Light works

On 01/09/14 10:14, jbeattie wrote:


Anyway, I almost bought a SP hub on sale over at Western ($131 USD
disc model) which I would build on a Velocity A23 or a DT450, which
is cheaper, but I want to try the A23. Go to Danscomp and get spokes
for maybe $30, so I'm at about $200 - $200, not too bad. So, I go
look at lights (at Universal, because they're within riding distance
and cheap): http://www.universalcycles.com/search.php?q=supernova
Crap, man . . . $240 USD for an 800 lumen light. The little brother
for $195 puts out 305 lumen with a five minute stand time (probably
good enough). The deal is my current battery light will do 350
lumens for 10 hours, and I already own it.

Lumens don't mean much - especially if the light comes from China.


More light shopping would be needed. And yes, I do blast my light on
my commute because it involves unlighted trail at least 3-4 nights a
week, and in pouring rain, 800-1400 lumens doesn't seem that bright.

Again, lumens mean little. They don't scale linearly either. Double
the lumens from a given light source and reflector does not give you
double the perceived brightness.

http://www.candlepowerforums.com/vb/...ightness-Index

--
JS

On Sun, 31 Aug 2014 13:50:27 -0400, Frank Krygowski
wrote:

I'm short on time tonite. Please excuse the lack of details and
coherence.

A good headlight beam is much more difficult to construct than most
people realize. I'd suggest reading this article:
http://www.peterwhitecycles.com/plight.asp
then taking a modern automobile out and observing its lighting of the
road, then parking the car facing a white vertical wall and observing
the pattern on the wall. Then thinking a lot. An ellipse is a very
poor approximation for that beam shape.

The problem with such an observation is that I have difficulties
comparing intensities. So, I did a bit of image enhancement and
produced:
http://www.11junk.com/jeffl/pics/bicycles/Front-Light-False-Color/index.html

This is:
http://www.11junk.com/jeffl/pics/bicycles/Front-Light-False-Color/slides/Schmidt%20Edelux%20II%20prototype.html
one of the original photos from:
http://www.peterwhitecycles.com/headlights.asp
I ran it through ImageJ editing software and produced a grey scale
image[1] :
http://www.11junk.com/jeffl/pics/bicycles/Front-Light-False-Color/slides/grey-scale.html
I then took this grey scale image and converted it to a variety of
pseudo color (false color) images. These are the best so far:
http://www.11junk.com/jeffl/pics/bicycles/Front-Light-False-Color/slides/pseudo-color-01.html
http://www.11junk.com/jeffl/pics/bicycles/Front-Light-False-Color/slides/16-colors.html
In theory, if I scale the image so that the maximum light intensity or
hot spot is a fixed value for all front lights under test, I can
compare the light distribution between lamps. Note that I'm not
really interested in which light is brightest. Just which one does a
better job of distributing the light where it's needed. If the light
creates an awful hotspot, it will be very obvious in false color.

There are other false color tables available:
http://www.11junk.com/jeffl/pics/bicycles/Front-Light-False-Color/slides/LUT.html

More later. Sleep beckons...



[1] ImageJ program
http://rsb.info.nih.gov/ij/

Convert to grey scale.
Image - Color - Channels Tool
Set it for Grayscale, which should also set it for only Channel 1.

Convert to false color:
Image - Lookup Tables
Pick an LUT (look up table). I suggest either 16 colors or BRGBCMYW.



--
Jeff Liebermann
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558

On Sun, 31 Aug 2014 20:02:46 -0400, Frank Krygowski
wrote:

On 8/31/2014 5:32 PM, Joerg wrote:

The rider is suspended mass. I would not mind a hub dynamo and would
tolerate the weight. But it's a lot of cost and effort, well north of
$100 for the part, then it has to be spoked up and of course the old
spokes won't fit. Then you have to get the cable up the fork and past
the stanchions. Ok, one can use the front caliper brake line for that,
maybe. But then you are still stuck with realistically 4-5W max which
won't be enough to supply my dream lights. And if I go through all this
effort to mod and experiment and build in a light system I want 8-10W up
front (with high-beam feature) and 2W in back.

I notice you consistently state your desires in terms of Watts, not in
terms of lumens, lux, candela, footcandles, visibility distances, beam
quality, etc. I think that's a sign of a potential problem. Watts are
_far_ from the most important measurement standard.


On the other hand, watts are very easy to measure :-)
--
Cheers,

John B.

