Jump to content

Roland Octa-Capture Interface - Now with Conclusions


Anderton

Recommended Posts

  • Members

To say that Roland has been putting a full-court press on promoting this interface would be an understatement – they think they really have something on their hands, and this Pro Review is just one example of the confidence they have in the product. After all, any company that would expose their baby to the kind of scrutiny has to be pretty sure there aren’t any hidden issues.

But really, I’m not surprised. I’ve used Edirol interfaces in the past, which come from the same lineage, and was always curious why their drivers seemed a little faster and a little more solid than other interfaces of that time. I’m also using the V-Studio 700 currently with Cakewalk Sonar X1 (as well as Sony Vegas and several other programs), and while it has a few rough edges in terms of integrating the control surface with all the changes in X1, the interface section has been solid. (FYI – I’ve heard that the next X1 update will include optimizations for X1, so no worries there.) I was particularly impressed with the V-Studio mic pres, and from what I understand, the ones in the Octa-Capture are based on the same basic technology.

As is traditional with Pro Reviews, I like to start with a photo tour and with links to the product landing page as that will give you some background on what Octa-Capture is all about. Incidentally, I like the way the landing page isn't afraid to get all techy - it doesn't insult your intelligence with "The greatest achievement in the history of Western civilization!!!"-type lines, preferring instead to explain the technology behind the product. Nice.

BjHYS.jpg

The above image shows an overall view of Octa-Capture. This is a compact unit, measuring approximately 11.25” x 6” x 2”. It not only fits in a 1U rack space, but the package includes rack ears if you prefer not to go the tabletop route. Power comes from an external transformer, packaged as a “line lump” (not a wall wart) so the power cable takes up only one space on a barrier strip.

Octa-Capture cannot be bus-powered (I tried), which isn’t surprising given that the eight mic pres are all Class A types. If you’re not familiar with Class A technology, it is far less efficient than Class B or Class AB amplifier structures, which is why Class A is used for preamps, headphone amps, and other low-power applications while Class B and Class AB (as well as other classes) are used for power amps. However, the advantage of Class A is the total elimination of crossover distortion as there is no crossover, and exceptional linearity as the signal always remains in the linear portion of the amplifier.

If that’s too much information, the bottom line is that Class A provides the most accurate amplification of lower-level signals, with the tradeoff being more power consumption.

Now let's take a look at the front panel.

tKr6m.jpg

Note that four of the inputs are on the front, as shown in the following closeup of the panel's left side.

rQPlz.jpg

Another four inputs are on the back. This "split input" approach helps keep the size down, but it also means that more “permanent” patching can be on the rear, while the front is where you plug in devices as needed. The two left-most inputs can also serve as hi-Z instrument inputs.

Now let's check out the front panel’s right side, with the panel navigation controls.

Be2uH.jpg

Wait – why would you need that, given that there’s an included mixer applet? Simple: You can also use Octa-Capture stand-alone, which in conjunction with the small size, makes it a natural for live recording (or even for something like a keyboard mixer). We’ll be explaing what these are all about later on, but as you can see, there’s direct monitoring and a headphone jack.

Link to comment
Share on other sites

  • Replies 350
  • Created
  • Last Reply
  • Members

Now let’s turn our attention to the rear panel, going from left to right.

mVwiy.jpg

This shows the left side, which shows where the AC adapter plugs in, the USB 2.0 connection, MIDI I/O (yes, you also get a 1x1 MIDI interface), and the coaxial S/PDIF connectors.

Aep2A.jpg

The above image shows the middle of the unit, with the eight balanced outputs (two main, and six “auxes”). Now let's look at the right side.

IVlPg.jpg

This picture shows rear panel’s right side, with the additional four mic/line balanced inputs.

One obvious omission: No optical I/O (then again, I’m not sure where it would fit!). Many interfaces include an ADAT “lightpipe” connector so you can add another eight mic pres via a second unit that feeds the ADAT in, so this might seem like a problem if you need more than eight mics. However, there’s a twist here in that you can cascade two Octa-Captures, via digital sync, if you want 16 high-quality mic inputs. Those who bought mic pre-to-ADAT devices might be disappointed they can’t be used with the Octa-Capture, but to compensate, if you like the Octa-Capture you’re not limited to using only one.

Not a lot of companies can aggregate ASIO interfaces (the only ones I know of are Echo and CEntrance) so this a useful feature, assuming you can use both in ASIO mode, not just WDM. (Geek sidebar: Apple’s Core Audio allows for aggregation, as does WDM; it’s ASIO that’s hard to deal with.) I don’t have two here for testing, however, Roland claims that the Octa-Capture can expand the V-Studio interface so we’ll be able to test that – and that’s probably a more demanding application, as we’ll be combining two nominally dissimilar units instead of two identical ones.

Okay...that’s enough for today. We try installation next, and just to present as much of a challenge as possible, we’ll install under 64-bit Vista.

