Comments by "Lawrence D’Oliveiro" (@lawrencedoliveiro9104) on "Technology Connections" channel.

  1. 0:29 I think those are only DVD-DL discs. I don’t see the two-halves thing on any of the DVD-R discs on this spindle on the shelf.
    2
  2. A lot of the generation loss copying between consumer VCRs was not so much actual degradation of the video signal, but poor timebase sync. There were gadgets you could buy (I have one) that would clean up the sync and give you much better copies. Would that have made a difference to the comparison?
    1
  3. I just realized the, ahem, connection with Burke’s “Connections” series. Thanks for pointing it out.
    1
  4. 7:21 Why jumble up the data? Because there was an expectation that a scratch on the disc, for example, could destroy a whole lot of bits in a row. But because of the shuffling, these destroyed bits end up spread out among a bunch of other good bits, lessening the impact of the run of errors.
    1
  5. 7:49 OK, you mentioned that point. ;)
    1
  6. 23:57 It was never going to work with Windows, put it that way.
    1
  7. I can only see tiny thumbnails, with no option to make them bigger.
    1
  8. 3:25 A quick ferret around the dub-dub-dubs for “jvc rx-222” suggests late 1980s-vintage. Which would be consistent with my Sony VCR from 1989, which has similarly pretty displays.
    1
  9. Reed-Solomon has never been improved upon. ;)
    1
  10. 6:53 That’s mono audio. The hi-fi stereo audio is recorded (I think) by extra heads also on the spinning drum, along the same tracks as the video. How do they keep from interfering? Through a magical feature known as “depth multiplexing”. The audio is recorded first, and because its frequencies are lower, the magnetization penetrates deeper into the tape. Then the video is recorded, which erases the audio close to the surface, but leaves the deeper part alone. During playback, each layer can be picked up by heads attuned to its particular range of frequencies. Yes, it all really is insanely clever...
    1
  11. 14:59 Conversely, Sony’s Video-8 format for camcorders decisively wiped the floor with its VHS-C competitor. The latter was a kind of hacky adaptation of VHS to a smaller cassette, with consequently shorter recording time, while the former was designed specifically to maintaining a higher time/quality tradeoff in the smaller format. Even the much-hyped ability of VHS-C tapes to be played in a VHS VCR (by putting them in a kind of adapter that mimicked the larger standard VHS case) failed to tip the balance.
    1
  12. FLAC is good, because it is a) lossless, and b) can include metadata (artist, title etc) that common players like VLC can recognize.
    1
  13. 2:12 Ayup.
    1
  14. 1:44 Blame your OS for the misrepresentation. My OS uses the correct units.
    1
  15. 19:13 Ecru to you, too! I like the maroons. I think purple, green and maroon would be my three favourite colours. Also, I think there should be more words for different shades of green. Every time I look at a forest by the side of the road, I am struck by how many different colours of leaves there are.
    1
  16. 4:47 As I recall, the BBC were experimenting with video recording onto steel tape in, I don’t know, the 1950s or so. The reels were huge and heavy. And the tape had to move at a very high speed to cope with the video signal bandwidth. The result of all this was moving parts with a lot of kinetic energy in them — dangerously so. And the edges of the tape were sharp ...
    1
  17. This was a frequently-commented-on drawback of CDs, that they had less room for that lovely album cover art.
    1
  18. That jumped-up new-fangled vinyl never sounded as good as authentic shellac.
    1
  19. 20:54 Yes, the light amplifiers have to be active devices. They need external power in order to amplify, after all. The pumping is just how lasers work. Remember what the “A” stands for in “LASER”? That’s right — “amplification”. And the “SE” is “Stimulated Emission”, aka “pumping”.
    1
  20. 11:58 I have a sudden vision of a time a few decades from now where giving unlicensed advice like this will be deemed illegal ...
    1
  21. You are a very patient man, all things considered. ;)
    1
  22. 0:55 People kept mistaking him for one. “I am not a Crookes Radiometer!”
    1
  23. Hearing is a sense. If you are not “hearing” it, then what “sense” are you using?
    1
  24. When I was little, back in the old country with the UK-style plugs, I used to be confused by the word “FUSED” embossed into their cases. Because adult relatives only used the word “fused” to mean “the fuse has blown”!
    1
  25. I love that name “Philo T Farnsworth”. It sounds so ... American. Can you think of another famous American with the initials “P T”? Another fun thing Philo Farnsworth was involved in was nuclear fusion. Imagine a nuclear fusion reactor simple and cheap enough to build in your home (though it won’t produce output power worth a darn). Look up “Farnsworth Fusor” to find out more.
