GSP-2101 FAQ: Tips/Tricks

GSP-2101 FAQ 3.0, Section 3: Tips & Tricks

Friday Tips:

Friday Tips was an email newsletter of sorts that Randy Thorderson of Digitech would periodically (each Friday, duh...) publish. This letter, however, contained 2101 tips rather than news. These tips have been archived by Steve Schow at:

Read Friday Tips

New Algorithm Programming Tip: [Oct 8, 1997 - erwin@bart.nl]

If you watch algorithms #9, #10, #11 and #12, you can see they all have the same scheme, where only the first module changes. This basis algorithm is actually a very nice one, some of my favourite patches use it.

You can easily make a new one with an effect that is not listed yet: select the alg. nr (or prefferably a patch which has settings you already like), go into edit mode, and delete the first module. Now add a replacement. Don't use autolink, just connect this new module to the inputs and the first mixer.

I used this method to create a patch with the reversed reverb (only possible with 2 SDISCS), it's a very easy way to create a completely new scheme. It applies to all standard algorithms of course.

Dual Master-Volume (Dist Off/Dist On): [Aug 28 1996]

The Master Volume Parameter controls the signal volume (mono) just prior to the end of the analog section of the 2101. The Master Volume actaully has 2 distinct values: one for Distortion on and one for Distortion off. This allows you to adjust the signal level so that the levels are equal (or unequal, if that's the way you want it) when you switch between Distortion Bypassed and Distortion On.

Unfortunately, this is not the case for the different distortion types. For example, if you switch from Grunge to Clean Tube, the Master Volume setting for Distortion On applies to both, and thus will not change. It would have been more appropriate (in my mind at least) to have provided 2 settings for the Master Volume, based upon a clean setting (Clean Tube and Distortion Off) and a distorted setting (Sat Tube, Dist Tube, OverDrv, Heavy Dist, and Grunge), much like the 2 channels in a typical guitar head.

Controling Master Volume for different Distortion Types: [Aug 28 1996]

To program the 2101 to change the volume when switching between two different distortion types, simply map a footswitch to a CC in your current bank. Then link this CC to the Master Volume parameter, so that the value changes as the switch is toggled. Next link this same CC (thus the same switch) to the distortion type, so that as the distortion type is changed, the Master Volume changes with it.

Unfortunately, this is not always possible with units 2.xx and earlier. A bug sometimes handles CC control of Master Volume improperly. For example, if you have set FS-1 (foot Switch 1) to toggle CC-1, which in turn toggles the Master Volume between -6dB and +3dB, you would expect it to perform as follows: -6dB +3dB, -6dB, +3dB, -6,dB +3dB. However, in some circumstances, the footswitch actually takes on 3 distinct states rather than the standard 2-state toggle. In this instance, sequential presses of FS-1 might produce the following results: -6dB, +3dB, +3dB, -6dB, +3dB, +3dB, etc. This has been reportedly cured in some cases by reordering the CC links to the paramters (perhaps placing the Master Volume CC before or after the Dist parameter CCs in all cases would prevent this). It is unclear if this has been fixed in version 3.00.00.

Distortion (Second Parameter Retention) [Aug 28 1996]

Notice that Clean Tube, Overdrive, Heavy Dist, and Grunge Distortion Types utilize only 1 gain parameter, whereas the Sat Tube, Dist Tube, Heavy Dist, and Grunge distortion types utilize 2 gain parameters. These gains are BOTH programmable regardless of whether the distortion type you have saved the patch with actually uses the paramter or not. Furthermore, the gain settings remain the same when you switch the distortion types (even from a tube type to a solid state type).

Therefore, if you design a patch that switches from Clean Tube to Distorted Tube, you can program the second gain parameter by changing the distortion type to Dist Tube, adjusting gain2 parameter, then resetting the distortion type to Clean Tube.

