Brightness Box?

xo-whiplock · Mar 30, 2013

I'm trying to sort though some ideas I have to make a brightness box for testing and/or adjusting the exposure meters on my used cameras. I'm hoping someone has built a brightness box and if so, could tell me what is needed to do it right? I'm looking for help in adapting what may already be laying around... dimmer switch, house light bulbs, old lamp w/socket, wood to make the box... etc. I only have an old GE light meter, and a DSLR to find light/exposure output from the box and then mark switch positions... Need EV 4 thru 15 if possible, but not sure of what bulb to use... this I can buy if needed. Pointers, links, schematics, anything to help would be appreciated. I tried both apug and google searches, but nothing I found helped.

Thanks in advance.

noacronym · Mar 30, 2013

I've had these ideas too, but dimmers and 3200K bulbs are not going to cut it. You'd do better to use daylight, a screen of some kind to focus on, and a known good meter to calibrate the daylight at that moment. Tedious, but workable. I've done this a million times. Forget the incadescent bulbs. 40 years ago film had a daylight and a tungsten rating. It still does, but they only publish the daylight speed in the technical literature. It still has a tungsten rating though, just like 40 years ago. We just don't have a clue what it is.

xo-whiplock · Mar 30, 2013

So I take it, it's best to just use a sunny day? I was hoping to come up with a DIY bench tester... I've seen commercial units, but can't afford one. I made a simple shutter tester, and just need to make a light source I can set EV's on to complete the bench testing. What about using a bulb from a movie projector and bouncing that off a white surface and using a dimmer switch?

noacronym · Mar 30, 2013

I have no qualifications to authoritatively discount your idea. I simply assert that film, meters, and associated equipment are set up for a basic color temperature. Any test procedures require at least one constant to build on. Daylight, no matter how wavering, is still a worthy constant, especially on this kind of work. As for me, I live in a 15 year old log cabin where the siding on thewest side of the house in summertime is a perfect neutral gray. So that is my standard. It works PERFECTLY. I've set up many meters on this, and measure against a Luna-Prp SBC I had professionally calibrated, A LunaPro that I converted to silver batteries with an onboard voltage regulator for 2.7 volts, and a Sekonic L308 S. I dare say my calibrations are dead on the money. Therein lies the fools folly. Accuracy without linearity is a fools paradise. But then, that's part of what you're bound to learn if you make up your mind to be proficient at this. GL

noacronym · Mar 30, 2013

To put my involvement to this thread to bed in my mind so I can move on to my own APUG interests, the linearity of which I spoke is contained entirely in the meter cell itself. In the case of CdS, it is entirely a resistance circuit. An SBC circuit is a form of simple amplification a lot like an audio amplifier. But in both cases your linearity is contained in the light-sensitive cell itself. Forget whatever ohm's law computations your mind dreams up. I tried that. A weak cell CANNOT be compensated for and hold linearity.

xo-whiplock · Mar 30, 2013

Okay... good point. A standard is good. I can use your method to verify a single EV between my handheld light meter and my DSLR's metering at any given time outside in daylight... I'm pretty sure they are within a stop of each other, and I can zero out my hand held to match my DSLR's metering, as the DSLR is most likely spot on. So, assuming that the accuracy of my DSLR is adequate, I would simply need to use a bulb that is bright enough at full power to meet or exceed an EV of 15, and then by reducing voltage, I would reduce luminescence... as my light meter reads a full drop in EV, I would simply mark that position of the dimer switch with that EV... and then re-test to verify... in theory anyway. Should I be worried about color temperatures? I would think just how bright and what my meter says the exposure is for shutter, aperture, at a given ASA/ISO and look up EV on a chart (work backwards really from EV desired, until I get shutter and aperture the same on my DSLR)... I'm just not sure a projector bulb will respond well to having a dimmer switch in series???

noacronym · Mar 30, 2013

Color temperature is everything.That is your only dependable standard outside of some multi-milliion dollar laboratory. It's linearity you are ultimately after. Film is set up for daylight. Meter cells are set up for daylight. The linearity curves have much in common. There's your common ground to establish one standard. From the one, you start looking for others. Each one is a variable that influences linearity. You're talking about 1 stop as your present accuracy rate available. I'm talking about 1/3 stop or less. And my talk still allows 33% inaccurate, which is terrible if we were working at Los Alamos, Kodak, Tek, NASA... My point is if I can hold linear to 1/3 stop, then certainly I'm a dummy compared to a lot of these guys. And remember, if your meter is within a whole f/stop of accuracy, the meter cell has lost strength and linearity is totally dead. The meter cannot be calibrated.

