Author Topic: Light testing - the results  (Read 40141 times)

Zipperhead

  • The cyclist formerly known as Big Helga
Light testing - the results
« on: November 30, 2008, 11:49:14 PM »
Part 1 - Introduction

This light test was Charlotte's idea. She put together all the hard work in organizing people and getting all the facilities together, and she and Liz worked very hard to get the garage set up.

All I did was put the camera on the tripod (Jonathan took the pictures). I just happened to have the sheet with the numbers on it, so I've put this bit together.

Any plaudits are entirely and completely due to Charlotte. Any mistakes are my fault, and I'm happy to take the blame for them.



The plan was to get a bunch of lights together, battery and dynamo powered, LED and incandescent and see how they performed. How hard could it be?

The lights were to be tested indoors, to allow the light output to be measured and to compare beam shapes, and outdoors to see what how they performed in "real world" conditions.

The battery powered lights all had freshly charged batteries. To ensure consistency with the dynamo lights, Charlotte mounted them all on her bike equipped with a hub dynamo. This bike was then put onto a set of rollers and pedalled up to a constant 15mph/25kph.

As you can see, this was a nerve wracking proposition



Outdoors, when the rollers & tyres were wet, was even scarier. I wouldn't have done it, and I briefed Jonathan to be ready to get out of the way if it went pear shaped.



For the indoors testing, the lights were shone along the length of the bunker onto a target on the end wall, then several things were done:

1. A light reading taken in the centre of the beam, and another at one metre away from the centre (for the incandescent lights this was taken horizontally, for the LED's either horizontally or vertically as their beam patterns tended to be more even). The light meter was set to display EV values, where an increase of 1 results from a doubling of the measured light. I then converted these to absolute values to allow the output of the lights to be easily compared.

2. A picture was taken of the target (obviously lit only by the light) with the camera set to auto. It was hoped that when these pictures were put together they would all be roughly the same brightness so that the beam patterns could be compared.

3. A picture was taken of the target with the camera set to manual exposure. For these, we hoped that when they were put together we would easily be able to compare the brightness of each light.

For the indoor pictures, the camera was set to 1000ASA.

Without pausing to even drink a cup of tea or scoff any Jaffa cakes, bikes were loaded up and a convoy set off (complete with rollers) for the local park where pausing only to scatter the doggers we set out to take more pictures.

The lights were all aimed along the path, which being nice and dark simulated the road, at a point which was about 20 metres away. A little way past that point we put a Brompton to have an example of a real world object that would reflect some light.

Again, two pictures were taken, one on auto to compare beam patterns, and the other on manual to compare brightness.

Although the lights could be easily aimed at the target indoors, outdoors it wasn't as simple. Because most of the lights didn't have a clearly defined centre spot (if they did, they would probably be useless as bicycle lights) they were aimed to try and maximise the amount of path that they lit up, but including the aiming point.

For the outdoor pictures the camera was set to 3200ASA (it's maximum) to reduce exposure time and also reduce the cameras exposure to the rain. That made the pictures a bit grainy, but not so much that it caused problems.

Cake never has a headache

Zipperhead

  • The cyclist formerly known as Big Helga
Re: Light testing - the results
« Reply #1 on: December 01, 2008, 12:20:42 AM »
Part 2 - Measurements





(If anybody wants the data as a CSV, let me know)
Cake never has a headache

Zipperhead

  • The cyclist formerly known as Big Helga
Re: Light testing - the results
« Reply #2 on: December 01, 2008, 12:28:08 AM »
Part 3 - Indoor pictures

Below are composites of the indoor pictures, first the auto exposures (to show beam shape), second the manual exposures (to compare brightness).

The name of the light and the exposure value are show on each individual photograph. Because they may be too small to read here, click on the picture to be taken to the Picasa album from where you can download a larger version.

Beam Shape



Brightness

Cake never has a headache

Zipperhead

  • The cyclist formerly known as Big Helga
Re: Light testing - the results
« Reply #3 on: December 01, 2008, 12:33:20 AM »
Part 4 - Outdoor Pictures

As with the indoor pictures, these are linked to the album where you can download the images.

Beam Shape



Brightness



Cake never has a headache

David Martin

  • Thats Dr Oi You thankyouverymuch
Re: Light testing - the results
« Reply #4 on: December 01, 2008, 12:45:40 AM »

For the indoors testing, the lights were shone along the length of the bunker onto a target on the end wall, then several things were done:

1. A light reading taken in the centre of the beam, and another at one metre away from the centre (for the incandescent lights this was taken horizontally, for the LED's either horizontally or vertically as their beam patterns tended to be more even).
How long is the bunker, so what angular deviation from center does 1m represent?

