Mel9 Tape Replay Pedal

[Shamanczarek]

The video demo is done with a guitar but presumably most any Keyboard could be used with it. They say no samples are used but don't tell you how it works.

 

[video:youtube]

[ABECK]

Could be an interesting way to blend in some additional sounds in a one keyboard rig. But, this really does seem tailored to guitarists. Though, I'd be curious to see what it would sound like on vocals, or even bass. Could be interesting.

[ElmerJFudd]

Neat. In this form factor it must be digital. It's not actual tape, and it's not samples being triggered. No audio to midi conversion then either...

 

I'd say it's taking the incoming waveform and altering its attributes to contain tell tale aspects of brass, reed, strings, and voice. So it's probably like a vocoder synth. The incoming instrument is the oscillator and the sound is being run through 9 different presets on the typical subtractive synth - filter/cutoff/resonance, asdr, some modulation, etc.

 

The brass and reed sounds would be killer in a guitar trio where no horn players or keyboards are available. Especially on tunes with great unison lead lines like Sir Duke or Jungle Boogie. I wonder if it would be as effective as a layered synth with piano or organ patch as the source sound.

[zoooombiex]

I think it uses the same general approach as their POG/HOG pedals, basically as you described - it takes the analog waveform and then severely processes it to emulate the different voices.

 

The POG works great with a rhodes (how I mainly use it), so I imagine piano patches would be fine. Never tried it on organ or synth though. I have one ordered, so I'll report when it comes in.

[Danny Linguini]

Cool looking little device. You should get two -- one for left channel and one for right, with Orchestra on one side and choir in the other. and of course 4 channels in your mixer just for one keyboard. I wonder how that would sound through a SSv3...

[GovernorSilver]

EHX hired one or more engineers that used to work for Eventide. I believe the HOG pedal was the first product of that hiring.

 

Then they brought Bill Ruppert onboard, and he quickly started cranking out the Effectology video series on Youtube. He's not an engineer, but a creative musician who's known for years how to simulate various iconic sounds (Mellotron, the "Won't Get Fooled Again" sound, the "On The Run" sound, etc.) with a guitar and mostly standard guitar pedals. It was years ago that he posted a video on how to do the Mellotron flute sound with just a POG2 and chorus pedals.

 

This MEL9 is just the latest brainchild of the combined genius of the ex-Eventide guy(s) on the engineering side, and Bill Ruppert on the creative user side.

 

It would be neat to see someone play a Rhodes, Wurly, or Pianet through one of these.

[JeffLearman]

I think it uses the same general approach as their POG/HOG pedals, basically as you described - it takes the analog waveform and then severely processes it to emulate the different voices.

I doubt it. I've done a lot of subtractive synthesis and that's not how you get those sounds. Furthermore, it doesn't explain sustain. It's pretty clearly triggering waveform generation (synthesis). The cello certainly sounds like a sample. Of course, it does use some subtractive synthesis, which is obvious from the trumpet slow attack demo.

 

I got a laugh out of this: "There are no triggering issues because it doesn't use samples." That implies that triggering problems only happen with samples, which is pure BS.

 

It's a cool little device, regardless. I bet it's limited to 6 voices.

[GovernorSilver]

I got a laugh out of this: "There are no triggering issues because it doesn't use samples." That implies that triggering problems only happen with samples, which is pure BS.

 

It's a cool little device, regardless. I bet it's limited to 6 voices.

 

Well, he did have to dumb it down for the guitarist audience. Out of the few guitarists who are using MIDI guitar devices, most of them are using romplers as the sound source - hence the reference to samples. Only an EHX employee would know 100% of the details but from what I recall of what Ruppert has said, it works like an octave pedal, except with much greater sophistication. The octave note on the octave pedal is synthesized based on the incoming pitch. It's a pretty old technology, appearing in 1970s fuzz pedals and the like. The advent of octave effects implemented in the digital domain led to more stable synthesized octave notes (eg. the original Whammy pedal), then eventually to working polyphonic octave effects like the HOG and POG.

 

I also recall asking Bill if the HOG sampled the input audio and he insisted it did not. I'd have to say he's right as I've run electric cello and viola through the HOG as well as guitar - no difference in effected sound.

 

Yet another reason I'd love to hear a Pianet, Harpejji, Stick, etc. through this pedal - it'll be a good test of this pedal's true "polyphony".

[ElmerJFudd]

Well. If you're triggering samples then the incoming pitch/frequencies have to be analyzed and converted into a midi note value so the appropriate sample(s) get called up for play back. No doubt this is where the most errors occur in tracking pitch. In some cases there is a history of opting for a unique "Midi pickup" for improved accuracy and speed, although most recently software has removed the need for it.

 

Synthesis makes the most sense, I think they are simply running the incoming audio through filters and an amp with adsr (and the sustain is being made in this stage). The filters are enough to shape the tone to reedy, stringy, brassy, etc. And there are 9 presets because it's a stomp pedal and I'm guessing marketing research suggests being highly programmable would not be appreciated. Simplicity is preferred.

[GovernorSilver]

There are a few guitarists who are using MIDI guitar devices like the Fishman Triple Play to play non-romplers like an Access Virus or Minimoog Voyager. But they are the minority of a minority . The majority of the minority use romplers because they're playing the keyboard parts in cover bands (and claiming they can't hire a keyboard player due to lack of availability or funding).

 

The incoming audio is analyzed only for pitch and envelope information to control the synthesized tones in the EHX pedals. This can be done so quickly in the digital domain when MIDI is left out of the equation, that a human won't hear any more latency that he/she would in a computer DAW. The full audio is not passed in. Otherwise, I'd get a difference in sound between playing guitar, viola, and electric cello through the pedal.

