Greetings from a bemused Ppuker

[Flyfisher]

Interesting - particularly FF's point about a digital signal being (my own word) "undegradeable".

A digital transmission is not completely robust and reliable, but it tends to be all or nothing (though when a digital link does fail it's usually due to the underlying analogue signalling). This is often called the 'waterfall effect' because digital transmission quality generally remains 100% until, suddenly, it falls over the 'waterfall' and disappears completely.

Anyone with digital television will probably have noticed this effect; the picture quality is never noisy or grainy as it could be with poor reception of analogue transmissions, but it remains at full quality until the digital decoder receives such a poor signal that it can no longer correct the data errors and the picture breaks up into blocks before disappearing completely.

 

[Flyfisher]

The point made in one of those links is that digital signals are not all or nothing...

it's a matter of definition, I think. That's why it's helpful to think in terms of the various 'levels' of any transmission link. At the lowest level, all transmissions are analogue, be they voltage levels on a length of wire, bursts of light along an optical fibre, or even the ring of a bell on a bicycle. "Digital data" is something that is layered on top of an analogue transmission link.

Digital TV and radio uses the same basic analogue transmission technologies as 'analogue' radio, but instead of transmitting a facsimile of an electrical waveform (something that is impossible to do with 100% fidelity), it uses the analogue transmittion to 'carry' digital data. The advantage is that 100% accurate digital data can be transmitted even though the underlying analogue signal may be quite severely distorted. it's only when the underlying analogue signals become so badly distorted that the digital data can't be carried that the digital layer of the transmission breaks down - hence the concept of digital being 'all or nothing'.

But that scenario only applies if there are no additional levels of error-correction or a reverse-channel link that allows the receiver to request a retransmission. I would guess that the underlying communication process that enabled me to make this post had to cope, at some point in the chain, with the complete loss of a packet of data. But this was corrected by the various 'layers' in the communication protocol so that the end result was a 100% perfect data transmission.

The analogy would be sending a letter to someone by recorded delivery. The letter might get lost for all sorts of reasons, so you don't receive notification of it's delivery, so you send it again. When it is delivered you receive notification and know it has arrived. The original letter will never be found, but the later copy arrives and is 100% identical. The "analogue" transmission mechanism might have been degraded, but the "digital protocol" ensured a 100% reliable transmission. This is the sort of "all or nothing" thing that is happening on the internet every day - though somewhat faster!

 

[Penners]

Thanks to all of you gurus for the above explanations. I confess that I assumed a SCART connection was digital - I didn't realise that all of this codec stuff was going on. That explains it.

 

[AMc]

Associated packeted internet protocols that allow the recipient to request the data again are indeed really very clever.

Until I read that Bluejeans link I didn't really comprehend that those forms of data correction aren't practical for high definition TV signals because the amounts pf data being chucked down the wire mean you can't pause, buffer and restart in the same way you accept for sending and email or even iPlayer etc. AFAIK digital TV transmission 'fills in the gaps' until there are more gaps than information which I guess is what HDMI receivers need to do too.

As a 1080p signal is (broadly) 4 times as much data the amount in transit is quite spectacular - dwarfing your normal data network. Given that you have 3 streams of 3.4 Gbps + a clock signal all of which need to arrive together and no time to wait it seems amazing to me that it works at all.

I vaguely understood about the amounts of data involved in video having edited my miniDV home video on my elderly PC the amounts of data are significant even compressed.

I still won't be spending more on my cables than my components though

 

[AMc]

Thanks to all of you gurus for the above explanations. I confess that I assumed a SCART connection was digital - I didn't realise that all of this codec stuff was going on. That explains it.

I'm surprised it helped in the course of this little thread I've discovered I knew less about the subject than I thought and managed to confuse myself too

 

[Flyfisher]

Until I read that Bluejeans link I didn't really comprehend that those forms of data correction aren't practical for high definition TV signals because the amounts pf data being chucked down the wire mean you can't pause, buffer and restart in the same way you accept for sending and email or even iPlayer etc. AFAIK digital TV transmission 'fills in the gaps' until there are more gaps than information which I guess is what HDMI receivers need to do too.

Not implementing a 100% guaranteed error corrrecting link for digital video is partly a technology (or rather the cost of it), but mainly I suspect the because of the application requirements. A single data glitch in a a financial clearning system could cause all sorts of havoc, but a single glitch in a digital TV display will only be visible for 1/25th of a second (at most) before the next frame of display information arrives - so it's no big deal (unless it was happening all the time, of course).

As a 1080p signal is (broadly) 4 times as much data the amount in transit is quite spectacular - dwarfing your normal data network. Given that you have 3 streams of 3.4 Gbps + a clock signal all of which need to arrive together and no time to wait it seems amazing to me that it works at all.

I vaguely understood about the amounts of data involved in video having edited my miniDV home video on my elderly PC the amounts of data are significant even compressed.

It's not a like-for-like comparison though. The very high data rate on an HDMI cable is due to the fact that it's carrying a continuous uncompressed digital video signal to refresh the screen in real-time, whereas a video recorder will operate at a much lower data rate because it's only dealing with compressed video data.

