Comparing lights
- Thread starter menacer
- Start date
big-palooka
Member
OK - will be looking out for it!
PC
PC
big-palooka
Member
Potholer - sent you PM - can you read & reply please asap
Paul C
Paul C
Jaz said:I've just ordered the hand torch mentioned above but I got it from a British supplier who keeps them in stock, www.lumenjunkies.co.uk I ordered it yesterday and it arrived this morning. Dealextreme were cheaper but appear to be a non-stock holding supply portal and the products get despatched from China. Depending on the exchange rate, you could save about ?5 by ordering from them, I opted to order from a UK supplier. I also ordered their ?15 headlightAfterall, anything has to be better than the Gellert I've been using and I need something to tied me over until we get the results for this comparison.
My first impression of the hand torch is the build quality is surprisingly good and altough the stated 900 lumens seems somewhat hopefull, it is certainly very bright. It's a useful torch for all sorts of applications, I'll using it at work, for cycling as well as to compliment caving.
...
I ordered one of these hand torches and here are couple photos taken on my Lumix DMC FT1 using only the torch for illumination. The torch is about 4 feet from the subject, the camera was hand-held so you can see a bit of noise because I guess the ASA ramped up to about 800.
big-palooka
Member
WOW!
New duo spot conversion arrived today - lit up my cellar like it had a generator powering a bank of truck lights.
Look forward to trying it out next week in some real caves!
Try and get you a bunch of orders from NZ if I can.
Thanks for super speedy delivery.
Cheers
Paul C
New duo spot conversion arrived today - lit up my cellar like it had a generator powering a bank of truck lights.
Look forward to trying it out next week in some real caves!
Try and get you a bunch of orders from NZ if I can.
Thanks for super speedy delivery.
Cheers
Paul C
Maggot said:
Following the discussion thread link from the above there are some interesting comments on the response of digital cameras to different lights. This might explain why some results do not seem to follow what the eye actually sees. There is a suggestion about using film to achieve a better comparison. I have absolutely no expertise in this field but it appears to me that unless the test unit responds the same to every light then a meaningfull comparison, using a digital camera, may well be flawed.
As I say I have no expertise and would like to see comments from others more knowledgable.
Jopo
Following this to its logical conclusion would suggest that there is no way of comparing lights at all. Even different films have different responses to colours. So do human eyes - which will also see something very different to a camera, having a much wider dynamic range, and the benefit of a huge instant post-processing facility attached to the other end of the optic nerve.
On its own, even the humble petzel tikka is adequate for caving, but is percieved to be a bit dim in the presence of, say, a Sten. The eye adapts more than any camera can, film or digital.
IMHO it's not how many raw lumens the light chucks out, it's how they are focused - a nice wide spill is just as important as the spot beam, especially when walking over rough surfaces. YMMV; the best way to find out if you like a light is to have a go with it!
On its own, even the humble petzel tikka is adequate for caving, but is percieved to be a bit dim in the presence of, say, a Sten. The eye adapts more than any camera can, film or digital.
IMHO it's not how many raw lumens the light chucks out, it's how they are focused - a nice wide spill is just as important as the spot beam, especially when walking over rough surfaces. YMMV; the best way to find out if you like a light is to have a go with it!
It depends what the desired end result is:Jopo said:
Measurement of brightness via an eyedropper tool in photoshop, etc (which would need a image with linear (or at least known) relation between actual brightness and image levels)
Comparison of beam patterns, which might be best done with 'normalised' images taken with aperture/shutter speed settings chosen to give the same peak image brightness.
Working out what lights actually do in practice, which is highly subjective, and depends on what other lights are around, how much a user has achieved optimal eye sensitivity for the level they're using, the extreme nonlinear response of the eye even ignoring dark adaptation, etc.
Using a light with about a 100:1 variation between highest and lowest levels, the difference to the eye seems nothing like 100:1 even on instant comparisons, let alone after allowing the eye to adapt properly to the lower level.
van the man
New member
Ok boys and girls
Lets get this back to the basics, I have built my own light built around a Cree. Total cost ?30 ish, its waterproof for my abilities (occasional duck). Did Prospect to Dow cave over the weekend and after 5 hours underground it was still going strong on the same rechargables. It works well in most of the restrictive caves we get in N Yorks but a bit weak in big passages. That said I love it and still blind most of my Duo using fellow cavers. Its not about Lumens its about useability!
Lets get this back to the basics, I have built my own light built around a Cree. Total cost ?30 ish, its waterproof for my abilities (occasional duck). Did Prospect to Dow cave over the weekend and after 5 hours underground it was still going strong on the same rechargables. It works well in most of the restrictive caves we get in N Yorks but a bit weak in big passages. That said I love it and still blind most of my Duo using fellow cavers. Its not about Lumens its about useability!
Maggot said:
Would a test using the same camera (with the same aperture and exposure) on the same film be more accurate? Given that a particular colour film may also respond differently to various lamps, would using black and white film be better? Or a digital set to B/W?
I was struck between the comparison between the KSE and the Myo, the only two lights in the test that I have. If anything - to my eye - the Myo gave slightly better results than the KSE in the test yet when comparing the two side by side (fresh charge/fresh batteries) the Myo is nowhere near as good. As I don't have any of the other lights to directly compare I have no idea.
Jopo
Bob Mehew
Well-known member
Jopo said:
The simple answer is the topic is most complex. There is a British Standard for miner's lamps, BS EN 62013-2:2006 which simply talks about using a calibrated photometric cell.
The key underlying challenge is the complexity of the spectrum of colours which will be emitted by an LED (high) and the ability of a digital camera to capture that spectrum remembering it only uses red, green and blue sensors to record the image (but I suspect it will be not that bad excepting the colour range problem of sRGB, AdobeRGB etc - I won't go further).