Link to comment
Share on other sites

  • Members

Well, someone at Roland must have taken grousing about their manuals seriously, because the instructions for installation are exceptionally precise - I almost expected to have a line in there saying “Please verify that your house is receiving electricity from your nearest power station before attempting to turn on your computer or install the Octa-Capture.” The irony is that installation was simple and painless, and I probably could have done it without any documentation at all. I just located the 64-bit drivers for Vista, double-clicked on setup, followed instructions...done.

Of course, before doing that I also checked whether there were updated drivers on the Roland site, but there weren’t (I guess this really is a new product!).

Anyway, installation seemed almost too good to be true in terms of how easy and fast it was, so I opened up Sonar X1 to see if the interface would show up. It did, was preselected, and ready to go. I loaded a loop, and it played using the WDM/KS drivers.

So I thought I’d experiment a bit. WASAPI was listed as one of the driver options, so I tried it as this protocol supposedly gives lower latency. The lowest latency I was able to get and retain rock-solid operation was 264 samples, or 6ms. Setting it lower with a loop sounded like it was okay at first, but the longer the loop played, the greater the odds of sound issues.

Then I investigated WDM/KS further. It also seemed happiest with 264; when I tried the next lower value of 132, playback worked without any catastrophes but there a few pops with file playback.

Next up, ASIO. This time I was able to get consistent operation at 128 samples, as well as with Sony Vegas (my other “true” 64-bit program installed on the same drive as Sonar). Granted that’s very respectable, but I was a little puzzled - I knew a Cakewalk clinician who was using a less powerful computer while doing clinics, and he was running an Octa-Capture with Sonar X1 at 96 samples.

The main difference between our systems was that he uses 64-bit Windows 7, and I’m currently using 64-bit Windows Vista. I talked to another friend of mine who’s a Sonar user but not involved with Cakewalk, and he basically said I shouldn’t take any latency specs seriously with any interface until I upgraded to Windows 7, which he said had far better performance with audio than Vista. Why am I not surprised?

So rather than pursue Octa-Capture’s "how-low-can-you-go" latency aspect right now, I’m going to wait until I upgrade to Windows 7. I already have a suitable hard drive for installing it, and the DVD itself should arrive in a day or two. Meanwhile, there’s plenty to discuss – like the signal flow, the mixer applet, the onboard DSP (including compression/gating and reverb), and more. To give you a little taste of what we’ll start to cover tomorrow, here’s a picture of the signal flow within the interface.

pSku2.png

Link to comment
Share on other sites

  • Members

You can get around the Octa-Capture user interface in a couple ways. The primary one is from the front panel itself, where you can access all functions. This employs the usual combination of pushing buttons, pushing knobs (two of the knobs have push switches), and turning knobs.

The front panel interface is neither particularly annoying or wonderful, but does make the selection process pretty obvious. As a test, I tried to see how far I could get in terms of programming advanced functions without looking at the manual, and to my surprise, I was able to dig deep into the functionality - from system settings, to channel settings, to setting up direct mixes. I also found that the more I worked with the interface from the front panel, familiarity came easily and before too long I felt totally confident that I could use the Octa-Capture for live recording and set it up fairly expeditiously.

The second way to work the interface is via a control panel that you can invoke within your DAW. The following illustration shows the page for the mic pres.

QYTWW.png

Not surprisingly, adjusting the parameters on-screen is easier than using the front panel. The main omission is that you can't adjust the Auto-Sens function (like a "learn" function for setting levels; more on this later) on-screen. However, someone at Roland was thinking ahead - adjusting Auto-Sens is one of the easiest-to-access front panel functions. Given the importance of level setting, particularly in a live performance context, being able to access it this easily is great. I wish it could be accessed from the applet as well, but the ease of adjusting from the front panel is definitely a mitigating factor.

Now take a look at the applet itself. You can see all eight channels at once, and a few cool features reveal themselves.

For starters, every channel has a compressor with attack, release, threshold, ratio, and gain controls, as well as an input gate. You can link adjacent channel compressors so that adjusting parameters on one compressor automatically adjusts parameters on the linked compressor.

One helpful visual feature is that the slider settings are visible whether or not the compressor is enabled, but when enabled, numeric values for the sliders appear. This makes it incredibly easy to see at a glance which compressors are active or not, along with both "ballpark" parameter values (the slider settings) and the exact numeric settings. Sure, you can also look at the compressor in/out switch, but that just tells you whether the compressor is active or not - the other info tells you exactly what the compressor is doing.

Also, note that you can kick the ratio all the way to infinity if you want to treat the input compressor more like a limiter to prevent overloads and distortion. While Auto-Sens takes care of for signals prior to the A/D converters, the compressor keeps levels under control "downstream" of the compressor.

Moving along, each channel can have +48V phantom power enabled individually. To me, this is a big improvement over interfaces where phantom power can be enabled only in groups of two or four inputs. Each input also has a low-cut filter and phase reverse. Finally, the first two channels can be configured as high-Z inputs for electric guitar, bass, etc.