Wet/Dry Mix (how it affects the use of volume changes) [Aug 28 1996]

The Wet/Dry setting control the amount the digitally processed signal is mixed with the analog signal (Fx Loop returns in the case the Fx Loop is not bypassed). A setting of 100% wet implies that the signal at the outputs of the 2101 are purely digitally processed (although you may have some dry element already mixed in the signal depending upon how you mixed the signal in the digital algorithm). If you choose to use a Wetness setting of anything other than 100%, you cannot completely control the overall volume of the patch from within a digital effect (whether that is a mixer, an EQ level, or the Master Mixer output levels). For example, lets say you have the wetness at 50% (1/2 the signal is pure analog, the other 1/2 is from the output of the digital section). Now, lets say you have linked the FC-1 pedal to the Master Mixer Left and Right output levels to control overall ouput signal volume. With the Pedal at 100% (full down - toe down), the signal strength will be 50% analog and 50% digital. With the pedal at the middle position (50%), the signal strength would be 50% analog and 25% digital (thus only 75% of its original strenght). At full heel down (0%), the signal mix would be 50% analog and 0% digital (thus only 1/2 the volume as when the pedal is at 100%). Thus, in this configuration, the pedal does not control overall volume; it only controls digital ouput volume, and as a result of the Wetness setting, the wet/dry ratio changes as the pedal is changed from (50/50 to 66/33 to 100/0).

Controling Overall Volume Realtime [Aug 28 1996]

As shown above, full control of the signal volume cannot be accommodated within the digital section unless the Wet/Dry setting is at 100% wet. This is the best case scenario.

If you have a patch with 100% wetness, then take advantage of volume control within the digital section. Prefereably, try to control only one gain or level setting if you can. For instance, many times I will start my algorithms off with a 2x1 mixer to mix the Fx Loop Left and Right inputs into a mono signal chain. This allows a single CC link to control the final overall stereo volume. Likewise, if a mono signal is being introduced to the digital section, a mono effect can be utilized (e.g., a PEQ3 has a level setting which would work well for this control). If you cannot perform the volume control with a single parameter, then use the Master Mix levels, assuming you have 2 CCs you can allocate to this purpose. If not, you'll need to rethink your gameplan.

If you do not use a full 100% wet setting, or need to control the volume in the analog section for another reason, you will be forced to use the Master Volume control. While it is possible to use the distortion gains, this results in some very obvious tonal variations which are often not wanted. The drawbacks of using the Master Volume Control are as follows:

1. Possible bug as reported in the section "Controling Master Volume for different Distortion Types"

2. Changing Distortion On and Off would render the CC link useless, unless you have CC links to both Master Volume parameters.

3. Low resolution volume changes. As you increase the Master Volume in +/-1dB increments, the gain from the digital section amplifies this increase to about +/-3dB increments which are very large volume jumps. This is NOT a smooth way of adjusting your sound.

4. Low volume settings result in low level signals appearing at the digital inputs of your Fx Loop devices and the 2101, which in turn, reduce the signal to noise ratio, and the effective resolution of your A/D converters. Note that this is generally unavoidable in many cases due to the dynamic nature of guitar signals, and that the sonic degradation is generally imperceptible.

Finally, any attempt to utilize the 2101's CC links to control overall signal volume will work with a special caveat. As a new patch is loaded, its volume settings are reset to the levels that were saved - not the levels in accordance with the pedal position. Therefore, you better have all of your patches set to the correct levels so you don't deafen yourself when switching from one patch to another.

Let's say you find this aspect of the 2101 annoying as I do. I would prefer that, when loading a patch, the 2101 read the value of the pedal(s) and made the appropriate setting changes prior to generating sound. This way, if I am playing at 50% pedal and I switch to a clean sound that was saved with the same overall level, it too will play at 50% volume rather than starting up at 100%, twice the volume I had been playing at!. There is a simple solution: Purchase a Stereo Volume Pedal and run the 2101's outputs through it. You could buy a cheaper mono vlume pedal, but then you'd be feeding a weaker signal to your Compressor (altering its threshold repsonse), your distortions (getting different distortion qualities), and to your noise gate (BIG problem here). Unfortunately, you now have 4 cables running to a pedal at your feet. A better solution would be an effects unit that was daisy chained after the 2101 which provided this type of control.

