Hardware vs. software reverb and effects

Discussion in 'rec.audio.pro' started by David Nobel, Aug 28, 2003.

  1. knud

    knud Guest

    > I'm not sure "staying completely inside the DAW" makes for quality mixes.

    When you're on the cheap it puts you into a whole different league than
    what you can get with crappy budget boards.


    blahblah
  2. David Nobel

    David Nobel Guest

    kludge@panix.com (Scott Dorsey) wrote in message news:<binnd5$b55$1@panix2.panix.com>...
    > David Nobel <dnobel@sympatico.ca> wrote:
    > ><sigh> It is becoming painfully obvious to me through reading the
    > >discussion on this thread that I have Champagne reverb tastes on a
    > >Cheese Whiz budget, and that quite probably "I can't get there from
    > >here."

    >
    > So, build a chamber. Not very expensive, though it might take up some
    > considerable attic space. Sounds like the real thing, or better than
    > any of the digital simulations anyway.


    That's an interesting idea. If I weren't in the process of selling my
    home and moving, I might pursue it. However, now I will have to wait
    for a bit.

    > Why do you want to run 96 ksamp/sec? I'd pick a 44.1 system that sounded
    > good over a 96 system that sounded bad, any day.
    > --scott


    That's a good point. However, I have good-sounding 96 k and happen to
    be in the camp that believes that working with acoustic music at
    higher sample rates before burning can improve the quality of the
    audio CD. It irks me to contemplate not being able to use that
    capability.

    The chamber is not practical for me now, but something else which
    occurred to me is running a round trip analog conversion solely to
    post process to an older, affordable, classic analog hardware reverb
    unit. Some folks believe in running digital tracks out to analog
    boards for mixing, so why not for reverb? This would of course require
    good ADA converters. So is it a stupid idea?

    — David
  3. David Nobel

    David Nobel Guest

    Thanks, John and Johnston, for those two very clear explanations.

    — David
  4. HenryShap

    HenryShap Guest

    How much space does one need for a chamber? Are there any websites/previous
    discussions on how to go about building one?

    Henry


    >So, build a chamber. Not very expensive, though it might take up some
    >considerable attic space. Sounds like the real thing, or better than
    >any of the digital simulations anyway.
  5. David Nobel

    David Nobel Guest

    > > Chambers are great and Altiverb rocks... it will eat quite
    > > a bit of CPU however.


    No kidding! On my 733 G4 Quicksilver, Spark XL just refuses to run any
    96 k files through Altiverb. It will process 48 k files, but becomes
    extremely unstable and crash-prone--and that's a single stereo mix,
    Altiverb set at high latency/low CPU load, with no other plug-ins
    loaded. I'm going to try it in AudioDesk tomorrow with hopefully
    better results.

    --David
  6. Scott Dorsey

    Scott Dorsey Guest

    David Nobel <dnobel@sympatico.ca> wrote:
    >The chamber is not practical for me now, but something else which
    >occurred to me is running a round trip analog conversion solely to
    >post process to an older, affordable, classic analog hardware reverb
    >unit. Some folks believe in running digital tracks out to analog
    >boards for mixing, so why not for reverb? This would of course require
    >good ADA converters. So is it a stupid idea?


    Well, it defeats the whole purpose of running at the high sample rates,
    and it adds additional converters into your signal path, but try it and
    see.

    If you don't like it, you can always downsample your stuff to 48, run
    it through the hardware reverb, resample it back up to 96, and then mix
    it in with the original tracks. Be a pain in the neck but if you like
    the sound, go ahead.
    --scott
    --
    "C'est un Nagra. C'est suisse, et tres, tres precis."
  7. "Ethan Winer" <ethan at ethanwiner dot com> wrote in message
    news:-ZWdnSr_eLLl89KiU-KYuQ@giganews.com...
    > Ricky,
    >
    > > I'm not sure "staying completely inside the DAW" makes for quality

    mixes.
    > <
    >
    > I've never experienced quality problems mixing entirely in a DAW, and I

    know
    > many others who also use DAWs successfully. What kind of quality

    degradation
    > do you mean, and what specifically / technically would cause it?


    One the best digital systems (which I have not used) it might be OK but at
    my level of equipment it sounds too 2 dimensional if done completely in the
    DAW. Or at the very least some analog (especially tape) warms it up.
  8. David Nobel

    David Nobel Guest

    > Well, it defeats the whole purpose of running at the high sample rates,

    Why? I thought one of the main reasons for running at high sample
    rates was to minimize the audible effects of artifacts/distortion
    created by digital sound processing. Here we are going 96/analog/96,
    doing some intensive processing in the analog stage in the outboard
    gear in which this is not a problem (though analog distortion might
    be).