It's also worth pointing out that the applet and interface communicate with each other, so if you vary the gain on the applet it shows up on the front panel display and conversely, if you change a parameter value on the front panel and that parameter is available in the applet, it will change there as well.

And that's it for mic preamp adjustments. There's a second page for setting up direct mixes (particularly useful for monitoring), and we'll cover that next.

Link to comment
Share on other sites

  • Members
Quote Originally Posted by Anderton View Post
WASAPI was listed as one of the driver options, so I tried it as this protocol supposedly gives lower latency. The lowest latency I was able to get and retain rock-solid operation was 264 samples, or 6ms. Setting it lower with a loop sounded like it was okay at first, but the longer the loop played, the greater the odds of sound issues.
If you are testing it under Vista 64, using the WASAPI driver, then you should test it with a few 32-bit apps. This is important because there is a known issue in Vista 64 whereby a 32-bit app using WASAPI doesn’t work; streaming never begins.

This has been an issue with Vista 64 from the beginning, and apparently Microsoft has no intention of ever fixing it.

Here’s my source – from Microsoft’s pro-audio developers’ forum.

http://social.msdn.microsoft.com/for...-ed4bec6c5b2b/

http://social.msdn.microsoft.com/For...8-c9f4ff6ebdf2

Quote Originally Posted by The horse’s mouth (Microsoft)
This is consistent with the known issue on WOW64 exclusive-mode event-driven streaming on Vista. The event handle is not being thunked correctly between the x86 app and the amd64 driver.
Link to comment
Share on other sites

  • Members
Quote Originally Posted by This View Post
If you are testing it under Vista 64, using the WASAPI driver, then you should test it with a few 32-bit apps. This is important because there is a known issue in Vista 64 whereby a 32-bit app using WASAPI doesn't work; streaming never begins.

This has been an issue with Vista 64 from the beginning, and apparently Microsoft has no intention of ever fixing it.
Thank you VERY much for saving me from frustration thumb.gif Actually I don't have 32-bit apps installed on my Vista-64 drive, I've been using it only to test true 64-bit apps. I still use XP for 32-bit apps, but will segue over to Windows 7 32-bit to replace that at some point...

I guess this is one more reason not to work too seriously on testing specs until I get Windows 7 installed. It was supposed to arrive today, but when I went to track it on the Fed Ex site, I got the following message: "High winds and thunderstorms at Memphis hub may cause some service delays and disruptions within the U.S. today" with a new estimated delivery date of March 2. Oh well. At least I can test the functionality of the interface itself - the dynamics, reverb, direct mix options, etc.

Thanks again for the heads-up. I'm sure that if the folks at Roland don't already know this, it will be very helpful info for when someone contacts their tech support people and says "Your interface sucks because it doesn't work with 32-bit apps in Vista." Of course, that assumes anyone out there is still using Vista... icon_lol.gif
Link to comment
Share on other sites

  • Members

Let's look at the Direct Mix functionality, which is what makes the Octa-Capture a mixer as well as an audio interface.

ofWrf.png

In a nutshell, there are four Direct Mix tabs (Mix A - Mix D). Each one has a 10-channel input mixer, which combines the various signals present at the inputs, and a 10-channel output mixer, which is fed from your DAW. Both mixers are available simultaneously so you can combine live inputs with DAW outputs, for example, to listen to DAW tracks while doing zero-latency monitoring of the inputs.

As to where these outputs go, you can assign each Mix to a specific output pair. For example, Mix A could go to outs 1-2, Mix B to outs 1-2, etc. What's more, one Mix can feed several outputs. The matching of Mix to output happens within the Patch Bay (overlaid on the main mixer interface), a modest pop-up window with drop-down menus to select which Mix feeds a particular output.

In the studio, these four mixes allow creating up to four different cue mixes for different musicians - the vocalist wants to hear more vocals, the bass player wants to hear more bass, etc. For live recording, you could ignore the Output Mixer entirely, and use the Input Mixer to set up four different monitoring scenarios for wedges or in-ear monitors.

Let's dig deeper. Of course, all channels on both the input and output mixers have volume and pan controls, but they also have mute and solo switches - very helpful when setting levels and such - and channel pairs have link switches, which makes life easier when you're recording stereo sources.

However, there are two main limitations. First, for the millions of people using 192kHz recording (kidding!!), only Mixer A can be active - B-D don't do anything. Second, Mixer A has send controls for feeding the onboard reverb, but not mixers B-D. In many cases, that won't be a problem with a little forethought; for example, it's generally vocalists who want to hear reverb in their headphones, so you'd just use Mixer A for creating the vocal mix. (Don't worry, we'll check out the reverb soon.)

However, perhaps the most important aspect to all this is that you can save complete setups that include all settings - preamps and direct mixes. If there's a particular setup you use for miking drums, save it. Narration? Save it. Recording guitars with amps and a direct signal? Save it. Granted, sometimes there will be session-to-session variations that require tweaking your settings, but in any event you've have a ballpark point of departure that will save you time.

I don't have to speculate that this is useful, because the V-Studio 700's eight preamps have the same basic feature set, and I've been taking full advantage of being able to store particular setups.

