I figured there was, but I haven't been in a studio in years. I was pointing out that the consumer grade Beta & VHS isn't suitable for direct use. The early Ampex digital recorders were just becoming popular enough to have some competition when I left the day to day TV Broadcast business. I have requested a tour of several stations to see the newer hardware, but was told to 'F' off.
It looks like the old 35 mm film is the only long term storage so far. I have a portable Sony Umatic deck in storage, & I've seen a lot at hamfests. A couple people who work post production in New York told me they were throwing away their U-matic equipment last year because they couldn't even give it away.
For a while. Other than custom ICs, any other part can be made, including rebuilding the heads. I'll bet parts for the old 1" R-R are hard to find, as well.
--
http://improve-usenet.org/index.html
If you have broadband, your ISP may have a NNTP news server included in
your account: http://www.usenettools.net/ISP.htm
Sporadic E is the Earth\'s aluminum foil beanie for the \'global warming\'
sheep.
OK Joerg, here are some snapshots I just took, the rai-uno shows all those brown grades that you complain about, The rai-tre just makes for an example of their studio stuff. And the Russia-today is actually live studio I think, but they got the aspect wrong, For other stations I would have to wait for the news, a bit late here. ftp://panteltje.com/pub/snapshots/ But this shows there is nothing wrong with colors in mpeg2 digital, in fact mpeg2 digtal colors are _not_ affected by tranmission problems, only by wrong enoders (very unlikely) or decoders. Also after traveling 40000 km x 2 to / from space, no noise, no errors, not even counting the land based connection to the sat station, and it is raining here too.
Nope. Companies that have atrocious products, usually have excellent support and support web sites. That's because they use support to compensate for product and documentation deficiencies. It's often cheaper to answer the same dumb question 5,000 times, than to fix the product.
Similarly, companies that have superior products don't need an elaborate support structure. An amazing number of products literally work out of the box and without any complications. In other words, if the product were any good, we wouldn't need support, and I might be out of some work. I would prefer having the manufactory spend their time and effort on better products, not on better support.
One of my customers (during the dot com boom) insisted that literally everyone involved spend at least half a day on the phones answering support calls. That included contractors, so I was included. Prior to that experience, I had a rather smug attitude and presumed that I could do no wrong. Also that the product, documentation, and software were perfect. It took 4 hours on the phone to demolish that theory. At that point, damage control was impossible because the next 2 generations of products were already in the pipeline. Still, it was a rather sobering experience. After that, I erratically would listen in on support calls, and absorbed what clues I could obtain. It helped.
Incidentally, 90% of what I was asked and later, what I heard, really was in the documentation. However, most of those calling did not have possession of the documentation. IT would install the product and lock up the disks and manuals. Sometimes the dealers would simply "forget" to supply the disks and docs. If there was a problem, IT would have to find the docs and drivers. Of course, nobody had an organized filing system so after a few months, the docs and disks were declared missing. Little wonder users were calling for help. Also, many of the telephone calls were to explain buzzwords, acronyms, and abrevs.
[Q]: What's the attention span of someone doing a search for troubleshooting info on the web? [A]: About 30 seconds. If it takes longer than 30 seconds to read an online troubleshooting document, the reader will usually start skimming or give up and see if they can find something else. So much for splattering all kinds of useless navigation and advertising crap on a support page.
--
Jeff Liebermann jeffl@cruzio.com
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558
There are companies with excellent product and great web sites. ON Semi, for example. Which gets rewarded with lots of design-ins, at least from my side.
[...]
--
Regards, Joerg
http://www.analogconsultants.com/
"gmail" domain blocked because of excessive spam.
Use another domain or send PM.
