Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[burakkucat]

To find the answer, we need to go back to the year 2011. I had recently re-homed myself here and was communicating with asbokid. The Bald_Eagle1 had landed with a very long tale of woe about his erratic VDSL2 (ITU-T G.993.2), profile 8a, service and Walter was attempting to improve the Internet access for Ewhurstians, in Surrey.

I was given a Huawei HG612 and once it was unlocked, by carefully following the instructions provided by asbokid, there was a mass of data that could only be understood by creating what has become known as the "snapshot plots". They are, of course, the Bit Loading, SNR, QLN and Hlog plots (all versus the sub-carrier index (or tone number)). The suggestion was to make use of GNUplot . . .

At that time all xDSL services made use of frequency division duplex (FDD), splitting the available bandwidth into upstream (US) and downstream (DS) bands. Back then, my initial proof-of-concept code was created as a Bourne again shell (bash) script which was then given to Bald_Eagle1 to develop. The US and DS bands needed to be in different colours to aid the visualisation for each snapshot plot. And so to the question: "Why did the kuro neko choose the colours green and red for the US and DS bands, respectively?"

I think it is quite common that if a person needs to think of colours, to label or otherwise mark some entities, the triplet "red, green and blue" will come to mind in that order. Why that order? I have no idea. If we think about that short list of colours we can see that they are in order of decreasing wavelength and increasing frequency. Which is a good enough reason for that ordering. Another set of colours that easily comes to mind (at least for some of us) is "blue, orange, green, brown and slate (or grey)". Finally, there are the seven colours of the visible spectrum "red, orange, yellow, green, blue, indigo and violet".

It is well known that the kuro neko wears a red collar and the reason for that colour collar is identical to the reason for psychopomp1's choice of avatar. But that still does not answer the question "Why Green and Red?"

Let us think about a FDD xDSL circuit. It is just a pair of transceiver units at either end of a radio frequency transmission line. One transceiver is the DSLAM or MSAN, i.e. the xTU-C, and the other is the EU's CPE, i.e. the xTU-R. The labelling of the FDD bands, US and DS, is relative to the xTU-R. If it helps, the US band can be called the transmit band and the DS band can be called the receive band. Those with experience of the telephony transmission (core) network, repeater stations, four-wire circuits, etc, (be they either analogue or digital) may well refer to such as the Go circuit and the Return circuit. And so the initial question has been answered.

Perhaps the following two lines will clarify things --

US    Transmit    Go        Green
DS    Receive     Return    Red

[sevenlayermuddle]

Worth mentioning also that arbitrarily coloured leds have not been available for nearly as long as red/green.   Istr the guys who first achieved a blue led, even though red and green were already common, eventually got a nobel prize for their invention.

https://www.bbc.co.uk/news/science-environment-29518521

[burakkucat]

On the subject of LEDs, I believe that tri-colour devices are still significantly more expensive that bi-colour devices. Also, as far as I know, there are no red/blue or green/blue bi-colour LEDs available.

[sevenlayermuddle]

Since we’re off topic, I think one reason for cheap bi-colour leds is, they only need two wires.   With one polarity the red device will be forward biassed and hence conducting and emitting red light, whilst the green is reverse biassed and not conducting.   And vice versa if you reverse the drive polarity, the green led glows but not the red.   I did not know that bi-colour rarely offered blue, but I have no reason to disagree.

Personally, I think I find the colours red and green tend to be much more commanding of my attention, vs blue.   I do not know whether that is normal human psychology, but have no reason to doubt that my vision and colour perception are normal.

[4candles]

A very interesting dissertation b*cat.   

I think it is quite common that if a person needs to think of colours, to label or otherwise mark some entities, the triplet "red, green and blue" will come to mind in that order. Why that order? I have no idea.

As in TV, monitors and photography etc, it's RGB innit?

[sevenlayermuddle]

For purposes of print, by my limited understanding, and dating back to paint colours, RGB does not work.   Instead, you resort to “subtractive” primary colours known as Cyan Magenta Yellow.   

Or as they were known to my childhood contemporaries, with limited educational facilities and standards, Red Yellow and Blue.   My recollection is they were usually recited in that order (RYB), but I may be wrong.

[roseway]

With light sources like monitors the levels of the primary colours - red, green and blue - add together to create the colours which we see.

With printed material, what we see is reflected light, so when we shine white light on an image, the pigments absorb some parts of the spectrum and reflect others. So a yellow pigment (for example) absorbs blue light, and we see what remains in the reflected light, i.e. red and green which add up to make yellow.

When talking about paints, people often used to refer to red, yellow and blue as approximations for the true subtractive primaries, probably because these terms were more familiar than cyan and magenta.

[sevenlayermuddle]

these terms were more familiar than cyan and magenta.

Confession time... I have been using printers with individual cartridges for at least 15 years or so.  But even  now, when my printer pops up a message telling me a cartridge is empty, if that cartridge is cyan, I get a sinking feeling.   

For I know that when I go to the cupboard to get a spare cartridge to install, I will find myself unsure which one to reach for... the one with a red label, yellow label, or blue label.

[burakkucat]

. . . I think one reason for cheap bi-colour leds is, they only need two wires.   With one polarity the red device will be forward biassed and hence conducting and emitting red light, whilst the green is reverse biassed and not conducting.   And vice versa if you reverse the drive polarity, the green led glows but not the red

That got me thinking about the number of leads required for a tri-colour LED. Three or four? Both are feasible.

And that latest query prompted me to think about the typical electricity distribution delta-star transformers: input 11kV to the delta, primary, windings, output 415V from the star, secondary, windings with the mid-point of the star connected to earth.

[sevenlayermuddle]

I can imagine do not doubt that it might be possible, for a tri-colour LED to economise on number of pinout wires.

But from my hazy memories of days when I used to design circuitry, I suspect that the added complexity of drive circuits to control such a device might outweigh the cost savings of LED manufacture.   

Genuinely unsure here, my memories really are hazy, despite a degree certificate from a time in the distant past, that predated a career misspent writing software bugs.   And in any case, is circuit complexity still a problem, in these days of massive scale integration?

[burakkucat]

Yes, I have to agree. The drive circuitry for a three wire tri-colour LED would certainly be more complex than that required for a four wire tri-colour device.

[Alex Atkin UK]

Once you go RGB I think its more common to find SMD implementations (5050) or even more so the ones with built-in IC controllers (eg WS2811/WS2812B) as they are mass produced for lighting purposes and large LED displays where needing only three wires greatly reduces your PCB and controller complexity.

If you search eBay for RGB LED its pretty much exclusively those.

Amazon on the other hand have plenty of standard 4 pin RGB LEDs from as little as 5.2p each if you buy 100.

Back to the topic, isn't it actually Red for upstream and Green for downstream?  I'm sure that's how the bash script worked for me, same as dslstats.

[roseway]

DSLstats has red for downstream and green for upstream, which I've always understood to be the normal convention.

[burakkucat]

Back to the topic, isn't it actually Red for upstream and Green for downstream?  I'm sure that's how the bash script worked for me, same as dslstats.

As I wrote the original shell script, its subsequent developments and then the binary package (all as a proof of concept, for Bald_Eagle1 to ultimately use) I have to reply with an emphatic "No". 

[ejs]

Having never used any of the pre-existing software like RouterStats or DSLstats, I've been using green for downstream and red for upstream for the last 6 years. I've now swapped them round and re-drawn each daily graph.