Need to buy a new recording interface, looking for suggestions

    Hey, I need to buy a new recording interface. I had narrowed down on a few options:


    1) RME Fireface UFX
    2) Apogee Ensemble
    3) Apollo Audio 8p (there's an offer where the are throwing in a free Octo Satellite)
    4) Antelope Audio Orion Studio


    Anyone have some inputs on these options? I used to have a Fireface 800, which has lasted me for 10 years but died suddenly last night, RIP. I'm hoping for similar longevity with as good quality.


    I am on an iMac.

    Quote from ashtweth

    High End Audio interface recomendation 4 recording kemper?


    :D


    Ash


    Thanks, Ash. I saw that thread when it was current.


    To be clear, I had a Fireface 800, so I know RME is great. But with a Mac, I've heard there are better options, such as the Apogee and the Apollo, which translates into lower latency. The new UFX+ looks sick, I may look into that as well.


    Also, Monkey Man's enthusiasm notwithstanding, I haven't heard of the MOTU's spoken about in the same league as the RME and other stuff. The specs are great, but we all know that just because the Axe FX has great DSPs, that doesn't make it the world's greatest guitar device. I have no experience, just pointing out what's happening on most threads on the net about audio interfaces.

    Quote from digbob

    antelope discrete 4 seems to be a cool option for 899


    Thanks. that's a great option, but I need more inputs. I plan to start recording real drums for my electronic project, so I need more inputs, at least 8, though more would be better.



    Has anyone tried these interfaces or own them? I was also looking at the Slate VRS8, but I'm not sure whether that is stable enough. Also heard some bad things about the Antelope Audio stuff.


    So looks like the race is between the Apogee, the Apollo and the RME.

    An Apollo 8 Duo is what I just bought the other day :)


    You can't beat the plugins and the fact they send you a free quad/octo satellite worth 1-1.5K approx. I don't need the 4 xtra pre's because I have a Focusrite octo-pre mkii dynamic for my kit. At the current price point it can't be beat IMHO.


    Good luck with the computer shit, man whatta drag, I really feel for you :cursing:


    Cheers,


    Shawn

    If it was me i think i would have gone for RME and just hook up some extra inputs over adat if i had to record drums. I am under the impression that you don`t need to throw that much money in to pre`s for a drumrecording so maybe save some cash on the adat preamps. But that is up to you:)

    Quote from Xander

    If it was me i think i would have gone for RME and just hook up some extra inputs over adat if i had to record drums. I am under the impression that you don`t need to throw that much money in to pre`s for a drumrecording so maybe save some cash on the adat preamps. But that is up to you:)


    There are other benefits to the higher end interfaces, mainly lower latency. For example, some of the Apogee stuff has a latency of 1.5ms. That's incredible.



    Woah, awesome! Are you bonding with it? I really like the idea of getting some free pres, though I doubt I am going to spend more money on UAD plugins... I think.



    Quote from nakedzen

    What's the budget? Something like the Focusrite Red 8 Pre (not the Clarett) should last you a long time. But it's twice the price of UFX II.


    Jeez, they only have the Red 4 pre here and it is too expensive for me.


    I was hoping to top out below $3000. I could use whatever money I save on new studio monitors.

    Quote from nightlight

    There are other benefits to the higher end interfaces, mainly lower latency. For example, some of the Apogee stuff has a latency of 1.5ms. That's incredible.

    Don't get too excited about the "low latency marketing" by manufacturers. :)


    1. If you record external instruments/vocals, you'll likely make use of the "direct monitoring" of your interface which cuts down monitoring latency anyway (maybe 1 - 1.5ms for AD/DA and the internal mixer). System roundtrip latency won't matter at all in these cases.


    2. The system's roundtrip latency only matters if you play virtual instruments (via MIDI). Be it keyboard/synth plugins or drums plugins. In this case the required buffer (latency) settings are determined by the required processing of the given plugin(s) in use. Practically it's pretty much impossible to keep your interface's buffer setting to the lowest possible setting in these cases. A buffer setting of as low as 64 at 44.1kHz (without crackles and dropouts) can be considered a success and will give you a roundtrip latency of roughly 5.5ms. If you go below 64, you'll very likely experience dropouts, crackles ... no matter if the interface gives you the option and had been advertised as "lowest latency in the market". :)


    3. If you happen to play virtual instruments, you'll be surprised how much latency the use of the good old MIDI cables introduce, compared to MIDI via USB. So if you play virtual (MIDI) instruments, make sure you always use USB-MIDI and try to forget about the old MIDI cables altogether.


    Imho, it's very important to understand the limitations of achievable latency. There's no point in getting overly excited about the technical possibility to run an interface at 192kHz with a buffer size of 32 .... your computer (and the given virtual instruments and plugins) will likely fail to process the audio and fill the buffers at this rate.