How about they can't find the documentation to read? Ever try to actually read the help files? I mean linearly start at one end and read to the other end. It's not easy, rough reading, totally disjointed, and fairly difficult. Yet, the few users that actually tried the built in help did exactly that. They expected it to read exactly like the printed documentation and were instantly lost when confronted with something different. Similarly, the documentation on disk is often difficult to find and use. I recently installed some wireless card that had the docs on disk. The problem was that I had to slog my way through 2 splash screen, several "click here to do something obvious", and a list of obscure menu choices on a garish page, in order to get to the docs. To insure failure, the search feature would only work on single words.
My guess(tm) is that most of the called did try to find or use the help, but failed.
Well, I don't want to read the docs either. In fact, I don't. I consider it a personal disgrace to need to read the docs. If the product were any good, it wouldn't need any docs. My customers also fail to appreciate the merits of me reading the docs in their presence. The usual question is "do you know what you're doing?" implying that reading the docs is somehow a mark of technical ineptitude. Until we find a way to eliminate the stigma from reading the docs, NOBODY is going to read them.
I have quite a few opinions on this topic. I'll supply just one observation. Manuals that are full of illustration and examples work far better than big thick wordy creative writing projects. In other words, a picture is literally worth 1000.0 words.
What has happened in my chequered past is that the docs mirror the personality of the tech writers. If the writers have a military background, the docs will resemble at military field manual with lots of decimal numbered sections, huge parts lists, monstrous indexes, and very little in between. If the writer has a flair for illustration, it will resemble a picture book or cartoon book. If the writer has a classical education, it will be as if Shakespeare had written the instructions.
Worse, it is literally impossible to scribble a manual that will satisfy everyone. I previously had an old cartoon from Mad Magazine on my office wall. It was a scene from a TV drama, as edited by all those involved in the production. All were radically different and slanted in the preferred direction of the author. There was also a similar cartoon in EDN(?) magazine, showing a child's swing, and the various ways different engineering and field service departments butchered it into something useless.
The same applies to the target audience of a manual. If you're writing for a couch potato, with a grammar skool education, the copy is not going to make a professional programmer very happy. Similarly, targeting the programmer, is going to make the manual unintelligible to Joe SixPack. I have some ideas on how to partially solve this dilemma, but that can wait for another rant.
I don't have enough experience playing support to do the same. However, I once had access to the support logs from a rather large domestic support organization. I vaguely recall that reading the manual to the customer over the phone was a common activity, which implies that they either didn't have the manual, didn't want to read the manual, or couldn't understand the manual. I have no numbers to separate the probable causes.
Agreed. I still occasionally use vi and nroff for word processing, but am slowly switching over to GUI based processors. Many years ago, I challenged a fast typist to a race between her using MS Word 97 and me using vi, nroff, spell, and such. We picked 5 formatted pages of text from a book and started typing. She was faster on the typing, but I caught up when it came time to format and print the result. I lost because I had to lookup a few nroff dot commands. Both were full of mistakes but that was to be expected.
What has really happened is that the users of function key driven word processors (IBM Displaywriter, WordPerfect) and key driven word processors (Wordstar), have been displaced by GUI driven word processors. Back in the days when there was still somewhat of a choice, various tests showed the GUI driven variety to be the slowest, most error prone, but easiest to learn, of the bunch. Whether this constitutes an improvement, is subject to debate.
--
Jeff Liebermann jeffl@cruzio.com
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558
I've seen this happen with Amazon.Com book reviews -- some get crappy reviews for what seems pretty clearly due to the reader just not having the appropriate background to understand/appreciate the text, even when in general it's quite highly regarded.
That's the main problem with LaTeX, nroff, and similar markup languages -- unless you use them on a regular basis, most people aren't able to remember some of the more obscure commands. I spent plenty of hours fighting with LaTeX due to all the exceptions to the usual rules as well...
Yes.
I'm not surprised.
It's an improvement if the program is still "expert-user friendly," that is, all those -key driven functions are still available; most people who use such a piece of software will naturally gravitate towards the shortcut keys. (...although I've definitely run into the opposite sort of people who seem to thing the more clicks the better and don't understand why you don't like having to click a half-dozen times to before a simple function...) Sadly, more and more software seems to only be "beginning-user friendly" and not "expert-user friendly."
