How is "Gliding" different than delay modulation or interpolated delays?

Discussion in 'The Rack Space' started by Saint Luminus, Nov 15, 2017.


  1. Saint Luminus

    Saint Luminus Member

    Messages:
    211
    Joined:
    Aug 18, 2013
    Location:
    Los Angeles
    I'm really trying to understand my PCM 81 much better.

    I understand the Gliding parameters and how to program it. But I'm trying to understand exactly how this is different than a unit with delay modulation. Or a unit which has interpolation so you can assign a modulator without any digital glitching.

    For example, in my simplistic way of thinking, the user assigns an LFO to delay time and the boundaries and speed of modulation are set by the LFO. So why are there Glide parameters as well? Isn't delay time interpolated so as to avoid digital artifacts? What exactly is the benefit of Glide?

    My only experience with programming is my PCM 81 & my MPX-1. As you know I can read & read about other units, but actually using one and programming it is a different story.

    I hope you experts can help me understand this "Gliding" concept. Itching to understand it better
     
  2. hydroquebec

    hydroquebec Member

    Messages:
    1,341
    Joined:
    Feb 7, 2012
    Location:
    Montreal, Canada
    I look at glide as portamento.

    It seems like a measure of the time it takes to get from value 1 to value 2. I've also tried to understand the difference between LFO modulation and glide modulation. From what I can understand, LFO modulation goes in steps from 0 to 127, where gliding gets you to the spaces in-between. That's probably not correct and certainly less than a scientific definition, so I'll hope that someone drops in to give the definitive.
     
  3. italo de angelis

    italo de angelis Member

    Messages:
    3,657
    Joined:
    Mar 21, 2011
    Location:
    Bibarasse in CassoPipa
    Forget about interpolation. That's not what makes gliding or traditional modulation different.
    Both techniques are applied to an interpolated delay anyway... or you'd have all sort of noises and cracks.

    Some explanations...
    https://www.thegearpage.net/board/i...ide-what-the-manuals-do-not-tell-you.1823856/
    https://www.thegearpage.net/board/i...tions-tips-tricks-all-about-it.1761299/page-2
    https://www.thegearpage.net/board/index.php?posts/22785076/

    Lfo runs a mod signal with values between 0 and 1. Bipolar sweeping goes between -1 and 1. YOU set the delays boundaries, not the LFO.
    In those LFO sweeping ranges you get a delay sweeping from delay time x to delay time y at the speed of the LFO.
    In a glide you get to have a glide speed parameter (or something close to that) where you can decide how fast or slow a delay change will be glided (reach) to the desired value.
    On an Eventide you'd get a glide speed parameter, like 0 to 1000 ms/Sec. A value of 1000 ms/sec means immediate/instant change of a delay to whatever other desired value you want.
    The sweeping LFO will take the delay time to a desired value very quickly. a 1 ms/Sec means that the delay time will change only 1 ms in a second... so you would need VERY long times to sweep a delay from a value to another. A delay set to 10 ms and set to sweep between 10 and 15 ms... with a glide of 1ms/Sec... would take 5 seconds to reach 15 ms.

    The confusion comes from the fact people are used to use an LFO, which has its own speed.
    Les't rewind time back...
    GLIDE was created to make realtime delay changes FROM A KNOB with the ability to avoid pitch artifacts.
    The bigger and faster is the delay change... the more you're going to get pitch effects... and I really mean dives... usable in special fx.
    So, by setting an amount of ms/Sec changes, one could minimize those pitch artifacts and get significative delay changes in time.
    Then comes the guitarist and his sick love for chorus/flanging effects... and he finds out he can patch an LFO to a gliding delay to create those effects.
    But this type of chorusing will sound different. Yes... the similitude with portamento is a good one. It takes time to reach the top parameter value and to fall back... and gliding sets how slow/fast that happens and at which pitch warble "costs".
    But we have a problem there... because we have an LFO which is sweeping at its own rate which has nothing to do with the glide rate.
    So we hardly get a gliding delay really going from value A to B as it could... because the LFO runs the sweep always much faster than the glide time.
    But we like this type of chorus as it has a special detuning flavor, clean character and what not.