Incidentally, the actual data rate of a now-pretty-much-obsolete miniDV camcorder is 25 Mbps (megabits/second) whereas the data rates of high-definition camcorders are actually quite a bit lower. The reason is that the HD video data is more highly compressed. This has advantages and disadvantages, but we'd better not get into that whole bag of worms.

 

[Penners]

I knew less about the subject than I thought and managed to confuse myself too

Welcome to my private club. I didn't know there were other members, too. :wink:

 

[Flyfisher]

The thing is, almost any technical subject (probably any subject actually) has layers and layers of complexity. Just as you (think you) understand it at one level, someone else can peel off another layer to reveal a whole new level of complexity.

Mrs FF was talking recently to the parents of a friend of FF junior and it transpired that the father works in electronics/IT. When I was finally introduced, on the basis that we'd have a common interest, it turned out that he was a silicon chip designer and spends his time worrying about the atomic physics of nano-scale devices. I could only barely understand the basics of some of the things he was working on because he was so much further down into the details of such things than I've ever been.

It made me realise that 'layers of complexity' and 'working models' are vital to our everyday lives. If we had to understand everything at the level of quantum mechanics, we'd never get anything done!

 

[Penners]

That's a fair point. I'd also add that, given the speed and scale of our technological advance, it must surely become increasingly the case that each of us will have ever more vertical specialisations.

My grandfather was a General Practitioner in the first half of the 20th century. In those days, a GP really was general - as well as the usual GP stuff he also did some dentistry, some optometry, even some minor surgery.

My father was an orthopaedic surgeon during the second half of the 20th century. He dealt with every part of musculo-skeletal system, but nothing else.

Mrs P has recently had a hip replacement, carried out by a specialist who does nothing but hips. A friend has had a knee replacement done by a surgeon who only does knees.

 

[Flyfisher]

Ah, but was Mrs P's surgeon a left-hand or right-hand hip specialist? :wink:

(Good example of ever-increasing specialisation though).

 

[Penners]

Ah, but was Mrs P's surgeon a left-hand or right-hand hip specialist?

Down here in Suffolk, the hip tends to be regarded as part of the leg mechanism, rather than the hand.

 

[Flyfisher]

That must be one of the advantages of seeing a specialist specialist.

 

[Gareth Hughes]

Ah, but was Mrs P's surgeon a left-hand or right-hand hip specialist?

Down here in Suffolk, the hip tends to be regarded as part of the leg mechanism, rather than the hand.

A simple guide for the non-specialist
http://www.youtube.com/watch?v=mVoPG9HtYF8
This is intended as general advice only, and no responsibility is accepted for the results of acting on it.

 

[Penners]

Nevertheless, Ezekial sounds like the man to see for any orthopaedic problems.

 

[FamilyWiggs]

Ah, but was Mrs P's surgeon a left-hand or right-hand hip specialist? :wink:

(Good example of ever-increasing specialisation though).

Ah, that reminds me of the old joke. What's the definition of a "holistic" orthopaedic surgeon? One who treats the whole bone.

Sub-specialisation in medicine is a real and growing problem. On the one hand it is great when you need an elective operation to see the chap (or chappess) who is an expert in the 3 inches of bone (or gut or artery) you have a problem with. The difficulty arises when you have an acute problem - the same chap on call in the middle of the night doesn't know so much about the other bones.

This is less of an issue for orthopaedics as you don't usually die from orthopaedic injuries, and can be transferred to a specialist the next day if necessary. It is a problem for general surgery, where time may be more critical; vascular and breast surgeons are increasingly stopping doing general (abdominal/gut) surgery on call, and emergency general surgery is being provided by fewer and fewer consultants. As sub-specialisation increases still further, upper and lower gut surgeons will become separate specialties further compounding the issue.

Why is this relevant? Well it isn't to speaker cables and digital data compression 8) , but it does have an impact upon the number of A&Es the NHS can safely support in the future. Quality of outcomes will improve, but access will become less local. Watch that debate get wrapped up in the rhetoric of the cuts over the coming years :roll:

 

[FamilyWiggs]

Ah, but was Mrs P's surgeon a left-hand or right-hand hip specialist? :wink:

(Good example of ever-increasing specialisation though).

Well he was probably a specialist in first or revision hips, and there are left and right specialists that exist - in foot surgery I understand. I was told of a hospital near me who recruited a new "upper limb" orthopaedic surgeon. When he arrived, they were somewhat troubled by the fact that he did shoulders, but not hands!

 

[worms]

Ah, that reminds me of the old joke...

and...

Does a doctor’s doctor, doctor the doctored doctor according to the doctored doctor’s doctrine? Or does the doctor, doctoring the doctored doctor, doctor the doctored doctor according to the doctoring doctor’s doctrine?

 

[FamilyWiggs]

:lol:

 

[Flyfisher]

 

[Penners]

there are left and right specialists that exist

Well, the chap who did Mrs P's left hip also did her right hip some four years ago. So he's obviously exceptionally versatile in his field. :wink:

He's an interesting chap - was once Regimental Medical Officer to the SAS, so one of the very few people on the planet with the courage to hack Mrs P about.