Dismiss Notice

Psst... Ready to join TalkBass and start posting, make new friends, sell your gear, and more?  Register your free account in 30 seconds.

E-MU 0404 and latency

Discussion in 'Recording Gear and Equipment [BG]' started by ole Jason, Jul 28, 2004.


  1. ole Jason

    ole Jason Supporting Member

    Apr 3, 2003
    Louisville, KY
    I'm looking to upgrade my soundcard. Right now I'm using an SB Live! card w/ KX drivers running with 2.66ms latency. I'm happy with the real time VST performance but I'd like to get something with a better s/n ratio and overall better clarity.

    The E-MU site just says it has 'ultra-low latency'. What is generally considered ultra-low, 2ms? 5ms?

    Also, will the higher resolution converters work better for recording bass? When I record bass it seems like the card can't hear all the frequencies, kinda like if you record a bass cab with a cheap microphone.
     
  2. nonsqtr

    nonsqtr The emperor has no clothes!

    Aug 29, 2003
    Burbank CA USA
    Latency is a big problem for computer recording, no matter how small it may be. My advice is, never use the computer as a monitor. Always monitor "before" the computer. Likely that means you'll need either an external mixer or something like a direct box. If you monitor from the DI or the mixer, you'll never have to worry about or care about how much latency your computer/sound card has.

    That being said, yes it is almost certainly true that better A/D converters will make your bass sound better. Low frequencies are the hardest to sample in the digital domain. The SB, nice as it may be, uses cheap converters (that's probably part of why it's so popular, 'cause it's cheap). I'd definitely go for a sound card with 96/24 capability, if at all possible. Then if you run it in 44.1/16 mode, it won't tax your system, and your bus and backplane will be a lot happier too.
     
  3. basshag

    basshag

    Oct 17, 2001
    Santa Cruz, CA
    E-MU's Digital Audio Systems actually offer zero-latency monitoring through its hardware-accelerated PatchMix mixer (so the monitoring signal gets routed straight to the outputs in hardware without having to go through the software application) - in the same way as an external audio mixer.

    BTW, I work at E-MU and have been a TB'er for several years.
     
  4. Josh Ryan

    Josh Ryan - that dog won't hunt, Monsignor. Staff Member Supporting Member

    Mar 24, 2001
    Yep, latency will not be an issue.
     
  5. Mandobass

    Mandobass

    Nov 12, 2002
    Raleigh, NC
    if you can raise the funds for the 1820m, i highly recommend it. i sold my sadowsky pre-amp and RNP in favor of the e-mu's pres.

    using the 1820M as a bass DI is simply incredible. i was totally blown away the first time i plugged in. i still get impressed by this box.


    on a side note i'm having a problem with my 1820m. for the first time ever, i get no sound output in winamp, games, media player, anything. however, when i load a project in cubase SX, it plays perfectly...kinda odd huh? if anyone has had similar experiences, lemme know.
     
  6. Josh Ryan

    Josh Ryan - that dog won't hunt, Monsignor. Staff Member Supporting Member

    Mar 24, 2001
    No idea about the winamp thing, but here is a link to a forum where you might find the answer.

    http://www.productionforums.com/emu/default.asp
     
  7. nonsqtr

    nonsqtr The emperor has no clothes!

    Aug 29, 2003
    Burbank CA USA

    So, why not just use a mixer? Why tie up your computing resources for monitoring purposes? I don't get it...

    If what you're saying is that you can actually set levels for your monitored channels, then I submit that it IS a mixer. The diff is, that now you're chewing up output channels for monitoring purposes. Not the ideal solution IMO.
     
  8. Josh Ryan

    Josh Ryan - that dog won't hunt, Monsignor. Staff Member Supporting Member

    Mar 24, 2001
    It is a mixer, it's called patchMIX. -you don't chew up output channels, you can simply add a new strip with the click of a mouse, which is an advantage over a hardware mixer. You don't chew up PC resources.
     
  9. Passinwind

    Passinwind I Know Nothing Supporting Member

    Isn't the whole point here that the original poster is using soft synths, realtime EFX, and the like? Outboard monitoring is going to do zilch for that, eh? :eyebrow:
     
  10. Josh Ryan

    Josh Ryan - that dog won't hunt, Monsignor. Staff Member Supporting Member

    Mar 24, 2001
    OK, then find me a system with a lower latency from the asio out monitor. Also, just re-read the original post and I'm not sure where you got that anyway. He just says:
     
  11. nonsqtr

    nonsqtr The emperor has no clothes!