    1
  26. Anything involving boiling water coming into contact with coffee is a bad idea.
    1
  27. Digital filters are quite capable of implementing a “brick wall” cutoff.
    1
  28. To make it more relevant to the younger kids nowadays, how about: Sony → Apple IPhone VHS → Android
    1
  29. 09:18 Actually, the real reason for EFM encoding is to reduce the frequency bandwidth of the signal that the circuitry has to cope with. Remember that each 1 represents a waveform transition. So two successive 1s in a row represent a frequency of 14 bits per encoded byte × 2352 bytes per sector × 75 sectors per second ÷ 2 transitions per cycle = 1234800Hz. By insisting on at least 3 0s between each pair of 1s, this is reduced to a quarter of that. At the bottom end of the frequency range, the maximum number of 0s you can have in a row is dependent on how stable you can maintain your clock sync.
    1
  30. Sorry, since it’s a minimum of 2 0s in a row, that’s a third of that frequency, not a quarter. The number of combinations of 14 bits that satisfy these constraints is 267, just slightly more than the 256 needed to encode 8 bits of data. (Which explains why 14 bits are used, and not 13 or 15, say.) Some are used for special marker purposes. There is also a 3-bit “packing period” between each of these 14-bit words, which can contain an extra 1-bit as necessary to even out the total numbers of 1s and 0s to avoid the overall signal having a DC offset. Source: John Watkinson, The Art Of Digital Audio , of which I have the 2nd edition from 1994.
    1
  31.  @thehoeslaper  Which country was this? I was paying NZD80 for a box of 5 DVD-R discs for use with the “Superdrive” in my 2002-vintage G4 Imac, and never had to “show ID”.
    1
  32. 18:10 Or, just plug your camcorder into the TV and let it play the tape directly. A Video-8 camcorder is also a VCR, after all.
    1
  33. 3:32 Pixels are not little squares. You hereby incur the wrath of Alvy Ray Smith.
    1
  34. 12:20 I think they have to be arranged diagonally, not vertically and horizontally. That way, straight horizontal motion can be interpreted as a mono signal (mix of both channels, rather than just one channel), for compatibility with equipment that can’t handle stereo.
    1
  35. 12:41 Ah, I see you noticed.
    1
  36. 4:52 Has anybody ever seen any 16-RPM discs? Are they in fact 16⅔ RPM, being ½ of 33⅓ RPM?
    1
  37. 10:42 OK, you have. It figures. ;)
    1
  38. 10:25 Only true for matte displays, which are the best at reproducing colour. Some people prefer glossy ones, for some reason—maybe because they look prettier, even though their colour reproduction is worse.
    1
  39. 14:28 Bring back the pristine sound of shellac! That new-fangled vinyl never quite sounded right.
    1
  40. 10:31 A moment of silence for the dear departed Cosmos man, please.
    1
  41. That description applies to sign-magnitude as well. The difference is, consider 16-bit integers: unsigned they can represent values from 0 to 65535. In twos-complement, you interpret the encodings after 32767 to actually represent that value minus 65536. So the encoding that would be 32768 becomes -32768, 32769 becomes -32767, and so on up to 65535 which now represents -1. The clever thing is, if you add two such numbers as though they were unsigned, you can interpret the result as signed and the answer is correct. Conversely, to subtract two numbers a and b, just add a and -b instead.
    1
  42. This idea that inaudible frequencies can make an audible difference ... the only rational explanation I can come up with involves some non-linearity of response of the ear. Perhaps not an explanation that audiophiles would like ...
    1
  43. 9:20 Interesting coincidence that cranking up the gain of a digital sensor increases the noise, which also manifests as a form of grain. Or is it?
    1
  44. Re the up-and-down versus side-to-side motion: stereo records make use of both: motion along one axis gives you one channel, along the perpendicular axis gives you the other channel. I think they’re actually oriented diagonally, so each channel gets both vertical and horizonal motion.
    1
  45. 13:03 Apple were quite keen on PhotoCD. They built support for it into QuickTime and MacOS, so that when you inserted a PhotoCD disc into the drive, it would automount and the images would appear as files (in a range of resolutions) that you could just drag and drop to your hard drive. I still have the CDs that came bundled with my 1993-vintage Mac Centris 650, one of which is the Kodak PhotoCD sampler disc. But yes, I remember asking a Kodak guy at the time why they didn’t just adopt JPEG compression (which was the hot new thing back then). Their format allowed for a range of resolutions to be encoded into the same file, but that didn’t seem to be important at the time, and it seems even less important now.
    1
  46. There was a GIMP addon that provided that capability some years before Adobe figured it out.
    1
  47. You didn’t mention THD? Yes you did.
    1
  48. 8:10 Ah, but what about the thin horizontal line running across the screen about two-thirds of the way up? I think that was a wire that was needed to reinforce the aperture grille. Almost unnoticeable -- until someone points it out to you.
    1
  49. 9:16 Ah, you did mention it.
    1
  50. Further thought: you put the wire at one third of the way up, whereas I seem to remember with the Apple monitors it was two thirds of the way up. I think Apple felt it would be less noticeable if the tube was flipped upside down.
    1