Tone changes with volume

Has anyone ever noticed how tonality changes depending on volume setting. I thought it was just me until I saw this review of the LW-1 passive preamp:
LW-1 review
"The X-coupler is designed for a wide range of combinations and can drive amplifier loads as low as 10k ohms. This, in theory, should help reduce the shift in tonal balance that results from the changes in impedance caused by the varying positions of the volume control."
Maybe I'm not crazy after all :-)
There's also the little matter that your own hearing perception alters with volume. As the volume goes up, you should hear more highs (and somewhat more lows). At low volumes, your hearing very efficiently concentrates on the all-important midrange. This effect will swamp any impact that a volume control would have.

Sounds like you're not crazy at all. You're normal.
Like Bo says: you're hearing the well known Fletcher-Munson curve effect, which is a bell-curve peaking in the midrange area. Your hearing seems quite normal to me too.
Thanks for the thumbs up.
True. Our ear canals, by their internal shape, are optimized for 2k to 4k frequencies - any thing above or below will very likely cause a change in tonal perception and (definitely) how we perceive relative loudness of certain frequencies. Of course, there is still such a thing as a poorly designed volume pot - but I don't believe the "... shift in tonal balance that results from the changes in impedance caused by the varying positions of the volume control" argument holds any water.
It has been well documented that some gear DOES alter frequency response / tonal balance as drive levels are altered. Moncrieff measured and documented this in IAR over 20 years ago and a relatively recent ( within the last year ) review of a Rotel amp in Stereophile documented the same thing. The reviewer stated that the amp seemed to change tonal balance as the volume was varied and sounded like two different amps depending on if you were listening quietly or had things cranked up a bit. JA's measurements pretty much verified these sonic characteristics when taking electrical measurements on his test bench. For sake of clarity, this was a stand alone power amp and not an integrated. The same thing could take place in an integrated amp or even a receiver though, so don't think that those are any less / more susceptible to design flaws either.

As far as preamps and line level sources go, there are different ways to build & design gain stages and attenuators. Most all products make use of some type of analogue volume control with a potentiometer of some type. Due to the variances in design, source components can either see a widely varying input impedance or one that is relatively stable when loading into a preamp. This has to do with the design of the attenuator / potentiometer aka "volume control". As such, i would not doubt that some sources could be "loaded down" or change their frequency response / tonal characteristics when mated with specific preamps or attenuator designs. The wider the range of impedances that the source component sees during normal operation, the more likely the sonics are to change slightly. This may account for a lot of the sonic variances that people experience when using various passive devices between line level sources and power amps i.e. the lower the listening level, the more resistance that the source sees. Cranking up the volume control would introduce less resistance in series with the signal, changing not only the volume but also the load that the source sees.

The same thing can happen in an active preamp too, but there are ways to maintain relatively constant impedances that a source sees. This can be done by building the attenuator in a specific manner. By controlling the amount of signal that is passed to the amp through a series resistor and how much is resistively shunted to ground through a separate series resistor , the source would always see a relatively constant load via what is called voltage dividing network. Many "audio purists" and "tweakers" feel that this is the the "best" way to do things as it only puts one resistor in series with the signal at any given time. The problem with this approach is that it is VERY time consuming and labor intensive to build, so this type of design has not found its' way into too many commercially produced products. Most other designs place some type of resistive trace or multiple resistors in series with the signal, which can only add up to increased signal degradation.

There are a LOT of paths to go down when building / designing a product and that is why some sound SO different from one another and respond differently when mated with various other components. Not all circuits are the same nor are they all equally stable. Sean