Flashing red signals.

[Bayum]

On a YouTube video shot at Church Fenton, when the signalling aspects towards Leeds turn red - they flash.

Any particular reason for this? What does it mean?

I've had a search on the forums and the internet, but the only results I can seem to find are flashing yellow aspects and flashing greens on the ECML.

 

[CallySleeper]

Care to link the video?

 

[Railsigns]

Most likely, the signal is fitted with LED inserts. When viewed on digital video, these often appear to flash because the camera's sample rate is so close to the frequency of the supply to the signal. The light looks steady to the naked eye.

 

[westcoaster]

Never heard of flashing reds, are you sure it's not just an illusion caused by the camera moving, same as it's hard to photo buses with certain destination blinds, that do not come out on camera.

 

[Nym]

I feel I am about to be corrected, but I beleive this is linked to returning a controlled signal to danger after it has been approached locked (or any kind of lock on a signal for that matter) by use of the emergency release or release lever.

I believe the flashing red is in place for the approach lockout time, I've known up to 120 or 180 seconds, when it will then return to a solid red aspect.

For the duration of the flashing red aspect, the approach locking remains in place and the route locked, so that if any train cannot stop in time and SPADs the signal under approach lock release, the route can continue to be taken. In more modern book wirings I have seen, this can cancel the emergency release.

Or

As above, apparent flashing could be due to the supply frequency.

 

[Darandio]

Never heard of flashing reds, are you sure it's not just an illusion caused by the camera moving, same as it's hard to photo buses with certain destination blinds, that do not come out on camera.

Also, on some videos it is often hard to distinguish between a yellow and a red.

 

[ralphchadkirk]

Nym: that's how it would appear on the panel or VDU, it doesn't actually flash on the ground.

 

[Bayum]

https://www.youtube.com/watch?v=CjKIvYs9JwI

4mins 50, 7mins 30, 10mins 40 and loads of others but I can't be bothered to find them.
--- old post above --- --- new post below ---

Also, on some videos it is often hard to distinguish between a yellow and a red.

It isn't a yellow - I've noticed it when trains have passed through the station as well and under the bridge.

 

[Nym]

Nym: that's how it would appear on the panel or VDU, it doesn't actually flash on the ground.

Ah, thanks.

Only ever seen it on a panel on Network Rail.

Out on the line on LU it's a solid red.
--- old post above --- --- new post below ---
Looking at the video, what frame rate where you sampling at?

It could be that you're sample lines up nicely to make that 'pulse rate' on the signal.

 

[Bayum]

Ah, thanks.

Only ever seen it on a panel on Network Rail.

Out on the line on LU it's a solid red.
--- old post above --- --- new post below ---
Looking at the video, what frame rate where you sampling at?

It could be that you're sample lines up nicely to make that 'pulse rate' on the signal.

Oh I've got no idea? How do I find that out?

 

[Nym]

Oh I've got no idea? How do I find that out?

Check the metadata (advanced properties in Windows) for the file.

I suspect it will be 25fps, 29.97fps or 30fps.

Probably one of the later two.

Meanwhile, anyone know what the signal supply frequency is for that area? I'm suspecting 33.3Hz but that would mean the lights would be pulsing around 10 times faster than they are.

 

[Bayum]

Check the metadata (advanced properties in Windows) for the file.

I suspect it will be 25fps, 29.97fps or 30fps.

Probably one of the later two.

I'm on a MacBook

 

[westcoaster]

The thing with signals is, they are sighted for drivers to see them, for example near London blackfriars some signals sit about 45 degrees from the post, so it could be the case you are filming the sides of the led's and not the front. This may cause the pulsing you see in the video's.

 

[Nym]

I'm on a MacBook

What did you film it with?
--- old post above --- --- new post below ---

The thing with signals is, they are sighted for drivers to see them, for example near London blackfriars some signals sit about 45 degrees from the post, so it could be the case you are filming the sides of the led's and not the front. This may cause the pulsing you see in the video's.

I wouldn't expect the viewing angle to affect it in such a regular way like that though, if the camera where moving in a regular manner it easily could.

 

[Bayum]

What did you film it with?
--- old post above --- --- new post below ---


I wouldn't expect the viewing angle to affect it in such a regular way like that though, if the camera where moving in a regular manner it easily could.

It's not my video - it's one from YouTube.

 

[DarloRich]

Very odd - A fault?