    On a Lexicon GLIDE has different parameters...
    If you look at the PCM80... just skip the GLIDE>HALL algorithm. That kind of GLIDE isn't as good as GLIDE can be. The delays can only be glided across a very chort range.
    Their point is the finer resolution.
    Take the Chorus>Rev algorithm as it's the only one where both regular chorusing and glide are available.
    You also get glide in the 4 voice REVERBs algorithms.
    In the Chorus algorithm you should kill the regular LFO chorusing, nulling the voices rates and depths. You can use the MASTERs to do that quickly.
    To try the glide here you need to patch a delay time (or the Master Delay) to the LFO and tweak GlideResponse and GlideRange.
    These two parameters can be very confusing. The GlideRange (0 to 1365ms) determines the delay range that will be glided when changing a delay from a value to another.
    The manual has a good example:
    "For example, if GldRange is set to 100ms, and the PstDly L and R
    times are varied over a range of 100ms, those delays will glide smoothly from
    one delay time to another. Suddenly changing the delay time to a greater value,
    such as 200ms, will cause the audio delay to change instantly to the new time
    without gliding. The available range is 0-1.365sec, with 0 causing all delay time
    changes to be "instantaneous", and 1.365 causing all delay time changes to
    glide." TRY IT!

    The GlideResponse (0 to 100 ms) sets how glide responds to delay changes.
    "The range is 0-100, with a typical default
    setting of 50. This setting provides a good, real-time glide. Changing a delay time
    from 0 to 100ms with a glide response of 0, causes the glide delays to take a
    minute or more, allowing subtle changes in delay times and echoes with no
    noticeable pitch shifting. A glide response of 100 is ultrafast, causing high speed
    pitch shift, and even chirping effects with large delay time changes."

    When an LFO is gliding delay one doesn't get to hear the whole gliding range as the LFO is usually running at faster rates than the long gliding times possible in those settings.
    There are presets though where glides are triggered by the KNOB or by an Envelop or ADSR... which can take long times and show the full potential of gliding delays.
    Be warned though... this potential is basically getting thunders and diving/rising pitch sweeps mostly useful in movies and sound effects works... not exactly chorusing.
    We do use a VERY small range of delay glides... because of the way they sound...

    Do you have the DualFX card? There's a preset showing Glides at its full power... "And the Gods"... perfectly simulating Tape Echo abuse.
    If you don't have the card... posted the preset parameters values and its MIDI SysEx files below, in the next posts.
    An important detail implied... in "normal" delay modulation you have a base delay and you are applying a sweeping amount to it. In glide you are sweeping the base delay itself.
     
    Last edited: Nov 15, 2017
  4. hydroquebec

    hydroquebec Member

    Messages:
    1,341
    Joined:
    Feb 7, 2012
    Location:
    Montreal, Canada
    That glide thread is awesome. It taught me that I can modulate the SIZE parameter. Modulating a reverb's size + delays and adding just a little but of post-processing can create some really swirly stuff. Thanks @italo de angelis.
     
    ctreitzell likes this.
  5. italo de angelis

    italo de angelis Member

    Messages:
    3,657
    Joined:
    Mar 21, 2011
    Location:
    Bibarasse in CassoPipa
    This is a reduction of the DualFX card "..And The Gods" preset. It won't sound exactly the same as on the DualFX version there's a tight reverb into the chorus. I can't do that on the regular PCM80... but it's quite close as I exaggerated the diffusion and size to simulate that. The GLIDE magic is exactly the same... so you'll get the idea.

    Here we go... USE THE CHORUS ALGORITHM!

    MIX = 100%WET FX ADJUST = +0dB
    *IN LVL L = Full IN LVL R = Full IN PAN L = 50L IN PAN R = 50R
    HIGH CUT = 8.00KHz FX MIX = 0% REVERB FX WIDTH = +45 STEREO
    *LOW LIMIT = 0 *HIGH LIMIT = 1 <<<<< VERY IMPORTANT!!!!
    LOW RT = 0.2X
    MID RT = 0.29 SEC
    CROSSOVER = OFF
    RT HC = OFF
    PRE DLY = O MS
    *REF LVL L = OFF REF DLY L = 0 MS REF LVL R = OFF REF DLY R = 0 MS
    *EKO FBK L = +0% EKO DLY L = 0 MS EKO FBK R = +0% EKO DLY R = 0 MS
    SIZE = 76 METERS DIFFUSION = 63% ATTACK = O% SPIN = OFF LINK = ON RVB OUT = OFF