Now, you might think "Okay, I can see where this would be handy for miking drums and whatever, but I'm just a singer-songwriter and never use more than a few inputs at a time. I don't need eight inputs, let alone the ability to store them." But what I've found is that I can leave everything pretty much set up, normalized, and ready to go, with specific saved setups for different applications. For example, I tend to use dynamic mics when miking guitar amps, and condensers with my voice. But I also use condensers on acoustic guitar, and a combination of dynamic and condenser on percussion. With Roland's approach, I can just leave the mics plugged in to the interface, and call up different presets when I change their positions for different miking scenarios.

We'll look at the onboard DSP in a little more depth next, and then move on to using our trusty RightMark analyzer to generate some real specs. It will be interesting to see if the numbers match up with my subjective impression of the sound quality.

Link to comment
Share on other sites

  • Members

So I spent the evening installing Windows 7 64-bit, installing Sonar X1, and of course, the Octa-Capture drivers.

I KNOW THAT ALL CAPS IS SHOUTING, BUT I'M SHOUTING!! I'M GETTING 100% RELIABLE PLAYBACK WITH ASIO AT 48 SAMPLES!

Now, that's with a single WAV file, and I'd be really surprised if once I start loading things up with amp sims and virtual instruments and such, I can maintain that figure. But even if I have to bump it up to 64 or 96 samples, that would still be very impressive.

The larger lesson here is given the same computer, same program, same interface, pretty much same everything except the OS, with Vista 64-bit I was getting 128 samples but it always seemed like it was on the verge of instability. With Windows 7, I'm getting 48 samples and rock-solid playback.

Guess who's probably never going to boot up into Vista again smile.gif Well, except to pull any remaining data off the drive.

And major, major props to Roland for getting those drivers so effing tweaked. That's really pretty phenomenal.

It's late, so I'll wait until tomorrow to give some stress tests and see what it will take to break it. I'll check out WDM as well. But for now, color me VERY impressed with the Octa-Capture latency when used with Windows 7. Freakin' great!!

Link to comment
Share on other sites

  • Members

This is the part where manufacturers always get a little worried, because we're not dealing with what someone thinks, but with cold, hard numbers. What makes it even more of a nail-biter is we've subjected other interfaces to the same testing, so there's a source of comparison. But as you'll see, Roland didn't have much to worry about.

Here's the frequency response graph (all tests are done at 44.1kHz, 16 bits).

s1gnc.png

The response is down only -0.5dB at 20kHz and more significantly, down only -0.5dB at 10Hz and a mere -1.6dB at 5Hz. That's phenomenal low-end response by any standards, and bests the Phonic Firefly 808 (also the subject of a Pro Review) by quite a bit. It's also somewhat better than the Mackie Blackjack (there's a review in the Articles Library) and Focusrite Pro 24 DSP (featured in yet another Pro Review), although by not as much as the difference compared to the Phonic.

Now let's look at A-weighted noise level.

WyQtY.png

It's pretty much down around -120dB, but in any event, is well below -110dB. Compared to the other interfaces I've measured, this is in the same general range as the Phonic Firefly 808 and Pro 24 DSP, but a couple dB noisier than the Mackie Blackjack. Remember that all of these measurements include the mic pre, and when you're talking noise levels below -110dB, that's not going to be the limiting factor in your system - for example if played back on a CD, the CD's maximum theoretical (not real-world) dynamic range is about 96dB.

Now consider intermodulation distortion.

sgm4Z.png

Here you can see distortion products at 120Hz (-78dB) and 180Hz (-95dB) at the low end, and at 14kHz (-100dB) at the high end. (In case you're not familiar with how IM measurements are taken, the spikes at 60Hz and 7kHz are intentional, and are what produce any distortion.) This is not as clean as the Phonic, whose high-end distortion product is down about -108dB; at the low end, 120Hz is down -100dB and 180Hz is down about -115dB. Unfortunately I didn't do a similar measurement on the Mackie, so we don't have that for comparison. Compared to the Pro 24 DSP, it's an unusual situation because the Pro 24 DSP had many more visible distortion products - not just 120 and 180Hz, but 240, 300, and 420Hz, and even some just below and just above 1kHz. However, these were all quite low in level - the first two were around -98dB, and the subsequent ones around -110dB. The Pro 24 DSP had distortion at 14kHz that was down around -108dB, but also had another spike at 21kHz that was around -110dB. This would imply to me that the Octa-Capture's output filtering is more rigorous. In any case, the Roland IM+Noise spec is about average.

Next up, THD (Total Harmonic Distortion).

9atiK.png

Once again, you can see the low noise; distortion products are at 2kHz (-70dB), 3kHz (-90dB), 4kHz (about -108dB) and 5kHz (about -111dB). Past that, any distortion gets lost in the noise floor.