    > and it adds additional converters into your signal path,


    Yes, and also may well degrade the SNR, depending on the analog
    hardware.

    >but try it and
    > see.


    I will if I can find an analog reverb to rent/borrow. If anyone has
    already tried this, though, I'd love to hear from them.

    > If you don't like it, you can always downsample your stuff to 48, run
    > it through the hardware reverb, resample it back up to 96, and then mix
    > it in with the original tracks.


    That would be easier for me to experiment with, since it's fairly easy
    to rent digital reverbs. If I don't need to mix the tracks too wet, it
    might be more promising. But heavy-duty processing at 48 k is
    precisely what I am trying to avoid by being in the 96 k environment,
    so to my mind it would be this that starts to defeat my purpose.

    Thanks.

    -- David
  9. Mike Rivers

    Mike Rivers Guest

    In article <42329f4f.0308291651.489615c6@posting.google.com> dnobel@sympatico.ca writes:

    > I have good-sounding 96 k and happen to
    > be in the camp that believes that working with acoustic music at
    > higher sample rates before burning can improve the quality of the
    > audio CD. It irks me to contemplate not being able to use that
    > capability.


    Believing and being irked aside, can you hear an actual and worth
    while difference (given that anything you change will make a
    difference) in your own work? And is that difference attributable only
    to sample rate? Certainly it makes a difference in some instances, but
    I feel humble to be in the company of so many people who believe that
    it's important to them.

    > The chamber is not practical for me now, but something else which
    > occurred to me is running a round trip analog conversion solely to
    > post process to an older, affordable, classic analog hardware reverb
    > unit. Some folks believe in running digital tracks out to analog
    > boards for mixing, so why not for reverb?


    It's certainly worth while if you want the sound of that reverb
    processor. It's the only way you're going to get it, unless you
    consider modeling it based on a sample of its impulse response to be a
    reasonable replication - and still, you'd need a trip through D/A and
    A/D converters in order to get that sample. Today's converters are
    fine for the purpose once you get above the "entertainment" level.

    As far as your comparison to mixing through an analog console, that's
    not the same idea at all. Usually the reason why this is done is
    because of the ease of mixing and of applying other outboard effects.
    The basic sound of the console (but not it's built-in EQ or dynamics)
    is trusted to be pretty transparent from line level in to line level
    out and is usually not the reason to mix a digital recording through it.




    --
    I'm really Mike Rivers - (mrivers@d-and-d.com)
  10. Mike Rivers

    Mike Rivers Guest

    In article <7ecb413a.0308291408.51d2700@posting.google.com> johnston_west@hotmail.com writes:

    > No Problem........ Say you're recording your main vocal track onto
    > your computer. You sing into your micorphone, the signal goes into
    > your mic pre-amp and then into your computer right?
    >
    > Well in adition to that, as you're recording, take a tap, (bus, split
    > signal, whatever), from your mic preamp, run it though your favorite
    > hardware reverb (Lexicon PCM 91 in my case) and record the output of
    > the unit (reverb only, no dry signal) onto two seperate tracks into
    > your computer at the same time.


    The flaw with that procedure is that you have to decide at the time
    that you're recording the vocal just what settings to use on the
    reverb unit. This works as long as the reverb (or more commonly, a
    chorus or other time-based effect) is an integral part of the sound,
    for instance on an electric guitar. However, when it comes to reverb
    for mixing, you often don't know what characteristics you'll need
    until you get more tracks recorded and see how they fit together. You
    can shape the frequency response of a pre-recorded reverb track, but
    you can't change the reverb time or diffusion characteristics.

    Life was simpler when we just had "the reverb" (such as the chamber
    we're talking about in this thread) and the only adjustment was the
    amount that you mixed in. While it's certainly possible to go down to
    the chamber and move the mic(s) relative to the speaker(s) to change
    the characteristics, generally these were placed where everyone
    thought they sounded the most natural and fixed in place.

    > The other alternative is to record the track dry, and then at any time
    > later, run that track though the reverb unit and record the output
    > onto two additional tracks. You can do this on as many of the
    > instruments as you like, and then automate the tracks like any other
    > tracks.


    This can indeed be done at mixdown time.



