A Impromptu/Cheap Luxmeter Alternative: Measuring Lux With a Canon Camera and CHDK

[csshih]

[FONT=&quot]If someone could provide me actual Lux measurements for the [/FONT]minimag 3AA, fenix E1 (not E01), lumapower f1, that would be helpful.
[FONT=&quot]
Hey All![/FONT]
[FONT=&quot]I was browsing the web yesterday, and I happened to stumble upon a Canon Camera Firmware addon![/FONT]

[FONT=&quot]It's called CHDK, [/FONT]http://chdk.wikia.com/wiki/CHDK

[FONT=&quot]A picture of the startup screen (disable-able)[/FONT]
[FONT=&quot]

[/FONT]

[FONT=&quot]Now.. What does this CHDK have to do with Flashlights, you ask?[/FONT]

[FONT=&quot]Lets see here…[/FONT]

[FONT=&quot]Main menu…[/FONT]
[FONT=&quot]

[/FONT]
[FONT=&quot]--> OSD parameters..[/FONT]
[FONT=&quot]

[/FONT]
--> [FONT=&quot]Miscellaneous Values[/FONT]


[FONT=&quot]AND what do we see here?! Show scene luminance in cd/m2? Aka lux!![/FONT]
[FONT=&quot]

[/FONT]

[FONT=&quot]Now, I can't let this opportunity pass up![/FONT]


[FONT=&quot]This little short post here is to see if this camera's lux o' meter is accurate at all![/FONT]
[FONT=&quot]I won't be doing any of them fancy measurements; this is just an accuracy test! [/FONT]
[FONT=&quot]Distance from camera, zoom, etc, will not matter, but will all be the same![/FONT]

[FONT=&quot]From ersanada's excellent review here: (https://www.candlepowerforums.com/threads/185431), we get [/FONT]
[FONT=&quot]
Romisen Cree Q5 RC-N3 RCR123 - 3950 lux @ 1 meter[/FONT]
[FONT=&quot]Romisen Cree Q5 RC-N3 2AA Nimh - 2780 lux @ 1 meter[/FONT]

[FONT=&quot]Ok, I won't be doing the 1 meter thing. I'm just trying to get as close to the measurement as possible, then swap to 2AA mode without moving anything.[/FONT]
[FONT=&quot]If the values match up, then the camera is accurate.[/FONT]

[FONT=&quot]Here's a picture of my setup with 123as, (fully charged protected ultrafires)[/FONT]
[FONT=&quot]

[/FONT]


[FONT=&quot]And here a zoom in of the screen[/FONT]
[FONT=&quot]

[/FONT]
[FONT=&quot]
[/FONT]
[FONT=&quot]Here's a picture of my setup with 2AAs, (fully charged eneloops)[/FONT]
[FONT=&quot]

[/FONT]
[FONT=&quot]
[/FONT]
[FONT=&quot]And here's a zoom in of the screen[/FONT]
[FONT=&quot]

[/FONT]


[FONT=&quot]Summary:[/FONT]
[FONT=&quot]
Ersanada's Luxmeter:
[/FONT][FONT=&quot]rc-n3 - Cr123a: 3950[/FONT]
[FONT=&quot]rc-n3 - 2XAA: 2780

[/FONT][FONT=&quot]Camera:[/FONT]
[FONT=&quot]rc-n3 - Cr123a: 3945[/FONT]
[FONT=&quot]rc-n3 - 2XAA: 2789[/FONT]
[FONT=&quot]
[/FONT]Nitecore D20 -
lightreview:3570 lux
camera: 3414 lux

Minimag 3AA
camera: 1646 lux

Fenix E1
camera: 169 lux

Lumapower F1:
camera: 545 lux

[FONT=&quot]Verdict: With a "calibration", the results are remarkably accurate compared to Ersanada's measurements.[/FONT]

[FONT=&quot]So.. what do you guys think?[/FONT]

anyone need help? shoot a PM or post in thread!
[FONT=&quot]
[/FONT]

 

[MWClint]

Re: A cheap Luxmeter Alternative: Measuring Lux With a Canon Camera and CHDK

http://chdk.wikia.com/wiki/CHDK

nice..i was in the market for a new camera..looks like i'll have to
try a cannon.

 

[csshih]

Re: A cheap Luxmeter Alternative: Measuring Lux With a Canon Camera and CHDK

Thanks for the link.
I was too lazy to post it up.


by the way, I'm using a SD1000.
don't base your camera buying decision on my test, though canon does make decent cameras.
chdk improves usability for advanced users quite a bit.. also for basic users..
there is now a battery meter!!

hmm.. no one else interested in this concept?