The specs are not as good compared to the Blackjack and the Phonic, although the Blackjack has a curious anomaly where the 5kHz distortion product is louder than the 4kHz one. Compared to the Pro 24 DSP, although the Octa-Capture has a lower 3kHz distortion product (-90dB), the Pro 24 DSP's 3kHz distortion is also at -90dB, and unlike the Octa-Capture where distortion products above 3kHz pretty much disappear into the noise floor, the Pro 24 DSP has a -95dB spike at 5kHz, and lower-level distortion products (averaging between -105dB and -115dB) all the way up to 20kHz.

Finally, let's close with the stereo crosstalk spec.

HD2dn.png

This is outstanding at high frequencies. At 20kHz crosstalk is less than -84dB, compared to about -75dB on the Blackjack, about -81dB with the Phonic Firefly 808, and -72dB with the Pro 24 DSP. In fact the Octa-Capture's crosstalk is pretty much better throughout the entire range than the Phonic or the Pro 24 DSP; the Blackjack is a tiny bit better in the 300Hz - 1kHz range, but otherwise, Octa-Capture gets the nod.

So what does all this mean? First of all, you're going to get great low end, with the flip side being you might want to add highpass filtering on some tracks to eliminate any spurious low-frequency material from getting through (it's very convenient that the Octa-Capture's DSP has a highpass filter on every channel).

As to the distortion, it is somewhat higher than comparable units but we have to be careful here - 2nd harmonic distortion is often associated as having a certain "warmth," which may explain why people comment on liking the preamp "character." I have noticed that some boutique mic pres (like the PreSonus ADL600, the subject of a previous Pro Review) seem to have just a little distortion to add "sparkle" - it's part of their "secret sauce," and maybe that's what's happening here. But, we also need to be realistic; a distortion product that's sitting at -78dB below full volume is not going to be audible - if anything, it might be "perceptible," but even that's a stretch.

The crosstalk figures explain why the stereo imaging and overall soundstage is excellent, but the fact that crosstalk is exceptionally low at high frequencies - which are, of course, the most directional frequencies - might explain why percussion sounds very "open" with the Octa-Capture. A lot of times percussion is panned off-center, and Octa-Capture is capable of preserving that with a high degree of accuracy.

We're left with the usual conclusions of doing bench tests: No one unit is better in all possible respects than other units, especially when you take cost and functionality into account. For example, none of the other units have programmable mic preamp gain, and while the Pro 24 DSP has compression built in, it's for only two channels, as opposed to the Octa-Capture's eight compressors (BTW all tests were done with compression bypassed).

The bottom line is that despite the high level of Octa-Capture functionality, especially in light of the price, there are no significant compromises in terms of specifications and in some respects, it's considerably above average.

Link to comment
Share on other sites

  • Members

I just had to test this with Pro Tools 9 to see if you really can use other ASIO interfaces with Pro Tools 9. Also, I thought that perhaps testing with Sonar X1 was a bit unfair - after all, what with Roland and Cakewalk being in the same "family," I assumed Roland made sure that the Octa-Capture worked really well with X1.

My first attempt was a fail - Pro Tools said I wasn't using a suitable interface, and quit. Hmmm...so much for being compatible with other ASIO interfaces.

But then I realized that the sample buffer was set to 48 samples, and thought maybe that was just stressing things out too much. So I changed the buffer setting to 64 samples, and voila - it worked perfectly.

So I experimented a bit more. I figured I'd bump the sample buffer up to 96 samples in anticipation of throwing on a variety of latency-sucking virtual instruments. Again, it didn't work - but did work when the sample buffer was 128 samples.

I think the take-away is that at least with the Octa-Capture, if you want to use it with Pro Tools 9, the sample buffer needs to be a multiple of 64 samples. Just for grins, I tried running the same experiment with the V-Studio 700 interface, and it too needed to be set to a multiple of 64 samples.

Oh and BTW, PT9 worked perfectly at 64 samples - no dropouts or other issues.

Now, one more possibly related comment. The Octa-Capture and V-Studio are the first interfaces I've used with Windows 7, so I don't know if the dramatically lower latencies compared to what I was getting with Vista-64 are a function of the OS, the way the Octa-Capture integrates with Windows 7, the drivers themselves, or a combination of factors. I'll try some other interfaces at some point (in fact, I'm going to start a Pro Review on the new Mbox very soon) to see if I can get them to work with the same low number of sample buffers as the Octa-Capture (as well as to see if they too need to be set to a multiple of 64 samples to work with PT9), but I gotta say one thing: If you're playing virtual instruments or guitar through amp sims, 48 or 64 samples ROCKS compared to 256 samples smile.gif

I'm starting to see why Roland is so hyped on this interface. It performs really, really well.

Link to comment
Share on other sites

  • Members

Hey, great review! I have one point of confusion. I own an Octa-capture, and while the user manual, the catalog descriptions, and your review here all imply that the compressor section can be used prevent clipping of the A/D converters, I do not find that to be the case. It seems like the compressor is merely DSP added on to the signal once it is converted to digital. In my experience, any sound that clips the input without the compressor engaged clips the input with the compressor engaged as well.