    LEVELS MASTER = 100% VOICE1 = -3dB VOICE2 = -3dBINV VOICE3 = OFF VOICE4 = -3dB VOICE5 = -3dBINV VOICE6 = OFF
    *DELAY TIME MASTER = 100% GLD RESP = 61 GLD RANGE = 61 MS
    DELAYTIME VOICE 1 = PATCHED
    DELAYTIME VOICE 2 = PATCHED
    DELAYTIME VOICE 3 = 0 MS
    DELAYTIME VOICE 4 = PATCHED
    DELAYTIME VOICE 5 = PATCHED
    DELAYTIME VOICE 6 = 0 MS

    ALL MASTER RATE/DEPTH = 0 %
    ALL VOICES RATES & DEPTHS = 0 Hz & 0 MS

    FEEDBACK MASTER = 100%
    FEEDBACK VOICE 1 = -95%
    FEEDBACK VOICE 2 = 0%
    FEEDBACK VOICE 3 = 0%
    FEEDBACK VOICE 4 = -96%
    FEEDBACK VOICE 5 = 0%
    FEEDBACK VOICE 6 = 0%

    PANNING MASTER = C
    PANNING VOICE 1 = 50L
    PANNING VOICE 2 = C
    PANNING VOICE 3 = 50L
    PANNING VOICE 4 = 50R
    PANNING VOICE 5 = C
    PANNING VOICE 6 = 50R

    *LFO RATE = 0.08 Hz SHAPE = SINE P WIDTH = 50% DEPTH = 100%
    *AR ENVELOPE ATTACK 4.000 SEC AR ENVELOPE RELEASE 4.000 SEC THRESHOLD = 1 T SRC = ADJUST

    PATCH 0
    *SRC = AR ENV DEST = DLY TIME VOICE 1 *000 = 66 MS *100 = 528 MS MAKE SURE 127 = ...................

    PATCH 1
    *SRC = AR ENV DEST = DLY TIME VOICE 4 *000 = 66 MS *100 = 528 MS MAKE SURE 127 = ...................

    PATCH 2
    *SRC = LFO DEST = DLY TIME VOICE 1 *000 = 0 MS *127 = 20 MS

    PATCH 3
    *SRC = LFO DEST = DLY TIME VOICE 4 *000 = 0 MS *127 = 20 MS

    PATCH 4
    *SRC = AR ENV DEST = DLY TIME VOICE 2 *000 = 86 MS *100 = 548 MS MAKE SURE 127 = ...................

    PATCH 5
    *SRC = AR ENV DEST = DLY TIME VOICE 5 *000 = 86 MS *100 = 548 MS MAKE SURE 127 = ...................



    As the preset description says:
    "Inspired by the first track of Electric Ladyland. Hit it with a percussive sound (gun shot, tympani, snare) then click ADJUST
    from 0 to 1. A wall of thunder crashes down 4 octaves and undulates rhythmically between the speakers. Hit it again
    with a similar sound, and click ADJUST from 1 to 0. Now, the explosions rocket up into space."

    Just try a low E hit, then click the ADJUST knob and listen. Wait a few seconds then click ADJUST back.
    That's the magic of GLIDE!!!
    If you want to experiment more with it... try the AR ENVELOPE attack and release to higher values, like 10 seconds for both....


    Have fun!


    Here's the MIDI SysEx file for any PCM80/81 without a DualFX card:
     
    ctreitzell and AnalogKid85 like this.
  6. bosetuno

    bosetuno Member

    Messages:
    305
    Joined:
    Jan 23, 2012
    Thanks a lot italo!
     
  7. jaykay73

    jaykay73 Member

    Messages:
    292
    Joined:
    Jan 28, 2014
    Great thread. This is the kicker...

    My question is: can you have glide modulation without the LFO running at all, i.e. LFO RATE = 0? Can other "sources" alone make the base delay sweep?

