No Pain No Gain?

As far as the ML No. 32 is concerned, I don't think that either the gain setting or the input offset affects the sound. I believe that these controls are used to match relative levels of each input. I don't think the different settings engage different number of stages of amplification; they just set the selection of the resistors in the stepped attenuator. So, for any given volume level--whether set by going +6 on the gain and -6 on the offset, or just increasing the volume knob by 6db-- the same set of resistors are switched into the circuit.

By the way, I do miss the incredible flexibility and terrific implementation of volume/balance control of my Ref. No. 32 (currently relegated to back-up status).

Greg, I was not referring to my system's noise floor, which is very low. Only if you put your ear on the tweeter and turn the volume past 3 quarters you start hearing a little hiss. If you sit back 8 feet, you can't even hear it anymore. I was referring to theoretical S/N. Anyhow, I don't think the ML is faulty in any way.

Thanks Larry for your explanations. You're right about the matching of inputs, and currently I have only one. So much of it has to do with my perceived comfort of volume level and where the "dial" sits.

Only if you put your ear on the tweeter and turn the volume past 3 quarters you start hearing a little hiss

No worry, looks like minimal thermal noise fm components. Natural.
For yr 32, optimise the gain setting by ear, for each source component. Use simple music, if possible a classical piece with cello. Adjust for maximum energy, dynamics, etc. Disregard "details", "soundstages" etc, at this point.

I don't know how EMMlabs implements its high/low gain option. I would guess that they too are using an attenuator. The issue would come down to which attenuator to utilize. I know that ML uses a series of discrete resistors for its attenuator of very high quality. I suppose EMMlabs also uses high quality components, so the choice probably doesn't matter.

The ML's use of discrete resistors switched in by relays means that it is highly accurate (channels are matched to within .1 db) and minimum number of components are switched in at any gain level. With regular volume knobs that use a potentiometer, one wants to operate at the high end of its range (least amount of attenuation) because that is where the two channels track most accurately. With discrete resistors switched in by relays or dial-type stepped attenuators with discrete resistors, channel tracking is not an issue.

In short, it probably does not matter where you operate the ML No. 32, within its volume range. Also, with the ML you don't have to worry about even accidentally twirling the knob too far or accidentally pushing the volume up button on the remote, because you can also program in the maximum volume for the system.

Greg/Larry, thanks for your contributions. My experience so far is that with most music it really does not matter, its more with my personal taste on what volume level I'm used to and comfortable with. However, with some recordings there are significant variations from the "usual" volume level and at least its a plus to be able to adjust. I guess that it really becomes valuable when you connect different components, as you wouldn't have to continually adjust volume to match the components input level.

Happy Easter
Henry