On 9/1/2014 1:01 AM, Jeff Liebermann wrote:
On Sun, 31 Aug 2014 13:50:27 -0400, Frank Krygowski
wrote:

I'm short on time tonite. Please excuse the lack of details and
coherence.

A good headlight beam is much more difficult to construct than most
people realize. I'd suggest reading this article:
http://www.peterwhitecycles.com/plight.asp
then taking a modern automobile out and observing its lighting of the
road, then parking the car facing a white vertical wall and observing
the pattern on the wall. Then thinking a lot. An ellipse is a very
poor approximation for that beam shape.

The problem with such an observation is that I have difficulties
comparing intensities. So, I did a bit of image enhancement and
produced:
http://www.11junk.com/jeffl/pics/bicycles/Front-Light-False-Color/index.html

This is:
http://www.11junk.com/jeffl/pics/bicycles/Front-Light-False-Color/slides/Schmidt%20Edelux%20II%20prototype.html
one of the original photos from:
http://www.peterwhitecycles.com/headlights.asp
I ran it through ImageJ editing software and produced a grey scale
image[1] :
http://www.11junk.com/jeffl/pics/bicycles/Front-Light-False-Color/slides/grey-scale.html
I then took this grey scale image and converted it to a variety of
pseudo color (false color) images. These are the best so far:
http://www.11junk.com/jeffl/pics/bicycles/Front-Light-False-Color/slides/pseudo-color-01.html
http://www.11junk.com/jeffl/pics/bicycles/Front-Light-False-Color/slides/16-colors.html
In theory, if I scale the image so that the maximum light intensity or
hot spot is a fixed value for all front lights under test, I can
compare the light distribution between lamps. Note that I'm not
really interested in which light is brightest. Just which one does a
better job of distributing the light where it's needed. If the light
creates an awful hotspot, it will be very obvious in false color.

There are other false color tables available:
http://www.11junk.com/jeffl/pics/bicycles/Front-Light-False-Color/slides/LUT.html

More later. Sleep beckons...

This looks really interesting, at least to a technically oriented optics
freak. I wonder if others are doing this image processing?

If they're not, I think some of these guys might be interested: Peter
White (major US headlamp importer), Andreas Oehler (of Schmidt or SON),
Chris Juden (tech guy at Britain's CTC) and David Gordon Wilson (author
of _Bicycling Science_ etc.).

I've thought for a while that it would be nice to have a standard way of
comparing headlight beams. The best we have now is beam shots, but with
different cameras, operators, backgrounds, etc. the results can't be
easily compared. This might be a step forward.


--
- Frank Krygowski

On 9/1/2014 7:43 AM, John B. Slocomb wrote:
On Sun, 31 Aug 2014 20:02:46 -0400, Frank Krygowski wrote:

I notice you consistently state your desires in terms of Watts, not in
terms of lumens, lux, candela, footcandles, visibility distances, beam
quality, etc. I think that's a sign of a potential problem. Watts are
_far_ from the most important measurement standard.


On the other hand, watts are very easy to measure :-)

"I lost my keys somewhere in the dark over there."

"But why are you looking for them here, underneath this streetlight?"

"Because I can see better over here."


--
- Frank Krygowski

On 8/30/2014 9:49 AM, Joerg wrote:
SMS wrote:
On 8/28/2014 12:33 PM, Joerg wrote:
wrote:
not so.....night riders in the park are seen with midlevel systems
both on bike and helmet....good enough for 25 mph on pavement.


25mph on pavement ... yawn

What I meant is something that also holds up at 25mph on a bone-rattling
trail. Like it does on cars. No super-expensive boutiques stuff, regular
affordable gear, just like on cars or motorcycles.

I use such trails regularly also for commutes and pretty soon it'll be
dark when I get back. So right now I sometimes have three different
lighting systems front an back in order to be able to switch to the next
after one fails.

No market for it at the prices such devices would command.


It can be made at regular price levels. I now have a front light that
costs $65 (from Magnus). The manufacturer made a few mistakes in the
design but only one is left after I cured one and they cured another:
The battery was rattling inside, a cardinal mistake almost all
battery-operated light bicycle light manufacturers make. So I punched
myself four neoprene washer for compression and, voila, no more rattle,
no more light flicker. The original swivel holder is junk but they also
have a very sturdy non-swivel holder that is the best I've evern seen.
This leaves only the switch. If they fix that one day they'd have a good
light at a reasonable price.

I am still looking for something brighter though. 1000 lumens plus, and
that can't run on little batteries as this would take LEDs in the 5-10W
range.


