What do you look for in damping factor?

It's been decades since I was up to speed on damping factor, so I'm interested in knowing what all of you look for. And—if you're so inclined—why. I may be in the market for a new amp sometime soon, and I'm no longer sure how much weight to give to this spec. THANKS.

P.S. Speakers I use are Maggie 2.6/R, Spendor SP3/1P, NHT SuperTwos and PSB Stratus Minis (hey, I like them all, to one degree or another!).
anything close to a 100 is good, above this you are unlikely to gain significant audible improvements (diminishing returns)
There are many answers to this seemingly simple question! There are some that say that a low number is just fine - I think that I read on the Gamut website that 25 is sufficient. http://www.gamutaudio.com/faq/details.asp?id=16

There are others that say, a large number is a must - the higher the better - several s.s. amps claim very high damping factors such as 500 or 800 or even 1000.

There are still others that say that DF is a complicated matter & that there is no correct answer that'll satisfy everyone. Since DF is defined as speaker impedance divided by amplifier output impedance & since speaker impedance varies with frequency, DF will also be a function of frequency. In fact, several amplifier's output impedance is also a (different) function of frequency. So, you have a ratio that has both numerator & denominator with 2 totally different functions of frequency. So, the DF curve, is at best, not drawable on a piece of paper when it comes to a real-world operating condition i.e. a particular amp driving a particular speaker UNLESS you know the exact impedance vs. frequency response of both items.
If you do, you have 1 more hurdle - what is the impedance vs. frequency response of your speaker cables? This fellow features in the mix too!

usually, a DF number is spec'd in the manual using 8 Ohms speaker input impedance. However, are any of your speakers 8 Ohms input? If they are, do you really think that they maintain 8 Ohms from 20Hz to 20KHz? what if your speakers are 6 Ohms input? then, is the DF spec'd in the manual any use to you?

So, best is to use your ears to judge if the amp connected to your speakers is providing you the bass response you are looking for. Forget chasing this spec - I believe that it has little info to offer you.
I never purchase on specs including damping factor, just sound quality. But I do know the higher the number for damping factor, the better the amp can control the low frequency drivers. A high damping factor no doubt would result in a tighter more controlled bass and would be an important spec for someone that is considering an amp to drive a subwoofer.
Damping factor is just part of the overall gestalt of a sound system. I used to look at it in isolation, and assume higher is better. But high damping factor usually calls for large amounts of global negative feedback, which introduce high-order harmonic distortion, which has been shown to be pschoacoustically detrimental even in very small amounts. Low damping factor is typically associated with little or no global negative feedback and relatively high total harmonic distortion specs, but that tends to be benign distortion that the ear is very forgiving of. In fact, one of the leading researchers in the field of distortion perception told me that there's a negative correlation between harmonic distortion and subjective preference - meaning that we tend to prefer higher distortion amplifiers, because they tend to have a more psychoacoustically benign distortion characteristic.

So high damping factor can imply sonic trade-offs in other areas.

Some speakers (typically high efficiency speakers) work best with a low damping factor amplifier. Some work best with a high damping factor amplifier.

What happens is, with a very low damping factor amplifier, the frequency response is modulated by the speaker's impedance curve. Almost always this is detrimental. However, a few designers (Silverline comes to mind) use this to give their speakers a smoother frequency response when used with low damping factor amplifiers. The ideal would probably be an ultrasmooth impedance curve, so that there are no impedance bumps to modulate the frequency response.

Now if a speaker has impedance peaks in the bass region, a low damping factor amplifier will often deliver more power into those impedance peaks, making them louder. So the designer can use the bass impedance peak/amplifier damping factor interaction to warm up the bass a bit in what might otherwise be a bass-shy design. That's one reason why high efficiency speakers sound often better on SET and OTL tube amps - such amps will deliver a little extra juice in the bass region where the impedance is high.

In my own designs, I worked very hard to get a smooth impedance curve from about 80 Hz on up, so the speakers should work well with a wide variety of amps. I allowed the bass impedance to peak pretty high so that a low damping factor amp would give me some free bass extension. The port tuning can be varied - I tune it higher for high damping factor amps. I mention this to show how a speaker designer might take amplifier damping factor into account.

Most of the speakers you list work best with medium (20-80) to high (80+) damping factor amplifiers. The Maggies are an exception - while they have a low impedance, their impedance curves are amazingly smooth, so they will work well with low feedback, fairly low damping factor amplifiers. Unfortunately, they also like a lot of power - which is rare in a low feedback amp.

