It wouldn't be that hard to separate that from the other changes like with one of the other on/off loads like the toaster or fridge or a hot plate coming on.
Even with the small number of other appliances which do vary the current they take in a continuous way, those changes should be a lot slower than the backlight current varying with the program content.
But I don't believe it would be possible to separate out the individual backlight currents when you have two TVs running at once showing different channels.
Yes, you have never been able to spell out how you work out what the backlight current is with a particular TV when there are two of them showing different channels.
I haven't changed anything I have said on that issue.
It isn't actually, most obviously with the most recent movies being watched on something like netflix etc.
Yes, that is a small subset of what isn't being broadcast live, but may well be of interest to movie makers. BUT it would be a lot more viable to get the stats on who is watching what from netflix instead of trying to get it from smart meters.
You clearly can't manage even the most basic comprehension.
House next to me when it was all privitised turned out not to have any metering for several years afterwards. It changed hands, and after about a year, the new-ish neighbours knocked on my door and asked if I got any electricty bills, because they'd never had any. They were an elderly couple, scrupulously honest, and really scared they were going to suddenly get a giant bill. It took months before anyone could sort it out - they couldn't choose a supplier because their house didn't have a meter point reference. It seemed to only be possible to get a meter point reference when a supply is installed, which had been there for decades. They checked with the previous owners who said they never had a bill either. It did get sorted out eventually.
The set's consumption, apart from the backlight, will be about 1/3 of the maximum. The remaining consumption, which we are measuring via processing the smart-meter's readings, is the backlight.
Which has different brightnesses, just like a colour movie.
Not nearly as much as the backlight does.
Those are all the other things happening, with their own temporal patterns. They will be (digitally) filtered out.
I could stick a meter on the AC cord, but that would be cheating.
So I'll report the numbers sent to the wireless household power meter I have which is fed from an aftermarket clamp-on device at the main fusebox.
It reports a new number every four seconds.
The "noise" in the house with the TV turned off is 404 watts to 415 watts.
If I turn on the TV, with a black screen it jumps to 504 to 516.
Plug my laptop (already running) into the TV, with a largely bright white screen display, and it's 522 to 533.
[All observations made over about two dozen readings]
So I'm going to conclude that the backlight is somewhere in the range
30-40 watts.
Irrelevant, when we are measuring on particular type and size of display. (It's a ten-year old cold-cathode 26" set, with rather more electronics inside than I'd expect for a modern one).
In message , at 06:16:54 on Sat, 14 Mar 2015, john james remarked:
You are spectacularly missing the point.
The object of the exercise isn't to find out what people are watching on TV (whether that's live or streamed or pre-recorded).
It's just an example of one of the appliances whose footprint can be separated out from the others. If you can separate out a TV whose backlight is varying by 20watts from one scene to the next, you can separate out people turning on/off 20watt light bulbs and so on.
There are codes of practice for how far back the utility company can bill you in those circumstances, as well as allowing time to pay.
Long ago I lived in a rented house for a year, and never got a water bill. The water company *insisted* the house didn't exist, and therefore couldn't be billed. But I did try...
Sure, but that isn't what was being discussed. What was being discussed was whether it would be feasible to work out what was being watched on TVs that do vary the backlight with what is being displayed on the screen to work out what channel they are watching.
Its trivially easy to work out what is the TV and what is a light from the different power levels and the fact that there are only a small number of TVs in the house compared with the number of lights and from the way they are turned on and off time wise. And that works for all TVs, not just the ones that vary the backlight with what is being watched.
But not relevant to what you can deduce from what the TV is showing.
See above.
Only with a single TV in use, not when more than one that modulates the backlight depending on the channel being watched.
No, not with being able to do anything useful about working our what is being watched on more than one TV at a time.
I have always agreed that its would be possible to work out what is being watched on a single TV that modulates the backlight level depending on what is being watched.
What I have always disputed is whether that can be done when more than one TV is being used at the same time.
and if you have reference patterns to compare to, you can easily filter out one from the other by trial and error, by saying "let's try every pattern that each live TV channel could be creating, subtract from the total and see if what's left matches any of the other patterns that another TV channel might be creating. Although in practice you do it with maths inside a DSP.
I don't believe that it is. Its just not possible to work out at any particular instant what current a particular TV is taking when there are two of them that have the current they are taking varying with the program content.
Easy enough with just one TV because there are no other appliances that are likely to be varying like that that take anything like that variable a current over such short time intervals. Even a stereo doesn't.
But there is no way to go from the last one to the first two.
I don't believe that that is feasible even if the two TVs were identical and they hardly ever would be, particularly with the hundred or more channels that the TVs could be watching.
And you'd have to record all the broadcasts forever because of the varying ad content.
All to find out what isn't even of any great interest, who has their TV showing what channel maybe with no one actually taking any notice of what is currently being shown on that TV.
Usually the retailer will allow you the option to retain flat rate pricing or take the time-of-use plan by default. The hidden catch is that some of them measure peak in 15-minute windows, so boiling a kettle is quite expensive.
In message , at 11:31:23 on Sun, 15 Mar 2015, john james remarked:
There is. You take the "pattern" from a TV programme currently being broadcast, and subtract it from the total. You ten see if what remains matches any of the other TV programmes currently being broadcast.
If it doesn't batch, then try a second TV programmes... iterate until you find two whose patterns do add up to the total.
HomeOwnersHub website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.