Is there actually a way to use the compressor to prevent overs while tracking?

Link to comment
Share on other sites

  • Members
Quote Originally Posted by Prontold View Post
Hey, great review! I have one point of confusion. I own an Octa-capture, and while the user manual, the catalog descriptions, and your review here all imply that the compressor section can be used prevent clipping of the A/D converters, I do not find that to be the case. It seems like the compressor is merely DSP added on to the signal once it is converted to digital. In my experience, any sound that clips the input without the compressor engaged clips the input with the compressor engaged as well.

Is there actually a way to use the compressor to prevent overs while tracking?
Welcome to the thread...and good catch! The compressors are after the A/D, and they can prevent overs "downstream" from the compressors and within the DAW, but not "upstream," prior to the A/Ds. I've corrected my original post so that people coming into this thread for the first time get the right information.

However, the Auto-Sens function is indeed before the A/D converters, and that's the element that can help prevent overs at the A/D converter itself. The thing to remember about Auto-Sens is it's like the "learn" function for levels in plug-ins like, for example, Guitar Rig. It depends on sensing the absolute maximum volume level and while most of the time you'll be safe, if there's a sudden increase in level and you don't "Auto-Sens" the levels taking that extra level into account, then you can indeed overload the A/D converters. I suspect that maybe a copy writer somewhere along the line confused the Auto-Sens function with the dynamics...

So I guess the conclusion is that while the Octa-Capture makes it difficult to have distortion due to overloads, it's not bullet-proof.

As you have an Octa-Capture, what do you think of it? I'd like to hear some other opinions in here smile.gif, but in any event, thanks for the input and the opportunity to figure out (and clarify) what's going on.
Link to comment
Share on other sites

  • Members
Quote Originally Posted by Anderton View Post
As you have an Octa-Capture, what do you think of it? I'd like to hear some other opinions in here smile.gif, but in any event, thanks for the input and the opportunity to figure out (and clarify) what's going on.
Sure, thanks for checking it out and confirming my experience with using the compressor!

I've had the octa-capture for about two months. At this point I record myself and sometimes my friends as a hobby, and I haven't had as much time as I would like for recording because of grad school. In my experience, the sound quality seems very transparent, especially in the low end. DIing a modified fender jazz bass into the first channel gives a tone that is almost too fat on the bottom, but not what I would call woolly (the bass is also outfitted with big nordstrand pickups, fat flatwound strings, and a series/parallel switch, so there can be a lot of bias toward the lows/low mids from all that). I'm not too into the onboard compressor or reverb, but I can imagine they might be useful for monitoring purposes during tracking. Part of why I got this interface in the first place was because sometimes I try to record myself on a drum kit, and setting levels in that situation can be a nightmare. The Auto-sens function is simple and works really well for preventing most clipping with minimal headache in stupid recording situations like that. All in all, the features are very well executed, and for $500 you get a lot of them, so I would feel comfortable recommending the Octa-capture as an excellent deal for a mid/high level interface
Link to comment
Share on other sites

  • Members

 

Quote Originally Posted by Prontold

View Post

In my experience, the sound quality seems very transparent, especially in the low end.

 

The specs certainly bear that out - I've seen hardly any interfaces that are essentially flat down to 10Hz.
Link to comment
Share on other sites

  • Members

 

Quote Originally Posted by rockdigital

View Post

If you're referring to recording 8 tracks simultaneously with USB, I don't believe this is possible. USB isn't fast enough and usually only allows you to record 2 instruments mono, or 1 stereo. I'm only aware of firewire that is capable of recording multiple tracks at the same time.

 

Wow. If you said this like 6 or 8 years ago, you might have something. USB 2.0 is technically faster than Firewire (burst speed, anyway). But there are PLENTY of USB 2.0 interfaces that can record 8+ channels. Shoot - the V-Studio 700 that Craig uses is good for 21-inputs, and Apogee's new interface does 32 over USB. The computer itself may have some issues if you're tracking with too many soft-synths & effects, but there's nothing wrong with USB 2.0 as a bus.
Link to comment
Share on other sites

  • Members

The reverb is far better than I expected - it's clean, smooth, and transparent. However, note that the only variable parameters are algorithm (with your choice of echo, room, small hall, or large hall) and for all algorithms, a maximum pre-delay time of 160 ms and maximum decay time of 5 seconds (see the attached image). Also, remember that this is for monitoring only, and only through direct mix A - I recorded the audio examples by taking the headphones output and feeding it into Octa-Capture inputs 3+4.

As to the audio exampes, do check them out - this is a very fine reverb, without periodicity or flutter.

The first audio example is the Room algorithm, the second the Echo algorithm, the third the Small Hall algorithm, and the fourth is the Large Hall algorithm.

Link to comment
Share on other sites

  • Members

And let's finish off our trip into DSP-land with the compressor. Unlike the reverb, this is available for recording, not just monitoring. The audio example plays the uncompressed sound of talking through a mic, then the compressed version, with a -20dB threshold and 8:1 ratio (note that this also brings up the background fan nose of my computer - sorry about that). Overall, the compression is more than utilitarian; it doesn't add artifacts, and it takes some work to make it "pump."