If you were starting from scratch, I'd suggest taking a wholistic approach - choosing speakers from among those models that will work well with the type of amplifier you like, and then making your final amplifier choice.

Specific to your Maggies, damping factor is a factor as the panels really do need to be controlled. In my day of selling audio we took in a pair of Maggies and they sounded dreadful with everything in the store they had tried. When I found out we had them I brought in my GAS Grandson the next day and it kicked them hard and well. Damping factor. The Hafler 120 did the same. Solid current mated with damping factor makes Maggies sing.
Thanks so much for all your thoughtful, complete answers. Rest assured, DF is just one of the things I would consider should I decide to make a change. I was just curious on what your thinking was, and found your remarks very helpful. Thanks again.
Around two or three is good. No, I'm not kidding!
Like others suggested, I suspect a high or low damping factor is likely neither good nor it is bad, rather it may render an amplifier more or less suited for a given speaker. For example, while Sean has commented elsewhere on this forums that he found the Vienna Mahlers always to display a sloppy bass, I listened repeatedly to these speakers at the recent Denver AudioFest, and found them to have an extremely deep tight and tuneful bass while being driven by a Rowland 312. I then checked the damping factor of this amp and found it to be a wapping 1000. One possible explanation of the apparent 312/Mahler high synergy is that the Mahlers have a fairly low impedance of 6Ohms, dipping to 3 Ohms in the bass. This relative low and uneven impedance likely requires a high damping factor, which is found in the 312.
Damping factor plays a role in system performance, just not a large one.

Damping factors above about 15 or 20 to 1 will result in no audible improvement, but such amplifiers that do have such damping factors have proven themselves incapable of playing definition and body in bass instruments, and for that matter punch is a problem too. IOW things get worse.

This is because such amplifiers employ large amounts of feedback to this effect. Feedback, while lowering output impedance, simultaneously restricts dynamic peaks and impact. Additionally as Duke mentioned above, odd-ordered harmonics that the ear uses as loudness cues are actually enhanced, causing an amp with feedback to be harsher.

The solution is less feedback.

We do perceive a different phenomena as the output impedance of the power amplifier is raised, and approaches that of the speaker- the amplifier will make more distortion.

We have words for this type of distortion (in order of increasing distortion): warmth, bloom, fatness, turgid and finally, muddy.

So the trick is to have an amplifier that ideally has at least 4:1 over the speaker in use. For a 4 ohm speaker this puts the amplifier output impedance at about 1 ohm, 0.5 ohms is likely to be near ideal.

Note that the output power of the amplifier is what places the diaphragm of the speaker at any position in space that it is capable of. The idea that the speaker diaphragm 'falls' back to zero and can ring really does not happen- the speaker diaphragm is powered to full excursion and then powered back to zero- power is on the driver at all times.

Due to this reality, damping plays a minor role.
Ralph, I am slightly confused. Are you trying to say that:

A. The perceivable synergy of the Rowland 312 and the Mahlers exists in spite of the amps high damping factor?

B. The 312's damping factor is immaterial to its synergy with speakers in general.

C. The perceived extreme degree of synergy between the two devices cannot be, and is likely a case of self delusion on my part?

By the way, if I recall correctly, the output impedance of the 312 is 100K Ohms.
The numbers Paulfolbrecht quotes are typical of many good tube amps; if (as others say) you just chase a high DF spec. you are going to eliminate quite a few good amps. I'm using VTL's with my Maggies and the DF is certainly under 5 (they sound great.)
Guidocorona, you may be thinking of the INPUT impedance of the Rowland 312. Usually, power amps’ output impedances are very low, hopefully below that of the speakers--especially for SS amps.

If the Rowland 312 is a Class D or switching amp--and I believe it is, the manufacturer may be quoting the figure in a different way because of the different topology. The Rowland site does not list the impedance figures of the 312 under products, 300 series---unless I missed it.

Typical tubes amps use transformers to lower their output impedance below that of the 4-8 Ohm speakers they usually have to drive. Atmasphere amps are, of course, different (being OTL's) and do necessitate being mated with a bit higher impedance speaker than most “normal” amps do. SET amps also have their impedance limitations.

A 100K Ohm amplifier output just sounds wonky to me.
Ralph, I am slightly confused. Are you trying to say that:

A. The perceivable synergy of the Rowland 312 and the Mahlers exists in spite of the amps high damping factor?

B. The 312's damping factor is immaterial to its synergy with speakers in general.

C. The perceived extreme degree of synergy between the two devices cannot be, and is likely a case of self delusion on my part?