The sad thing is that we're going the wrong way. Look at an old Belt
Beacon. Those were very rugged. You could bolt them on and they would
not go anywhere.

You could use rugged trailer lights but you need flash circuitry.


Or a motorcycle light. If I find nothing else I'll do that. Making a
4-6V to 12V converter or a constant current converter plus a flash
circuit is no big deal (I am an electronics engineer) but those lights
are big and get in the way on an MTB. The only mounting place you have
is the seat post and ideally nothing should be wider than 2-3".


The most rugged light I've seen are the Aervoe and VISIONX strobes.
These can withstand being run over by a truck. The deal is that you'd
have to fabricate a mounting bracket to attach them securely to a
bicycle in the proper orientation.

http://www.amazon.com/dp/B00MXL0H2U
http://www.amazon.com/dp/B005TFQHAS
http://www.amazon.com/dp/B003GAXXUU


Flares are a good idea but their light goes a bit in the wrong direction.

A bit, but many bicycle tail lights don't have sufficient light off to
the side. You can get some that do, but they are costly. The CatEye
TL-LD1100 was very good, but it's discontinued.

"sms" wrote in message
...
On 8/30/2014 9:49 AM, Joerg wrote:
SMS wrote:
On 8/28/2014 12:33 PM, Joerg wrote:
wrote:
not so.....night riders in the park are seen with midlevel systems
both on bike and helmet....good enough for 25 mph on pavement.


25mph on pavement ... yawn

What I meant is something that also holds up at 25mph on a
bone-rattling
trail. Like it does on cars. No super-expensive boutiques stuff,
regular
affordable gear, just like on cars or motorcycles.

I use such trails regularly also for commutes and pretty soon it'll be
dark when I get back. So right now I sometimes have three different
lighting systems front an back in order to be able to switch to the
next
after one fails.

No market for it at the prices such devices would command.


It can be made at regular price levels. I now have a front light that
costs $65 (from Magnus). The manufacturer made a few mistakes in the
design but only one is left after I cured one and they cured another:
The battery was rattling inside, a cardinal mistake almost all
battery-operated light bicycle light manufacturers make. So I punched
myself four neoprene washer for compression and, voila, no more rattle,
no more light flicker. The original swivel holder is junk but they also
have a very sturdy non-swivel holder that is the best I've evern seen.
This leaves only the switch. If they fix that one day they'd have a good
light at a reasonable price.

I am still looking for something brighter though. 1000 lumens plus, and
that can't run on little batteries as this would take LEDs in the 5-10W
range.


The sad thing is that we're going the wrong way. Look at an old Belt
Beacon. Those were very rugged. You could bolt them on and they would
not go anywhere.

You could use rugged trailer lights but you need flash circuitry.


Or a motorcycle light. If I find nothing else I'll do that. Making a
4-6V to 12V converter or a constant current converter plus a flash
circuit is no big deal (I am an electronics engineer) but those lights
are big and get in the way on an MTB. The only mounting place you have
is the seat post and ideally nothing should be wider than 2-3".


The most rugged light I've seen are the Aervoe and VISIONX strobes.
These can withstand being run over by a truck. The deal is that you'd
have to fabricate a mounting bracket to attach them securely to a
bicycle in the proper orientation.

http://www.amazon.com/dp/B00MXL0H2U
http://www.amazon.com/dp/B005TFQHAS
http://www.amazon.com/dp/B003GAXXUU


Flares are a good idea but their light goes a bit in the wrong direction.

A bit, but many bicycle tail lights don't have sufficient light off to the
side. You can get some that do, but they are costly. The CatEye TL-LD1100
was very good, but it's discontinued.

Today I saw a bicycle chained outside a shop I went to, with a strip rear
light about 6" long and mounted vertically behind the seat post.

IWHT that probably has abundant side illumination and the length (height) of
it would give some safety margin if you have a rucksack that might hang down
and obscure a typical size/shape rear light.

Didn't have time to try and get a look what brand, but there's probably
pictures on the web somewhere.

"Joerg" wrote in message
...
Ian Field wrote:


"Phil W Lee" wrote in message
...
Joerg considered Thu, 28 Aug 2014
12:33:45 -0700 the perfect time to write:

wrote:
not so.....night riders in the park are seen with midlevel systems
both on bike and helmet....good enough for 25 mph on pavement.


25mph on pavement ... yawn

What I meant is something that also holds up at 25mph on a
bone-rattling
trail. Like it does on cars. No super-expensive boutiques stuff,
regular
affordable gear, just like on cars or motorcycles.