Link to comment
Share on other sites

  • Members

First, the standard disclaimer – this doesn’t mean the review is over, just that I think I know the unit well enough to provide some conclusions. I still want to try cascading it with the V-Studio 700 just to make sure it works as expected, but aside from that, I think we’ve covered pretty much everything that’s important. However, if anyone has any questions or comments (particularly from those who own the unit), please continue to post.

Also for those wondering about Roland’s participation in the latter part of this thread (or lack thereof), unfortunately their Product Manager was in Japan during the recent earthquake. He’s safe, and we wish him well; fortunately, reports are that Roland, being located in a more southern part of Japan, has not suffered serious damage. But I can certainly understand that there are far more pressing issues right now then checking in on a pro review. Hopefully when the situation settles down, he’ll be able to add some comments.

Now, about the Octa-Capture.

The primary characteristic that stands out to me is cost-effectiveness. For under $600, you’re getting eight high-quality mic pres, onboard DSP, a very nice mixer application, Auto-Sens level setting, and a relatively intuitive front-panel user interface. That’s quite a deal.

Here are what I consider the main limitations:

  • Only one headphone out. Granted, I don’t know where they would have been able to fit another one on the front panel, but sometimes it helps to be able to have someone else listen on phones while tracking. Of course you can do this, but you’d need to add a headphone amp to a couple of the monitor outs.
  • No optical I/O, either S/PDIF or ADAT. There is coaxial S/PDIF, so you’re not totally lacking for digital I/O. The ADAT input is perhaps less of an issue than it appears at first, because the main reason for adding ADAT I/O is to provide expansion - being able to cascade Octa-Captures is a mitigating factor.
  • The reverb works only with Direct Mix A, and you can’t record with it. Actually this is kind of a flattering limitation, because I wouldn’t care if it weren’t for the fact that the reverb sounds very good.
  • The compressor can’t go in front of the A/D converter to prevent overloads. However, see below for my comments on Auto-Sens.

And now, what I consider the main strengths.
  • Eight combo jack inputs for line inputs or XLR mics. Even better, four are on the front panel and two of these can serve as high-Z instrument inputs. A lot of interfaces have one or two front-panel ins, but having four is really convenient, as is having eight really good mic pres.
  • Auto-Sens. Why doesn’t every interface have this? This is a really useful and time-saving feature that’s easy to use and effective. Big props.
  • Compact size. This could definitely work for mobile, laptop-based recording applications. It’s not tiny, but I don’t see how it could be any smaller and still provide equivalent functionality.
  • Individual +48V phantom power for each input. I think this is very important, and otherwise surprisingly rare in the world of interfaces – most let you enable phantom power for all mics or none, or particular groups of mics (e.g., one group of four and another group of four, or for a couple inputs but not for the others).
  • Front-panel programmability. Of course the mixer application is much easier to use, but you can do everything you need to do from the front panel. Although pushing buttons and turning dials can be tedious, the Octa-Capture “operating system” isn’t hard to deal with.
  • Universal (100-240V) “line lump” power supply. It only takes up one space on a barrier strip, and works anywhere with the right plug.
  • Cascadable. I like that you can add another eight mics if needed, while staying within the same interface context.
  • MIDI in and out. I’m always surprised at how many audio interfaces don’t include MIDI. It’s still a part of my studio, that’s for sure...
  • Stable, trouble-free drivers. Installation was totally painless, and the performance is exceptional. I attribute part of that to Windows 7, but Roland shares in the credit.

I feel that at this point, there’s a certain level of “interface fatigue” - there are so many of them out there, with such varied feature sets, that there really is something for everybody. Roland is entering a crowded market at what are still sketchy economic times, but it would be a shame if the Octa-Capture got “lost in the noise.” This is a really solid interface, with some unique features, at a righteous price.

I’m sure many will look at the Octa-Capture and say “nice, but I already have an interface.” However, when you get your hands on one, the Octa-Capture solves a lot of problems faced during typical recording situations – from being able to handle multiple mics, to giving the singer some really nice reverb in the headphones, to adding light compression on the inputs, to automatic level-setting. Roland definitely did their homework with the Octa-Capture. For their first foray into pro-level interfaces, they’ve made the right moves.
Link to comment
Share on other sites

  • Members

Hi!

Thanks a lot for this great review I was very thrilled while reading it!

I was asking myself, regarding all the specs and graphs, if the preamps, converters etc are significally different from those in the VS700R? Is the VS700R an Octa-Capture with a built-in Sonic Cell ( sorry, maybe more an octa-Capture with ADAT, less the 4 sub-mixes and reverb and a Fantom XR with ARX) or is there something more regarding sonic performance as an audio interface? Have you already done the graph thing with the VS700r that you own? Is there any differences at all?

Thanks!

Link to comment
Share on other sites

  • Members
Quote Originally Posted by Dyonight View Post
Hi!