 

[Nym]

Ahhh, well I've done a bit of a dig on his profile, seems he's using a highish quality camcorder, so it could be anything.

What I suspect it is is the light signal going from in phase to antiphase with the sample frequncy of the sensor on the camera being used. This would create a 'beat frequency' that can be seen by going brighter and darker.

This frequency is determined by the modulus (magnitude) of the difference between the frequencies.

I suspect but again may be corrected, that this camera will be sampling close to the frequency of the signal head.

This could be that the signal is running at mains frequency (50Hz) and the camera is sampling at 50fps (nearly) with around a 0.3Hz difference due to calibration, but I don't think 50Hz would be anywhere near signalling equipment, unless it was internal within the signal head only.

25fps and 125Hz might be it...?

----------

Thinking more about it, when I was looking at high power LED drive circuits, they did drive at arbitrary frequencies to maximize performance, (hence only doing it on red due to a different p-n junction material for red compared with the other colors).

This would put it at any old frequency and would explain a beat frequency of around 0.3Hz if the LEDs are driven at, for example 49.7Hz?

With it being a gloomy day, the exposure time on the sensor would be slightly higher, as would possibly the sensitivity although it doesn't look much more than 200ISO, this would allow the capacitive nature of the LEDs being driven to show as 'dim' rather than off for the antiphase part of the signal observed.

 

[DarloRich]

@Nym - that is what i attributed to the first segment suggested (7m 30) as the signal doesn't go "black" but the other two segments are different.

Curious

 

[Nym]

@Nym - that is what i attributed to the first segment suggested (7m 30) as the signal doesn't go "black" but the other two segments are different.

Curious

What I find unusual though, and why I don't think the independent driver theory is correct is that they are exactly in time with each other. For this to be true, each LED array would need to be driven at the same frequency, and exactly in phase with each other, making it seem to me like it's the supply.

EDIT:
Just watched at 5:07 and they're not in time with each other there, so it could be the independent driver theory.

 

[Railsigns]

Just to add, on some videos I've seen, the apparent flashing of the LED signal lights has been far more pronounced, i.e. signal goes completely dark for a few seconds then appears lit for a few seconds, and so on. I've only ever seen this effect on signals with LED inserts, not proper LED signals.

 

[westcoaster]

@Nym - that is what i attributed to the first segment suggested (7m 30) as the signal doesn't go "black" but the other two segments are different.

Curious

I still think it's the way the signal is filmed/viewed in the link I'll post you have a double yellow led signal but the top yellow is much darker than the lower one. http://www.rogeramos.co.uk/signal.jpg

The signal head is not flat but curved, and unlike colour light bulbs there are no markings on the signal head. If you pass under a new led signal in bright sunlight you can not see what colour is shown apart from the upper ring/ single arc of LEDs for that colour,

 

[Class172]

It's not my video - it's one from YouTube.

You may be able to download the video off YouTube using a website or a program such as ClipGrab (runs on Windows/OSx/Linux), then view the metadata when you have downloaded it.

Ahhh, well I've done a bit of a dig on his profile, seems he's using a highish quality camcorder, so it could be anything.

What I suspect it is is the light signal going from in phase to antiphase with the sample frequncy of the sensor on the camera being used. This would create a 'beat frequency' that can be seen by going brighter and darker.

This frequency is determined by the modulus (magnitude) of the difference between the frequencies.

I suspect but again may be corrected, that this camera will be sampling close to the frequency of the signal head.

This could be that the signal is running at mains frequency (50Hz) and the camera is sampling at 50fps (nearly) with around a 0.3Hz difference due to calibration, but I don't think 50Hz would be anywhere near signalling equipment, unless it was internal within the signal head only.

25fps and 125Hz might be it...?

If I am not mistaken, isn't it derived from the theory of superposition?

 

[Bayum]

Ahhh, well I've done a bit of a dig on his profile, seems he's using a highish quality camcorder, so it could be anything.

What I suspect it is is the light signal going from in phase to antiphase with the sample frequncy of the sensor on the camera being used. This would create a 'beat frequency' that can be seen by going brighter and darker.

This frequency is determined by the modulus (magnitude) of the difference between the frequencies.

I suspect but again may be corrected, that this camera will be sampling close to the frequency of the signal head.

