-staging allows creating more obvious effects. The question for me was whether “cranking the color” a bit more would make the sound better or worse.


The following screen shot shows settings for adding saturation that was very beneficial with program material. Note cranking the input and output gain a bit. After matching levels carefully between the bypassed and enabled settings, enabling the saturation added a bit of “sparkle” and presence that definitely added some worthwhile mojo to the overall sound. I would not hesitate to use this plug-in for mastering on suitable material; with complex rock tracks, the extra saturation actually made the overall sound more, not less, defined.




Due to the emphasis on mic and line operation, I was curious what happened when using the Hi-Z input. Basically, it just meant the mic-specific controls (impedance and pad) had no effect, otherwise everything else worked as expected. I tried it with one of Gibson’s new HP Standard guitars, and while you could crank it for “tube” distortion, as I’ve often said what guitarists associate with a “tube” sound is really more about the amp system, where the cabinet and speakers form a sophisticated filter to tame the tube distortion. I’d rather use one of the amp sims instead of the 610-B to get a guitar’s “sound” however if you just want to add a preamp and some gentle EQ, the 610-B will do that very well with guitar.


I was somewhat surprised that the 610-B was so effective with program material, because I often feel “character” preamps and channel strips make a sound that’s “different,” but not necessarily “better.” In this case, it was better. Overall, this is a useful plug-in that adds value to the Apollo Twin USB package.




Link to comment
Share on other sites

  • 4 weeks later...
  • Members

The "Raw" Distortion Plug-In


This is another plug-in that comes with the Apollo Twin USB's software suite, and emulates the Pro Co Rat distortion. Like almost all fuzzboxes, the Rat has a fairly aggressive, raw sound that's normally "tamed" by running it through a guitar amp. Although you can do the same thing by adding a cabinet plug-in, the Precision Channel Strip EQ can tame the raw sound effectively by tightening it up and giving it more "focus."




The Raw plug-in's Filter control already has good tone-shaping abilities, but as a single control it can't tailor the sound with the same precision as a multi-band EQ. The Distortion control determines the fuzz intensity, which you can really crank up for leads, or dial back a bit for chords.


The Precision Channel Strip includes both EQ and dynamics, but to provide the most "honest" comparison, we'll use only the EQ in this example. There are five bands; the middle one is a traditional parametric stage with variable Q. The two lower bands have frequency and gain controls, but instead of Q, the response changes from low shelf to peak ("bell") to high pass. The two upper bands are similar however the response changes from high shelf to peak to low pass.




Because of the way the two high and low bands "double up," that means stacking the low pass or high pass filters allows for creating steeper rolloff slopes than just one band by itself. It also means you can do a "stepped" kind of shelf. I find these options very useful for guitar.


The audio/video example plays a guitar part through the Raw plug-in by itself, then the Raw plug-in followed by EQ (and explains why the EQ settings were chosen, and what they accomplish). As you listen to the second example, you might think it doesn't really sound all that different from the Raw plug-in by itself. But if you go back and compare to the beginning, you'll hear a definite difference: the EQ helps focus the sound, reduce high-frequency "fizz," and tighten up the low end.





  • Like 1
Link to comment
Share on other sites

  • 3 weeks later...
  • Members

RealVerb Pro


This is the reverb bundled with Apollo Twin USP, and it's unique not just among Universal Audio's reverb offerings—some more specialized than others—but among reverbs in general. The RealVerb Pro’s goal in life seems to be the ability to synthesize a variety of different rooms and spaces, as opposed to emulating classic reverbs or being a convolution reverb. It’s been around for a while, but that doesn’t diminish its usefulness.


As such, it’s very versatile. The “Big Deal” controls here are Shape and Material, because they’re what give the most precise control over the nature of the reflected sound. The 36 room materials include “real-world” options like hardwood, audience, brick, air, cork, carpet, and the like, as well as 12 artificial materials with fixed decays. You can adjust the reflective and absorptive properties, as well as use two materials to create a “hybrid” room (e.g., hard marble surfaces but with an audience inside).




The 15 shapes aren’t strictly acoustical spaces, but include plate and spring algorithms. Although you can use a single shape, like materials you can use two different shapes—and adjust their sizes—to create a blend of early reflections. One of the coolest features is when you morph between two materials or shapes, there’s no glitching, nor do you have to wait for the sound to “settle down” before you can evaluate the difference. However unless you know how materials like fiberglass and plywood affect reflected sound, you’ll need to spend some quality time with the documentation.


The rest of the options are somewhat more standard, including a three-band parametric to control the reveb’s overall response, although it can also be configured as a shelving EQ or an EQ with elements of both. A panel for Timing provides a graphic interface for adjusting the early reflections amplitude and predelay, as well as the reverb tail’s amplitude, pre-delay, diffusion, and decay time. If you were raised on convolution reverbs, you’ll find the flexibility of algorithmic-based reverbs a whole other world. While arguably not as “real” as convolution reverbs, there’s far more flexibility.


The Positioning panel determines the stereo spread and imaging for the direct, early reflection, and reverb tail components. There’s also the expected Mix control but also an unexpected Distance control, which sets the distance from the source sound.


Finally, there’s a control for morphing between two presets of your choice. Although this limits the number of user-adjustable controls, presumably you’ve already made your settings for the presets, and are more interested in exploring the various intermediate states.


If you just want to slap some reverb on vocals, RealVerb Pro will certainly do the job but to me, its main value is if you need to create environments for sound design work. If you need to hear footsteps in a garage or in a bathroom, you can do it.


At this point you probably want to hear the sounds, so let’s go for it. The following audio/video example steps through 21 presets so you can hear the reverb applied to a quick handclap transient, so there’s no decay in the source sound that could obscure the effect of the reverb. As the different presets change, check out what happens with the parameters—you’ll get a pretty good idea of at least some of what RealVerb Pro can do.





Link to comment
Share on other sites

  • Members

Softube Amp Room Half-Stack


Softube is of course well-known for their amp sims, and the bundled Apollo Twin USB software includes Amp Room Essentials. This consists of the Amp Room Half-Stack for guitar (the subject of this post) and the Bass Amp Room 8 x 10 for bass.




The big deal here is being able to play through these amps (and record/monitor them) with virtually no latency; but if you want to do VST or other automation, you can run them as standard VST2 plug-ins.


Amp Room Half-Stack's sound will be familiar to anyone who's played around with Marshall amps. Tone controls are Presence, Bass, Middle, and Treble, and there are also controls for Preamp and Master gain. The audio/video example starts with a Les Paul Traditional (neck pickup) playing through the default setting, but with a bit more preamp gain. However, note that you can also vary the mic position - distance from the cabinet, and when close-miking, you can also vary the mic angle.


Since this is all about the sound, check out the audio/video example and you'll get a good idea of the amp's basic sound.



Link to comment
Share on other sites

  • Members

Softube Bass Amp Room 8 x 10


This is also bundled with Apollo Twin USB, and of course like the Amp Room Half-Stack you can play through it with near-zero latency. Another element they have in common, which I forgot to mention in the previous post, is that you can bypass the cab and amp separately. So if you have a different amp sim where you like the amp but not the cab or vice-versa, you can use the half you want from the plug-in.


The Bass Amp Room is a bit more sophisticated as it includes both D.I. and amp sections, as well as the ability to blend between the two. The amp section is your basic bass amp (switchable between high and low gain channels); the D.I. has a fair amount of tonal flexibility, but also includes a limiter.




But if a picture is worth a thousand words, an audio/video example is at least worth a magazine...so here ya go.



Edited by Anderton
Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.


  • Create New...