Darker repeats for 1776 Multiplex?

[Craiz]

Hey folks - I've got a 1776 Multiplex in the works, and I'm super pumped to have it done. However, one thing I was considering was that the trails seem to have a pretty significant high pass going on, so basically all of the bass is getting removed after a short amount of time. I tend to like darker repeats, so I was wondering if there were any modifications (part swaps, simple add-ons, or whatnot, considering it's a prefabbed pcb) that I could do to make it have a less severe high pass (or even better, a slight low pass)?

Here's the build doc for reference. http://1776effects.com/wp-content/uploads/2015/03/Multiplex-Build-Doc-1.7.pdf

Thanks!

[cortezthekiller]

I would think that raising the value of C12 to 0.1uF or 1uF would help to add bass into the repeated signal. If you wanted to cut more high frequencies as well I would try raising the value of C20 perhaps.

[notnews32]

Did you build up the pedal from a printed PCB or from a stripboard/perf layout? There are more than a few ways to accomplish what you're after, and they'll each be more or less difficult  to implement depending on how you build up the circuit. It sounds like you're looking for a way to dial in your desired amount of high- and low- end frequencies in the delayed signal and not the dry signal, right?

You can add a passive tone control circuit to the wet signal path, with a tone control pot that essentially sends either the high or low frequencies to ground (or using a dual-gang/two pots to control high and low freqs independently). Depending on how it's implemented you may experience a bit of volume loss as signal to ground is lost, which is intrinsic of the passive tone control scheme. You can tap into the power lines and add an active tone control, and again there are tons of different flavors here too. A bit more complicated on a PCB than a stripboard or perf, but potentially do-able.

If you read the PT2399 datasheet and compare it to the Multiplex schematic, you should find a few circuit components responsible for filtering the delay lines. Simply swapping those for gradually increasing/decreasing values you just keep reiterating that process until you land on the sound you like. Unless you implement a pot as a variable resistor here though, you won't be able to tweak these settings once the circuit is boxed up. Unless I use the same amp and same guitar and same effects every night, I find myself wanting on-the-fly control over things like EQ/filtering, so I use pots and external controls as much as I can. The pins of the PT2399 designated by "LPF" are involved in the filtering of the delay lines.

Let me know what you end up doing and how it turns out! Just figured I'd provide some generic info.

[PRR]

> raising the value of C12

C12 could be a LOT bigger. As Killer says, several tenths of uFd. 1uFd may not be excessive.

"Too much" bass reverb is muddy. (Especially on mechanical springs.) However not-enough reverberation in the bass is un-musical. "Good music rooms" tend to have more reverb below 500Hz than above (seat stuffing absorption), and a decline above ~~~2KHz (seat upholstery fuzz absorption). That 22nFd suggests a bass-cut as high as 1.5KHz, which may be a fine high-halo of reverb, but isn't going to fill up the bottom.

[Craiz]

Thanks so much all! This has been really helpful.
> C12 could be a LOT bigger. As Killer says, several tenths of uFd. 1uFd may not be excessive.

Does this apply to all pt2399 delays/reverbs? So for instance, if I wanted to have a lighter/darker repeats switch on a delay, I could switch between a smaller/bigger value cap in that placement?

[Mark Hammer]

"Darker" repeats would imply that not only is it NOT getti8ng thinned out at the bottom, but that treble is being trimmed at the top on successive repeats.  If that's the goal then not only increase C12, but raise C25 up to 68nf or even 100nf.

[Craiz]

Thanks very much, mark - that's exactly what I was wanting.

[Mark Hammer]

My pleasure.  This is a mod I do on every delay that permits it.  I will insist to my dying breath that "in the real world", repeats lose top end as reflections are eaten up by imperfectly reflective surfaces.  Progressively "dulling" repeats not only makes them more realistic, but also prevents the sound from getting too cluttered, by moving the repeats to the psychological background.  In many respects, it's like the way one adjusts one's tone so as to not conflict with the other instruments in the band, and very much like the way birds tend to adopt pitch ranges and songs that can each be distinctly heard despite how many are being chirped at once.  Having the repeats stake out different parts of the spectrum than the main signal prevents the "now" from getting confused with the ambient sound.

Don't forget to tell us whether you felt this component change was an improvement.

[PRR]

> insist to my dying breath that "in the real world", repeats lose top end as reflections are eaten up by imperfectly reflective surfaces.

We can imagine, and build, rooms with dead bass and live treble. But this usually requires specialized absorbers. And sounds uncomfortable. Like your head inside a plate-steel boiler.

Well-tuned live-music rooms have rich (reverberant) bass and clear (low reverb) treble for clarity. Concert Halls and Opera Houses, Leo Beranek, discusses this but at great expense ($50!). This is a massive revision of a 1962 text, which I can't find. Concert and Opera Halls: How They Sound may have much the same thoughts for $7 used. Dear Leo is apparently still alive, 101 years old. (I worked very briefly with his senior associate on a project where Leo may have done the broad strokes.)

Fig 2 and 3, page 10, of this paper shows target and typical reverb times:
http://www14.in.tum.de/konferenzen/Jass06/courses/4/Flegontov/Flegontov.pdf

Large room for symphonic music, 2.7sec @ 125Hz, 2sec @ 500Hz, 1.7sec @ 2KHz, and typically sloping to 1sec @ 8KHz. This may not be many dB of treble loss per-loop. But bass-loss is probably wrong for full-range reverb.

Guitar spring-reverb usually has heavy 500Hz bass-cut, because springs can't take big bass, and because small clubs can be muddy.

I'm not sure what is "ideal" acoustics for modern rock/pop. Chamber halls "need" rich bass because acoustic bass and cello are not strong instruments; 1,600 Watt bass amps may have changed this. I was always happy, for jazz and singer/guitar acts, to have a room you could "speak" in; not rolling in reverb so much that wordssoverrlapppeddd; traditionally 0.6 to 1.0 seconds mid-range, and no shrill top-echoes.

[Mark Hammer]

You've essentially outlined why, despite the typical implementation of 6 and 8-pole (or more) lowpass filtering to keep audio grime out of the delay path, I find a simple 1-pole lowpass in the feedback loop does the job.  One still needs to leave something at the upper end of the passband to continue to trim away upon successive repeats.  Too steep a filter in the feedback loop and it probably would sound "unnatural".  The objective is to make each repeat just noticeably duller than the previous one.

[PRR]

The many-pole is needed to keep out whine. Ideally it will be far above the audio band, but this is only the 21st century so we don't have audio memory any bigger than a 1980 keyboard buffer (why is that??) so we have to under-clock to get any real delay at all.

On top of that, a 1-pole from midband takes the zing off the top, like carpeting the walls. For many "nice (fake) rooms", you may want *less* than a full-pole hi-cut. Shelf-filter is a start. If you have 30mS of delay and want 2 seconds RT60 bass, 1 sec treble, you want 60dB off in 30 passes which is only 2dB per pass at the top of the band, less in midrange. Perhaps 5K series, 22K+5nFd shunt. That's unity-gain for bass and -2dB above midrange, with -1dB somewhere midband.

However with a just-too-short delay, a simple 1-pole does put a small slant all across the band and may be enuff to deaden whine.

[Mark Hammer]

Well now you're talking true reverberation and ambience.  I'm just talking about making a simple echo box sound a teensy bit more like the real reverb world.

[Craiz]

So I finally got back to a computer - Thanks again for the help all. However, raising the value of either C25 or C20 to try and get successively darker repeats does not change the character of the repeats at all, but rather higher values lower the feedback. That's the only discernible change as far as I can tell. Would another cap value change do the trick instead?