 

[crocodilo]

Re: A cheap Luxmeter Alternative: Measuring Lux With a Canon Camera and CHDK

Definetely a great way for comparing brightness between two flashlights, or a rough way to manually do simple runtime graphs (10 minute reading interval, for instance)!

EDIT 23MAR09: Already tried it, to ascertain the difference between high and low modes on some of my lights: fantastic!

 

[csshih]

Re: A cheap Luxmeter Alternative: Measuring Lux With a Canon Camera and CHDK

Added nitecore D20, minimag 3AA, fenix E1 (not E01), lumapower f1

 

[electrothump]

Re: A cheap Luxmeter Alternative: Measuring Lux With a Canon Camera and CHDK

Thanks for the link.
I was too lazy to post it up.


by the way, I'm using a SD1000.
don't base your camera buying decision on my test, though canon does make decent cameras.
chdk improves usability for advanced users quite a bit.. also for basic users..
there is now a battery meter!!

hmm.. no one else interested in this concept?

Heck yeah I'm interested. I have the sd 1000 also. Can't wait to try it.

DN

 

[Any Cal.]

Very nice. I don't have much to add or help with. I ran CDHK in my Canon, but took it off. Maybe I should reinstall it now that I know what incredibly important things it can do.:nana:

 

[darkzero]

Awsome, thanks for posting what the menu looks like. I came across CHDK on a machining forum & downloaded it for my SD950IS but never got around to trying it. This motivates me to try it now! I've been very happy with Canon cameras although there's a few things they lack. I'm on my third Canon now. :twothumbs

 

[csshih]

lux measurements of minimag 3AA, fenix E1, lux lumapower F1 anyone?

unless someone wants to donate a light.. I'm trying to see just how accurate this thing is. :nana:

 

[geek4christ]

That is a really cool find! Thanks for posting.

I haven't messed around with CHDK much, but my wife has a 590 that I will now be using to do some rudimentary lux measurements. :twothumbs:

 

[Viper715]

I have two Cannons the A530 and the SX100. After reading this I installed it on both. It works great and seems to be fairly accurate. I have tested it against known performers and it is spot on within a few lux. I'll post some shots later. Great find.:twothumbs

 

[MrGman]

The light meters in any good camera should be calibrated as that is what determines the exposure based on distance, f stop and ISO settings. I am quite certain that the calibration of the meter in the camera would be set at the factory quite well and that the calibration of the camera circuitry around that meter to get accurate and consistent exposures of the images would work. So it should not be surprising that a camera that actually will give you lux readings should read the same lux numbers as a lightmeter for the same light source at the same distance.

 

[polkiuj]

Thanks a lot dude!!

I've tried it and it is wayy cool! (A530)

One thing I wanna add is that you should use it in manual mode, lock all variable settings like ISO, exp., etc.

Try to use SPOT exposure instead of the default matrix.

Also, I've noticed that you have your compression quality set to superfine. (If you want) Change it to normal and you will have extremely small file size with (almost, not to my eyes or even widely zoomed in) no difference in picture quality. =D You'll have almost 4000++ shots on 1 card (although I dunno why anyone would need so much xD)




MrG, unfortunately this is not the case. My camera seems to be much less sensitive to light compared to a lightmeter. I'm only getting about 400+ lux @ 1m on my LD10 (suppose to be 2700 lux or so). You'll have to be much closer to get the same readings. I usually go further until I get 270 lux and work from there. Maybe I'll try going closer today. =D

 

[656nm]

MrG, unfortunately this is not the case. My camera seems to be much less sensitive to light compared to a lightmeter. I'm only getting about 400+ lux @ 1m on my LD10 (suppose to be 2700 lux or so). You'll have to be much closer to get the same readings. I usually go further until I get 270 lux and work from there. Maybe I'll try going closer today. =D

Consider reflected vs direct illumination. A light meter is measuring the light directly illuminating the sensor and at a pre-defined distance standardised at 1m, while a camera would be measuring reflected light with several variables at play. That reflected light will only be a small percentage of the total light falling on that point, when you take factors such as scatter due to surface roughness into consideration. You will need to come up with a multiplier to use to come up with a true lux number. That multiplier will depend on the surface which you have the light pointed at, as well as your distance from it.