This could be that the signal is running at mains frequency (50Hz) and the camera is sampling at 50fps (nearly) with around a 0.3Hz difference due to calibration, but I don't think 50Hz would be anywhere near signalling equipment, unless it was internal within the signal head only.

25fps and 125Hz might be it...?

----------

Thinking more about it, when I was looking at high power LED drive circuits, they did drive at arbitrary frequencies to maximize performance, (hence only doing it on red due to a different p-n junction material for red compared with the other colors).

This would put it at any old frequency and would explain a beat frequency of around 0.3Hz if the LEDs are driven at, for example 49.7Hz?

With it being a gloomy day, the exposure time on the sensor would be slightly higher, as would possibly the sensitivity although it doesn't look much more than 200ISO, this would allow the capacitive nature of the LEDs being driven to show as 'dim' rather than off for the antiphase part of the signal observed.

Could you please translate into English?

I'm a teacher, not a physicist.

 

[DarloRich]

Could you please translate into English?

I'm a teacher, not a physicist.

not a physics teacher?

I am lost on more than the most basic explanation. Frequency rates. stuff. conflicts, camera to blame, etc. stuff. head hurts

 

[Nym]

If I am not mistaken, isn't it derived from the theory of superposition?

Yes it is.
--- old post above --- --- new post below ---

Could you please translate into English?

I'm a teacher, not a physicist.

not a physics teacher?

I am lost on more than the most basic explanation. Frequency rates. stuff. conflicts, camera to blame, etc. stuff. head hurts

What does Bayum teach then I wonder?

It's basically waves add together sometimes, and not other times, and this happens at a regular interval for constant waveforms (and some very special non-constant waveform patterns).

 

[Bayum]

Yes it is.
--- old post above --- --- new post below ---




What does Bayum teach then I wonder?

It's basically waves add together sometimes, and not other times, and this happens at a regular interval for constant waveforms (and some very special non-constant waveform patterns).

I'm a primary teacher.

Annoyingly the primary Curriculum has little physics or chemistry, but more focus on biology and physiology.

I remember waves and the like from triple science - but GCSE was such a long time ago.

 

[Class172]

Could you please translate into English?

I'm a teacher, not a physicist.

You may want to read the article I linked above.

Basically as the light and camera have different frequencies, the two, when represented as a wave 'superpose' (essentially they are added together by the vector components). See the picture below.

When you have two different frequencies, you will end up with points where the waves are in/out in phase, that results in a node (small) and antinode (large), which could correspond to the bright and dark red light respectively.

Hopefully this is a bit easier to understand, using theory from my AS physics course.

 

[westcoaster]

http://upload.wikimedia.org/wikipedia/commons/1/17/Two-aspect-signal.jpg
Viewed from the side you only see a strip of light at roughly 45 degrees, but viewed straight on at cab height will be 100% full.

 

[ryan125hst]

It must be something to so with the frame rates on camcorders and the frequency of the alternating current powering the signals, as suggested. However, it affects older signals with filament bulbs as well as LED's:

When I went in the Cathedrals Express at the end of last month I took my camcorder. You can clearly see the red signals dimming and brightening again very slowly. I can't give you a link on my phone and my computer's Internet connection has decided to stop working properly tonight, so I can't post a link, but if you search for Cathedrals Express Sir Nigel Gresley on 29th March 2014 and look for the channel "Ryan Evans" (linked to the Google Plus account I had to make to get a You Tube account- anyone know how I can change it to ryan125hst to be consistent? ), you will see I have uploaded a three part video. You should be able to see the same problem occurring a few minutes into the first part.

(I'll post a link for you tomorrow if I remember)

 

[Bayum]

You may want to read the article I linked above.

Basically as the light and camera have different frequencies, the two, when represented as a wave 'superpose' (essentially they are added together by the vector components). See the picture below.

When you have two different frequencies, you will end up with points where the waves are in/out in phase, that results in a node (small) and antinode (large), which could correspond to the bright and dark red light respectively.

Hopefully this is a bit easier to understand, using theory from my AS physics course.

So yet again, physics is being a general nuisance and complicating matters forevermore?
--- old post above --- --- new post below ---

http://upload.wikimedia.org/wikipedia/commons/1/17/Two-aspect-signal.jpg
Viewed from the side you only see a strip of light at roughly 45 degrees, but viewed straight on at cab height will be 100% full.

Why would viewing a signal from a vantage point other than the driver, have the affect that it is having upon the video at the moment?