    Quote from Xander

    If it was me i think i would have gone for RME and just hook up some extra inputs over adat if i had to record drums. I am under the impression that you don`t need to throw that much money in to pre`s for a drumrecording so maybe save some cash on the adat preamps. But that is up to you:)


    That is an interesting idea, but I also have a synth set up where I think it would be really helpful to have nice preamps for it. The ones on the Fireface 800 were very transparent, but a bit sterile.


    Also, wouldn't good preamps be even more critical with something like drums? I'd be looking at something like an Audient if I was buying separate preamps, but those are at least $1000, and I wouldn't want to have to spend more and make more connections.


    Too true. At the same time, I think it is erroneous to assume that extra fire power has no use. It comes into play when you start processing your audio as well. I had a great interface, it was just very old. Could probably still outlift many of the prosumer solutions in the market. However, whenever I tried to process stuff in either Cubase or Logic, it would start to stutter and I'd have to increase the buffer. Not saying that the workaround isn't valid, but I used to have all kinds of problems with even applying processing with heavier plugins that I have.


    I'd be happy to run at 44.1 kHz provided I can process the audio the way it needs to be processed.


    Also, for the record, I play drums, so having a lower latency will be like a godsend. I can only see that translating into more accuracy when I play and then later sync other stuff up.


    I wasn't aware of this midi over USB thing being lower latency than midi cables.

    Quote from nightlight

    Also, for the record, I play drums, so having a lower latency will be like a godsend

    I completely understand that ... my point was: The interface won't give you any "extra fire power" in this regard. The buffer setting requirement is not determined by the interface but by the performance of your computer system and the used virtual instrument (e.g. Superior Drummer 3) and/or plugins.
    This is very important to understand.


    For example I own a Focusrite Liquid Saffire 56 which allows me to set the buffer as low as 32 @ 44.1kHz which would (in theory) result in a calculated latency of 1.5ms. The actual roundtrip latency would be 4ms. But it's absolutely impossible to ever run e.g. Superior Drummer 3 at this buffer setting. Although my system is pretty fast and fuly equipped with SSDs and 24GB of RAM and a 6-core CPU at 3.6GHz ... no way ;)
    If Cubase is the sole active application and the sun and moon phase are right, then it works at a buffer size of 64 but for the sake of being safe at all times I rather keep the buffer at 96 or even 128. A monstrous audio interface with heavily marketed low latency glory wouldn't change that AT ALL. That's what I tried to make you understand. :)

    Quote from lightbox

    I completely understand that ... my point was: The interface won't give you any "extra fire power" in this regard. The buffer setting requirement is not determined by the interface but by the performance of your computer system and the used virtual instrument (e.g. Superior Drummer 3) and/or plugins.This is very important to understand.


    For example I own a Focusrite Liquid Saffire 56 which allows me to set the buffer as low as 32 @ 44.1kHz which would (in theory) result in a calculated latency of 1.5ms. The actual roundtrip latency would be 4ms. But it's absolutely impossible to ever run e.g. Superior Drummer 3 at this buffer setting. Although my system is pretty fast and fuly equipped with SSDs and 24GB of RAM and a 6-core CPU at 3.6GHz ... no way ;)
    If Cubase is the sole active application and the sun and moon phase are right, then it works at a buffer size of 64 but for the sake of being safe at all times I rather keep the buffer at 96 or even 128. A monstrous audio interface with heavily marketed low latency glory wouldn't change that AT ALL. That's what I tried to make you understand. :)


    Ah yes, I get that completely. It is probably possible to record one channel or a stereo channel of audio at that buffer setting. I can and have done that, with latency of about 2ms using Cubase and my old interface. But as the project complexity increases, more elements will add crackles and other artifacts.


    The difference between devices that permit the same buffer and sample rates is that some of them will be able to handle more project tracks than the other. These devices are also likely to be able to record more tracks simultaneously. For example, the latest RME can handle 94 audio channels simultaneously. So there is a bit more going on under the hood than what the buffer and sample rate indicates.


    It just isn't as apples to apples comparison between different interfaces just because of the sample rate and buffer sizes. Brute force counts when you have more tracks and more processing. It also permits more tracks to be recorded simultaneously.

    Quote from nightlight

    The difference between devices that permit the same buffer and sample rates is that some of them will be able to handle more project tracks than the other.

    This is not due to the interface, as the tracks are processed and summed inside the computer, not the interface.
    HOWEVER, if you use an interface where you can offload the DSP (like the Apollo 8P), you make the plugin processing tasks easier on the computer - so in that case (where some DSP takes place in the interface), you are of course better off, if you previously hit the ceiling on the computer itself. This is ONLY true for the plugins that can actually be offloaded, of course :)

    Quote from Michael_dk

    This is not due to the interface, as the tracks are processed and summed inside the computer, not the interface.HOWEVER, if you use an interface where you can offload the DSP (like the Apollo 8P), you make the plugin processing tasks easier on the computer - so in that case (where some DSP takes place in the interface), you are of course better off, if you previously hit the ceiling on the computer itself. This is ONLY true for the plugins that can actually be offloaded, of course :)


    For Pete’s sake. It’s like any sound card.