..d
"By creating we think. By living we learn" - Patrick Geddes

Zipperhead

  • The cyclist formerly known as Big Helga
Re: Light testing - the results
« Reply #5 on: December 01, 2008, 07:38:46 AM »
How long is the bunker, so what angular deviation from center does 1m represent?
..d

Good question, but I don't know, we'll have to wait for Charlotte to tell us. I can tell you that the outer ring on the target was 2m diameter.
Cake never has a headache

Re: Light testing - the results
« Reply #6 on: December 01, 2008, 07:41:01 AM »
Hi Guys,

Thanks for all your effort in making this happen.
It looks like a brilliant resource. :)

Sorry to be a bit "nit picky" but, is the Cyo the 60 Lux version, rather than the standard 40?
I thought, I read somewhere in here recently, that Charlotte bought the higher powered version?


This has certainly encouraged me to help the economy, by spending my way out of the gloom, both financially, and literally! ;)

Cheers, andy

Gandalf

  • Each snowflake in an avalanche pleads not guilty
Re: Light testing - the results
« Reply #7 on: December 01, 2008, 08:20:18 AM »
What an awesomely cracking effort, many thanks to everyone involved  :thumbsup:

Re: Light testing - the results
« Reply #8 on: December 01, 2008, 08:59:45 AM »
Great photo of Aunty C there  :thumbsup: ;D

clarion

  • Tyke
Re: Light testing - the results
« Reply #9 on: December 01, 2008, 09:03:36 AM »
Wow - good work, chaps!

That Betty is insane, but the one I would want, were I to be offroad in the dark...

Cyo is a good unit.  Not the brightest, and not the best beam shape, but I think I know what I'm saving my pennies for... :)
Quote from: Charlotte
Total hero!

Re: Light testing - the results
« Reply #10 on: December 01, 2008, 09:04:49 AM »
Very interesting. Thanks for all the hard work; sorry I couldn't be there.

I'm glad no one came off the rollers...I wouldn't want to be sued.

Re: Light testing - the results
« Reply #11 on: December 01, 2008, 09:05:04 AM »
remarkable.  Amazing work, everyone.

We should contact the different magazines see who wants to pay for them...  (I say 'we', I'm claiming absolutely no responsibility or credit :D)

ZH, if you email me the data sheet for the graph, I'll try and do a cost / brightness graph on one of my dull conference calls this morning.

Re: Light testing - the results
« Reply #12 on: December 01, 2008, 09:16:50 AM »

Superb. Thankyou Charlotte and Team.

chris

  • (aka chris)
Re: Light testing - the results
« Reply #13 on: December 01, 2008, 09:17:29 AM »
Excellent stuff. It is interesting to note how much things have moved on since the Cateye Opticube, which was pretty good in its day. Pity you didn't take a NEver Ready along to use a a reference!

The thing with modern lights is how quickly the technology is changing, and keeping at the cutting edge of said technology is an expensive business. Thankfully these days, even modestly priced lights give more than adequate performance.


Re: Light testing - the results
« Reply #14 on: December 01, 2008, 09:19:44 AM »
I assume no colour correction was done on those photos, and it's interesting to see how blue the Dinotte 200L is.  I don't recall it as looking particularly that colour, but the human eye is very good at compensating for such things.  The Bikehut is also a bit blue, but that one was a lot dimmer.  The Hope shows some blue-ness, maybe this reflects a previous generation of LED technology still being used on the slightly dimmer LED lights.
Actually, it is rocket science.
 

David Martin

  • Thats Dr Oi You thankyouverymuch
Re: Light testing - the results
« Reply #15 on: December 01, 2008, 09:25:19 AM »
I assume no colour correction was done on those photos, and it's interesting to see how blue the Dinotte 200L is.  I don't recall it as looking particularly that colour, but the human eye is very good at compensating for such things.  The Bikehut is also a bit blue, but that one was a lot dimmer.  The Hope is also a bit blue, maybe it reflects a previous generation of LED technology.

That is an interesting question. LEDs will have a very specific spectrum that approximates to white when seen by the human eye. Whether that matches the colour selectivity of the filters in a dSLR or the frequency response of a light meter (one would presume the light meter is panchromatic) has not been established.

It is a monumental test.. With a good write up it could make a magazine very happy..

If I am down in London, I'll bring the Ay Ups along to see how they fare. Having dual heads may be a bit challenging for the setup.

