The fascination of the bucket brigade delay

[spargo]

I've been intrigued lately on analog delays, and have looked at a couple schematics to see what's out there, such as the Rebote, Echo Base, etc. among others.

My main question is in regard to the bucket brigade delay.  In searching around, there seems to be a certain sense of "awe" when the BBD is mentioned.  For those that have experienced the BBD as well as other delay methods, such as the PT2399, what would you say the difference is in terms of sound?  Is the BBD justified in everyone saying it's so awesome, or is it just simply the fact that it's "vintage" so people think it is that much cooler?  It seems to go without saying that BBD devices are in short supply, and for sourcing and cost other chips like the PT2399 would be much more preferred.

Any thoughts on this topic in general?

[Taylor]

I think you might already know this based on your second paragraph, but to be clear, neither the Rebote nor Echo base are analog (BBD-based) delays.

I am not one of those people who hears miniscule differences in audio gear. So, you can take it from a non-golden ear, non-audiophile, that analog delays really do have a unique sound.

The reason for that is simply to do with the technology, and the explanation can be superficial (the sound comes out distorted, with limited bandwidth and dynamics) or it can be quite in-depth, and some good analysis has gone on on this forum (not by me).

A good portion of what people like about analog delays is captured in PT2399 delays like the Rebote and Echo Base. I think of these as halfway between really clean digital and analog. They get you very much in the ballpark. But BBDs can be quite special and it basically has to do with all those stages of transistors distorting just a little bit, all the tiny changes that happen when copying a voltage a thousand times, etc.

I say "can", because a lot of what makes a particular circuit or processing type sound good is the circuit design itself, rather than solely the technology used. A bare bones analog delay, stretched past its reasonable time range, without filtering or companding will sound horrible (maybe in a fun way, but maybe not). An analog delay can be made very clean (that was my impression of the Moog delay) and a PT2399 delay can be quite lo fi, gritty, whatever descriptors people use for analog. I'll paraphrase RG here, to say that rubbing BBD on something doesn't instantly make it awesome.  

[Johan]

there is sertainly a bit of "corksniffing" about the worship of BBD delays( noisy, bandwith restricted and short delaytimes...ofcourse people love them.. )...If you think about it. the echoes in most situations where a band is playing, is only something that fills the background and adds some atmosphere. unless you're The Edge, the echoes themself are not the sounds that would be in focus
if you mainly sit at home and listen to what happens when you are not playing, BBD's are great...personaly I've stuck with a BOSS DD3 for over 20 years now and see no reason to change(crappy digital delay, if you ask some corksniffer..)..it echoes in the background
...clearly this is just an opinion...please dont flame me..
J

[R.G.]

I have to say that I run into this dichotomy  (literally "cut into two pieces") a lot. People who play on stage want the show to be good and the audience to be happy. People who play off-stage want subtle tone colors and textures. On stage, you generally don't have time to worry about whether the 15th delay is losing some midrange and sounds a little grainy or dull, even if you can hear it over the audience noise and the singer, who has a cold and is about to sneeze into the microphone.

They are two different environments. I can tell in two seconds whether someone who is trying out pedals ever steps out on stage to play for an audience of more than one.

[Mark Hammer]

Oh sure, bring context into it! 

BBDs and digital approaches both apply a sampling strategy.  As such, they both suffer from the same concerns: namely the bandwidth/delay/sampling-rate tradeoff, aliasing (which comes from "stairstep" reproductions of audio), and clock noise.

As it happens, direct clock noise is less of an issue with digital chips because the clock frequency is so high you can't hear it (unlike BBD clock rates which can be well into the audio range).  Indirect clock noise (i.e., heterodyning or simply line noise) can be an issue with digital, though not always.  The sampling-rate/delay/bandwidth tradeoff is not as big an issue with digital because you don't need to ever clock down into the audio range to get appreciable delay times, and those delay times still come with pretty decent bandwidth.  BBDs, on the other hand, require some soul-searching about bandwidth because aiming for longer delay times generally forces some sacrifices with respect to bandwidth.

The bandwidth considerations occur for two reasons.  One is that longer delays can result in the clock whine being more audible.  The other is that the slower the clock, the more audible and obvious the aliasing is from all that stairstepping.  Getting rid of both of those requires some serious lowpass filtering to keep them well in the background.

Where BBDs can have an edge over digital is with respect to resolution.  In the otherwise so-so Dave Hunter book on effects pedals, Roger Mayer makes some interesting points in an interview, and notes that the tails of delays have fewer bits to encode the signal.  So, the A/D may be 24 bit, but for those tiny quieter parts of the signal there may be only 10 or 8 bits to encode them.  BBDs, have essentially infinite resolution because they are analog.

As RG astutely points out, that potential difference in audio quality for the nuances means diddley squat in a gigging situation.

BBDs are not really warmer than digital chips.  The warmth is essentially a byproduct of the lowpass filtering they require to make them behave.  On the other hand, that has some psychoacoustic benefits by moving the dry signal more to the foreground by virtue of having more harmonic content.  Easily applied to digital chips as well.

[spargo]

Some interesting thoughts, thanks guys! I really appreciate this community.

In terms of actually sourcing legitimate BBD chips, is SB the only way to go?  I hear they're really sketch on eBay.  For availability, if I were to make a BBD delay would it be unrealistic to expect to find the same chips a couple years down the road to do another build of the same pedal?

[Mark Hammer]

Although SB is a preferred source, it is not the only source of BBDs, assuming one is talking about current-production chips.  Cool Audio is making clones of MN3205, 3207 and MN31032 chips for Behringer`s massive pedal output.  I know Beiling WAS making MN3208 clone chips and MN3207 clones, though I have no idea if Coolaudio has bumped them or not.  If we are looking at earlier generation chips, like any of the MN30xx series, SB and e-bay are some of the only places to get whatever unused stock remains in the world.