If you want consistancy with published values, you will need to maintain the 1m distance between the light source and the wall (to ensure the same beam profile), and also always use a fixed, pre-determined distance from the camera to the wall. Then, you will need to come up with a calibrated multiplier to use to convert your camera's number to true lux, by comparing your readings on standardised sources to published results.

Keep in mind that emitters vary in brightness, and you can see up to 30% difference in spot brigness between different lights in the same run, based solely on the randomness of the orange peel in the reflector, emitter efficiency, emitter placement, etc.

Minimising ambient and reflected light will also be key to acheiving accurate results.

Also, avoid using unregulated lights in your tests (minimag 3AA comes to mind) as they will also perform differently depending on battery charge levels and battery chemistry.

 

[electrothump]

Well, I can't get my camera to tell me the firmware number, nor can I get anything to load into my camera. What am I doing wrong?

DN

 

[polkiuj]

Consider reflected vs direct illumination. A light meter is measuring the light directly illuminating the sensor and at a pre-defined distance standardised at 1m, while a camera would be measuring reflected light with several variables at play. That reflected light will only be a small percentage of the total light falling on that point, when you take factors such as scatter due to surface roughness into consideration. You will need to come up with a multiplier to use to come up with a true lux number. That multiplier will depend on the surface which you have the light pointed at, as well as your distance from it.

If you want consistancy with published values, you will need to maintain the 1m distance between the light source and the wall (to ensure the same beam profile), and also always use a fixed, pre-determined distance from the camera to the wall. Then, you will need to come up with a calibrated multiplier to use to convert your camera's number to true lux, by comparing your readings on standardised sources to published results.

Keep in mind that emitters vary in brightness, and you can see up to 30% difference in spot brigness between different lights in the same run, based solely on the randomness of the orange peel in the reflector, emitter efficiency, emitter placement, etc.

Minimising ambient and reflected light will also be key to acheiving accurate results.

Also, avoid using unregulated lights in your tests (minimag 3AA comes to mind) as they will also perform differently depending on battery charge levels and battery chemistry.

Wow! That is so true! What I'd do is to use an established source (hopefully, my LD10) and set the distance so that my camera measures ~2700 lux. Then I compare from there. =)

Well, I can't get my camera to tell me the firmware number, nor can I get anything to load into my camera. What am I doing wrong?

DN

Try reading the FAQ. Use the available programs to get it onto ur camera. =)

 

[electrothump]

Try reading the FAQ. Use the available programs to get it onto ur camera. =)[/QUOTE]


Try reading the FAQ??? DUh! Been there, done that. Had what I read worked, I would not have asked for help.

DN

 

[csshih]

Also, avoid using unregulated lights in your tests (minimag 3AA comes to mind) as they will also perform differently depending on battery charge levels and battery chemistry.

Got a point there.. fully charged NI-MHs..

but, as for calibration, I am doing what polkiuj says, using a light whos lux we know, and positioning it so the camera reads that value.

we then position another light at the exact same place!
(as said before, this a rough measurement)

Try reading the FAQ??? DUh! Been there, done that. Had what I read worked, I would not have asked for help.

DN

It's in the FAQ, item #3.
http://chdk.wikia.com/wiki/FAQ#Q._How_can_I_get_the_original_firmware_version_number_of_my_camera.3F

or, if you don't want to read it,

• create a file called ver.req in the root directory of your SD card. If you're using Notepad on Windows, be sure you don't name it "ver.req.txt" by changing "Save as type" to display "all files".
• IMPORTANT: Note that it is spelled ver.req, as in VERsion REQuest, that Q is NOT a G, Type it correctly!!!
• switch on the camera in playback mode and wait for the main screen (don't switch on in rec mode and change to playback later)
• press the func set+disp. buttons together (press set first, hold it down, then press disp.)

 

[chmsam]

Be sure that the file is in the root directory of the SD card and not in a sub directory.

Another thought is that you might want to try a new or reformatted card just in case.

Be sure there is a photo on the card and that the last image is not a video.

If that fails could you describe step by step what you did?

 

[electrothump]

It's in the FAQ, item #3.
http://chdk.wikia.com/wiki/FAQ#Q._How_can_I_get_the_original_firmware_version_number_of_my_camera.3F

or, if you don't want to read it,[/QUOTE]

Thanks, I got it working. I bought another card. I went to a friends house that also has a camera like mine, and we could do it on his, but not mine. So, I bought another card, and it worked seamlessly. Dunno!?? Heck, I had read, and re-read all that stuff for what felt like the twentieth time with no luck. Seems like a pretty cool ap.

DN