Can a variac be used to maintain constant voltage?

Hi: I amusing a variac to control the power going to our Threshold S/500 II. Over the past year that I have had the amplifier connected I have only had to adjust the variac voltage within +/- 1 volts perhaps thrice in order to maintain a constant 120 volts. I simply wonder if in fact the variac is designed to maintain a consistent voltage despite incoming voltage swings. I read one forum thread in which a member stated that a variac can not compensate for swings in incoming voltage, for instance if the incoming voltage is 125 volts, the variac will swing to 125 volts despite having the set the dial to a constant 120 volts. Thank you for your assistance.
What you read is correct. For a given setting, an ordinary Variac will provide an output voltage that is proportional to input voltage.

There are some Variac-like devices, such as this one, which regulate output voltage by using a motor to automatically vary the setting as the input voltage changes. When the motor runs, however, this particular series of regulators, and I would imagine others that are like it, makes far too much noise to be usable near an audio system.

-- Al
There are also Constant Voltage Transformers based on transformer saturation effect

The problem with them is poor efficiency and often some distortions of the sinewave.

Even if your voltage changes by 10%, your amp has, most likely some form of feedback making it regulated. Only max peaks will be affected and only if voltage drops. Lets assume drop of 10% making drop in available (peak) power of 20% it will change perceived loudness only by 5%. Just set your loudness not higher than 95% (to prevent clipping) and regulation will take care of the rest. Preamps, CDPs etc are all regulated.
Perceived loudness = k^(1/3.5) where k is ratio of powers.

Regulator will make it worse, in my opinion, since amplifier's supply current is in form of narrow spikes of high amplitude causing voltage drops - unless you use greatly oversized transformer or device that provides line AND load regulation.
So, if an ordinary variac does not maintain voltage at its fixed setting, what is the purpose of using such a variac, i.e. is it simply to revive an old amp from storage for instance by slowly increasing the voltage to nominal levels?
That would be good application. They also bring supply voltage to regulation range of electronics in case of wide variations (constantly too low or too high).
Variacs are not specifically designed for audio applications, but rather for industrial, and general electronic testing. They operate very much like a large potentiometer, only- using inductance, rather than resistance, to divide voltage. Further- like a pot; they can only reduce voltages and cannot increase what is presented at their input, or maintain a set voltage, if a drop occurs at the line.
Actually, a Variac is a variable Autotransformer, which can be, and often is, designed to step voltage up as well as down. That would be done as shown in the first figure of the linked writeup, where the input voltage is placed across a segment of the winding that encompasses fewer than the full number of turns. Variacs that are designed for 120 volt single-phase inputs are commonly designed to output up to 140 volts, and sometimes more.

One other benefit they might provide, in addition to those Kijanki mentioned, is some reduction of high frequency noise that may be present on the incoming AC. However, as Rodman indicated Variacs are not specifically designed for audio applications, and an audio-oriented power conditioner (which in most cases won't regulate voltage) or a power regenerator (which will regulate voltage) can be expected to be a better alternative, albeit at a considerably higher price.

-- Al
The Monster AVS2000 voltage regulator is a giant automated variac like device. I had one. The rheostat inside has a mechanical wiper which is electronically controlled to maintain the 120 volts. regardless of the incoming voltage.

So it keeps the voltage output steady with variable incoming.

Just curious why you think you need to run the amp off the variac. On the back of the amp what is the input AC voltage rating? 120V? 125V line voltage will not hurt the amp.

Depending on how big, (VA rating), the variac is the variac could be starving the amp of power.

Plug the amp directly into the wall and set back and listen..... Post back your results.

