why does preamp improve sound quality?

I recently listened to a Mark Levinson no.390s CD processor directly connected to a pair of Quad II-forty five tube amps. When a matching Quad preamp (much cheaper than the Levinson) was placed between the Levinson and the amplifier, the sound improved dramatically even at moderate volume- it became much more clear and transparent. Why would this be the case? Wouldn't adding an extra piece of equipment add more distortion?
A good pre-amp doesn't improve the sound quality, it just does not degrade it as much as a lesser quality pre-amp.
Post removed 
I have tried going direct several times but having the preamp in the chain has always been better. Arthur

I'm sorry if I'm not as clear on how impedance and other factors affect your situation. I've always thought that introducing the gain of the preamp between the two means that the amplifier has more power in reserve to accurately transmit the signal, with particularly better dynamics.

There would have to be more distortion introduced with the extra stage, but it sounds like there is more distortion when your power amp takes the whole load itself.
Elizabeth, you addressed the issue of impedance mismatch between the source and poweramp. Can you explain what kind of impedance mismatch? In terms of electronics, what does a preamp do with the signal before it enters the poweramp?
A good preamp can improve the impedance matching between sourced and amp, and also give the music more body and life, primarily from extra gain. I've tried it both ways and have preferred a good preamp every time.
It probably has to do with matching the output impedance of the source with the input impedance of the amplifier. If you have all of the manuals, find the specifications for:

Mark Levinson no. 390s "output impedance"

Quad preamp "input impedance"
Quad preamp "output impedance"

Quad II-forty five amp "input impedance".

After looking at these, see if the numbers suggest that "CD to AMP" would be better matched with the preamp inline.
Bufus, I Mark Levinson/Quad experience took place in a store demo. I have a similar Mark Levinson (no.39)/Jadis (SE845) setup, and am considering a Jadis preamp (JPL2), so wanted to know if the Jadis preamp would make my system sound better. I won't have a chance to test it out, so wanted to know if adding a preamp in this kind of situation generally improves the sound. I certainly don't want to buy an expensive and unnecessary piece of equipment, and find out it degrades the quality of my system! Unfortunately, I won't have access to any manuals before my immediate purchasing decision, since I'm in a different city.

Thanks for the explanations about impedance matching and extra gain, it really begs the question about my understanding how a preamp works in the first place. Does anyone know any website was a 'primer' or 'basics' about hifi explaining this? I wish audiogon had a section explaining technical basics for beginners like me.

Does a preamp actually add any more gain than the volume control on my CD player does? Or is it 'better quality' gain? I heard somewhere that the volume control on my CD player can both attenuate and add gain to the signal. So for low volumes, why would having a preamp be better if the improvement were primarily due to gain? At low volumes, would the direct sound be better, if it weren't for impedance mismatching?
Did you know that the signal comes out of your components at 'full volumn' and that your volumn control knob does not increase but rather it reduces the strength (volumn) of the signal?

A really good pre will do this better than some. My own Cd player has a very sophistaced volumn control so it serves as a pre even for signals from other sources, such as my DVD and satellite. Tone and bass controls are not not needed.
Getting the book Robert Harley's "The Complete guide to High End Audio" would give you a good introduction to the basics of electronics and many other useful bits of information. For me pre's have always improved the sound. Without pre might be a bit cleaner but with is richer.
Yes, a preamp will add some distortion or cause the signal to "lose" something. But like any other active component the power supply, regulation and output devices will determine everything. A top-notch preamp will overcome any loss that occurs with adding another compnent to the chain. The quality of the output of a good preamp in terms of dymanics, frequency response and voltage swing will be much better in many cases than the output of cdp's all because good preamps have better power supplies.

With respect to impedance matching, the lower the output impedance of the preamp means that more voltage is transferred to the amp. This is because power is eaten by resistance - and the more output resistance a preamp has, the more power it consumes from the signal, leaving less voltage at the amp. The amp is now forced to swing a higher voltage difference in a fininte amount of time making things more complicated and affecting the sound quality. So a preamp with a lower output impedance than a cdp is an overall plus - even though there is a slight loss in signal transfer.

