Advantages of s/pdif for outboards

Discussion in 'Microphones (live or studio)' started by Bobby Yarrow, Feb 10, 2004.

  1. Bobby Yarrow

    Bobby Yarrow Guest

    I'm looking at reverb units, and many of the ones I'm thinking about have s/pdif I/O's. Seems nice, as I could use those now unused sp/dif I/O's from my DAW, but I want to make sure I'm drawing the right conclusions.

    First off -- I really should know this -- if I use s/fdif, am I avoiding the extra cycle of d/a/a/d that's I'd ordinarily get when using outboard gear on a DAW mix? Seems obvious that I would . . .

    Relatedly, wondering whether this would improve the latency problem. I get extremely low soundcard latency, but I have the idea that there's some measurable latency just associated with the digital/analog conversions and the processing within the outboard unit. Would using s/pdif help this at all? Can I expect to get in the neighborhood of my internal monitoring latency?

  2. clintrubber

    clintrubber Guest

    Hi Bobby,

    That's correct.

    Difficult to say in general for all units how much of each, but let's say that all three steps (so A to D, the actual digital processing and lastly the D to A give a certain amount of latency (delay)). So you'd get rid of two of three, but may still have troubles because of the delay of the digital processing. If that;s really a problem depends of course on the amount of delay and the actual application. And if your DSP-task _is_ only delay, it'd be easily compensated for. Same for reverb I'd guess, you might get by by compensating it by means of a bit less pre-delay.

    One thing that should be mentioned though is that now the issues w.r.t. synchronization can pop up, but that's only in case of feeding the S/PDIF-out of the hardware box back to your DAW.

    In this case your DAW would be the clock for the reverb-box and the DAW on its turn listens to the clock it's getting back from the reverb-box.
    I guess all should be OK.

    (As you'll know in the S/PDIF-format the actual sounddata (the samples) and the clocking is 'mixed').

    Hope his helps,

  3. Bobby Yarrow

    Bobby Yarrow Guest

    Thank you Peter, very helpful of you. I am thinking primarily of a reverb application, and I'm ashamed to say that dialing back the predelay hadn't occurred to me. An excellent point, and the sort of plain thinking I sometimes miss.

    Thinking of it now, as long as I've got only wet signal coming back from the reverb, latency really isn't an issue. I can't imagine that I'd run more than, say, 16 ms or so, and an extra 16 ms of predelay on reverb really isn't going to amount to much. Could always adjust it manually once it's recorded, or use a plugin to move it forward, but I imagine that'd be a pretty rare case.

    The W.R.T. synchronization issue is frankly unfamiliar to me. I can't make out if you're saying it would be applicable to me or not. I do plan to use s/pdif both out to reverb unit and back to the DAW. My quick google search turned up a lot of strange stuff on WRT synchronization, but it all seemed a little out of my depth.

    Thank you.
  4. AudioGaff

    AudioGaff Well-Known Member

    Feb 23, 2001
    Silicon Valley
    It is typical to have up to 6ms for A2D latency and a little less for D2A. Add a few ms for any DSP processing.
  5. Bobby Yarrow

    Bobby Yarrow Guest

    Outstanding. So, by bypassing the d>a>a>d loop, I'm eliminating something near 12 ms of latency, and avoiding an extra pass thru the converters. Sold.

    I will still lose my soundcard latency as I monitor the return from the reverb, but I won't be losing 2X that latency, right? (No latency on the way out, just on the return?) And then, assuming that the reverb unit's processing latency is static, my total latency will be static, right? That is, no extra latency working off a heavy mix.

  6. lorenzo gerace

    lorenzo gerace Active Member

    Jan 27, 2002

    I use digital reverbs and FX processors connected digitally to may DAW every day; I lock them via SPdif or ADAT (Lexicon MPX-1 and Eventide Eclipse) and they work like a dream; the quality of the reverb is still far more superior than any plug-in I have tried so far, and going I/O digitally I cannot spot any latency problems, just go for it, you'll be glad you did it.

    Hope this helps

  7. Courtney

    Courtney Guest

    I use an M-One XL on an SPDIF loop from my Multiface. Works like a charm.

    To incorporate the verb into the Cubase audio export, I record it in real time. Once this is done, it is easy to correct for any latency, although to tell you the truth I hardly ever have to do this.


  8. Bobby Yarrow

    Bobby Yarrow Guest

    Nice, nice. I think my question's been answered, so I'll go ahead and elaborate. My sometimes-trusted advisor, who's used to working on much bigger and better rigs than my own, is real discouraging about this. Basically, his position is that once you get the audio into a DAW, there's no taking it out til mixdown -- that it won't work right, latency is a worry, and it's not worth it. He counts the using outboard units as one of the advantages of not using a DAW. But you only want to trust your advisor so far, yeah?

    As plugins get better and better, I'm increasingly pleased with results working within the DAW. Some of my plugin compressors in particular seem incredibly good, and for everything short of the master bus I can't hear any reason to go out of the DAW. But reverb in particular still poses a problem. First off, I still don't have a reverb plugin that I'm actually impressed with -- some better than others, but none that compete with the better outboard units. Add to that the fact that the better ones are huge resource hogs, and that they can't be sensibly printed til the mix is pretty well done, and reverb is bit of a problem.

    Seems to me that for the cost of a Waves bundle I pick up a reverb unit that would please me much better. And by the reports here, it sounds like it'll be no trouble at all.
  9. lorenzo gerace

    lorenzo gerace Active Member

    Jan 27, 2002
    Originally posted by:

    Couldn't agree more: I once did a test with one of my mixes, swapping a pair of plugins (Waves RenVerb) with the aforementioned Lexicon MPX-1 (not top of the line, but far better than a plug-in for several duties) and Eventide Eclipse (truly a one of a kind unit, my current fave) and the depth and densness (for lack of a better word) of the mix completely changed.

    Yes, it's probably better to spend in just one harware unit that has that killer sound than an entire plug-in suite that doesn't, better, it would be best to have both :D ;) ; plug ins have the ability of multiple instances, while an harware processor only allows for one stereo send (or two in cae of machines like Lexicon 480, Eventide Orville or Tc 6000), so...anyway a hardware processor thrown in really adds a lot IMO.


  10. Courtney

    Courtney Guest

    Another option is to use a reverb plugin such as SIR, that can use impulses from high-end hardware units. Although you lose some tweakability, the sound quality is superb.

    More good news - this plugin is free!

    Voxengo also have a non-free version out that looks like it could be even better.

    The impulses are freely available on the net, and there are a lot to choose from.


  11. Bobby Yarrow

    Bobby Yarrow Guest

    I like SIR, and when I'm using Samplitude I like their version even better. They're huge resource hogs, as you know, but very good. In a full mix, I find I basically have to print the reverb to use them, and that's a drag. And, at the same time, well, I'm not convinced that a good outboard unit wouldn't do better.
  12. Courtney

    Courtney Guest

    A good hardware unit would be better, but unless you only need one instance, and mix down to stereo in real time, you'll have to print the reverb anyway. I actually don't mind doing this because it allows for loads of control in the mix, and creative use of the reverb.

    I personally use both - reserving the hardware unit for important mix elements like vocals.



Share This Page