When the DL4 Delay Pedal first shipped in 1999, Eric Klein was working at a guitar store in Tucson, Arizona, and, like many guitarists, couldn’t wait to get his hands on it. Today, he’s the chief product design architect at Line 6 and is part of the team behind the DL4 MkII, an updated version of the familiar green box that sometimes seems as if it’s a requirement on every pedalboard. As Klein admits in our conversation, it was a serious challenge to both keep the sometimes-quirky vibe of the original and update it with new delays, “secret” reverbs, MIDI control and more to suit the needs of today’s players at the same time. Judging by the response from musicians everywhere, Line 6 hit a home run.

How long have you been at Line 6? What are your day-to-day responsibilities there?

Eric Klein: Twelve years, and I attempt to oversee the product design process for Ampeg, Yamaha Guitar Group and Line 6 products. Sometimes it’s trying to spin a lot of plates to make sure that things are consistent, easy to use and look good, so the customer is happy. No Papyrus font or things like that.

You’re a guitarist yourself?

I’m a hack but, yeah—I play guitar, bass and keyboards, but just well enough to get them into a computer, and then I chop them up.

The original Line 6 DL4, which came out in 1999, was part of a larger family of other pedals that included the FM4, MM4, DM4 and AM4. What was the idea that drove the DL4 and the other pedals?

At the time, there just weren’t any delay pedals that had a multitude of different types of delays built in. To my knowledge, it wasn’t going to have a looper initially. But they added the looper, and it just kind of snowballed into this thing, and halfway through, the team realized, “Wow, this could actually be a big, big deal.” And it was.

I wasn’t here for when the DL4 and other pedals were released, but I imagine it was really exciting and probably a little nerve wracking at the time. There are stories of people testing them out, and seeing how the tape delay, for example, really behaved like a tape delay. When you cranked up the feedback and changed the time, you’d get that self-oscillation, and it squealed the way you’d expect, and that coming from a digital delay was really sort of new. Then the ability to do all these other different types of delays as well was pretty magical. The whole notion of the expression pedal—and morphing all five knobs and being able to do things that were really unique then, but that are considered standard in a lot of pedals these days—I imagine it must have been nuts.

At the time, I was working for a guitar store in Tucson. I had ordered one for myself and was waiting for months, and when I finally got mine in, I plugged in, and it totally changed my sound. I mean, I was running keyboards through it, organs through it, guitar, clearly, drums. Whatever I could, I would get it through the DL4. I actually had it inserted into my audio interface at one point so I could route it through at any time. But it was a huge deal to me. Like a decade before I was even considering going into this side of the industry.

The DL4 was a very creative and inspiring tool, because, for a lot of players, they’d never had access to all of those different types of delays.

Absolutely. I mean, I always like to say this about any multi-effects or modeler: If you’ve never had a chance to really play the real thing, it allows you to find out what you like. If you’re not a hundred percent sold on the modeling thing, they’re still great for sort of educating you on what type of sounds you like. Like, “Oh, this is awesome. I use nothing but the Roland Space Echo sound. I’m going to go buy a real Roland Space Echo, because I love it that much.” And that’s actually what I did.

Originally, it was all about having a gigantic palette of sounds at your fingertips. Since then, it’s morphed into more of a performance aspect, where all of the other advantages to digital products have come to the forefront, because digital has been there for years now. The ability to switch between delays, or to be able to hit a footswitch and toggle between multiple parameters, or to recall four pedals via MIDI instantly—things like that are really allowing people to perform in a way that they haven’t even tried to in the past. We have some bands, like Twelve Foot Ninja, Spotlights and others, who claim they really can’t do what they want to do live without this type of technology. They just wouldn’t be able to pull it off unless they had a whole army of stagehands backstage trying to switch everything in and out. Now it just kind of just happens automatically.

That’s where I think we take for granted some of that utilitarian quality of it—of just getting the job done—let alone the inspiring sonic quality of it. You talked about the looper coming a little later in the development initially. Do you think that any of the original designers anticipated the looper aspect of it becoming such a big deal?

