A dim question

No, energy use doesn't reduce at anything like the light output. To give an example, I measured a 500W halogen dimmed down to about the same output as a 40W lamp -- it was consuming 300W. So to a first approximation, assume there's no reduction in the energy usage (i.e. don't bother dimming filament lamps in order to same electricity).

That's to be expected. When you run a lamp dimly, the colour temperature of the filament reduces - with the result that much more of the radiation produced is at infra-red rather than visible wavelengths. You're still consuming electricity, but you're producing heat rather than light.

What an interesting observation!

My parrots love to get as close as poss to a rack of three incandescent spots (small ones - 240v) with reflectors - you know, the kind of early

Would there be more or less total IR/heat output dimmed or full voltage?

Just how long IS that piece of string?

There'll be less of everything when dimmed, but the visible light will be reduced by a greater factor than the IR.

What I wonder is if there is a discrepancy between what a dimmed lamp consumes and what you get billed for, I mean can the eddy-current disk in my meter accurately follow the triac switching on at the latter part of each half cycle?

No. the light output goes down roughly as the square IIRC.

i.e. half power 1/4 the light.

I could be wrong here, it may even be worse..a lot depends on the spectrum of the original bulb.

Yes, your meter accurately follows it.

That's a blow, for many years I have thought SCRs were a legal fiddle! What if it's1 cycle on amd 9 off?

Likewise.

Your meter simply integrates the instantaneous product of voltage and current, so whatever nasty waveform you throw at it and whatever the power factor, it will read out the energy used.

Some larger commercial customers get penalised for low power factor, but domestic customers don't.