By the way, if I recall correctly, the output impedance of the 312 is 100K Ohms.

Guidocorona, the answers are yes, yes and no. The synergies you hear have nothing to do with 'damping factor'. BTW the impedance you refer to is input impedance, not output impedance.
Negative feedback is not normally a problem unless the device is unstable or is over driven (bad design, misuse). Excessive damping factors may be indicative of a design that is not balanced with regard to other factors, such as stability, distortion at clipping etc.

Extremely low damping factors will cause frequency response to vary with speaker impedance.

Audio is represented by a voltage signal. In order to preserve this you should have low ouput impedance compared to the speaker input impedance, which equates to a high damping factor (at least greater than a nominal value of 10 in order to be inaudible when matched with a load that varies significantly with frequency,such as most speakers).
A solid state amp with a damping factor of 25 or lower will sound more tube-like. I have a Onkyo A-9555 Integrated amp and the damping factor is only 25.......sounds very tube-like too!
I'm not sure you can equate a high damping factor with large amounts of global feedback in all instances. Digital amps may be the exception in that they have a high DF of a 1000 but they don't seem to suffer in sound quality, the lack of openess or transparency. I know for sure when a conventional amp is designed with large amounts of global feedback it is usually involves a poorly executed circuit design. So it is my opinion that if your looking for an amp with a high damping factor that results in visegrip control of the low frequency drivers without sacrificing sound quality you may want to go digital.

For example I have used the McCormack DNA-1 Rev B. This amp uses no global feedback, has a low damping factor of 150 and seems to have unlimited power reserves. Because of the low damping factor the bass at times can be a bit mushy & loose depending on the preamp used but definately excels in soundstaging & sound quality, one of my all time favorites.
Having Zu Audio Presence 16 Ohm 101 Db I have been told by Zu DF must be <10.

I just bought the ASR Emitter 1 Ex. With a DF @ 600. Zu says the bass will sound bad. I don't get way having it's own sub amps in each speaker the Presence speaker will not work with the ASR amp.

I hope Zu is wrong and I don't need to sell the speakers.
Audiokinesis, high damping factors don't automatically indicate high global feedback.
So high damping factor does not necessarily imply sonic trade offs in other areas.
As you've posted in earlier threads, it's easier to make a speaker with a low impedance to have an ultra smooth impedance curve, so there are no impedance curve bumps to modulate the frequency response. Typically speakers with such a low impedance are driven with amplifiers with high damping factors.
Are you suggesting that your bumping the impedance in the bass region to correct for a fundamental flaw? Using a port will of course compromise phase and time accuracy.
Th impedance curve of the Maggie's will work well with high damping factor amps well.
I would suggest choosing a speaker that fits into ones budget, room accommodations, listening tastes, then choose the appropriate type(s) of amplification, and then make ones final amplifier choice.
Unsound, what Duke was alluding to is that speakers can be designed using either the Voltage Paradigm or the Power Paradigm.

In the case of one of the speakers he mentioned, he set the port resonance *below* the cutoff of the speaker. Now with a Voltage Paradigm amp, this would have been a roll-off in the bass. But with an amp with a higher output impedance (Power Paradigm), it instead extended the bass response another half octave- to 25Hz. I've heard him call this his 'free lunch'.
Atmasphere, thanks. That appears clever, but I still think the use of ports come with a price. So much for free lunch?
Unsound, for that you would have to hear them. I know one person that used such speakers to replace a set of Carver Silver Amazing loudspeakers- he likes Duke's speakers better because they are smaller, are easier to drive, sound better overall and of course go deeper in the bass.
Of course those are all good reasons to like a speaker, but, amongst other things, ports can compromise time and phase fidelity. Obviously, some feel the trade off worth it.
" What do you look for in damping factor? "

Nothing. Its almost meaningless number, largely explained by Athmaspere in his 1st post.

A point about damping factor which is commonly not realized, pertaining to dynamic speakers, is that the ability of the amplifier to "control" and damp the woofer will be limited by the dc resistance of the woofer voicecoil itself (as well as by the resistance of the crossover inductor, the speaker cables, and the amplifier's output impedance).

That voicecoil resistance will typically be several ohms, and with respect to woofer control/damping will have exactly the same effect as if the several ohms were in the form of amplifier output impedance rather than in the speaker, corresponding to an amplifier damping factor in the rough vicinity of 2 or even less.

One more reason why the difference between amplifier damping factors of, for example, 20 and 2000, is in itself insignificant.

-- Al