    --
    I'm really Mike Rivers - (mrivers@d-and-d.com)
  11. Ethan Winer

    Ethan Winer Guest

    Ricky and all,

    > On the best digital systems (which I have not used) it might be OK but at

    my level of equipment it sounds too 2 dimensional if done completely in the
    DAW. <

    Of course I've heard these comments, but I just don't buy them because
    logically they make no sense to me. Depth and space are ambience and
    left/right timing issues. How can adding a bunch of numbers digitally affect
    image width and depth? And why wouldn't the same thing happen with analog
    summing? I guess my real question is, "What TECHNICALLY AND
    SPECIFICALLY occurs in the digital realm that could cause such a phenomenon?

    Here are my two theories as to why some people prefer analog tape and why
    some people blame mix busses:

    People that prefer analog tape refer to its natural compression, gentle
    suppression of transients, a small but desirable increase in distortion, and
    a noticeably wider image than digital recording offers. That last one is
    significant because it's actually caused by a defect in analog tape
    recorders - jitter! If you've never set the bias on an analog tape recorder
    it's a real eye-opener. Briefly, you record a 10 KHz. sine wave on a track,
    and adjust the bias while watching the playback level coming off the tape.
    But just watching the playback meters shows you the enormous number of
    high-frequency dropouts that occur constantly. I am not talking about a
    junky old Tascam deck that packs 8 tracks onto 1/4-inch tape. These severe
    dropouts occur with the most expensive professional multi-track recorders
    using 2-inch tape. And people argue that digital jitter compromises stereo
    imaging! Digital jitter is nothing compared to the constant drop-outs on
    analog tape, and all the constantly changing phase shift caused by the tape
    wiggling around as it travels past the heads.

    As for the mix buss, I'm sure all of you know from experience that the most
    significant thing that happens when you sum tracks is psychoacoustic: A
    track that sounded clear all by itself may now be masked by another
    instrument having a similar frequency range. A soloed electric bass track
    where every note can be distinguished clearly can become rumbling mush when
    you add in a distorted rhythm guitar or chunky sounding piano played by a
    ham-fisted rock and roller. I am convinced that this is the real reason
    people wrongly accuse the "mix buss." Yes, such masking effects happen with
    both analog and digital summing. But it's easier for some folks to blame
    "digital" when they can't get some instruments to sit well in a mix.

    --Ethan
  12. Mike Rivers

    Mike Rivers Guest

    In article <42329f4f.0308300318.3db7b16b@posting.google.com> dnobel@sympatico.ca writes:

    > Why? I thought one of the main reasons for running at high sample
    > rates was to minimize the audible effects of artifacts/distortion
    > created by digital sound processing.


    That's entirely too broad a statement. The reason for running at a
    higher sample rate is to increase the available audio bandwidth. There
    is no other reason to do so.

    Some people can give you convincing reasons for providing a wider
    bandwidth for the audio channel. Others can provide equally convincing
    reasons why the bandwidth provided by 48 kHz sampling (or even 44.1
    kHz) is adequate for human beings. Take your pick, but base it on
    listening experience, not someone else's writings.



    --
    I'm really Mike Rivers - (mrivers@d-and-d.com)
  13. Bongolation

    Bongolation Guest

    Timo Haanpää <thaanpaa@sci.fi> writes:

    >> Reverb is computationally expensive. This means that the cost
    >> can be mitigated by changing details of the underlying hardware
    >> platform.


    > What puzzles me is the Lexicon 224XL was released 20 years ago
    > and is still considered the best digital reverb ever by many
    > people. I would guess a modern PC would have 1000 times the power
    > of those boxes. Why should there be a problem mimicing
    > the exact sound of it with a PC? I don't know.


    This makes no sense to me, either, and I hear this objection to software
    plugins a lot.

    The remarkable rate at which resources improve in home/office computers
    seem logically to outstrip whatever resources were available in
    state-of-the-art digital rack units a few years ago. Sure, the basic
    processing architecture can be customized for specific tasks, but
    ultimately and by definition any digital process is just processing ones
    and zeros. At some point, a newer generic computer is going to be able to
    handle it, no sweat, assuming the plugin software is well-written.

    >> Plus, anybody develops a nice reverb plug is gonna get it
    >> purated to hell and back, and you can't really pirate a box.


    This is certainly a more persuasive argument! :cool:

    =*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*
    E-mail: bongolation<AT>mail.md - Change <AT> to @ symbol to reply.
    See COMPLETE headers for more info. Headers are good - view them.
  14. Bob Cain

    Bob Cain Guest

    Mike Rivers wrote:
    >
    > In article <42329f4f.0308300318.3db7b16b@posting.google.com> dnobel@sympatico.ca writes:
    >
    > > Why? I thought one of the main reasons for running at high sample
    > > rates was to minimize the audible effects of artifacts/distortion
    > > created by digital sound processing.

