Fun fact: female voices both have more spectral information than male voices (typically), and are easier to understand in noisey environments. There’s a reason most automated PA recordings are female speakers.
Huh, according to the wiki, my knowledge is out of date. It’s conventional wisdom that female voices are more intelligible than male voices, and are easier to pick out of a noisy environment, but that looks like it could be an artifact of doing the early human factors testing with pretty much exclusively male pilots. When you run the same testing with equal proportions of male and female pilots the bias disappears.
Try an android phone next time
Edit never mind beaten to it.
I thought they nailed it in “Her”.
Can’t they just get Peter Cullen or Frank Welker to do it?
Fuck hell no. I have had speaking computers for thirty years and listening computers for fifteen. The reason I use computers is precisely because they are not emotionally manipulative. I think that people worrying about how others feel about them is a fatal flaw, and I think that it is terribly counter-productive to instil it in tech which does not suffer from those problems. People could use such tech to get over this rather than indulging it further. Trying to sound “friendly and approachable” is both sleazy and a waste of time.
I see emotional incontinence as being the most significant obstacle against establishing any sort of civilized society.
Funny thing is (and as per the article), Scarlett Johansson didn’t record her lines with Joaquin Phoenix. He did all his acting with Samantha Morton, then Spike Jonze changed his mind and replaced her post-production.
Any voice would be highly annoying when it’s being used to lecture you about using it incorrectly, along with the non-existent “unexpected items in the bagging area” rather than get on with checking out your shopping.
Maybe Arnold’s Terminator voice would work?
Please wait for an attendant.
Please remove the last item from the bag and scan it before placing it in the bag.
Majel Barrett-Roddenberry’s voice is the only suitable computer voice.
The whole concept of self service checkouts is annoying as fuck. If we just refused to use the bastard things, they’d take em out again.
They can be pretty quick if you just got one or two things and have memorized the station’s programming. It also depends on whether you can cutoff the thing’s jabbering mid-word.
At my local grocery store, I can buy a can of Rockstar on the self-checkout machine in about 15 seconds, since I know the exact button placement, and the order of operations. I’m gone before the receipt even prints.
They get really obnoxious and slow when you bring your own bag, have more than 3 things, don’t follow the instructions it gives you, and want to pay in some weird exotic currency like a personal cheque or Western Union wire transfer.
I only use them when I want to change a £20 and am buying something costing 50p. It saves me having look apologetically at the person on till duty.
I love them. I can pack at my own pace, and in my own way. The only thing I dislike about them is the frequency that they object to some item I’ve scanned and placed in the bagging area with no indication of what the problem is. I’m beginning to suspect that they’re programmed to do it to give the staff something to do.
May be a problem with the load cells under the bagging area. They may lie if the object is placed off center, or is too light. If the sensed weight difference of the object placed on the area is outside of some tolerance range, the thing will then complain.
I’m fairly sure that I don’t want people to have access to computer generated voices successfully engineered to be ‘likable’. I’m not looking to bleed from the ears or anything; but ‘bland and affectively flat’ is perfectly adequate for that.
Given the typical situations where synthesized voices are deployed(AVR phone-tree-hellscapes, lousy kiosk experiences, assorted consumer products demanding my attention) I’m hard pressed to think of why I’d want any of them to be more successful at manipulating my emotions. It’s bad enough that people can put social pressure on you for the cost of a minimum wage phone drone, it will only be worse if that can be done automatically for less.
Voice “D” was my favorite but I am a fan of things that are both cute and creepy. My house is full of goth kitsch and my favorite leggings are covered in cartoon dead panda heads wearing pink bows.
It would be a fun voice for a personal item but terrible for anything general or public.
They also don’t work if you’re buying alcohol.
That only matters in states / countries where they let you buy liquor in grocery stores. >_<