    JK
     
  8. hydroquebec

    hydroquebec Member

    Messages:
    1,341
    Joined:
    Feb 7, 2012
    Location:
    Montreal, Canada
    I know we've discussed this before, but man, I truly hate twisting that encoder so much. I'm so damn worried that at one point, it's going to stop working, even after having my PCM serviced. I should probably have more patience and balls.
     
    ctreitzell likes this.
  9. Saint Luminus

    Saint Luminus Member

    Messages:
    211
    Joined:
    Aug 18, 2013
    Location:
    Los Angeles
    I had read those first two links, but I was still confused. That last link there, ohhhhhhhhhh! That was very informative. Ok I'm starting to get it a bit more.

    The real proof is in the programming and then listening. Thanks a lot Italo! :) I'm going to try the preset you just posted. Time to get to work.
     
  10. Saint Luminus

    Saint Luminus Member

    Messages:
    211
    Joined:
    Aug 18, 2013
    Location:
    Los Angeles
    It was weird at first, but I've gotten used to it. :) So far so good!
     
  11. Saint Luminus

    Saint Luminus Member

    Messages:
    211
    Joined:
    Aug 18, 2013
    Location:
    Los Angeles
    That is a very interesting way of looking at it. Portamento I understand!!!! Thanks for that!
     
  12. AnalogKid85

    AnalogKid85 Member

    Messages:
    1,004
    Joined:
    Aug 7, 2008
    Location:
    Orlando, FL
    Yes, absolutely!

    Try patching EnvL or EnvR directly to a delay time...set a min/max delay times (000 & 127) and watch some crazy stuff happen ;)

    You can make "dynaflangers" like this too...and (correct me if I'm wrong) I believe @italo de angelis actually used something similar to this in the Korg DL8000R library, a new kind of take on the 1210 where there was a still a "normal" stereo chorus pair of voices, but also another pair that were swept directly by an envelope.

    You could also patch an expression pedal directly to delay time, or even just an "on/off" switch (like a Boss FS-5U). The mere changing of one delay time to another—with the right glide response/glide range settings—is enough to produce smooth glide modulation (that includes tap tempo changes too, if the delay line is set up for "echoes:beat").

    This really is one of the greatest features of this box. Just about anything—or combinations of things—can be a source for glide modulation. It really turns delay time into a "stretchable," flexible raw material for building all kinds of things, in ways that you just don't see in most other boxes.
     
    Last edited: Nov 16, 2017
    ctreitzell likes this.
  13. AnalogKid85

    AnalogKid85 Member

    Messages:
    1,004
    Joined:
    Aug 7, 2008
    Location:
    Orlando, FL
    @italo de angelis , I can't remember where you posted it, but there was a modified factory preset you put up with a "walkthrough" of sorts for a multivoice glide chorus in a 4-voice algorithm that used two different glide responses (one for the V1-V4 glide delays, and one for PostDlys L&R). That would be a good one for this thread :)

    Can't remember if it was HRI or the early days of TRS, but it's around here somewhere...
     
  14. AnalogKid85

    AnalogKid85 Member

    Messages:
    1,004
    Joined:
    Aug 7, 2008
    Location:
    Orlando, FL
    There's really no way out of it. I've thought this through...even if you had a Peavey 1600X, that would cut it down somewhat, but you're still going to be doing a LOT of turns of the encoder (you might say to yourself: "I've finally got some 'scenes' set up, so I can take the load off the encoder with some of these sliders"—fine, but now, what about if you want to see what those sliders are doing? guess what you're going to have to use to traverse the matrix left-to-right, to see all those parameters? ;) ).

    From time to time, the encoder on mine has acted up just a tad, and by that I mean, it might do 2 clicks for every 1 sometimes...so I'm expecting that it won't ever go out "suddenly," and I'll have ample to time prepare/back up my presets before I had to get any repairs or replacements done on it (hopefully I'll be in a position to get a 2nd PCM at some point—I'd like to know I could keep working on stuff while one is "in the shop" :) the prices are exactly prohibitive anymore for owning multiples).