 

 

 

[ElmerJFudd]

Oh, is that the case? The timbre of the synth is the same regardless of inputting instrument? That's intriguing. Cool, not as cool as if it was actually using the source audio, but cool in its own right. Then I assume when running through the pedal, there's a parallel send and/or true bypass for the guitar sound to amp.

 

Hmm, this is an interesting approach for synth sounds where the source keyboard doesn't have Midi out. But that's a rarity with the oldies like Rhodes, Clav, and pre midi synths being the most likely candidates.

 

 

 

 

[GovernorSilver]

Oh, is that the case? The timbre of the synth is the same regardless of inputting instrument? That's intriguing. Cool, not as cool as if it was actually using the source audio, but cool in its own right. Then I assume when running through the pedal, there's a parallel send and/or true bypass for the guitar sound to amp.

 

That's what Bill Ruppert said in his explanation of how the HOG works, and of course my experience seems to bear that out. I can only speculate the MEL9 works on similar principles (analysis of audio input for pitch, amplitude, and envelope, which is then applied to a synthesized tone instead of processing the audio). It'll be interesting though if someone can post proof of the MEL9 sounding differently depending on whether guitar, electromechanical keyboard, or bowed instrument is input into the pedal.

 

The MEL9 has a Dry knob sets the output level of the dry signal. I don't know if it's true bypass. My EHX DMM 550TT (delay) has a simple Mix knob for wet-dry mix control - interesting they went with separate level controls instead on the MEL9 and other 9 series pedals.

 

EHX also makes the Switchblade/Switchblade+ pedals for more signal routing options. They're marketing it of course as a solution for guitarists who want to run the MEL9, KEY9, and other 9-series pedals in parallel, to parallel amps.

[zoooombiex]

I also recall asking Bill if the HOG sampled the input audio and he insisted it did not. I'd have to say he's right as I've run electric cello and viola through the HOG as well as guitar - no difference in effected sound.

 

Interesting - I run my POG with rhodes & guitar and there is definitely a difference in the wet sound when playing the same notes. The guitar has more harmonics in the dry sound, and even when you turn the dry level off on the POG those carry through to the wet sound.

[zoooombiex]

Got mine today. Some clips are below. I've only used it with keys so far (EDIT: guitar comments added at the end), but in terms of first impressions:

 

1. Some of the simulations are pretty close - cello, strings, and voices. The sax and brass not so much. I have a M4000D and I did a quick comparison to the Mel9. Clip below. i went for the most similar patch I could find on the M4000D, not necessarily my favorite patch. Note in the brass comparison there is a patch that's pretty close to the Mel9, but then at the end I briefly played another patch on the M4000D that is much more natural sounding, just for contrast. And I don't have an "orchestra" patch on the M4000D, so I skipped that.

 

I divided the instruments between the left and right channels. The M4000D starts first for each instrument:

 

 

2. It has some polyphony issues. It's not so much the number of notes, but the complexity of the chord. Anything with a half-step interval is problematic, as is a natural 7th. Sometimes it gurgles and sputters out, particularly with a half-step, and sometimes (particularly with a 7th) it generates a tone an octave lower in addition to the octave up, as though it is confused about the lower note you were already playing. I made a clip showing some of these issues. For most basic chords it doesn't have any issues though.

 

 

3. In terms of triggering and tracking, it can be a little touchy. I'm using a rhodes that's decently - but not perfectly - regulated. Volume variations that are fine with the natural sound of the rhodes can sometimes lead to drastically different levels (or no triggering at all) with the Mel9. Also, on lower notes, it takes a fraction of a second to catch the right note - see the beginning of the polyphony clip where there is a slight slur to the individual notes at the bottom.

 

Tracking clip:

 

4. It sometimes has a high-pitched clicky-sound. You can hear that a lot in the polyphony clip. It's worse on the complex chords that seem to confuse it, but it also shows up in other places. And it doesn't seem to be clipping, because (a) the output level is at a normal instrument level that doesn't clip my other effects, and (b) it doesn't always happen at the beginning of a sound when it's loudest - sometimes it comes in after a second or so. I need to investigate this further.

 

Overall, a nice pedal so far within its limitations & design parameters. I don't really use these sounds to try to mimic the actual instruments, but more as an abstract/alternative voice, often manipulated through effects. I think for that type of use, it will be great. (E.g. if you hold quickly cycle through the rotary knob of voices while playing, you get some very interesting results.)

 

When I get another chance I'll see how it does on guitar.

 

EDIT: I had a little time to try it on guitar, but not to get any recordings. I think most of the issues noted above are still present on guitar, but not as prominent due to the nature of the instrument.

 

-If you play chords down low on the neck with a lot of close intervals, it will still gurgle and fizzle. But higher up on the neck it handles them better. You almost have to try to confuse it on guitar, whereas on keys (more simultaneous notes, easier to combine close intervals) you almost have to avoid it.

 

-tracking is about the same. on very low notes, you can still hear some 'slur' right as it grabs the note. but about the same speed overall as a POG or Ravish, as I suspected. interestingly, if you half-release a note, you can often get it to slur back down a little, like you pushed it flat.

 

-the high-pitched clicky noise is still there sometimes. not sure what that is.

 

-i wish the sustain control had more range on the low end. a mellotron has an extremely abrupt cut-off when you release a note, but on the Mel9 even with the sustain all the way down and muting notes as fast as possible, there is still a little decay. I noticed this on the rhodes, but thought maybe the dampers weren't muting the note fast enough, but it's the same on guitar.

 

-some instruments (i'm thinking of sax in particular) have a window on the neck where they sound pretty similar to the mellotron, and then outside that they get a little funky.

 

-from a overall playing perspective, i was left with the feeling that the pedal mated better with the guitar, but with a lot of interesting potential on both.