Bottom line - good active preamps will be a plus because if they weren't, then who would buy them? We would all have passives or cdp's with volume control.
I just read reviews of Robert Harley's "The Complete guide to High End Audio" on amazon.com, and there was some pretty serious criticism there about his lack of technical knowledge and techincal errors in the book... makes me wonder if it would be the best source to learn about impedance matching? Any other suggestions for a good primer? It would be great if audiogon had a FAQ. Another option would be to use wikipedia... here are a couple of links:

One thing that no-one seems to be mentioning is that the line section of a preamp also has to control the interconnect cable. The better they do this is often also part of the measure of quality a preamp can portray. Passive controls do not control the cable at all (hence the difference in sound from low to high volume) so if a preamp is effective at controlling the cable *and* is competent at the other tasks of a line section (low distortion, wide bandwidth, etc.) then it can *easily* sound better then a passive.

Some might look at this as an impedance issue, and to a large degree it is, but it is not to do with how the unit drives the amplifier so much as whether or not the construction issues of the interconnect are adequately bypassed enough so as to be minimized.

And mind you, not all line sections are up to this task. If not, a passive will be better...
Harleys book is what several members of gon recommended to me as I was getting back in the field - I found it very useful admittedly not so much for the nitty gritty electronic analysis - as more general overview of each components issues and what to look for in putting together a system
Thank you for posting, Ralph!

I, for one, understand the preamp role better from having read your comments.
I agree, thanks for the post! Could you explain how the preamp controls the interconnect? Thanks in advance.
An attenuator (passive) doesn't control anything but the volume function (voltage usually). In a passive attenuator, it's the source component that effectively drives the load via the volume control + connecting wires. If one has a source with a strong output stage &, say, a compatible impedance TVC (or a carefully designed ladder) there should be no problem in driving the load without an active stage inbetween...
Of course, change the source & you may have problems.
OK- Here's how the preamp line stage should function.

The problem is that the interconnect cable has electrical parameters, capacitance for the most part, that interacts with the signal. The higher the impedance, particularly of the source, the greater the interaction. Reducing the impedance of the amplifier input also serves to reduce the interaction.

The problem is that amplifier inputs have to fairly high to accomodate various preamplfiers, some of which would be unhappy with a low input impedance. Thus most amplifiers are about 100kohms at the input. This is generally high enough not to interact with preamps, but it will interact quite a bit with interconnect cables. The input connectors themselves can play a role at this impedance also.

Thus it comes down almost entirely to the source to control the capacitive (and other) effects of the cable. A passive control tends to have a significant impedance in series with the cable at low volume settings and this causes audible interactions, which are reduced as the volume setting is increased. A line stage OTOH has a fixed impedance (hopefully fairly low) that has less interaction with the cable, and if low enough impedance will 'swamp' most of the adverse effects the cable may have.

IMO, this is one of the more important functions of the line stage- to provide a low impedance buffer to control the cable, while allowing for some gain and use of a proper volume control in such a way that the setting of the control does not interact with tonality, only volume. Its a bit of a task.

The input impedance of the amplifier could be reduced, which would help, but many tube preamps in particular would suffer low frequency loss as the input impedance of the amplifier is reduced, due to the coupling cap at the output of the preamp.

Balanced lines tend to reduce some of the noise pickup problems that are an issue with single-ended cables. To really take advantage of what balanced lines can offer, a low source impedance from the preamp is really nice to have. Then it is possible to send signals over 50-100 feet without degradation.

The recording and broadcast industries have known this for decades, but it is not that well known to audiophiles. But if you think about it, most of the significant recordings from the late 50s and early 60s that are still revered today were done with mic cables that carried the microphone signal as far as 200 feet! And this before the age of exotic cables. None of it would have been possible with a passive control interjected- you need a robust low impedance to make it happen.