What is Amp Bloom?

[FredB]

What is amp bloom and does anybody have any sound clips of an amp with good bloom?

[Steve Conner]

It's when flowers sprout out of it in the springtime. It can't be cloned, you need 60 years worth of dirt in the cabinet. If you get potatoes, that's not so good.

[Jazz P Bass]

Sonic Bloom

[DrGonz78]

When an amp is just on the thrust of being used as fertilizer!! Then and only then will it be in total bloom... Usually around spring of the given year!!

[Jazz P Bass]

Seriously, 'Amp Bloom' sounds like an audiophile phrase.
Whatever Happened To The Audiophile? : NPR

[Steve Conner]

Seriously, I think of "bloom" as being a dynamic effect where the note fizzles out a bit immediately after a hard pick attack, then "blooms" back in during the sustain phase. Can either be caused by blocking distortion, or really bad power supply sag, or the two effects working together.

[Resonator Guy]

I believe amp bloom is the stuff that falls out of the cabinet after you have removed the chassis.

[loudthud]

This is something I've been looking into, I don't know if it qualifies as "Amp Bloom". It is a plot of amplifier output level verses input signal level otherwise known as voltage gain. The volume control of the amp is adjusted so that the output starts to clip at 100mV. As the output increases, the gain falls off a little because the power supply is sagging. I haven't done this on very many amps yet, but the 5E3 seems to reduce gain more than the others. You can see that in the 20dB (10x) below clipping, the gain falls off 3dB. Other amps were on the order of 1dB.

dBg means dB in reference to the generator so gain is computed automatically. I got the idea from reading reviews in Bass Gear Magizine by Tom Lees. He uses the Audio Precision Compute-Linearize function which I don't think works correctly. It often shows a slight increase in gain of about 1dB before gain starts to fall off.

[FredB]

I found a few other threads on amp bloom, and noticed that they are filled with humorous replies. Evidently amplifier bloom is closely tied to comical bloom.

What did you use to make that graph? That's interesting. It shows the overall effect of sag in the amp without getting complicated with the mechanics of the sag itself. The plot seems to indicate there is about 3db of continuous curve compression before clipping and much more after the amp is driven into clipping.

I scoped my guitar with PAF pups and the E string seems to put out an initial attack peak of up to 2Vpk and then a slowly decaying signal of about 250mVpk for several seconds. About half of the serious comments on bloom suggest that it results in an increase in volume after an initial attack peak and quickly following dip in volume, and about half of the serious comments suggest that bloom results in a change in the harmonic content of of a note.

It would seem these two characteristics could occur either alone or simultaneously. Most all of the comments suggest that bloom is related to sag, although posts add that too much or too little sag can prevent bloom from occurring.

[loudthud]

Thud, what did you use to make that graph? That's interesting. It shows the overall effect of sag in the amp without getting complicated with the mechanics of the sag itself. The plot seems to indicate there is about 3db of continuous curve compression before clipping and much more after the amp is driven into clipping.

The graph is created on an Audio Precision System 1. Note the "AP" in the upper right corner of the graph. Newer versions of the instruments/software have a slightly different logo, it looks more like "Ap)". You'll see lots of graphs with the logo if you look at any audio equipment reviews. It is essentially an audio generator and distortion analyzer controlled by a program running on a PC. Newer and more elaborate instruments than mine can do an FFT and digital I/O to test DACs and A/D converters.

This graph is simply sweeping the generator amplitude (rms volts) and then reading the rms volts at the output of the amp. The output is plotted with units set to dBg so gain is automatically calculated. You can't really tell from the graph where the amp starts to clip so before the sweep I set the amp's volume control to clip when the generator is 100mV. After the amp begins clipping, the peak amplitude at the output doesn't increase but the rms voltage does increase. I'm still learning how to use this thing. I'm not sure if the sweep speed or reversing the sweep would show a different result.

I scoped my guitar with PAF pups and the E string seems to put out an initial attack peak of up to 2Vpk and then a slowly decaying signal of about 250mVpk for several seconds. About half of the serious comments on bloom suggest that it results in an increase in volume after an initial attack peak and quickly following dip in volume, and about half of the serious comments suggest that bloom results in a change in the harmonic content of of a note.

It would seem these two characteristics could occur either alone or simultaneously. Most all of the comments suggest that bloom is related to sag, although posts add that too much or too little sag can prevent bloom from occurring.

Clipping is almost always occuring in a guitar amp. Even on what sounds like a clean signal, the pick attacks are being clipped off. The dynamic response of a guitar amp is very complex and difficult to evaluate in electronic terms. But when it's right, you know it's right.

[pdf64]

When considering sag, don't forget squish - no bloom should be without it.
For information regarding squish, see sag Sag
My feeling is that with a 5E3, the screen supply sag is 'as or more' significant than the main B+ supply sag.
But squish is up there also.
Pete

[Dave H]

This is something I've been looking into, I don't know if it qualifies as "Amp Bloom". It is a plot of amplifier output level verses input signal level otherwise known as voltage gain. The volume control of the amp is adjusted so that the output starts to clip at 100mV. As the output increases, the gain falls off a little because the power supply is sagging. I haven't done this on very many amps yet, but the 5E3 seems to reduce gain more than the others. You can see that in the 20dB (10x) below clipping, the gain falls off 3dB. Other amps were on the order of 1dB. off.

I think my 5E3 type amp (actually 2 x EL34) compresses even more than that. A while back I checked the amp's gain at just below clipping (<5% distortion) and at 1/10 input below that and I think the gain fell off by about 5dB at full power. I didn't take notes. I'll do it again to check (my memory).

[FredB]

pdf64,

Excellent, I noticed the output reduction due to the screen supply sag, although it slipped my mind. I've had to start making notes on all the particulars involved with amp dynamics.

I think I'll refer to output reduction due to bias shift from either fix or cathode bias shifts as squash to differentiate it from screen squish.

I noticed that plate supply sag increases clipping, while bias shifting reduces clipping. I wonder about screen squish. I'll have to check, although I expect that it reduces clipping like a bias shift. Perhaps that's why they both were referred to as squish.

The fabled amp bloom, I conclude comes from just the right balance of the various aspects of dynamics within an amp. Now, to figure out just how to optimize that balance...

* * *

When checking an amps gain vs input level, are you checking at any particular frequency?

[loudthud]

When checking an amps gain vs input level, are you checking at any particular frequency?

I used 1kHz and didn't see any significant change at 200Hz in the 5E3. The guy in Bass Gear Mag uses 200Hz.

[eschertron]

Sonic Bloom

But of course you have to use organic free-range capacitors in your 'sound magic' amplification and distribution system.