Thing is, the energy budget is far higher on those, as they literally
throw the stuff away (via cooling systems).
On a bicycle, you either have to carry the energy with you in the form
of batteries, or generate it using your own muscle power, so the
energy available is much lower.

There's not much point in carrying batteries big enough for car
lighting on a bicycle in order to be safe at 25mph, if the weight of
the batteries slows you down to 15mph.

Lithium batteries are about as light as batteries get.


They pack a punch. A couple of 18650 Li-Ion cells are small and light
weight yet can easily power a 10W set of LEDs for over an hour. That
would rival the illumination of a serious truck head lamp.

http://www.amazon.com/3000mAh-Rechar.../dp/B004LHFU62

One does not need that much power all the time, for most of the ride
2-3W will be fine. But it sure helps to switch to full power high beam
during a gnarly section of trail for just a few minutes. Or when a car
driver is about to screw up right in front of you.


But unless you *REALLY* know your electronics - its advisable to buy
manufactured charging equipment!


Oh yeah, and once a Li-Ion has gone exotherm you may not be able to get
the fire put out. It is very tricky to safely charge them when they are
in series for a higher voltage.

Never hurts to repeat the warning - *NEVER* throw water on a lithium fire!
Smother it in sand if you can.

Its far better to use a single cell and a buck/boost converters - charge
equalising circuits for series chains can get real complex.

Jeff Liebermann wrote:
On Sun, 31 Aug 2014 13:50:27 -0400, Frank Krygowski
wrote:

I'm short on time tonite. Please excuse the lack of details and
coherence.

A good headlight beam is much more difficult to construct than most
people realize. I'd suggest reading this article:
http://www.peterwhitecycles.com/plight.asp
then taking a modern automobile out and observing its lighting of the
road, then parking the car facing a white vertical wall and observing
the pattern on the wall. Then thinking a lot. An ellipse is a very
poor approximation for that beam shape.

The problem with such an observation is that I have difficulties
comparing intensities. So, I did a bit of image enhancement and
produced:
http://www.11junk.com/jeffl/pics/bicycles/Front-Light-False-Color/index.html

This is:
http://www.11junk.com/jeffl/pics/bicycles/Front-Light-False-Color/slides/Schmidt%20Edelux%20II%20prototype.html
one of the original photos from:
http://www.peterwhitecycles.com/headlights.asp
I ran it through ImageJ editing software and produced a grey scale
image[1] :
http://www.11junk.com/jeffl/pics/bicycles/Front-Light-False-Color/slides/grey-scale.html
I then took this grey scale image and converted it to a variety of
pseudo color (false color) images. These are the best so far:
http://www.11junk.com/jeffl/pics/bicycles/Front-Light-False-Color/slides/pseudo-color-01.html
http://www.11junk.com/jeffl/pics/bicycles/Front-Light-False-Color/slides/16-colors.html
In theory, if I scale the image so that the maximum light intensity or
hot spot is a fixed value for all front lights under test, I can
compare the light distribution between lamps. Note that I'm not
really interested in which light is brightest. Just which one does a
better job of distributing the light where it's needed. If the light
creates an awful hotspot, it will be very obvious in false color.

There are other false color tables available:
http://www.11junk.com/jeffl/pics/bicycles/Front-Light-False-Color/slides/LUT.html

More later. Sleep beckons...



[1] ImageJ program
http://rsb.info.nih.gov/ij/

Convert to grey scale.
Image - Color - Channels Tool
Set it for Grayscale, which should also set it for only Channel 1.

Convert to false color:
Image - Lookup Tables
Pick an LUT (look up table). I suggest either 16 colors or BRGBCMYW.



The downside with false colour tables is that, while it is great for
contrast, nobody has a consistent internal sense of what order colours
should go in, so in your examples, the user is left thinking "is cyan
greater than magenta?" And other imponderables.

On 9/1/2014 1:36 PM, Phil W Lee wrote:


I wouldn't want to bet on the consistency of camera settings between
makes though, particularly if they have different sensors, so choosing
a readily available camera that can be obtained almost anywhere at
reasonable cost yet is advanced enough to be able to manually set all
the relevant parameters is probably vital.

If camera sensors varied, could things be standardized by having the
test include two photographic "grey cards" in the image?
http://en.wikipedia.org/wiki/Gray_card
Say one card 90% white, the other 18% as mentioned in the link, and
placed at some agreed-on distance?

Hmm. It's beginning to sound like we need a standards organization.

--
- Frank Krygowski