Does anyone build buffers into their stompboxes?

[duck_arse]

...... he has a day job as well?

[drolo]

measurements...
(courtesy of Jack Orman, 2012)

I am surprised (and worried) by those measurements
Did Jack maybe just happen to have pedals with really bad buffers or is there no way to avoid this?

Could increasing input and output caps and raising the gain a little on buffers counter the effect he has measured ?

[R.G.]

measurements...
(courtesy of Jack Orman, 2012)

I have this memory of a cartoon from something I read in my childhood. It showed a newspaper front page with a picture of a man in a small boat pointing at the enormous, undamaged side of a battleship. The caption on the picture in the newspaper was "Reporter point to the place where disaster struck, almost."   

Quoting myself,

There are some esoteric issues with buffers that I've delved into by necessity. To the best of my knowledge, these issues are never mentioned in the "buffers are the work of the devil" rants on the internet.
...
Good buffers, as opposed to something just tossed in, require some attention to design,

In no particular order, my thoughts on reading that set of measurements include:
- Boss pedals use a single NPN follower for a buffer - hardly a paragon of good buffer design
- 2.5db down is audible, but just barely
- the frequency response measurements are artifacts of choices about coupling caps and such, not much to do with the buffers themselves
- pretty pictures
- Jack sells relay bypass boards? I didn't know that.

Designing a good buffer does take some work. "Designing" a buffer takes no work. You cram in a JFET or NPN follower and have done with it. The Boss/Ibanez bypass setup is designed for acceptable results in isolation from a mass-produced product. Cheap and cheerful, and a relic of a time with small, uncompleted pedal setups; setups so small that they usually didn't have pedalboards, but were hooked up loose on the stage.

The article does allude to some of the issues - noise accumulation and distortion of the switches. Maybe we'll see some more on that.

Back at the office, we spend time worrying about what I call the "systems architecture" of effects systems as they're used today. This necessarily assumes that the player will put onto their pedalboard any mixture of commercial pedals, paleo-pedals from the Pedal Primitive Age, boutique wonders and DIY stuff with wires hanging out. And lots of them. Ten pedals is a small pedal setup these days. I remember a time when three was a plethora. The pedals are what they are, but how you use them can make a difference.

A good buffer, placed in the guitar or better yet before the tone/volume controls, stops the most egregious issues with signal transfer from a guitar, that being treble loss due to loading. If that were on every guitar, a buffer in every pedal would be largely unnecessary. Even with unbuffered guitars, only the first pedal in the chain, the one that the raw guitar signal touches, really needs buffered. But you can't design a pedal ignoring buffered inputs because the pedal just might be the first in the chain.

Better systems design would put in a buffer bypass switch, as artifus suggested, so that you would only have the one buffer that was actually needed in the signal chain. But there is remarkably little systems design in pedal setups, other than perhaps the guys who are paid to build pedalboards and wire them up.

As I said, there are issues with buffers. But there are issues with true bypass as well.

I am surprised (and worried) by those measurements

That may have been the desired effect.

Did Jack maybe just happen to have pedals with really bad buffers or is there no way to avoid this?

The article says he didn't pick, just had four Boss pedals at hand. And there are definitely ways to avoid this.

Could increasing input and output caps and raising the gain a little on buffers counter the effect he has measured ?

Yes.
It is entirely possible to design buffers with responses literally from DC to daylight; likewise buffers with gain. Likewise low noise, likewise whatever else you want a buffer to do. But it does take work and skill. Perhaps some luck. 

As far as I can tell, the real issues with buffers remain unexplored, even among the people who thump the tub for true bypass.

In an ideal world, one would have one buffer, on the guitar, to sidestep the issues of treble loss from loading. Any following effect would not need its own buffer.

[slacker]

It's all a matter of spin, Jack's pictures show that using 4 Boss pedals gives a flat response at any frequencies guitarists care about, there's a very slight loss of volume which can be recovered by turning your amp up a whisker.

