Fragile Harmonics - Not so crazy after all? =D

[dave g]

I'm sure plenty of you have already read this paper. I had forgotten about it for a while, but re-read it again last weekend. The section talking about low amplitude harmonics of the vibrating guitar string are quite literally "squashed" by the hard limiting effects of BJTs, but "survive" under the soft-limiting conditions of a vacuum tube got me thinking about none other than Dr. D...

For those of you who don't have time to read the whole paper, well, you should make time But anyway, this paragraph from page 4 nicely sums it up (the pictures really help, too. Look at the one on page 4, as well as the ones on pages 8 and 9 to see the difference between hard and soft limiting):

All amplifiers have zero incremental gain during some portion of an input modulating waveform at
sufficiently high overload levels. A riding waveform mixed in with the modulating waveform can not get
through when the modulating waveform induces a hard-limit condition. Soft-limiting tube amplifiers,
which are sought after for their pleasing sound, let the small waveforms from emore portions of the modulating waveform (i.e., there is less hard limiting).

Essentially what this portion of the article is saying is that tube amps seem more "touch responsive" because they make the guitar behave less like a MIDI controller. Depending on how you attack and massage the string it's going to sound different because you have a different superposition of low amplitude harmonics riding on the fundamental. Soft limiting still allows an amplifier stage to have a degree of incremental gain even when clipping, so all of those riding harmonics still pass through. Hard limiting basically kills all that because you've got one very high energy fundamental frequency driving the gain stage into a type of nonlinearity where everything riding it literally gets "squished"!

I've always held the opinion that one of the largest reasons we like tubes is because of their soft-limiting (and even though cutoff is eventually a hard limit, the transition is rather slow), but for the reason that it results in a predominance of low-order harmonics when a waveform is clipped. It's possible to achieve a similar effect in a solid state amp, but you have to resort to using capacitive filtering to take the "edge" off of the clipped wave. This of course has the unfortunate effect of absolutely devastating the high end response, even when running in the linear region...sort of a throwing the baby out with the bathwater type thing. Put another way, tubes simply don't dirty up the bathwater to begin with However, even with any amount of tone shaping you can't get a transistor to soft-limit.

Anyway, I'm convinced that no matter how crazy it sounds, Dr. D's statement actually has a lot of truth to it. Whether or not he actually realized all this is up for debate

Then again, this paper was written in 1984 and I'm sure HAD was/is a member of the Audio Engineering Society...

Oh, and by the way - 3000th topic in Dumble Discussion. Hooray!

[martin manning]

Dave, thanks for posting this paper. I hadn't come across it yet, and it looks like I've got a nice read coming up in the very near future. I think you may indeed have found the inspiration for HAD's quote, and yes it does seem to be a succinct and accurate summary of the physics.

Some similar discussion is here:

http://spectrum.ieee.org/consumer-elect ... d-of-tubes

MPM

[Harald.Nowak]

Dave: very interesting paper, indeed. However I still think that soft vs hard clipping still depends more on the kind of circuit you use than on the kind of active component (like transistor vs. tube). You can of course easily see this in all the Tubescreamers (and their clones and derived pedals) - soft limiting can be as easily achieved as adding a resistance in series with a diode (or as complicated as using many diodes of different type, using mosfets in series with that aso - see for example the Zendrive).

Using capacitors in transistor- or opamp based- circuits to more closely emulate tube sound does imho NOT throw the baby out with the bathwater: the reason beeing, that tubes themselfes partially "rely" on their relatively big internal capacitance between grid and cathode and the miller effect to get that soft sound; with transistors (eg JFETs) you NEED to emulate this with a capacitor of the correct size - this additional capacitance will however NOT kill every higher harmonic (throwing tha baby out) but simply create the needed tube like transfer curve.

An additional fact to bring my point home: what about the snubbers in the so beloved ODS? Tube only amp and still in need of capacitors! From my experience I can only tell, that the snubbers are quite important to get soft enough clipping, so the OD is not too harsh (assuming the reason beeing that the OD tubes are biased rather cold).

All in all I think HAD mus have been very well aware of those facts and I still think his famous statement should be seen as an egineers joke.

[Structo]

Not necessarily as a joke but a sign of the far out times.

He was trying to make an artistic statement that would sound mystical/ magical, in an attempt to make himself seem even more accomplished.

Perhaps he was trying to state something that a lay person could understand but still leave enough mystery in it to sound vague.

[renshen1957]

Perhaps he was trying to state something that a lay person could understand but still leave enough mystery in it to sound vague.

Couldn't he have just poured epoxy over some of the words and accomplished the same thing?

