Let Me See Your Desktop!
Pages: 1 2
Clive Semmens (2335) 3276 posts |
True, but not relevant, since the picture doesn’t change on those double or triple exposures, it just flashes at 48 or 72 times a second instead of 24. A picture that flashes at 24 frames a second would be horrible, but one that doesn’t flash at all and changes 24 times a second is absolutely fine and completely smooth looking. Edit: almost true – but fast-moving objects do produce unnatural-looking juddering effects. This can even be noticeable at a hundred frames a second or more if they move fast enough. But not headache-inducing at any frame rate; just not quite natural looking. And exactly the same in the cinema and on TV of course, but I bet you’ve never noticed it. You probably will now… Interlace TV is really only 25 fps (with that old 50 fps flash rate) as well. |
Rick Murray (539) 13861 posts |
One can only imagine that early cinema did indeed run the shutter at 24fps. I have two 8mm/Super8 projectors. One has a fine blade that flashes the frame twice for each movement, and one has larger blades that move to cover the changing of the frame.
The point I was making about the difference between updates and redraws. As long as you weren’t watching a film or playing a game, one could run a flat panel monitor at something silly like 10fps – as the pixels stay as they were until updated. A 10fps CRT, on the other hand, is the tuff of nightmares.
There are numerous reasons for this. It is quite possible that a fair few movies of the DVD generation underwent three-two pulldown to convert a 24fps film to 29.97fps (NTSC). This will then either be slowed to 25fps for PAL, or if keeping the same speed, will require discarding about one frame every six, not to mention trying to find some way of interpolation to restore the 100 missing lines of resolution. Going from film to PAL is simpler, it is just sped up to 25fps and directly telecined at that rate. But if you’re using a computer? Well, you then need to take this source (at 25 or 29.97 fps), apply interlacing, and then map it into the display size (as there’s likely no longer an expected 1:1 relationship between video pixels and display pixels); all the while translating it temporally to a video system that is likely running at 60, 75, or 85fps. Which then gets sent to a video display panel that treats each pixel as “sample and hold”, and for a lot of the older generation of LCD panels, while they could accept a variety of input signals at various framerates, the actual LCD hardware updated at exactly 60fps (the sometimes quoted 16ms response time). With modern multistandard equipment, it is often left to the device to perform a lot of these conversions. Your display, when run at 24fps, and hooked to a media player outputting a film at 24fps, will still be running the display panel itself at 60Hz (or thereabout). Which means that it has to pleasantly map 24fps to the native device refresh rate. It’s quite likely that it buffers an entire frame and switches it in one go, however 60/75/etc divided by 24 is not a whole number, so there may be small hiccups. Even 100Hz displays, don’t divide by 24. And as for European domestic DVD players that are capable of playing back NTSC discs? Some of the built-in conversions from NTSC to PAL are quite horrible. I got rid of a player I had back in the UK as the initial DVD release of a series I bought was in NTSC (in the UK!). It played. It just looked awful. It’s a shame that people tend to associate high motion video with the home video camera era, and seem pathologically brainwashed into believing that a slightly blurred, muted, 24fps original is “cinematic quality”. Things will be a lot nicer when HD/60fps content is the baseline and 24fps is an anachronism of an older age. So, your cute little spaceship whizzing across the screen? Just take a moment to try to imagine exactly how many conversions that it went between camera and screen. :-)
No, I am too busy noticing the really bloody obvious compression artefacts in HD video. Which, it seems, not a lot of other people can see. |
Grahame Parish (436) 481 posts |
The worst compression artefact I notice is in broadcast TV and video playback shots where someone is walking past a striped background, like railings or fences. There’s usually a block that leads slightly in front of the movement and trails a bit further behind the movement. I find that so annoying and wondered why they couldn’t add these cases into the compression/decompression algorithms to minimise the effect. |
Clive Semmens (2335) 3276 posts |
I’d not even started to think about compression artefacts! But then most of my computer work is text or static images anyway. I like a sharp, hi-res screen with a good colour gamut, then as long it doesn’t flicker I’m fine. |
Chris Mahoney (1684) 2165 posts |
It’s not even just the old ones. I have a Sony Blu-ray player hooked up to a projector. It can play in true 24 fps where the projector projects at 24 (actually 120, with each frame played 5 times, like you discuss above)… and can also take native 25 fps content and munge it, resulting in ‘judder’ every second. Joy. |
Bernard Boase (169) 209 posts |
Desktops or desks’ tops? Here we have: ARMX6 Desktop The other side of the room is Desk 2 which has a Raspberry Pi 3B+ on the left, and on the right a Mac (not mine!) showing the ARMX6 desktop via VNC. You can also make out spare Pis and an AverMedia capture device that I would love to make more use of. |
Rick Murray (539) 13861 posts |
Whoo. Even at that screen size, your icon bar is full!