Alternatvely, you may be compelled to use a device in your Fx Loop to control your volume. This could be a bad idea since you may not always want your Fx Loop on (No Sum position). In the event you ever design a patch where you need the Fx Loop Bypassed or Summed, you have just defeated you volume control methodology, and its time to rethink your gameplan again.

In the end, it would have been preferable if the 2101 had a digitally controlled level for the final mixed output (both left and right, but not necessarily independant controls -- that would be unnecessary). Also, it would be nice if the 2101 had incorporated a Utility setting that would allow the user to determine if the 2101 loads programs with settings in tact, or it the settings are altered upon loading to acount for the position of the external and internal pedals.

Programming Bank Settings Faster: [Aug 29 1996]

I have a setup that uses the same settings for switches 1-5 in every bank. Rather than programming all 5 switches in bank 1 then reprogramming in Bank 2, etc.. ad nauseum, I have found a faster way to do it:

Start with Bank 1, Switch 1 and program it as needed.

Set the 3rd field according to your desires

Press 1 twice. This will increment the Bank Number.

Press 3 once and set this field appropriately. You will immediately notice that you need to press the 3-button X-number of times and turn the data wheel Y-clicks left or right to acheive your objecive. Remeber this, and repeat sequence for all Banks 1-10.

Once all banks have been programmed for a given switch, increment the switch number and repeat the process.

By remembering a consistent order of keypresses, this routine will ocur very swiftly.

Making XLR-to-1/4" connections: [circa 1995 - revised Dec 19, 1996 by Jim Wintermyre]

Simple cables can be constructed to connect the XLRs (with or without the cabinet emulator activated) to a 1/4" device; no low impedance to high impedance transformer is required. You'll need some shielded two conductor wire, a female XLR connector, and a 1/4" mono plug for each side (L and R).

XLR Pinout:

Pin 1 = Shield Ground

Pin 2 = Positive Balanced Signal

Pin 3 = Negative Balanced Signal

Connections:

XLR pin 1 --> 1/4" plug sleeve (cable shield ground)

XLR pin 2 --> 1/4" plug tip (cable center conductor)

XLR pin 3 --> No connection

Note that this is contradictory to common knowledge, but due to the fact that the GSP-2101 XLR outputs are servo-balanced this configuration must be used to acheive proper phase and also accommodates equal output levels between the TRS outputs (which are NOT servo balanced) and the XLR outputs (which are servo balanced) when both are used. For more on this servo-balancing techno stuff, see Jim's archived message on the subject.

XLR pin numbering: looking at the end of the female connector with the locking tap "up", the top left pin is #2, the top right pin is #1, and the bottom centered pin is #3.

Default Noise Gate Settings: [circa 1995]

With the factory default programs, the gate threshold is set to a fairly brute-force level. This is to compete with the typical music store's fluorescent lights and other noise sources. This can be reduced on a program-by-program basis to prevent the 2101 from chopping off decaying sound or quiet passages.

Wah pedal Smoothness: [circa 1995]

The wah effect's response to the FC's treadle is not instantaneous. To speed things up try the following:

1. Turn Display CC Changes OFF.

2. Use the Smooth Wah module in versions 2.00.00 and later.

3. Set the MIN and MAX wah values to something less extreme than 0 and 100. For instance 15 and 85 provide a good wah sweep without requiring as much CPU overhead due to the reduced resolution.

The Ducker: [Pete Lyall, Jan 11 1995]

The ducker is like an inverse noise gate. When a trigger is present, it opens, thereby blocking the input signal (or attenuating it). When the signal goes away, the ducker/gate closes and passes the input to the output.

Duckers (really nothing more than a compressor triggered by a separate signal/trigger) are big in voice overs (voice on top of music). This way when the voice starts, the music is 'ducked' or attenuated.

A practical application in the GSP-2101 is where you might want a nice delay, but you only want it to echo when you are not playing (letting the echoes ring out in the silence). You would feed the output of the delay into the ducker, and use the guitar's dry signal as a trigger.