The impression I’ve always had was that the looper was more like, “Hey, we can do this. We have enough memory. How cool is this?” And, “Oh, yeah, this is cool. We should do this.” Now, there are bands like Minus the Bear and others who use it as a musical instrument. It completely changes how they approach the guitar. It changes how they approach songwriting. It changes the sonic thumbprint of their band. It’s really exciting to see that kind of stuff.

We’re now 20+ years on from the original DL4. Why was it time to make the MkII?

I can’t remember if [Line 6 co-founder] Marcus Ryle came to me in the hallway, or I went to Marcus in the hallway, but I think it was 2017, and we said, “Oh man, we should do a DL4 MkII.” At the time, DL4 was still selling really well. There’d be little conversations in the hallway here and there, but we decided that if we were going to do one, we should do one for its 20th anniversary in 2019.

Then all sorts of things happened—projects got shuffled around, COVID hit, everybody’s working remotely. So, it didn’t actually drop until three years later. It was a long project that started and stopped here and there, but the original impetus was to have it be a big hurrah 20th anniversary thing that just didn’t happen, unfortunately. Nonetheless, it was a good excuse to make one.

The new one is built on HX technology, right?

Yes, with a lot of manipulation to sort of embrace some of the “tweakiness” of the original, too.

We have these pillars that sort of become the North Star of every project, and the number-one pillar for this particular product was that it needed to behave exactly like the original DL4 in many ways, or it had to be objectively better, so no one would complain, “Oh, I should have gotten the original one.”

We were really concerned about that. We wanted to make sure, when we included the original 15 delays on the MkII, that they sounded like the original DL4. The challenge was that the original DL4 doesn’t run on any type of modern SHARC or ARM architecture, and it wasn’t using the modern type of coding we use. Although the MkII is using a variation of the HX modeling that we have in HX Effects, HX Stomp and Helix, we had to do a lot of work to make sure those original delays were behaving the right way. “All right, what parameters need to be mapped? How do we map the knobs?” Because the knobs work differently on Helix products than they do on DL4 and other products. It wasn’t nearly as simple as just, “Okay, we’re going to just port these models into here and, oh, we have the old code. Let’s just bring them in.”

Speaking to those 15 original delays—was this a ground re-up revoicing against the original. For example, you’re sitting with the new algorithms and comparing them next to the original, and saying, “Does it have that same taper on the knob, etc.,” right?

Exactly. At one point, in lieu of the additional reverbs, we had planned on actually modeling the op amps and the A/D and D/A converters in the original DL4 to actually give it that tiny bit of grime that you get from a typical 1999 digital product. And we did enough listening tests, and we couldn’t hear a difference there. So, it’s objectively sonically cleaner in the MkII. There didn’t seem to be any magic or mojo in that old sound that we felt warranted the additional DSP hit. And then when we looked at it, we realized, “Oh, there’s enough DSP for reverbs in here now, so let’s do that instead.”

Can you compare what you were able to do with the modeling tech for the original versus what the HX modeling tech allows you to do today?

HX was built up from the ground up. The legacy effects … those didn’t come in until right around when HX Effects dropped. It took that long for team members to actually port all of our legacy effects to the new platform.

It’s an all-new platform, with quite a bit more DSP. That allows the DSP engineers to create tools that allow the sound designers to go in when they’re capturing the specific idiosyncrasies of old tape delays, or platter delays or bucket-brigade delays, to emulate the exact sort of behavior of that bucket-brigade chip. They’re looking for what happens when you crank the feedback and it goes into self-oscillation, “Does it actually clip the right places?” They’re doing that sort of thing.

The more DSP you’re able to throw at it, the more of those weird idiosyncrasies you can recreate. The things that don’t make a delay clean, the things that make it sound angry and unique, and cool and kind of vibey. Those were the hard things. Making a clean, pristine, awesome delay is relatively simple, it’s just DSP. But to really try to nail the weird, quirky things that were originally thought of as “bad” in older products, those are actually harder to do, and where all that DSP comes in.