    The only true way out of it I can think of would be to design editor software for it—then and only then could you really keep the encoder use to a minimum, I think (who knows how many times you'll have to use it while designing & testing the software though ;) ).
     
    ctreitzell likes this.
  15. italo de angelis

    italo de angelis Member

    Messages:
    3,657
    Joined:
    Mar 21, 2011
    Location:
    Bibarasse in CassoPipa

    That's what I said in my post. Look at the preset there, either the parameters values or just just download the SysEx file and install it. If you have the Dual FX card then you just need to load the "...And the Gods" preset and find out how the AR Env is used to modulate the 4 delays. The AR Env is an open control to modulation.
    You set the attack and release times and choose a trigger and a threshold for it to work.
    One can use the KNOB, as in the preset, to trigger the attack rise and the release fall on a second click. Trigger source could be a footswitch, a MIDI CC, the input level, the LFO itself, using the pulse wave to start A/R cycles. Or you can patch any other modulation source directly to the delay. Will it work? Yes. Will it sound good? Not always. Some sources are just not good for the job... that's why AR Env is a nice choice.
     
  16. italo de angelis

    italo de angelis Member

    Messages:
    3,657
    Joined:
    Mar 21, 2011
    Location:
    Bibarasse in CassoPipa

    You have no way around this, unless you find a MIDI SysEx editor and build your editor. Or a hardware MIDI controller with similar capabilities.
    And the thing is not over yet... because you may have troubles at fine resolution control of wide range parameters.
    A slider or knob may be unable to set a delay time to the exact value you want. The 1365 ms range of the Chorus delay can be broken into small parts, like 11 or 12 ms for each slider "click".
    No matter the fact you are using the correct MIDI SysEx string with the proper 4 byte, Nibs, low to high the PCM80 is expecting to see...
    A computer program should work better. PC1600... forget it.
    So... if you are really worried... know that the encoder will take many, many years of hard tweaking before it starts to be unusable, provided the machine lives in a non humid/dusty environment and you don't lean on the knobs (this may happen because of the unit height position in a rack).
    The only way you can be sure about the encoders is to get a new board NOW. That will make the swap very easy when the problem may start.
    If you wait... you may have a hard time getting the parts...
     
    ctreitzell and AnalogKid85 like this.
  17. italo de angelis

    italo de angelis Member

    Messages:
    3,657
    Joined:
    Mar 21, 2011
    Location:
    Bibarasse in CassoPipa
    Well... you won't really get musical glides on the PCM80/81 when sweeping the delays over a large time range. That's why many modulation sources aren't well suited for this.
    I wish the GLIDE on the PCM worked like on an Eventide, with much better and logical control over it. But it's different... so most of the time you get too much detuning or delay "chunks" being swapped one after the other... which boils down to a not so great implementation of any true smooth glide function.
    In the end... you are still forced to use glide with an LFO to get musical results... only with a new way of thinking... the LFO rate should make sense when compared to the glide speed... that is VERY slow... to make the complete glide run from min to max and back... or it won't be a proper glide.
     
    ctreitzell likes this.
  18. italo de angelis

    italo de angelis Member

    Messages:
    3,657
    Joined:
    Mar 21, 2011
    Location:
    Bibarasse in CassoPipa
    You can have the PCM showing the remoted parameter on the display. Yep... but you need to kill your life thru the MIDI implementation manual... and hope your controller can send multple SysEx strings for each slider, knob, etc...
    The point isn't so much about this... it's more about the fact you can't control the largest parameters by each single value in their resolution scale.




    That's your best bet. If you know what you are doing and how to use the PCM MIDI implementation manual it won't take long, nor so much testing really.
    So the real issue here... is finding such software.
    Maybe some more modern hardware MIDI controllers may have better resolution of their faders/knobs... but I doubt about that.
     
    Last edited: Nov 16, 2017
    ctreitzell likes this.
  19. italo de angelis

    italo de angelis Member

    Messages:
    3,657
    Joined:
    Mar 21, 2011
    Location:
    Bibarasse in CassoPipa

    Can't remember about that... but last night I was editing that "...And the Gods" preset to make a simple 2 voice glide delay for testing purposes... and finding out how limited the overall glide response is, unfortunately. I may post the file with some examples if I get to a worth result.
     
    ctreitzell likes this.
  20. italo de angelis

    italo de angelis Member

    Messages:
    3,657
    Joined:
    Mar 21, 2011
    Location:
    Bibarasse in CassoPipa
    That's exactly the whole point of it. Describing GLIDE with words is a very limited (and tiresome) way to deliver its concept.
    I might have a glide test bench preset in a while....
     
    ctreitzell likes this.

Share This Page