MIDI Guitar with the 2101: [Sept 27, 1996 Tom Owens]

Using MIDI guitar, run the outputs of the synth into the FX returns on the GSP. Put FX loop on "Summed" for whatever patches you want to use MIDI sounds and the guitar/synth are both effectively mixed thru the GSP.

I had tried this a couple months ago and couldn't seem to make it work. KNOWING that it would work, I tried it again and had no problem. Funny how your perceptions can effect success...

This allows me to eliminate a really cheap, crappy rack-mount mixer from my setup. Not only THAT, but all the digital effects in the GSP are now applied to the synth output.

B-3 with reverb! Flanged Trumpets!

I'm so thrilled. Everything sounds oh SO much better. ...with LESS equipment!

Bypass Mode & Signal Routing: [Oct 2 1996 - Jim Wintermyre]

Hi guys. I was messing around with the bypass mode on the 2101 last night and came across some behavior that I didn't expect. In bypass mode, all effects are bypassed (analog and digital). This much I already knew. What I didn't know was that:

1) The signal from the preamp (with analog effects turned off, of course) is still sent out the FX-SEND.

2) The signal at the FX-Returns is ignored, regardless of the FxLoop setting in the patch. Thus the signal into the digital section is bypassed analog section signal only.

Creating a Square Wave Tremolo Effect: [Oct 16, 1996 - Jim Wintermyer]

Fig 1: Jim's Square Wave Tremolo Block

Think of this as a BLOCK which performs the Square Wave Tremolo function. Just stick it in your effects alg like any other module (in much the same way the modulated delay module is really a "macro-module", i.e., a module which consists of more than one individual entities).

Fig 2: Jim's Square Wave Tremolo with a Flanger

Here are the the main settings:

Tone:

Level 100

Ref G3: 196.00 Hz

Duc:

Thresh -13 dB

HoldTm .010

Attn 100 dB

Attack 0.10

Release 0.10

Trem

Speed 8.3

Depth 100

Level 100

The Flanger works fine with the default settings. It's really just there to show how cool it sounds in conjuction with this block. Set input 2 of the 3x1 to 100, the others to 0 to hear the "square wave" output. INPUT can be the left input, or the output of some block in another custom algorithm. OUTPUT can go the left and/or right outs, or to another block. The input section analog settings can be set to whatever your heart desires. I would recommend setting the noise gate to 100 dB attenuation below the threshold, though. Otherwise you'll hear the "square-wave" effect on the input noise when no signal is there, which can be annoying.

Other cool options:

- Link the Ducker threshold to the C1 pedal. Vary it from -50dB to -10dB. This should cause the signal to go from completely off (-50 dB) to completely on (-10 dB). The steps in between change how long the signal is on in the square wave cycle, i.e. it changes the widths of the on/off portions of the square wave.

- Link the tremolo speed to the C1 pedal. Vary it from 4 to 16 Hz for an interesting effect. Link to Jim's original Post.

Creating a Ramp Wave Tremolo Effect: [Nov 6, 1996 - Jim Wintermyer]

A few weeks ago I posted a patch I called the "square-wave" tremolo (see diagram in section above). After playing with it some more, I found that you can get some interesting effects by playing with the ducker parameters. In particular, you can get a ramp or sawtooth type effect where the sound swells during each cycle of the tremolo by setting the ducker attack and release parameters as follows:

Attack = 0.00

Release = 0.60

For the opposite effect, use:

Attack = 0.60

Release = 0.00

These numbers work with the original tremolo speed settings I posted. Adjust to taste.

Link to Jim's original Post.

Link to Chris Sherwood's patch (ramp-tremmed chorus).

Playing Dynamics to Control Effects Mixing: [Oct 18, 1996 - Paride Zizzari]

I want to drive the sound effects with the playing volume. (play hard to get one type of sound/effects, play softly to get another one). Easy. Combine a ducker and a silencer. Since the first cuts on high volume and the other on low, why not use the silencer to drive the ducker? And here we go (sorry I don't have sysex dump), so you will have to program it yourself.