    Aug 29, 2003
    Burbank CA USA
    Yeah, what I meant was, output channels "on the card". Most cards only have a limited number of outputs. Typically 2, or 4, or if you get an advanced one, 8. My take on the original post was an effort to record live music, like bass for instance. I'm guessing, that there might be some overdubbing involved, and stuff like that. You know, just "regular" home recording, the kind you might do on a PortaStudio or something.

    So, let's say you're doing overdubs. What you'd like to do, is play along with the stuff you've already recorded. Well, in that case, you "have to" use your software. There's no choice at that point. In that case you "will" have a latency issue. No question about it. You can get that latency down to a very small value, with a fast computer and good sound card, but it'll still be there, and it'll still be annoying. So, if you have a good sound card like the EMU, what you can try to do is to send your recorded tracks out one set of outputs, and then use the built-in mixer in "direct mode" to send out your monitoring signal. In most cards you have to do that through separate outputs, 'cause you can only configure each output channel to be one or the other (tracked or direct), but not both.

    That's why I was suggesting an outboard mixer. That takes care of the problem, and it doesn't chew up any outputs on the sound card. Am I missing the point, or does that make sense?
     
  12. Josh Ryan

    Josh Ryan - that dog won't hunt, Monsignor. Staff Member Supporting Member

    Mar 24, 2001
    You can listen to the already recorded tracks, then listen to the tracks going in that you are playing via the patchmix monitor. You are able to choose at which point in the chain you listen at. You could use, for example, cubases output but why would you? I still don't get why an outboard mixer would be better than that. The E MU is doing the same thing, just in a different shape and with software sliders. You have monitor outs, headphone outs, etc. so you don't need to use the other outputs.
     
  13. Passinwind

    Passinwind I Know Nothing Supporting Member

    OK, then find me a system with a lower latency from the asio out monitor.

    Lower than what? Did I miss the answer to what the ASIO latency of the EMU card is?

    Also, just re-read the original post and I'm not sure where you got that anyway.

    Erm, from the part that says" real time VST performance"? I'm only asking a question here, not necessarily passing judgement on anyone else's input.

    But FWIW, I usually set my buffers (Gadget Labs card) such that the latency is 100+ ms. With a hardware mixer for monitoring (and preamping, which in my case is quieter than the noise floor of the card), overdubs work just fine. So, to me, latency would only be an issue if I decided to use software synths, or wanted to record with real-time EFX.

    I'm mainly interested in this thread because my card is orphaned, with poor XP drivers, and very problematic ASIO support. Real-time EFX could be very useful, epecially for live PA applications. So, not to hijack the thread, but what is the latency of the ASIO monitoring bus?
     
  14. nonsqtr

    nonsqtr The emperor has no clothes!

    Aug 29, 2003
    Burbank CA USA
    All right, let's break it down. Let's say you record the first track (let's say it's a guitar). That you can do without any monitoring. It goes into the sound card, and then into the computer, and through the software, and into a data file. The track now resides on a file in your computer. To play it back, you have to use your software, and your sound card. That means, you'll get "playback latency".

    Now, you record the second track (let's say it's a bass). You want to listen to the guitar while you're doing this. That means, you'll be playing back the guitar track, at the same time you're recording the bass track. The guitar track will have the "playback latency" attached to it, while the live bass won't. You now have three choices. Either a) you listen to the bass "after" the software gets it (ie using software monitoring), which means the "whole thing" (everything) will have playback latency, or b) you listen to the bass "before" the software gets it (ie using the EMU zero-latency mixer-monitor), which means your guitar will have playback latency but your bass won't, or c) you run the guitar through your monitor speakers, and just listen to the bass "as is" (ie without monitoring).

    Obviously, (c) is the preferred method. If you use (c), then it doesn't matter how much playback latency your guitar track has. A faster sound card doesn't benefit you in that case. But, there are some limitations with the (c) method. One is, you have to do a "live" recording (ie not "directly" into the sound card, you have to monitor "live with your ears"), because if you use anything inside the computer, you're dealing with latency again. Two is, you need some kind of external monitoring system (like a reference amp and speakers), because you can't really wear headphones while you're trying to hear a live guitar track (well, you "can", but it might be a little hard to hear the nuances of the guitar under those circumstances).

    I've tried all these different methods. To my ear, methods (a) and (b) are completely useless. I can hear a latency as low as a few milliseconds, and it disrupts my timing. The only one that works for me is method (c).
     