Is it the type of thing where you not only have more precision, because you have more DSP to work with, but also that you can model more stages? What I mean by that—on an earlier product, if you wanted to capture the sound of a delay, you had, let’s say, four DSP “blocks” to capture that character—for example, the input stage, the delay stage, the output stage, and one or two other little idiosyncrasies in between. But now, it’s like you have 10 different DSP “blocks” that you can devote to modeling even more of the components of the signal path to capture the unique qualities of each delay?

Yeah, that’s a really good way to describe it. Sometimes, it’s not just a specific component that needs to be modeled, but there’s actually behavioral modeling going on as well. Whereas before the processor might allow for a linear travel of a particular set of parameters, now we’re able to do complex curves that emulate how the original knob behaved, or the interaction between different components. All of that needs to be accounted for as well. It’s not just putting together Legos. It’s like Legos with a bunch of like, weird, sinewy, nebulous circuitous wires between them that all need to be accounted for.

With all this new DSP power, you can account for that?

Yeah, exactly. And sometimes we’ll be surprised—for example, you crank the treble on a delay, and then you crank down the time, and you get this weird chirp on the original delay. But that chirp doesn’t show up in our model. Alright, so what’s causing that chirp? Why is it somehow tied to the treble knob? Where is the disconnect? How do we recreate that? Sometimes, we have to tear it apart and figure out where it happened. Sometimes, it’s such a weird anomaly where we say, “Okay, that really doesn’t need to be there.” Or, “Wow, that takes up a lot of DSP to pull that off, and nobody’s going to care. In fact, it is sort of an objectively bad thing, anyway.”

So, we get rid of that little thing, because nobody is going to turn the knobs in that exact fashion and hit this chord in the right way that’s going to make the weird thing happen that we can’t somehow recreate. There is a bit of balance in what we’re trying to accomplish, and that’s where our sound designers are so great, because they know what’s appropriate and what’s not appropriate. They know what’s doable and what’s not doable. And the genius of our DSP team is that they’re able to tweak their code to allow our sound designers to do that sort of thing.

While we’re on the topic of unique idiosyncrasies, are you typically working with one “golden unit” as your point of reference, or multiples? So, if you’re modeling a Space Echo, are you working with five of them and looking for consistencies and inconsistencies from one to another? Do you run into one unit that does something special, even if it's wrong, and go with that?

I wish. There are cases where we have multiples of specific distortion pedals, sometimes multiples of different amps, and we’ll shoot for something right in between the two if we feel both of them are representative and sound great in their own way. In that case, we might take the best characteristics of one and apply it with the best characteristics of another. But original Space Echoes are really hard to come by.

Well, that’s just one example.

Totally. That’s one example where we just wanted to find one in really good working condition. Now, we’ll have to go in and fix things a lot of times, and we’re fortunate in that we live in a city [Los Angeles] where we have some extremely talented techs who can go in, tear things apart and make sure they’re in pristine working order.

When you’re creating your models, how often would you say you’re consulting actual schematics versus just kind of working within what you’re hearing?

If the schematics are available, a hundred percent of the time, because they can shorten the process quite a bit. It’s rare that there’s no schematic to go by.

On a simpler pedal, whether it’s a delay or a distortion, a lot of engineers will open it up, look at the traces and go, “There’s that op amp. There’s that diode. I know how to do that. We have a library for that.” And then we’ll check it against what we have in our library. “Okay, that is the diode we think it is, whether its gooped [encased in epoxy] or not.”

All of that said, we try to, at the very least, use a schematic as the initial roadmap at all times.

Let’s talk a little bit about decision making behind the models on the new MkII model. How did you do that? You knew you wanted to keep the originals in there, so was it a question of figuring out what would be most complimentary?

There are a lot of pieces that go into why we do or don’t do certain things. Sometimes, it’s just that an engineer has a really cool idea and says, “Hey, what do you guys think about this?” And we’re like, “That sounds really interesting.” So, they’ll come back the next Monday, “Hey, I worked a little bit of this on the weekend. What do you think?” “That sounds awesome. What if we did this and this and this?” “Oh, yeah, let’s do this,” and it just turns into something original and unique.