What’s the 45? I see, also, a copy of “The C Programming Language”. One could say that anybody that doesn’t own (or has read) that has no business calling themselves a programmer. ;-) Mine’s on the shelf beside the two ARM ARMs. |
Clive Semmens (2335) 3276 posts |
Okay, my hand’s up. I do own that, but haven’t read it all & certainly am not a C programmer – not much beyong Hello World in C. Well, a bit beyond, but very far from professional. Do I get a Get Out Of Jail Free card for having written the ARM Assembler Guide & the best part of the 2007 edition of the ARM ARM, the first Neon docs & the first Advanced SIMD docs? (Although apart from testing my understanding, I’ve done very little programming in ARM Assembler post-26-bit days.) |
Steve Pampling (1551) 8180 posts |
Well a copy of MiniDisc would tidy up the left-hand side I think and I’m not sure why both Edit and StrongED would both be loaded. |
Clive Semmens (2335) 3276 posts |
To fill up the icon bar? How about having Zap on there as well? |
Bernard Boase (169) 209 posts |
An unintended fool-ya. It’s just a black CD-ROM designed to look like an EP, even with grooves (though circles) moulded into its top surface. Even the data surface is black, but it seems to work. |
Bernard Boase (169) 209 posts |
And next to it, even more influential in my earlier programming days, Kernighan and Plauger’s “Software Tools” (bought December 1977) in which all the code is in Ratfor (Rational Fortran) that is translated into Fortran for standard compilation. A similar approach was taken by my employer at the time (Shell) who called theirs, er, Shelltran. The approach really did improve one’s code. |
Clive Semmens (2335) 3276 posts |
I no longer have my copy of Daniel D. McCracken’s Guide to Fortran IV Programming… |
Stuart Swales (1481) 351 posts |
McCracken! Blimey – a blast from the past – that’s where I learned FORTRAN from. The oldest book on my programming bookshelves still is Rodnay Zaks’ “Programming the 6502”, purchased 17 AVRIL 1979 for 68,50FF |
Rick Murray (539) 13861 posts |
But you have experience with assembler instead. That means you intrinsically know most of the stuff that needs to be pointed out to users of higher level languages. Oh, and you can laugh at those who think “GOTO” is an evil instruction. In processors, there’s only a hard branch (B in ARM, JMP/Bxx in 6502) or some sort of returnable branch (BL in ARM, JSR in 6502). So it’s all GOTO really! ;-)
What about the earlier ones? Here are mine:
Because they operate in subtly different ways? Certainly, I find it fairly easy to convert between BASIC and text using Edit. Something Zap makes a little more difficult due to working directly with tokenised code.
Pfft. :-p |
Stuart Painting (5389) 714 posts |
I thought to myself: “Yes, I’ve still got my copy of McCracken. It’s in this bo… wait a minute, it’s not there…” Oh dear, chalk up another item that presumably found its way into the skip in 2009 :-( |
Bernard Boase (169) 209 posts |
…and two to the left, with a less readable spine, Kernighan and Plauger’s absolutely indispensable “The Elements of Programming Style” (1974).
Edit for simple things like Rick says, StrongED for ListOfFound and column-wise selection among other functions. |
Clive Semmens (2335) 3276 posts |
I have that thinner one, but it predates my time at ARM. The other one? I’m not sure whether that was one I contributed to or not :-D – tell me the date of it & I can tell you.
I think mine was probably among the things that were trashed when a bunch of National Front thugs trashed my flat in Bradford while I was in India 1983-4. |
Tristan M. (2946) 1039 posts |
I thought I did this. I seem to have a photo already anyway. Sloppily bodged together standing desk. The junk on the desk is a little different to when I took the photo, and the case for the Pi Zero is different but it’s still roughly similar. |
Pages: 1 2