[R.G.]

I was avoiding the use of the word "spin". I was also avoiding the use of the term "gee-whiz"graph.

(see http://books.google.com/books?id=SICioQIKhiwC&pg=PA62&lpg=PA62&dq=%22gee+whiz%22+graph&source=bl&ots=7lUPfk5rkD&sig=VeWReCepsaqJs0NRF5ledPasLVE&hl=en&sa=X&ei=kSWyUez2JYO4yQHl1IDgBg&ved=0CEAQ6AEwBTgK)

Survival in my last career was often predicated on having a master's degree in statistics taught by Darrell Huff, both in reading other people's work and in creating your own, as a self defense. Truth and justice did not always win.

[artifus]

as vic reeves once said: "ninety nine point nine percent of all statistics are made up on the spot."

[Morocotopo]

I do! I do!

I made a true bypass Compulator clone, and since it´s usually the first pedal in the pedalboard chain, I added an opamp buffer, switchable in or out independently from the FX. So I use it when I need it, if not it´s switched out.
I also made another buffer in those small Hammond boxes, footswitchable, for another board. Want it? kick it in. Don´t want it? Kick it out.

Am I smart or what?

[R.G.]

Lest anyone get the wrong impression, let's be clear. Some buffers are better than others. Some switching schemes are better than others. The differences may not be in a single-dimensional sense. Better in one direction may imply slightly worse in two other directions.

One of the lessons of Huff's "How to lie with statistics" is that oversimplification of a multidimensional issue can lead to misunderstandings - fast! This can happen even if the presenter was trying their best to be fair, objective, and inclusive, and presenting their data absolutely truthfully. The traditional reference to this is the testimony of the blind men and the elephant. The blind men were absolutely truthfully describing all they knew about the elephant. I suspect that if the blind men had been put into a discussion group to describe the elephant to one another, heated arguments would have broken out.

And everyone, most assuredly including myself, is blind about something.  

[artifus]

Am I smart or what?

'he smart! why you no smart?!'  

hey rg you should check out more or less from the bbc's radio 4: http://www.bbc.co.uk/programmes/b006qshd - you'd love it if you don't know it already. (i really wish this division of the bbc ran its news service  )

[Keppy]

I don't build bypass buffers, because the first pedal in my chain is a buffered tuner. Many (not all) of the problems of true bypass can be largely avoided by having a buffer up front in the chain. Whatever problems can be caused by buffers can be largely avoided by not stacking them. For me, adding buffers would be wasted effort. For someone with a constantly-shifting setup, buffered pedals or a standalone buffer might be better.

[toneman]

I added a preamp/buffer to all of my Lovetone clones. 

Made them "even better" 

[R.G.]

Your ears will always tell your brain what they like.

The trick is to find ways to keep your brain from overruling them because of something you've been told.

[Jdansti]

A buffer is best **right after the pickups as close as you can get it to them** in terms of not losing any of what your pickup puts out. In the guitar is the best place. If you like the treble loss or resonances you can reintroduce them later. You can never get back signal or signal-to-noise ratio once lost to attenuation and cable capacitance rolloff.
Second best position is right after the guitar, with as little cable between guitar and buffer as possible. Third is on the pedalboard. Fourth is in the pedals, fifth is no buffers.

All of this makes sense, but what I don't understand is that when I take a Tillman preamp, which I understand boosts the signal level more than a unity gain buffer, and place it at the end of a 10-ft guitar cable, I get a much cleaner, "richer" tone than without it.  It "sounds" like it's making up for cable losses, although as you mention, you can't add the losses back in. So what's going on, and would it sound even better if were to place the Tillman inside the guitar?

[R.G.]

All of this makes sense, but what I don't understand is that when I take a Tillman preamp, which I understand boosts the signal level more than a unity gain buffer, and place it at the end of a 10-ft guitar cable, I get a much cleaner, "richer" tone than without it.  It "sounds" like it's making up for cable losses, although as you mention, you can't add the losses back in. So what's going on, and would it sound even better if were to place the Tillman inside the guitar?