The more unique delays in DL4 MkII were largely done by Dan Godlovitch up at our Victoria team. The Victoria team were all ex-DigiTech. I mean, they’re all just brilliant, and Dan has a really creative mind. He’s always coming up with really, really cool ideas that are unique, and in some ways, as complex as standalone boutique pedals. We’ll go back and forth on a lot of stuff.

Interestingly, I don’t think too many people noticed. And if they did notice, they didn’t reflect this online. But we were releasing a lot of delays in Helix updates. I’m really astounded there weren’t a lot of people saying, “Oh, there’s got to be a DL4 MkII coming because there’s so many delays coming out.” It’s like, “Guys, it’s obvious what we’re working on!”

That’s interesting that you bring that up. Obviously, those are things that were happening in parallel—the development of the DL4 MkII and the Helix updates. Were you doing any sort of testing the waters as you were releasing those delays?

Yeah, I think we knew we were going to do a DL4 MkII. We didn’t know exactly which delays were going to make it in. We built the infrastructure, and I think every delay was able to make it in. There were a couple of reverbs that just wouldn’t fit, so we couldn’t put those in. As far as the delays, in many cases, we were specifically creating delays that were unique and cool, and told a story for Helix-sustaining updates, so they could also end up in DL4 MkII.

We try not to do too much work that can’t be shared across a wide variety of products just because it’s expensive—engineers are expensive. The more love we can share, the better.

So, you’re talking quietly releasing new delay models out in the market with Helix updates, and the DL4 had been out now 20+ years. I’m sure there was no lack of feedback to work off of as you developed the MkII, right?

No. And part of it is negative feedback! [Laughs]

How did you factor all of that in? You’ve got web forums. You’ve got customer feedback. You’ve got all the artists you work with, as well as your own personal experiences using it. How do you bring all that feedback together?

We’re a big team at Line 6. When we decided we wanted to make a DL4 MkII, we knew what we wanted to do. We also knew what we didn’t want it to do. We knew it was going to probably look funky, because the original DL4 looked funky. And then decreasing certain sizes, it might make it even look more funky. But we wanted it to be unapologetically a DL4, for better or worse.

Then, you look at exactly what people complained about the most. The biggest complaint was the footswitches, because the footswitches would break. Most of that was in the first several years, but the reputation kept going for the bad footswitches, which was terrible. We hated that that happened. We were able to eventually fix it.

We also looked at the mod community. There were a lot of people who were modding their DL4s with specific features, like the ability to get extra presets, which was really just toggling the Expression pedal from heel to toe so you could get almost two snapshots per preset. Or a dedicated tap switch, things like that. We thought, “Okay, well, what if we take the most popular mods that people are already doing for their DL4s, and we somehow brought those into firmware, so the user could sort of pick which mods they wanted via global settings.”

The other big one, at least for me personally, and for a lot of people, was MIDI, because there was no MIDI in the original pedals, and we do MIDI on as many things as we can these days.

When it came to the looper, we didn’t want to screw with the looper too much. We wanted it to be pretty utilitarian. There are a ton of great, amazing loopers out there, and we’re not going to pretend to do what they’re going to do, and we’re not going to try. We want people to treat DL4’s looper like a DL4 looper. We wanted a longer time. We wanted the ability to store that loop. We wanted the ability to change delays while the looper is going, things like that.

We specifically looked at every little compartmentalized set of features, looking for where we could improve things, but with this sort of looming rule to not mess with a good thing. As soon as we started pushing it past where it became uncomfortable and not “DL4-like,” we pulled back.

You’ve added an XLR jack. Was that based on seeing someone like Reggie Watts and how he was using it?

It was, yeah. Yes, it was specifically Reggie Watts, but it was also in the hopes that we could reach out to a wider variety of musicians. For example, vocalists or people who mic a real cab. People were using [XLR to 1/4”] transformers into their pedals because they wanted to mic their cab. That’s how they could get post effects for amps that didn’t have effects loops. So, all these weird little things like that would totally work and fit within our budget … told a different story that we thought was fun.

As you said earlier, there was a little bit of, “First, do no harm,” but as you were working on the looper, what were the big things you felt needed to improve?