Thanks a lot for this great review I was very thrilled while reading it!

I was asking myself, regarding all the specs and graphs, if the preamps, converters etc are significally different from those in the VS700R? Is the VS700R an Octa-Capture with a built-in Sonic Cell ( sorry, maybe more an octa-Capture with ADAT, less the 4 sub-mixes and reverb and a Fantom XR with ARX) or is there something more regarding sonic performance as an audio interface? Have you already done the graph thing with the VS700r that you own? Is there any differences at all?

Thanks!
I have not measured the sonic performance of the VS-700R, but it does seem very similar (if not identical) in terms of sound quality. When I asked the same question to Roland that you're asking, the response was that the mic pre sections and signal path are essentially the same. There are some functional differences: The VS-700R has optical I/O, word clock, and the Fantom VS built-in, but doesn't have the Auto-Sense function, reverb for cue mixes, or the additional hi-Z input.

BTW if you have a VS system, I highly recommend downloading the latest drivers (the same ones used for the Octa-Capture). They have superior mojo than the previous ones for sure. Also, assuming you're using it with Sonar, you'll be happy to know that the X1b update includes tight integration between the VS-700C console and X1, like ProChannel support.
Link to comment
Share on other sites

  • Members

Okay, there's been one remaining loose end to this review - testing out the ability to cascade two Octa-Captures. While I don't have two of them here, Roland claims you can also expand the Cakewalk VS-100 or VS-700R interface (further evidence that the two are very similar) and as I have a VS-700R, I figured I'd give it a try.

My first attempt was not successful because I looked in the Octa-Capture manual about how to do expansion, and figured I'd make the Octa-Capture the timing master and slave the VS-700R to it. However, that's not the way to go. I did something that disconnected USB from the VS-700R, and thought I'd have to re-install the driver.

While looking for the driver in my last download of V-Studio updates, I stumbled upon a PDF about how to do expansion with an Octa-Capture. Long story short: The order in which you turn on the units and connect them is subtly different than connecting two Octa-Captures. When I followed the instructions for expanding the VS-700, everything worked perfectly - I didn't have to re-install anything or jump through any hoops, I just had to RTFM once I found it smile.gif And let me say once more to the Roland manual-bashers - something changed, because the documentation was totally clear, and holds you hand all the way through.

I booted up Sonar, and here's what I saw:

JrXi7.png

The top three entries are the last I/O of the VS-700R, and the next four entries are the first expansion I/O for the Octa-Capture. Pretty slick. This turns the system in a 27 in/32 out interface (technically 29/34, but you have to give up the coax S/PDIF I/O to provide sync between the two devices).

You can't run two units at 192kHz - I'm crushed, because I do ALL my recording at 192kHz (yes, I'm kidding - does anybody really use 192kHz for their day-to-day recording?). Seriously, for those who need lots of I/O, this is really pretty darn cool.

Anyway, I think I've pretty much covered everything now. To re-iterate, I've been VERY impressed by the Octa-Capture's performance and general design. Roland made a lot of very good moves with this interface.

Link to comment
Share on other sites

  • Members

Ooops...two other details I forgot to mention.

  • The package comes with two rack ears for rack mounting.
  • There's also a "Cakewalk Production Plus Pack" DVD included. It includes some virtual instruments (Cakewalk Sound Center, Rapture LE, and Studio Instruments Drums, all of which are cross-platform) and Sonar LE 8.5.1 (Windows only).
Link to comment
Share on other sites

  • Members

This review has been helpful. I'm planning on upgrading from an E-MU 1616m to either the Octo-Capture or the MOTU Ultralite mkIII. I'm leaning toward the OC more now because of this review.

I did not see a review of the Ultralite. Do you have an opinion/comparison about it?

Link to comment
Share on other sites

  • Members
Quote Originally Posted by Guitarpima View Post
This review has been helpful. I'm planning on upgrading from an E-MU 1616m to either the Octo-Capture or the MOTU Ultralite mkIII. I'm leaning toward the OC more now because of this review.

I did not see a review of the Ultralite. Do you have an opinion/comparison about it?
I reviewed the original Ultalite back in 2007, but the Mk 3 is quite different. The main differences I see compared to the Octa-Capture:

* Octa-Capture has 8 mic pres with individually enabled phantom power. Ultralite has 2 mic pres. Neither has ADAT port expansion for adding additional octal mic pres, so if you need more than two mics at a time, your decision just got made for you in favor of Octa-Capture.
* Ultralite works with both FireWire and USB. Octa-Capture is USB 2.0 only, so if you need to work with FireWire, again the decision was made for you in favor of Ultralite.
* Ultralite includes EQ as part of its DSP; both have dynamics and reverb, although the Ultralite has two compression options.
* Octa-Capture's Auto-Sens level setting is a very cool feature, which Ultralite doesn't have (no do other interfaces I know of).
* If you're a Sonar fan with a V-Studio system, you can expand it with the Octa-Capture for sessions that require lots of inputs.

Hope this helps!
Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.


×
×
  • Create New...