I see that you are talking about the sixties. NTSC itself has not preserved that, because the public objected. I will not bore everyone with the changes over time, though i kinda recall what they were, i would have to research it to give a decent answer. Color reproduction has been very good since the early 70's and is / has been nearly flawless since 2 years after the introduction of vertically integrated reference (VIR)
Don't know about you but my old sat box is 8-bit and i see it regularly. I think i am placing the "issue" in the correct place here.
Yes i have seen it. If i understand the specifications correctly (i cannot afford a set just over personal curiosity) it is actually an issue between excessive compression for the material or poor encoder implementations.
Some of the horizontal ones are probably due to the interlaced source material. But it is a bit odd that the better lit left hand edge of the rail is particularly badly damaged (about 4 pixels even vs odd shift), it is in sync in the middle, whereas on the right side of the frame the image out of sync in the opposite sense. A quick check in an image processing program shows that the even scan lines are roughly 8 pixels shorter than the odd scan lines based on looking at sharp vertical edges on the books/handrail. This might be a quirk of the capture hardware.
It is always going to depend a bit on which single frame you grab. But there appears to be something systematically odd about this image. Resampling the even frame lines to make them 8 pixels longer would improve the image quality at the edges considerably.
The test card is clean in this respect.
If Jan can capture the coefficient stream of an initial frame and save it as a conventional JPEG still I'd be interested to take a look and see what quantisation parameters they are using.
On a sunny day (Thu, 17 Jul 2008 15:07:47 -0700) it happened Joerg wrote in :
What edges??????
Well if you want reoltion of teh hannel you use a testcard, as I provided. This has nothing o do with 'resolution', it is just their picture composition.
wrong,
The images are taken withe the xine shapshot and I did not switch on de-interlace. Maybe I should have. As the hand moves fast horizonatlly you get this
--------- ----------
----------
Well I know that transmission has no effect on image quality other then causing blocks or display irregularities. Resolution, color, dynamic range ARE NOT AFFECTED. I have the impression you are still thinking way to much analog!!!!!!
If all else fails use test generator.
One thing that is fun here with sat is that the whole world is at you finger tips _without_ any picture degradation. For example here is China (Olympics?? They had also all Chinese space launches covered). And they know how to make a good picture too.
Let's record China CCTV9 from the eeePC:
grml: ~ # showr cctv9 showr: record mode show: selected program=cctv9 show: using hotbird 5 E show: current_angle=5 E same angle 5 E show: move_command=0 xine: no process killed dvbstream: no process killed xdipo: no process killed mplayer: no process killed show: moving dish executing /usr/local/bin/xdipo -c 1 -g '5 E' -f 11034 -p v -s 27500 -a 8192 || exit 1 xdipo: HAVE LOCK show: have LOCK show: no player specified, using xine as default player show: using cp /dev/stdin /mnt/hdd4/video/cctv9.hotbird.12:53_18_07_2008 show: using vpid apid executing /usr/local/bin/xdipo -c 1 -g '5 E' -f 11034 -p v -s 27500 -a 620 630
-o | cp /dev/stdin /mnt/hdd4/video/cctv9.hotbird.12:53_18_07_2008 xdipo: HAVE LOCK
See how easy it is from the command line? Here is a snapshot of the result: ftp://panteltje.com/pub/snapshots/cctv9_de_interlaced.png
On a sunny day (Thu, 17 Jul 2008 13:43:52 -0700) it happened Joerg wrote in :
Do they still make trinitrons? That was bad tube, and sony was even worse working around PAL patents in the past. Trinitron was bright, but for the rest junk.
On a sunny day (Fri, 18 Jul 2008 09:25:01 +0100) it happened Martin Brown wrote in :
Yes, all in fact.