    >
    > That's entirely too broad a statement. The reason for running at a
    > higher sample rate is to increase the available audio bandwidth. There
    > is no other reason to do so.


    It just occured to me that there might be some merit to
    David's statement. If a process that is applied to the
    recording is non-linear and generates distortion products,
    such as compression or limiting, then all of the products
    above 24 kHz will alias back below 24 kHz if working at a 48
    kHz sample rate. If working on 96 kHz sampled material,
    some of those products will remain in the 24 kHz to 48 kHz
    region which can then be eliminated when resampling back to
    a lower sample rate by the intrinsic anti-aliasing filter.

    I have no idea how audible this would be but at least there
    is some theoretical basis for remaining at higher sample
    rates if using non-linear processes.


    Bob
    --

    "Things should be described as simply as possible, but no
    simpler."

    A. Einstein
  15. Mike Rivers

    Mike Rivers Guest

    In article <3F510137.CF401991@arcanemethods.com> arcane@arcanemethods.com writes:

    > It just occured to me that there might be some merit to
    > David's statement. If a process that is applied to the
    > recording is non-linear and generates distortion products,
    > such as compression or limiting, then all of the products
    > above 24 kHz will alias back below 24 kHz if working at a 48
    > kHz sample rate.


    Not if we're talking about an analog processor, coming back into the
    digital system through an A/D converter. The converter will lop off
    everything above 24 kHz. It's not just a good idea, it's the law (in
    order to make the sampling theorem work right). Now in a digital
    system, it might be possible to create higher order distortion
    products and if you wanted them it would require a higher sample rate
    to let them through.


    --
    I'm really Mike Rivers - (mrivers@d-and-d.com)
  16. In article <_rudnTSSa-Z-jdOiXTWJjg@giganews.com>,
    Ethan Winer <ethan at ethanwiner dot com> wrote:
    >I'm surprised nobody mentioned what to me is the biggest advantage ALL
    >plug-ins have over hardware: you can contain the mix entirely in your DAW so
    >all settings are recalled. As soon as you have even one external hardware
    >device, you have to remember and set physical knobs to exactly duplicate a
    >mix.


    But, if you use an outboard processor with digital IO and can save the
    settings to a preset (or do a MIDI sysex dump), you can use outboard
    hardware and also have complete resettability. The digital IO removes
    the need to trim gain pots and saving a preset takes care of all of
    the rest.

    I do this with the NuVerb and TC M2000 and it works fine.


    Regards,

    Monte McGuire
    mcguire@theworld.com
  17. In article <20030829200700.06073.00000138@mb-m07.aol.com>,
    knud <symphonle@aol.comblahblah> wrote:
    >>I don't go near the Waves stuff,
    >>its boxy and crappy.

    >
    > More of a room simulator. Slapping some onto a simulation of organic
    >sampled instruments (orchestral emulation etc) can produce uncanny results.
    >
    > One trick with the ol' Waves True verb is to just defeate the early
    >reflections. Switch them out, every time.


    You could also try the Waves Renaissance Reverb, a box that sounds
    more like a traditional reverb than TrueVerb. RVerb sounds very good
    to my ears and it gets used occasionally even though I have nice
    hardware 'verbs to choose among.


    Regards,

    Monte McGuire
    mcguire@theworld.com
  18. In article <iET3b.301660$YN5.207776@sccrnsc01>,
    Ricky W. Hunt <rickywhunt@hotmail.com> wrote:
    >"Ethan Winer" <ethan at ethanwiner dot com> wrote in message
    >> I've never experienced quality problems mixing entirely in a DAW, and I

    >know
    >> many others who also use DAWs successfully. What kind of quality

    >degradation
    >> do you mean, and what specifically / technically would cause it?

    >
    >One the best digital systems (which I have not used) it might be OK but at
    >my level of equipment it sounds too 2 dimensional if done completely in the
    >DAW. Or at the very least some analog (especially tape) warms it up.


    Are you sure it isn't simply that your DAC can't handle the complexity
    of a full stereo mix, but it can handle each track separately?