As long as the workflow was identical, we wanted to be able to expand the looping time. So, we expanded that from 14 seconds to two minutes. For a lot of the more ambient musicians, they need way more than two minutes. So we added an SD card slot in the back, which increases looping time into many hours. And then that card will also store that loop. Now, it only stores one loop. We didn’t want to get too crazy. There wasn’t an elegant way to have a bunch of different loops stored with banks because, again, that’s not really what DL4 is.

It changes the interface at that point.

Exactly. If you turn it off and turn it back on, the loop stays in memory.

Traditionally, there was one delay tied to the looper bank on DL4, and we wanted the ability to be able to change delays at any time. We wanted people to be able to change presets while the looper is still going, so if they assign the tap switch to a one-switch looper, they can do that. If they happen to have a MIDI controller, they can do all of these things simultaneously from an external controller. I guess we wanted to keep the core functionality the same, but for power users, we wanted to give them functionality that didn’t get in the way of the traditional user. It’s kind of hidden in there almost like an Easter egg in some ways.

So, you can’t exactly trip over it and get confused. You really need to think about what you’re doing, but it is there.

Yeah. Exactly.

There are the 15 “secret” reverbs, right? Can you touch on that a little bit? Why were they “secret?”

Yep. Not so secret anymore! For a lot of guitarists, delay and reverb are kind of two sides of the same coin. They’re both time-based effects. Most of the time in a signal path, you have a reverb followed by a delay or vice versa. Looking at the presets that Helix users were making, the delays and reverbs were almost always tied together. Sometimes people would assign them to the same footswitch.

I asked our system architect Brandon Nelson one day, “Hey, does DL4 MkII happen to have enough DSP? If we don’t do the whole analog modeling of the [original DL4] circuits and stuff, could we use that DSP for reverb?” And he came back saying, “I don’t know if we can do every reverb, but we can do some of the reverbs.”

From there, we picked the best 15 reverbs that we could fit. We wanted it to be flexible enough to where the user can put it before, after or in parallel with the delay. But we knew that if we were to somehow add the functionality and the reverb selection to the top panel, the UI just would have been ugly, stupid-looking and weird, so we had to make a decision early on. “Well, let’s just call them secret and kind of hide them. Let’s put them in the manual and the cheat sheet, but not make a big deal out of them.”

For the first month or two of marketing, we didn’t really even mention the reverbs, and we let people discover them. What’s nice is that because there’s a separate mix parameter for both the delay and the reverb, you could have just a reverb preset if you turn the delay all the way down, and vice versa. Or, because the expression pedal now also automatically stores the reverb knobs, you can morph between reverb and delay. There are all sorts of really cool, interesting things you can do when you have two effects versus one. We just didn’t have the UI to kind of embrace the second one.

We’ve already touched on this a little bit, but just as far as functionally, what do the MIDI I/O and the USB provide for users?

USB is just used for firmware updates currently. We’re not sure how we’re going to use it, if we’re going to use it in other ways in the future, but it’s there in case we want to.

MIDI is new for DL4 and, besides the obvious—like the ability to select presets from standard program change messages—every knob has its own CC. Delay selection has its own CC. Reverb selection has its own CC. All of the looper functions have their own CC. Because this is a stereo pedal, we really wanted keyboardists to embrace it as well. A lot of the looper functions can be triggered from MIDI notes, which is something that the old Echo Pro rack unit actually did.

If someone was to walk into their local Guitar Center, is there a particular delay type that you’d recommend someone check out on the DL4 MkII?

What’s so nice about the DL4 MkII is that the defaults are set up to sound good right off the bat. As soon as you see that silkscreen with all the different delay types, when you turn the selectors, it’s going to automatically set everything, including all the knobs the way you want.

You may not know that Cosmos is based on a Space Echo, necessarily, but if you’re turning the knob and you can get to those 15 locations right away, you’re going to get a really good idea of what you get with DL4 right off the bat. And if it’s something that interests you, and you do a little bit more research, you may find, “Holy cow, this thing does a ton more that I wasn’t even aware of.” So that was sort of the story we wanted people to go through in their exploration of DL4 MkII.

Just turn the knob, hear the range and get inspired.