Best Regards,

Steve

[Harald.Nowak]

Couldn't he have just poured epoxy over some of the words and accomplished the same thing?
Steve

Thats propably what he did - or another way to phrase it! He was known for creating diversions - he did so with the epoxy and similarly with his words.

[TheGimp]

Ya know after reading through UL60065 and UL600950, there might be another reason to pour epoxy over a board.

[Cliff Schecht]

I agree with the above statements, at least in a more generalized sense, that SS can achieve the same soft clipping that tube preamps can achieve. Tubes follow a 3/2 power law whereas FET devices follow a squared law. I've seen more than one way of using feedback to achieve a 3/2 power law with FETs that give them a very tube-like characteristic.

Whether these really sound like their tube counterparts is up in the air of course as different construction techniques make even different FETs sound different, let alone competing active device technologies. But I've heard some pretty damn nice sounding FET based distortion and overdrive pedals and such..

[Bob-I]

Just a thought while reading this.

I can't tell you how many times folks have told me that SS is advancing to where it sounds like tubes. The first time was in 1966 with a Thomas Organ Vox compared to a Super Reverb. It's getting so common I laugh it off.

When the latest in the Line6 came out the salesman at Sam Ash tried once again and out the Twin Reverb Patch next to a Twin Reissue. I played the Line 6 and said, it sounds pretty good. Then the Fender and he just said... sorry... and walked away.

Thing is, these SS modeling devices DO sound damn good in their own right. Just don't expect to duplicate the nuisances of a great tube amp with them.

Who not just develop great devices that give a great variety of tones. There are many smart engineers working on these things.

[daydreamer]

Thing is, these SS modeling devices DO sound damn good in their own right

Couldn't agree more Bob, they are distinct sounds even though they may have some tonal signatures of the original. But beyond that they just don't have that 'wild presence' that all that voltage creates.

Sort of like the analog LFOs etc of a Moog vs they plugin in versions. It's all in the voltage I think, there is a real electromagnetic presence that can't be emulated digitally although they can 'ape it' pretty well in a recording these days.

i use both, digital/tube hybrids and all tube. Can't beat tube for getting the hairs on end and fill the room with that 'presence'.

[M Fowler]

I agree as well Bob. I tried using my Line6 Flextone XL100 a 212 combo live and it was a terrible experience, not only with the tone but the amps functionality.

I played a Peavey Bandit 65 for many years only because I didn't want to bring my tube amps on the road and the small combo was easy. But I finally realized what tone I had given up not using a tube amp.

Mark

[talbany]

Back in the early 90's I had a sound company and did a ton of Wedding/Hotel gigs in DC mixing all types of bands..The modeling amps had just appeared and the guy's doing those types of gig's really liked them..They were light and you could fill them with a ton of sounds and at whisper volumes no big pedal boards and when mic'd got the job done,especially the ones who knew how to program them right didn't sound half bad (for a wedding) from the board..(As long as you keep the power amp out of clipping)
Still not a tube amp!!

Tony

[Max]

Back in the early 90's I had a sound company and did a ton of Wedding/Hotel gigs in DC mixing all types of bands..The modeling amps had just appeared and the guy's doing those types of gig's really liked them..They were light and you could fill them with a ton of sounds and at whisper volumes no big pedal boards and when mic'd got the job done,especially the ones who knew how to program them right didn't sound half bad (for a wedding) from the board..(As long as you keep the power amp out of clipping)
Still not a tube amp!!

Tony

Hi Tony,

do you know this comment or even recommendation of one of the most experienced Dumble players and studio cats:

http://www.youtube.com/watch?v=cjSX5UvVGzI

If you watch this vid you should see more links on the right side of your screen to Carlos Rios' comments on some recording stuff if you should be interested. I'm a big fan of him as the great and versatile player he is and of his straightforward, practical and unmystical view on gear and especially of his always gentle and decent attitude despite the name he has in guitar players circles.

AFAIR he used such an amp for this tour as an example that is documented on a DVD that has been posted in several parts at YouTube:

http://www.youtube.com/watch?v=xTjU4PWQNWs

Here he's on stage with Stevie Nicks:

http://www.youtube.com/watch?v=emIb8sXQI-c

All the best and have a great week,

Max

[M Fowler]

There you go huge difference between using my Line6 for recording purposes then it is for stage use. Of course that is only my experience of owning a Line6 since company was first created until today.

I have the first generation of XL Flextone. Oh and the amp isn't light by any means it is 100w with 212 Eminence.

Mark

[Structo]

WHat Carlos says about the power tubes is right on.

They degrade so slowly that we don't really realize that the top end is starting to lose it's sparkle.
So what we think is a great, sweet amp tone may not cut it in a mix or work in a studio.

So when we finally change the power tubes, we don't really like the way it sounds right away until a few more hours of run time are logged.