Adjust the silencer sensitivity to clip on low signals ( ~ -40 db) and adjust the ducker sensitivity to the same level (-40 db) and the Attn to 100db. Thus at high signal values ( > 40db) the silencer drives effect1 AND cuts the ducker ! At low signal level ( < 40 db) the silencer cuts the signal AND the ducker drives effect2. Mix it all and you get dynamic response effects.

Duc: Thresh -40 dB

HoldTm .010

Attn 100 dB

Attack 0.00

Release 0.00

Silencer Thresh -40 db

Release 300 (adjust as wished)

I user a delay for the effect1 and reverb for effect2. So I can play rythm on harder strokes and lead or solo with soft play. It's amazing!

Eliminating Crackling from your Tube Distortion Sounds: [Nov 15 1996 - Wally & Randy]

It's typically the gate. That is what the hysterisis setting is for. It's probably set too low (meaning too close to zero dB). Try setting it to at least 6dB.

Try these simple tips:

1) lower the threshold

2) increase the hysterisis

3) increase the hold time

Generating Responsive Feedback on Command: [Nov 15 1996 - Matt]

I use a Roctron Chameleon and to get feedback on command I created two CC links. One adjusts the gain of an EQ band and the other adjust the overall volume. Pedal "up" everything is "normal". As I press the pedal down, the gain of the EQ band is increased while the overall volume is DECREASED. This keeps a nice balance on the overall volume while cranking a specific frequency to instigate feedback. I find the frequency by experimenting. One rule - nice narrow bandwidth helps to avoid drastically changing the basic sound.

Seems to me (a former 2101 owner), that the 2101 could do this easily and maybe even better! Using a PEQ and linking the gain to a CC and linking a CC to a level somewhere in your algorithm should work quite well. With the 2101 it would be easy to target a few bands whereas on the Chameleon you pay the price of limiting you basic EQing ability.

IIR Comb Filter Module: [Jim Wintermeyer - Nov 16, 1996]

Back in DSP classes, I learned about comb filters. I thought the 2101 would be the perfect place to experiment with them. Turns out it works great! Who would've guessed you could use a delay as a a *tone* shaping device! I was planning on writing a Tip (like Friday's Tips) about it, but since you asked...

Try making a simple algorithm with a delay, and a 2x1 mixer, like this:

Set the delay time to .001 sec (1 ms), and the regeneration to 75%. The 2x1 is there to let you hear the difference between the two. I like to link each mixer input to a footswitch so its easy to switch between them.

You should find that the path that goes through the delay gives it a slightly thinner, almost out of phase sound. The delay with the feedback loop (regen) is actually an IIR comb filter. You can even figure out its frequency response - since the delay is 1 ms, the filter generates resonances at multiples of 1/1ms = 1 kHz! If you set the regen to high, it becomes unstable.

Try changing the delay to .002 s - you can clearly hear the resonances change.

If instead you use a delay (with regen off) and 2x1 mixer, with the output of the mixer going through the delay and back into the mixer in a feedback loop, you create the same thing as the delay with regen on. So why do this? Well, you can insert things into the feedback loop. Like a phase inverter before the mixer! This has the effect of shifting the resonances.

In the previous example, if the resonances are at multiples of 1kHz, there are valleys at in between them (i.e. at 500 Hz, 1500 Hz, etc.) With the phase inverter in there, the feedback becomes negative, so you end up with *resonances* at 500, 1500,... and valleys at 1000 Hz, 2000 Hz, etc.

So all this means that with a delay time of x sec, with positive feedback you reinforce the *even* harmonics of the fundamental = 1/(2*x), whereas with negative feedback you reinforce the *odd* harmonics! Pretty cool, huh.

This is what I love about the 2101 - its so flexible in the algorithm department, that it can even be used as tool for learning about DSP! The uses never end...

But the main use is to create cool tones right? So how does this relate to cool tones? Well, the comb filter effect is one that can't be reproduced by any of the effects blocks in the 2101! (At least I haven't found a way) The flanger and phaser come close, but there's no way to turn off the modulation, so the delay time is always changing. So now you have another block for coloring your tone!

Plus the block I described takes up less memory than either the flanger or the phaser.