    Try turning off your interface and run a heavy project. Now turn on the interface.


    Are you able to tell any difference in how it is being handled? How much is being processed?


    It is a myth that your interface does not reduce the load on your Pc. That is how they handle more channels when recording as well.


    A helpful hint: it’s not because they have more inputs.

    Taking the argument further, you know that most audio interfaces use ASIO, right?


    You can run the same project on two different interfaces and also figure out that I'm not making this up.


    Helpful hint: It is not just the way the Asio is handled.


    Those interfaces take the strain off your PC. I don't understand how this myth is perpetuated that there isn't less latency/more processing power with the more expensive devices.


    EDIT: Just simple real world tests will work here. Just take the same PC/Mac, and use two different interfaces and test this for yourself. A $100 Behringer and an RME would be a good comparison. I understand that IN THEORY, there should be no difference. But you'll notice it for sure with a heavier project and the same sample rate. Heck, CPU usage will vary based on the interface used, I'm confident of that.


    It's not just latency with one track. Add something like 50 tracks. Now, are you able to record at the same buffer setting without drop outs? Basically talking about the buffer setting here, to be perfectly clear.


    I am not saying that just because I bought an interface. Easy tests, please run them and determine for yourself.


    EDIT: I should caveat this all by noting that CPU, RAM, SSDs make much more of an impact. But it's easy enough to determine whether a $100 interface will keep up with something like a $1000 interface with the same inputs. Drivers, ADDA, multiple simultaneous channels.


    Just try recording at the same latency using all the channels on an audio interface with two devices with an equal number of inputs. I guarantee one will perform better than the other. Why, all things being equal?

    Good call AJ!


    You'll be very happy with the unit! With the bonus bundle, whoa! Now you have 8 Kick Ass Pre's, 12 dsp accelerators total, wicked plugins {real-time}wicked gain structure, world class conversion etc... May I suggest some of the Neve stuff, especially for drums, vocal and bass. Manley is nice but it's brighter.


    These plugs are/can be dsp heavy but look at the legacy stuff. Here's a handy chart in case: Instance


    The Virtual mixing console (takes a bit of getting used to), it uses the interface cpu instead of the computers pre & post. I literally just started learning on it and haven't had much time to really record but so far it's amazing.


    Can't wait to hear the new and improved tracks! 8o:thumbup:8o

    Please read this extract from Sound on Sound on how the chips and circuit design used in an audio interface affects latency:


    https://www.soundonsound.com/t…ruth-about-latency-part-2



    Last month I explained how to measure real-world audio latency using an external synth as a signal generator, plugging this into one stereo channel of your soundcard, and then looping the monitored version via your soundcard's analogue output back into the other stereo input channel. The inter-channel timing difference is the total latency of the input and output software buffers, plus whatever constant delay is added by the A-D and D-A converters, and anything else in this signal path.
    Since I was rather surprised at the high value of 189 samples measured on my Echo Mia, which effectively adds about 2ms to both input and output at 44.1kHz, I contacted Echo for clarification, and received a fascinating insight into what goes on under the bonnet of a typical soundcard. The AKM AK4528 converter chip itself adds a 31-sample latency for the ADC oversampling digital filter, and 30 samples for the DAC. At under 1ms at 44.1kHz, this is about what I was expecting.
    However, for PCI buss efficiency, the soundcard DSP has tiny additional 32-sample internal buffers on each side as well, and it turns out that an additional 32 samples are used for real-time sample-rate conversion in each direction, when operating at 32kHz or under. This explains my total measured value (31+32+32+32+32+30 = 189). So now you know how some soundcards manage to support new sample rates after a driver/firmware update — it's extra DSP code that runs on the card itself.
    I've since measured the onboard latency of M Audio's USB Duo at 152 extra samples, while the ESI Pro WaMi Rack 192X's is lower at 91 samples, presumably because it has no SRC and uses smaller DSP buffers. With most modern soundcards the real-world audio latency is always going to be between one and two milliseconds greater than the software buffer size both when recording (input) and when playing back soft synths in 'real time'.
    By the way, if any of you want to take real-world audio latency measurements further, there are various other possible configurations. If for instance your card has S/PDIF I/O, you can use a loopback cable between its out and in to perform a similar test to my analogue one. You'll still need to somehow patch the left output to the right input to avoid howlround — I managed this using the Mia's Console utility, which has pan controls on the input monitoring — but the difference between this digital measurement and the analogue one should be just the delay caused by the ADC and DAC.



    Of course, this is an article from 2002, and things have improved. Still, I suggest running a test with your gear and measuring whether your reported latency is the same as someone with a different interface and the exact same computer using a loopback like in this test.