This was a snapshot from a clip that moved, and possibly zoomed, I just looked for a place that had least motion for the snapshot, at least in the middle, I am sorry I cannot put HD movies in full resolution online, and very few people would have the 1920x1080 monitor to display these, not even mentioning the soft. Also cannot do it for copyright reasons. But I can assure you that you do not see those lines in a de-interlaced movie.
He's right. Zoom in close and take a look. I thought at first it was interlace related but now I am not so sure.
The hard vertical edges at both left and right of the picture have a scale offset that depends on horizontal position. Centre of the image is fine. It is almost as if the odd/even scan lines have a very slightly different natural length. Look for example at the books spines or the roof decoration at the extreme left or right. There is a 2-3 pixel systematic error (it may be due to panning but I don't think so).
The Astra HD watermarks are not affected so I think it is coming from the camera and not from the broadcast encoding. There is also some lateral chomatic abberation visible in the corners as purple/green fringing. That is very definitely optical in nature.
On a sunny day (Fri, 18 Jul 2008 12:16:55 +0100) it happened Martin Brown wrote in :
Yes, I should have taken snapshots with de-interlace 'on' in xine.
The whole issue with mpeg2 compresion and DTV is, that interlace sucks sucks, and did I mention sucks,
Also it is difficult to take a 'snapshot' from just one frame, as when there is motion it will be fuzzy. I frames, B frames et. I did mention to Joerg to look up mpeg2 encoding, my impression is he is still thinking analog, There are 4 sure artifacts in mpeg2 (and related encoding systems), but the idea is that those artifacts are not perceived by humans as such. It is the same as with mp3 encoding, clearly one can hear the diffference between .wav and .mp3, as you can see the difference between .dv (series of jpegs) or mpeg[x] (with interpolated stuff), and even the jpegs have artefacts, but with sufficient high bitrate these become less and less perceptible. That is also why you should look at the movie, not the individual frames.
The problems Joerg describes, as far as I understand them, seems to be related to either a bad decoder (rotten TV set?), and / or the broadcaster using too little bandwidth or old analog tapes or something. The picture problems do not seem to be transmission channel related, except for the picture breaking up or disappearing altogether. And even that I would not directly blame on multipath, but on a weak signal.
DTV can be really good, as we have it here now for many years (satellite), anda year or so ago all went digital here terrestrial. There is no more analog here, and I did put my old analog portable with the trash last year, as it ony received noise...
I can only say picture quality, choice of stations have improved, and antenna requirements have become simpler. The added advantage of digital reording, for example directly to DVD is a big plus too.
Yes. That is entirely consistent. There is a roughly 0.2% zoom between odd and even scanlines.
Actually it would be interesting to see what happens it you tweak your software so that it will only capture pure I-frames.
That is what I hoped. I tried rescaling the odd even scanlines to cure it and sure enough the image looked fine, but the Astra watermark was now pixel offset for odd/even scan lines.
Probably not. Got this used, much better for CAD than LCD.
IMHO a good CRT. Best thing since sliced bread for CAD work. LCD do not render so well when you have lines that do not run exactly horizontal or vertical.
--
Regards, Joerg
http://www.analogconsultants.com/
"gmail" domain blocked because of excessive spam.
Use another domain or send PM.
On a sunny day (Fri, 18 Jul 2008 08:12:35 -0700) it happened Joerg wrote in :
But Joerg, trinitron has vertial stripes.
so any tilted horiontal looks like this:
- - -
or, if not much tilted:
--
---
----
Pretty much same as LCD.
The vertical stripes did let more of the electron beams pass through then
a normal shadow mask with holes, but it also made the trinitron mechanically
more sensitive to vibration (tap on the side of the set).
As the guns were horizontally in line some correction electronics ws simpler.
What counts though is dot pitch.
So, in short, it is not possible to display a true diagonal line through
_any_ grid or mask with holes.
You would have to use an XY display :-)
ElectronDepot 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.