    Regards,

    Monte McGuire
    mcguire@theworld.com
  19. David Nobel

    David Nobel Guest

    mrivers@d-and-d.com (Mike Rivers) wrote in message news:<znr1062298374k@trad>...
    > In article <3F510137.CF401991@arcanemethods.com> arcane@arcanemethods.com writes:
    >
    > > It just occured to me that there might be some merit to
    > > David's statement. If a process that is applied to the
    > > recording is non-linear and generates distortion products,
    > > such as compression or limiting, then all of the products
    > > above 24 kHz will alias back below 24 kHz if working at a 48
    > > kHz sample rate.

    >
    > Not if we're talking about an analog processor, coming back into the
    > digital system through an A/D converter. The converter will lop off
    > everything above 24 kHz. It's not just a good idea, it's the law (in
    > order to make the sampling theorem work right). Now in a digital
    > system, it might be possible to create higher order distortion
    > products and if you wanted them it would require a higher sample rate
    > to let them through.


    Should have made myself clearer here. My point about staying at the 96
    kHz sample rate after the round trip conversion to the analog reverb
    was contingent on proceeding to do other processing--EQ, compression,
    stereo enhancement, etc.--in the digital environment. As mentioned,
    theoretically, the finished CD should benefit sonically from having
    done this digital work at the 96 kHz vs. 44.1 or 48 kHz sample
    rates--even if the original recording was done at the lower rates.

    -- David
  20. In article <c_idnc2j2ZcIOs2iU-KYuA@giganews.com>,
    Ethan Winer <ethan at ethanwiner dot com> wrote:
    >Of course I've heard these comments, but I just don't buy them because
    >logically they make no sense to me. Depth and space are ambience and
    >left/right timing issues. How can adding a bunch of numbers digitally affect
    >image width and depth? And why wouldn't the same thing happen with analog
    >summing? I guess my real question is, "What TECHNICALLY AND
    >SPECIFICALLY occurs in the digital realm that could cause such a phenomenon?


    Truncation and low resolution math. Back in the stone age (v. 4.0 and
    earlier), ProTools TDM had this problem. The mixer would only accept
    20 bits from each channel and did the math at relatively low
    resolution and did not dither the result to the final word width. The
    result was a lot of truncation errors. This sounds like someone
    stripped away some ambience and space.

    People say that "well, 20 bits is way down there, so you can't
    possibly hear it". I say nonsense, because you aren't running full
    scale stuff into a mixer, and it's common to have a good bit of gain
    post mixing bus in the form of stereo limiting and compression.

    One has been able to use a TDM mixer since version 4.1.1 that accepts
    the full 24 bits from each channel and operates at 48 bit internal
    resolution. It sounds a lot better than the old mixer. There's even
    a version that does this and dithers from 48 bit back to 24 bit and I
    think it sounds better still.

    But, people seemed to have caught on to the 'bash the mixer' concept
    from the days when it really was a problem that had to be fixed by
    Digidesign (what... 4-5 years ago?). Regardless of what they're
    actually running now (and half of them have no clue what software
    they're talking about when they complain), they still point the finger
    at 'digital summing'.

    One other theory I have as to why some people like analog summing is
    that most folks' DACs are pretty poor quality, and can't handle a
    complex mix all that well. Breaking it down to smaller chunks and
    summing analog won't 'challenge' a crappy DAC design so much. In this
    case, sure, analog summing sounds better, but let's point the finger
    at the DAC, and not the summing bus, right?

    >Here are my two theories as to why some people prefer analog tape and why
    >some people blame mix busses:

    [snip]
    >As for the mix buss, I'm sure all of you know from experience that the most
    >significant thing that happens when you sum tracks is psychoacoustic: A
    >track that sounded clear all by itself may now be masked by another
    >instrument having a similar frequency range. A soloed electric bass track
    >where every note can be distinguished clearly can become rumbling mush when
    >you add in a distorted rhythm guitar or chunky sounding piano played by a
    >ham-fisted rock and roller. I am convinced that this is the real reason
    >people wrongly accuse the "mix buss." Yes, such masking effects happen with
    >both analog and digital summing. But it's easier for some folks to blame
    >"digital" when they can't get some instruments to sit well in a mix.


    This is also a big reason for 'blaming the mix bus'. It's easy for a
    beginner to produce a bad mix. Many of the tones and instruments that
    beginning musicians play with are not very mix friendly either. For
    example, kids love to use distortion boxes that sound great soloed but
    absolutely wreck a mix when you lift the fader. Their arrangements
    are sometimes pretty poor too. I see the same problems with live
    sound as well, and all contribute to a poor quality mix.


    Regards,

    Monte McGuire
    mcguire@theworld.com

Share This Page