My guess is that whilst it would not be precise, one might get something akin to reliable measurements using a digital camera in colour (but with a fixed white balance) taking a photo of a colour card target (a fancy printed card containing precisely colours including white and various greys) ensuring that the set of photos covering all the lamps do not contain any which are over exposed. Plus one also has to have each light adjusted to its optimum performance in terms of LED temperature and supply voltage as well as measuring the intensity across the half sphere into which the light is thrown / cast. There is then a potential problem of how a digital camera manipulates the sensor which will apply to raw file information though probably not as much as to jpg files.
Colour film would not be an improvement since it works on using several light sensitive dyes to record the image, though again it might not be that bad. I am not sure about black and white film but it also has a basic problem of having a small linear range of recording exposure against intensity of illumination. But as some one has said, you would then have to read the information from the film using another instrument.
A simpler approach would be to borrow a photometric cell. I doubt if an old light meter would be of much use since the range of light intensities are going to be mostly lower than those used in photography and so the meter reading would be very low. Even with a photometric cell, once would still have to look into its spectrum response which probably will not be that simple. This topic would be simple stuff to an old fashioned astronomer who uses a visual telescope, but I don't know of one to ask.
footleg
New member
I gather from the discussions that the argument here is that digital cameras are not the best way to measure the brightness of LED lights for comparison for caving (i.e. To compare them in terms of how they appear to the human eye).
But I would argue it is totally appropriate for measuring the brightness for digital photography purposes. So it you want to know what light is brightest for photography then this data is very useful.
If what you want to know if how bright the lights are for caving then the digital camera might not be the best measure, but what is important is not the absolute best measure, just a good enough measurement to rank them. If a light has a particularly strong colour tint then it might not be so great for caving anyway. So I would think the relative exposure times needed to produce a good exposure for a photograph would be a good enough measure to tell me the apparent relative brightness of the lights?
But I would argue it is totally appropriate for measuring the brightness for digital photography purposes. So it you want to know what light is brightest for photography then this data is very useful.
If what you want to know if how bright the lights are for caving then the digital camera might not be the best measure, but what is important is not the absolute best measure, just a good enough measurement to rank them. If a light has a particularly strong colour tint then it might not be so great for caving anyway. So I would think the relative exposure times needed to produce a good exposure for a photograph would be a good enough measure to tell me the apparent relative brightness of the lights?
Bob Mehew
Well-known member
I am not saying digital cameras are not the best way, but they do have a large number of factors which have to be taken into account. The simpler approach is photometric cell but even that has a fair number of factors to take into account.
The problem with the human eye is that it also has a bias in the way it sees various colours. But I accept that this is less significant for today's LEDs. Another variant is the angular range which the eye can see. The extreme example is tunnel vision that some people have such that they can hardly see anything outside a very small cone of vision directly in front of them. But I suspect that how wide spread the light cast is significant in how people feel a light is good or bad. So frankly I sense that measuring is a better approach that the Mark 1 eye ball. And even then, one would have to work out what is an acceptable spread of the light cast in order to make a judgement for ranking, or else find some means of presenting the information so people can make up their own minds.
In terms of using a light for photography, then colour bias is highly significant. (Just take a picture using a mixture of flash light and tourist cave lighting to see how different this can be.) And there is also ensuring one gets even lighting over a reasonable portion of the image. I think the only way to achieve good results is by careful measurement of both the amount and spread of illumination and also the colour cast. Which is where using a colour card as a target comes in. My small card (15 by 5 cm) cost ?20, an A4 sized one comes in at around ?75. Not things to take underground! And then you will need to use a telephoto lens to fill the image with the card, so the work gets laborious.
The work can be done above ground and I think it might make a good final year project for a student in a suitable degree subject or indeed for someone interested in the topic who has access to a good camera and telephoto lens, a large room and the lights plus other odds and ends. I can offer some help but as I am already up to my eyes in trying to sort out dynamic rope testing using a high speed load cell to monitor the influence of the knot, I am not prepared to attempt the work myself.
The problem with the human eye is that it also has a bias in the way it sees various colours. But I accept that this is less significant for today's LEDs. Another variant is the angular range which the eye can see. The extreme example is tunnel vision that some people have such that they can hardly see anything outside a very small cone of vision directly in front of them. But I suspect that how wide spread the light cast is significant in how people feel a light is good or bad. So frankly I sense that measuring is a better approach that the Mark 1 eye ball. And even then, one would have to work out what is an acceptable spread of the light cast in order to make a judgement for ranking, or else find some means of presenting the information so people can make up their own minds.
In terms of using a light for photography, then colour bias is highly significant. (Just take a picture using a mixture of flash light and tourist cave lighting to see how different this can be.) And there is also ensuring one gets even lighting over a reasonable portion of the image. I think the only way to achieve good results is by careful measurement of both the amount and spread of illumination and also the colour cast. Which is where using a colour card as a target comes in. My small card (15 by 5 cm) cost ?20, an A4 sized one comes in at around ?75. Not things to take underground! And then you will need to use a telephoto lens to fill the image with the card, so the work gets laborious.
The work can be done above ground and I think it might make a good final year project for a student in a suitable degree subject or indeed for someone interested in the topic who has access to a good camera and telephoto lens, a large room and the lights plus other odds and ends. I can offer some help but as I am already up to my eyes in trying to sort out dynamic rope testing using a high speed load cell to monitor the influence of the knot, I am not prepared to attempt the work myself.
linuxfanatik
New member
I have a lamp on my trusty bike 'Kielder classic' by Emmelle in red - anyway its a cat lamp and has a beam that makes a lighthouse look dim - it was only forty quid but it took me a while to source rechargble batteries that can keep up with the power drain without the resource of a thermo-nclear facility for backup power! linuxfanatik