For more info on comb filters and how they relate to the 2101s architecture and programming, see the archived mail message Jim posted in Nov 1996 related to Comb Filters, DSP, and the 2101.

Creating a Damped Delay Module: [Jim Wintermeyer/Randy T.- Nov 16, 1996]

Add the following structure to your algorithms o reproduce a realistically damped delay/echo effect:

Note: set the Regeneration to 0% and use the mixer levels to adjust the regeneration (feedback) levels. This can also be implemented with the LP Filter & Delay switched around. This effectively recreates the effect of high-end loss due to ambient conditions such as diffuse walls or curtains which normally absorb much of the high end components rather than reflecting them back to the listener.

Flanger/Phaser/Chorus Algorithm [Chris Townsend - Jan 3 1997]:

I thought some of you might be interested in a new flanger/phaser/chorus algorithm. You wonder what is the point since the 2101 already has these algorithms preprogrammed. One reason is to learn how these algorithms work internally, but most importantly the algorithm offers much greater control over certain parameters so you can create effects that were currently impossible. I was looking through some algorithms for reverb which consisted of IIR All-Pass filters. I remember hearing that Phasers are variable All-Pass, and I thought this could be programmed on to the 2101. After some experimenting I came with this basic algorithm. For you DSPer's I changed it to a Finite Impulse Response, since I found that it was much more stable and generally sounded better. I have no formal DSP training yet so correct me if I'm wrong about this.

Here's my basic Phaser Algorithm. This single algorithm has an enormous amount of tone shaping possibilities. It is shown here with modulation delays:

At this point lets forget the above algorithm and start out with a simple flanger algorithm. Keep in mind a flanger is just a variable comb filter or delay. I think there is an article on the 2101 web page about comb filters for those who are interested.

For a classic flanging effect set the Modulation Delay Depth between 0.50 to 2.0 milliseconds and the Modulation Speed between 0.1 to 1.0 HZ. This is very similar to the flanger module in the GSP2101. I think there are some noticeable differences in sound quality, but it is hard to tell. This my due to more granularity in the Mod Dly compared to the flanger. This can also be used as a Chorus effect by increasing the delay to about 10ms and increasing the depth to around 5 to 10 ms. The point is to understand that a Flanger is just a variable delay (aka Mod Dly). This knowledge is helpful in understanding the new phaser and flanger algorithms.

Now with a few changes to this algorithm you can create another very neat effect I call an "inverse flanger".

This algorithm subtracts the delayed and original signal so keep in mind that when the delay approaches zero the output will disappear. If this is not desirable try adding negative feedback or small amounts of delay to get some truly original sounds.

Here's my example, but remember there are many other possibilities.

Inverter = Out of Phase

Delay = 0ms

Fdback = -40

Speed = 0.50 Hz

Depth = 1.75ms

Wvform = Tringl

Also give this algorithm a try.

Now back to the original phaser algorithm, but implemented using Flangers to conserve memory. To create a standard phaser effect each flanger should have the same settings.

for example:

Delay = 0ms

Fdback = Off

Speed = 0.5 Hz

Depth = 0.75ms

Wvform = Sine

Also all mixer inputs and outputs should be set to 100, and the Phase Inverter should be set to "Out of Phase". This phaser is similar to the Phaser module built in to the 2101, yet there is definitely a noticeable difference in sound quality. I am curious as to the exact implementation of Digitech's Phaser module, but I would assume that's top secret. Remember that if you want to make it now sound little more chorusy (what a neat word) just increase the delay and depth about 5-10 ms.

Next we can create an All-Pass filter without any modulation. This one of the many effects that can't be done using the preprogrammed modules. For example:

Delay = 2ms

Fdback = Off

Speed = Don't Care

Depth = 0.00ms

Wvform = Don't Care

Remember to set both flangers with equal values or you will get something sounding more like a comb filter(Non-modulated Flanger). Some of you maybe wondering what is the point of an All-Pass Filter. If an All-Pass Filter passes all frequencies equally then it shouldn't change the sound at all. Well actually this only true for steady state inputs like a sine wave. But since most of us don't like to listen to pure sine waves all day there is a noticeable effect on the sound we listen to. Also it should be noted that All-Pass filters produce a non-linear phase shift, that is different frequencies are delayed different amounts(please correct me if I'm wrong about this). Though at this point in my college career I don't know how calculate exactly what the phase response is.