  15. Josh Ryan

    Josh Ryan - that dog won't hunt, Monsignor. Staff Member Supporting Member

    Mar 24, 2001
    I don't get the difference between B and C. Why is listening to the bass (to use the same example) on the way in (using only a mixer, whether it's the EMU or any other external) different than listening to it just "in the air"?
     
  16. ole Jason

    ole Jason Supporting Member

    Apr 3, 2003
    Louisville, KY
    This post had completely slipped my mind until tonight when I had a chance to start looking at soundcards again.

    For clarification, I'm using Cubase SX and recording both real instruments and virtual instruments via midi controller. For all my midi stuff I'm using VST's for the sounds. I also use VST's for tone shaping and effects on the real instruments (ie reverb, delay, compression).

    It's because of this that I was asking about that latency of the EMU card. I understand the issues with monitoring but I want to know how it performs, latency-wise, with real time effects and virtual instruments. I would like to have something that offers latency as low as the soundblaster however I think I could live with up to 5ms or so.

    With latency higher than that it would be impossible to play in overdubs with the midi controller in real time.
     
  17. Josh Ryan

    Josh Ryan - that dog won't hunt, Monsignor. Staff Member Supporting Member

    Mar 24, 2001
    Simply put, the E MU kicks the soundblasters ass.
     
  18. ole Jason

    ole Jason Supporting Member

    Apr 3, 2003
    Louisville, KY
    haha works for me. I was a little scared because I've seen people bragging about getting their latency "down" to 7ms or 10ms on some supposed high end cards.
     
  19. basshag

    basshag

    Oct 17, 2001
    Santa Cruz, CA
    A clarification on latency in regards to the E-MU cards. For the scenario that nonsqtr outlines above, E-MU offers zero-latency. This works because the PatchMix mixer runs on a DSP chip and not in software, and functions just like an external hardware mixer, "floating" above whatever recording application you use. When you record a guitar solo, for example, over a backing track, you are hearing the backing track coming from Cubase (or whatever), but hear your guitar coming from Patchmix before it hits Cubase (you have to make sure that you mute the recoding track in Cubase so you don't hear the delayed version). Whatever effects you use within PatchMix are also hardware-accelerated by the DSP chip, and can be inserted either before sending the audio via ASIO to Cubase or after it (or both). This also allows you to monitor your instruments/voice with effects, but recording them dry to Cubase if you choose, just like an external mixer. This sounds more complicated than it is, and there is a lot of documentation in the pdf manual that you can sheck out under the support section of E-MU's website.

    As for general latency, there is always latency on CPU-dependent software. E-MU's cards let you get down to about 2 ms, but this varies depending on your computer configuration (processor speed, memory, drives, buffer settings, number of tracks playing/recording), but is as low as any pro card out there. The only way to avoid software latency completely is to have a fully hardware-accelerated system like the ProTools HD system (not the Digi 001 or 002 which run on the CPU), with systems generally starting at around $10,000.
     
  20. nonsqtr

    nonsqtr The emperor has no clothes!

    Aug 29, 2003
    Burbank CA USA
    Exactly. You want the mixer fully "in hardware". The diff between (b) and (c) in my earlier post is, that people who use (b) are the ones who typically use their computer for some kind of effect. If you're onboard, and your mixer does that kind of thing, that's great. Otherwise, you're outboard, and you're tying up an output on your sound card. The good cards (like the EMU it seems) allow you to do a lot of stuff "in hardware", like hardware accelerated effects and so on. That's really what you want. The minute you leave your sound card, for any reason, and enter the computer (even just touching the bus, like getting stuff into and out of memory), you'll have a latency issue. I got my rig before hardware acceleration was really "sophisticated", and for that reason chose to use an outboard mixer. But Blisshead is right, if your onboard system does all that, you may not need an outboard mixer. It may depend partially on how many tracks you're running too, relative to the sound card's capabilities. We frequently do live drums on my system, which means I'm using all eight inputs at once, and at 96/24 that tends to put a little stress on the hardware. I'd want my onboard hardware mixer to siphon all that down to two channels, and run it right back to the outputs so I can listen to it. That's a fairly sophisticated capability that not many sound cards have. Or like basshag mentioned, the ones that do are typically more expensive. And they're usually the pro-type setups like ProTools, and not the consumer-type setups. It's nice that there are manufacturers delivering that kind of functionality to the home studio, at a reasonable price point. Very nice indeed. :)