..d
"By creating we think. By living we learn" - Patrick Geddes

andygates

  • Peroxide Viking
Re: Light testing - the results
« Reply #16 on: December 01, 2008, 09:26:39 AM »
Nice work; great test.   :thumbsup:
It takes blood and guts to be this cool but I'm still just a cliché.
OpenStreetMap UK & IRL Streetmap & Topo: ravenfamily.org/andyg/maps updates weekly.

Charlotte

  • Dissolute libertine
  • Coffee snob and pencil fetishist
    • charlottebarnes.co.uk
Re: Light testing - the results
« Reply #17 on: December 01, 2008, 09:27:34 AM »
Great write up Graham - I know how hard you worked on this yesterday - we all owe you a big thanks for that  :thumbsup:

Okay - the distance from the target was about five metres.  Basically, it was the length of the SEEKRIT BUNKER, with the target up on one wall and the rollers just inside the door at the other wall.  I'll measure it tonight so that if anyone wants to do angular deviation calculations, they can.

The IQ Cyo was indeed the 60 Lux version without the reflector, powered (as all the dynamo lights were) from a SON hub.

I didn't envisage this to be a commercial exercise, Mike.  Besides; we've put this all in the public domain now, which would make the material much less attractive if we wanted to sell it*

The real surprise for me is that the Smart rechargeable lights were so damned bright in the centre of the beam.  They're the ones that SJS used to knock out for £35 with two incandescent lamps and a lead acid battery.  I only put them in the test for a reference point, but it would appear that battery powered incan lights are pretty damned good still.  I wish we'd have bothered to get some Lumicycle halogens (or HID) in the lineup because based on what we've seen here, they'd still be the kings of sheer output.

*Incidentally; if any magazine editors are reading this and do want to reproduce it, speak to Zipperhead.  It's all his work and he holds the copyright.
Malo periculosam, libertatem quam quietam servitutem.

Re: Light testing - the results
« Reply #18 on: December 01, 2008, 09:35:31 AM »
I had to register to say well done! It's a really useful reference, thanks for all your hard work.

It would be great to hear more impressions from the people there. FWIW, I use an IQ Cyo, which I find more than adequate for commuting, but I wish it did the Lupine Betty thing. Can you get a tan off it too?

Zipperhead

  • The cyclist formerly known as Big Helga
Re: Light testing - the results
« Reply #19 on: December 01, 2008, 09:43:57 AM »
I assume no colour correction was done on those photos, and it's interesting to see how blue the Dinotte 200L is.  I don't recall it as looking particularly that colour, but the human eye is very good at compensating for such things.  The Bikehut is also a bit blue, but that one was a lot dimmer.  The Hope shows some blue-ness, maybe this reflects a previous generation of LED technology still being used on the slightly dimmer LED lights.

One thing that I missed was that I had left the cameras colour temperature on auto (because it makes such a good job of all my other pictures that I'm not going to screw with it).

It would have been interesting to set it to a fixed value to get a better comparison of colour - but I don't know if that would make much difference in the real world.
Cake never has a headache

clarion

  • Tyke
Re: Light testing - the results
« Reply #20 on: December 01, 2008, 09:46:20 AM »
In htose sort of light levels, the difference in colour temperature would probably be barely noticable.  Certainly, when your ambient could be anywhere between moonlight, street noise from illuminated signs, and sodiums, a few degrees would be neither here nor there - your eyes adapt quickly.
Quote from: Charlotte
Total hero!

LEE

Re: Light testing - the results
« Reply #21 on: December 01, 2008, 09:55:06 AM »
Really excellent and thorough piece of work.

 

chillmoister

  • King of Compton
Re: Light testing - the results
« Reply #22 on: December 01, 2008, 10:00:28 AM »
toptastic .....huge thanks to all involved !!   
appearing in a tea room near you

Woofage

  • Ain't no hooves on my bike.
Re: Light testing - the results
« Reply #23 on: December 01, 2008, 10:03:14 AM »
Let me add my congratulations and thanks on a superb study :thumbsup: .

I'm not sure if this has been noted but looking at the method it seems to be easily repeatable, perhaps even at a different location with different people. As Chris noted above, lighting technology has moved on rapidly in recent years so I can imagine there will be a desire to carry out a new set of tests in a few year's time with results comparable with this one.
Pen Pusher

Jaded

  • The Codfather
  • Formerly known as Jaded
Re: Light testing - the results
« Reply #24 on: December 01, 2008, 10:10:20 AM »


One thing that I missed was that I had left the cameras colour temperature on auto (because it makes such a good job of all my other pictures that I'm not going to screw with it).

It would have been interesting to set it to a fixed value to get a better comparison of colour - but I don't know if that would make much difference in the real world.

If you took RAW that wouldn't matter.
“The Silent have spoken.”