I stand corrected! The power supply I used at home(test purposes), would only maintain, or maintain at at a reduced voltage. The veriac(Ohmite) I had on the job could up to double the voltages, presented to it(but still directly proportional to the input). Gotta stop the early morning posting.
My own reason for having this variac in the first place is that prior to the Threshold, I was using a pair of vintage vacuum tube monoblocks running 8 el34's per side: when turned on the amps would spike to the maximum level as gauged by their ammeters and then settle back down to the nominal 6 amp running rate: so, rather than having to deal with this initial surge and its potential negative impact, I bought the variac to slowly bring up the amps to running voltage. The variac is selling for a few dollars short of $600.00 now and I do not if its worth keeping this expensive piece for any valuable audio uses -- i.e. reduction of high frequency noise as Almarg mentions above -- or sell it off to someone with a more suitable application. I still find it amazing that the variac is showing a consistent 120 volts as set by my original dial setting. I had a PS Audio Power Plant Premier at one time and the incoming voltage showed as high as 125 volts sometimes. So it makes me wonder if the variac is not in fact more useful in my current Threshold application as might otherwise seem.
With your voltage only varying by around 1-2 volts, plug your amp right into the outlet. I would love to have my voltage stay that close to 120 volts. Mine will vary up to five volts, and sometimes more, the same same day. I live in a straight residential area. In other words, no commercial equipment running off of my part of the substation feeding the power to our house. My friends that even live in a different state have similar variations.

One problem with a variac can be when your amp draws a lot of current (when it puts out a lot of watts), the voltage will drop a lot more through the variac. Another problem a regular conventional variac will usually have a contact (sometimes with a motor type brush), that can add noise to your power. With voltage that stable, plug your equipment in without the variac. You will be way better without it.

I have a couple of variacs. One of them will put out about 130 volts, the other one will put out about 140 volts (if memory is correct) with 120 volts in. More voltage output yet, with more than 120 volts in. With my voltage running low, and a variac boosting it, it could output way too much when the voltage would go back to 120 volts, or worse yet, a lot more when my voltage goes to around 126 volts, as it does sometimes. That would probably damage something. Another good reason not to run one.

Variacs are mostly used for testing. One would be to correct the voltage to 120 volts for testing and measuring, another reason is to simulate what some equipment will perform like on 114 volts to 127 volts. This is what an outlet can give out, and still be within power guidelines set by government regulations. They can be used for testing and simulating, or temporarily correcting other voltage swings too. But still, they are mainly used for testing.
I still find it amazing that the variac is showing a consistent 120 volts as set by my original dial setting.

If you have a descent true RMS DVM meter you should check the voltage at the receptacle the amp is plugged into at start-up. That will tell you what the true voltage drop (VD) caused by the inrush current draw of the primary winding of the power transformer of the amp.

Next voltage reading to measure is with the amp loaded. That is with music playing, volume set to where you normally listen. Play something with lots of dynamics.

Measure the voltage... If the voltage is fluctuating in rhythm with the music that will tell you the variac is voltage/current limiting. To verify the variac is causing the VD repeat the same test with the DVM at the wall receptacle the variac is plugged into.

Dynamics performance of the amp will suffer.
Your amp should sound best connected to a dedicated 20 amp branch circuit without being restricted, choked, by the variac.
On a related point what is the difference between a variac and an autotransformer (which is what I have)? Thanks.
On a related point what is the difference between a variac and an autotransformer?
The Wikipedia writeup I linked to earlier provides a good answer:
From 1934 to 2002, Variac was a U.S. trademark of General Radio for a variable autotransformer intended to conveniently vary the output voltage for a steady AC input voltage. In 2004, Instrument Service Equipment applied for and obtained the Variac trademark for the same type of product.
So a Variac is a kind of autotransformer. Other kinds of autotransformers may not be variable, and may be designed for purposes other than controlling AC power. The "autoformers" used in the output circuits of some McIntosh amplifiers are an example of an autotransformer that is not a Variac.

-- Al
Thanks all for your informative input. I guess a variac is simply not suitable for my purposes so I am selling it.
Kind of late to the party here, but I have used a product built by Furman called the AR-15 series II, which regulates AC output voltage. It works like some of the products mentioned above, where it has a multi-tapped transformer tapped every volt or so between 90 and 130 volts. It is electronically switched (as opposed to mechanical contacts). Regulates quite well, although I have stopped using it for my audio equipment. The dynamic output impedance of this device fights with the power supplies of the Cambridge Audio 840 CD deck. Still use for computers and other devices. Really works to minimize power brown outs or over voltages.