Now that you can create an "ideal" phaser, it is possible to take it another step further. For example what if you used different waveforms or Speeds for each flanger. This creates something that kind of sounds somewhere in between a flanger and phaser. Also you could change the settings on the mixers to create different effects. For example lowering input 1 on Mixer 2 will add some flanginess to the output. Or you could add a tremolo to vary the volume and create an effect that will "Morph" between a flanger, a Phaser, and everything in between. This seems noticeably different than "panning" between a phaser and flanger module, because it just mixes two signals. Also I have noticed some modules like tremolo and AWah seem to weaken signals even when disables by about one half, so you'll need to adjust the mixer levels to compensate. I'd appreciate it if someone could verify this.

Or you could try adding autowah.

Now if you take the basic Phaser algorithm and switch the inverter to "in phase" This will create a dual stage flanger, for some more interesting effects. Keep in mind this NOT the same as a two flanger modules in series. Then if you disable "Flanger 1" you will be back to a single stage flanger. possibilities are enormous. I would love to see some other inventive patches using this basic algorithm.

Another topic of consideration is feedback. I have found the best way to add feedback is using the feedback parameter on the Flangers. It is also possible to create an algorithm that sends a slightly delayed output back to the input. This does work, but it is very unstable even at moderate levels of feedback. It might be interesting to experiment some more with this.

Stereo should be fairly easy to implement also. I think the best way would be to use 4 mono flangers (as opposed to two stereo flangers) which would allow independent adjustment of each channel.

Some questions I have are about the LFOs in the modules. Mainly I am curious know whether they run in sync. For example if I change one of the speed to .53Hz and the other to .50Hz soon they will become out but of phase, I assume that if I change the speed so both are .50Hz one will likely be out of phase, but what if I save it and change to another patch and then switch back. Will the LFOs start in phase when I switch back to the first patch. I have done some listening test and they it sounds like they sync up when you change patches, but it is still hard to tell.

Also I have noticed that each module has a little bit of delay in it. For example when I disable Flanger 1 and make Flanger 2 have a depth and delay of 0.0, it should cancel and have no output. In fact it never totally cancels out. Mainly high frequencies get through which imply that everything is not in phase. It seems strange to me that a delay when set to zero or the mixers would have even the slightest bit of delay. This means that this algorithm will never be an "ideal" phaser, but maybe that is a virtue.

Have Fun, Chris Townsend

p.s., I wonder what this will sound like? Too little time, and so many possibilities.

Harmonizer Tracking Problems [Ryan - Feb 25 1997]:

A: Make sure you feed the pitch shifter a good tone to begin with. I had problems with it when using a scooped global or analog EQ. The Analog EQ comes BEFORE the harmonizer, and it had trouble locking onto the notes with significantly less midrange.

p.s., A "scooped" EQ is an EQ with the highs & lows boosted and the mids cut making the shape of a "V" (vee) or "U" (yu).

XLR Output Overdriving Mixer Inputs [Randy T. - Apr 9 1997]:

A Solution...if you don't have pads or MIC/LINE switches...

Make/buy XLR to Tip/Ring/Sleeve adapter and plug into the LINE inputs of your board instead of the XLR. REMEMBER there is nothing magic about the XLR plug...it is just a connector...if you use an adapter as mentioned, there is NOT any quality loss or added noise problems...it is still a balanced line level signal.

Secret 2101 Designer's Screen [Sven Peterson - Apr 9 1997]:

Press and hold the "2" button while powering up the 2101. A screen will appear with names of the Digitech GSP-2101 team members and some quotes! Not overly useful, but interesting!

Written by Curtis LeMay Pastor,
CLPastor@aol.com or clpastor@ee.tamu.edu
Last revised as of Oct 8, 1997.