Posts by NXM 122 posts • joined 12 Sep 2012
I ask the Pixies.
I know it sounds mad, but if I've lost something I ask the Pixies to help me find it. They don't hide things, its me who does that by accident, but they know where they are. Last time I'd looked for a pair of glasses for days, no sign of them. So I asked the Pixies, and 5 minutes later I spotted them in plain sight, right where I'd left them.
I'll try asking them for answers to difficult questions next!
The company I worked for in about 1989 had a luggable PC with an LCD screen and a VGA socket, but you couldn't just plug a screen in. Oh no.
You had to flick a set of tiny dip switches hidden on the side to a particular state with the machine off, then plug the screen in and boot it. If you didn't set the switches back to LCD mode afterwards, it killed the LCD.
That didn't last very long.
I'm still bashing away on my Model M which I got when I was ordered by the PHB to swap it out with a new one on a Glasgow shipyard in 1992 as the boss thought it was too dirty.
It'll still be used by the radioactive evolved cockroaches who'll replace us after the next war.
I think it probably would be a criminal offense, at least for right-pondians, if this gentleman's conduct and punishment is anything to go by:
https://www.theguardian.com/uk-news/2018/apr/18/hairdresser-daryll-rowe-given-life-sentence-for-deliberately-infecting-men-with-hiv
Yep, I saw. Very good video, particularly the advice about old equipment and analogue scopes - they're not digital, so they can't lie to you.
But his assertion that you don't get a stepped output from a DAC isn't totally right. The DAC takes the input value and converts it to a voltage which stays the same until the next update, which makes the stepped effect. All DAC's do this. Ways of overcoming it is by oversampling and filtering the output at the Nyquist frequency, which removes the steps and takes high-frequency step-change artefacts out. I find the idea that the samples represent a lollipop graph isn't true in the real world.
My experience is indeed with cheap DAC's, and oversampling the output really does help. More expensive DAC's can do this without any external circuitry. When I designed an audio player years ago when triggerable players were really expensive I did used 16-bit data, oversampled on the fly, with 11KHz sample rate to get decent quality voice output. 8KHz didn't work very well at all.
Most of my experience is with ADC's, and I still don't get the result that sampling near they Nyquist rate results in a good rendition of the original wave. If your input is slightly less than half the sample rate for example, you get a gradual increase then decay of the amplitude all the time. I've found that you can feed it up to 1/5 of the sample rate if you want a decent result, and even then it's ragged. I find the Nyquist rate tells you more about aliasing filters than input frequency limits.
He's right about the bit depth though.
Thanks for the comments though, other peoples views are always informative.
I feel I do know what I'm talking about, because I build these things and am telling you what I see whenever I try it.
You get a gain rolloff and worse rendition as you increase the input frequency.
If you don't believe me, make yourself an adc circuit and feed the results to a dac, ensure the gains match, then put both into a scope set to subtract one from t'other. You'll see.
" the sample rate determines the highest reproducible frequency"
Technically true, but if you send a sine wave into a sampler and observe the sampled wave compint out of the other end, you see the wave gets increasingly inaccurate above about 1/10 the sample rate. You can get a decent rendition of the original wave at half the sample rate, but only if you happen to sample at the peaks - ie it only works for certain frequencies.
So the higher the sample rate the better the result, depending on where your hearing tops out.
Following being at many very loud gigs in my youth I can no longer hear bats, but even crumblies like me benefit from higher sample rates than that offered by CD's.
I totally agree about the code bloat problem. I write real-time stuff in assembler for cheap processors because it's fast and small. Also, if I save £1 per chip that's extra profit in my back pocket, where it belongs.
I wouldn't get away with it in C because it just doesn't run fast enough. The downside it that it takes bloody ages to get it right.
I've still got an 8-pin Epson in the Dead Format Museum! Along with an RCA Selectavision player, Philips N1700 vcr's, BBC computers, a Nixie voltmeter, Atari computer, original Laserdisc players, etc etc. Young uns don't appreciate how difficult and expensive it was to do anything such as store & replay media then.
I've worked in hvac and believe me, they're not jokes. Not when you've seen a setup you left working perfectly under automated control to integrate hearing or cooling to keep rooms at set temperature, then gone back a couple of years later to find chillers festooned around the windows, all on full blast, fighting against the heating, also on full blast.
Because people apparently don't understand how thermostats work.
I live a few miles away from Carlisle.
There used to be a health food shop run by an elderly gent (a lot older than me, anyway) who was very helpful and courteous, but not snappily dressed.
He owned a huge chunk of the city centre. Don't judge by appearances! Except if it's me, as I'm shabbyly dressed and a total pleb.
If it were a parliamentary election I'd agree with you. But the brexit vote was one single issue presented to the entire population and not a 'safe seat' where the same party gets in every time.
There were effectively three choices: leave, remain, and don't care (represented by not voting). A third of people chose don't care.
I was on the receiving end of a text tsunami a few years ago.
The local neighbourhood watch (run by the police) had set up a text service that was supposed to alert people to the presence of bad 'uns, but hadn't made it clear exactly what you were singing up to. Lots of people joined. At some stage those running it sent an incomprehensible 'test' message from a number nobody had heard of. Some people responded with a 'what the hell was that's response, which was copied out to the entire group. Followed by more 'what the hell" responses which were also copied to the entire group. The result was constant txt pings from the tortured, and evidently non-monitored, server.
I had to turn the phone off as it became unusable. The only way I worked out what was going on was some fierce googling on the sending number which turned out to be owned by the plod.
Some time later they decided to simultaneously phone all numbers on their database at 4am because of another test message put in there months earlier by someone who forgot to remove it. Both my landline and mobile going off at that unholy hour almost caused me a coronary.
They scrapped the service after that.
The Register - Independent news and views for the tech community. Part of Situation Publishing
Biting the hand that feeds IT © 1998–2021