It's pushing your psycho-acoustic buttons.

The Tillman preamp is a simple single ended JFET with a gain of something like two to six, depending on the JFET. The sound is just a bit louder, which people nearly always describe as "cleaner". It also has a non-obvious touch of square-law distortion in about the right amount to add a subjective sweetness to the sound, hence the "richer" effect.

Whether it would be even better in the guitar is a solid maybe. Don Tillman sure likes it that way, and it's how the thing was intended to be used. The difference is that the cable is causing a slight treble loss before the boost/distort that the JFET does. It's very subjective as to whether this is better or not. Your ears will have an immediate better/worse/no change opinion upon hearing it, but you gotta let them hear it. I'd try it with a very short cable first, out of sheer laziness.

[Gus]

A buffer is best **right after the pickups as close as you can get it to them** in terms of not losing any of what your pickup puts out. In the guitar is the best place. If you like the treble loss or resonances you can reintroduce them later. You can never get back signal or signal-to-noise ratio once lost to attenuation and cable capacitance rolloff.
Second best position is right after the guitar, with as little cable between guitar and buffer as possible. Third is on the pedalboard. Fourth is in the pedals, fifth is no buffers.

All of this makes sense, but what I don't understand is that when I take a Tillman preamp, which I understand boosts the signal level more than a unity gain buffer, and place it at the end of a 10-ft guitar cable, I get a much cleaner, "richer" tone than without it.  It "sounds" like it's making up for cable losses, although as you mention, you can't add the losses back in. So what's going on, and would it sound even better if were to place the Tillman inside the guitar?

Run a sim of the circuit
figure out the source output impedance of what effect(s) is last in line when you hear this and make a model of it
Then add the 10 foot of cable capacitance a simple guess would be 39pf or so a foot 390pf total or measure it
The Tillman has an input resistance of  3meg from what found from a web search
Next model the system without the Tillman but with the input of what the Tillman was driving
ALSO you are adding a little gain and the input to the next stage is getting a larger signal and maybe changing the harmonics(distortion) you are adding to the signal

If you look at sims I have posted you will see I often add a simple guitar model and cable model to the sim to show the interaction guitar, cable and input of the effect

As R.G. posted you need to think about the "systems architecture"

EDIT R.G. we must of been writing about the same time

[Jdansti]

Thanks R. G. and Gus. So it sounds like part of the difference is "in my head" and another part is due to the harmonics caused by the non-linear distortion of the JFET.  I'll try the Tillman with a 6-in cable and compare the sound to the 10-ft cable.

[R.G.]

So it sounds like part of the difference is "in my head"

It's worth noting that "in your head" is real - to you. And the sweetening effect of soft asymmetrical distortion is shared by most humans. And the tendency to hear slightly louder as "clearer" is common to most humans. These are fairly well recognized psychoacoustic effects.

I'll try the Tillman with a 6-in cable and compare the sound to the 10-ft cable.

That would be a very interesting test. Post your results back here if you can.

[Thecomedian]

I think the reason louder sounds clearer is because it's easier to pick out or focus on individual objects. In the same way you have to speak louder to deaf people, they need all the subtleties of letter and word formation to be at higher volume in order to pick it out. Brains just seem to work like that. The way I describe it just now almost sounds like speakers not having enough amps driving them compared to having enough amps driving them. I know that I've played around with this in restaurants, if the restaurant is loud enough, I'd isolate and cycle through different table conversations just to see if I can pick their words out of the crowd. It goes without saying that very low volume restaurants are more difficult to do.

[artifus]

http://en.wikipedia.org/wiki/%^&*tail_party_effect

[Jdansti]

http://en.wikipedia.org/wiki/%^&*tail_party_effect

Agressive anti-curse feature snuffed your link.