noacronym · Mar 30, 2013

No a dimmer on a projecter bulb will totally kill color temperature and more likely burn out the bulb and cause the dimmer to meltdown. DON'T do it. Ohm's law and power equations--very important.

xo-whiplock · Mar 30, 2013

Okay... scrap the dimmer switch and projector bulb... I get what you are saying now with regard to "fixing" a camera's metering... can't do it if the photocell is worn out and lost its ability to produce liner output. So, if I am able to make a brightness box and my meter output falls outside of desired linearity, replace the bad parts with good ones. I guess we are back to the beginning of this thread, in that going outside and using daylight at different EV's is better then trying to mimic it in a box... without some specialized skills and parts. The old KISS principle. Thanks for helping me out.

Chan Tran · Mar 30, 2013

xo-whiplock said:
Okay... scrap the dimmer switch and projector bulb... I get what you are saying now with regard to "fixing" a camera's metering... can't do it if the photocell is worn out and lost its ability to produce liner output. So, if I am able to make a brightness box and my meter output falls outside of desired linearity, replace the bad parts with good ones. I guess we are back to the beginning of this thread, in that going outside and using daylight at different EV's is better then trying to mimic it in a box... without some specialized skills and parts. The old KISS principle. Thanks for helping me out.

I am very much interested in this subject. If it's not possible to calibrate an old worn out meter, it's still possible to check the performance of the meter. I am not sure as to which color temperature to use. Kyoritsu Electric Co. Ltd made a good number for standard light source as well as meter and shutter testers. Many of them have color temperature of 2800K although a number of newer LED based ones specified as only white LED and I don't know the color temperature. Since most film are daylight balance film it makes sense for the light source to be daylight equivalent but then I am not so sure about this. Anyone has any idea about this?

Using a dimmer changes the color temperature as you dim it down so it's not a good thing here.

What I did in the past is to use a dichroic color head. I used a Beseler 45 computerized color head which has accurate filter display. I turn all the filter off and then check the brightness at the diffuser which is a 6'' circle with a spot meter. It's about EV15 in this condition (I have to dial in equal amount of filters a bit to get it exactly EV15). Checking eveness of the brightness with the spot meter and found within 1/10 stop so it's ok there. Dial in equal amount of filtration of C, M, Y and checking the brightness with the spot meter I get to about EV9 with all the filter dialed in.

Bill Burk · Mar 30, 2013

Check out Post #36 of this thread for a good choice color temperature for meter calibration...

(there was a url link here which no longer exists)

Chan Tran · Mar 30, 2013

Bill Burk said:
Check out Post #36 of this thread for a good choice color temperature for meter calibration...

(there was a url link here which no longer exists)

Thanks Bill. According to Stephen in that thread the color temperature for the calibration light source changes from 2800K to 4700K in 1960. That's fine. So an LED with white LED would be a good choice. The color temperature has to be significantly off for the meter to read differently. Now I think the LED doesn't change color temperature as you dim it down. How would I dim the LED. There is current control and then pulse modulated technique. Something for me to work on...

Bill Burk · Mar 30, 2013

Chan Tran said:
Thanks Bill. According to Stephen in that thread the color temperature for the calibration light source changes from 2800K to 4700K in 1960. That's fine. So an LED with white LED would be a good choice. The color temperature has to be significantly off for the meter to read differently. Now I think the LED doesn't change color temperature as you dim it down. How would I dim the LED. There is current control and then pulse modulated technique. Something for me to work on...

Then there are physical modifiers (screens) and unit count (fewer / more LED's in a bank)

xo-whiplock · Mar 30, 2013

The following was in response to Chan Tran explaining his use of his optical color printer/enlarger... The issue of light meters and exposure meters is a different direction...

Color temperature changes throughout the day... from sun rise to mid day, to sun set again. So color temperature is not a constant in daylight, but a variable. You'll see in most editing software you have the option to select the temperature according to the effect you want or to match better the time of day you took the shot so it looks like what you had seen when taking the picture... sometimes just changing a photo to cooler or warmer helps with the effect you are trying achieve, that may not have been present when you shot the photo for affect. I'm not experienced in optical printing but I'm guessing you want to reach the exact or as close as possible to the film negative's color temperature in order to render accurate color reproduction initially and thereby reduce the need for post editing to fix it, and one may not be able to fix it, if initial setup is too far off.... just a guess. So, even though your EV is reduced after applying the color filters, it just means you have to expose the photo paper with the new EV. Am I correct in this?

So from reading post up to this point I take it that in order to check the exposure metering on a camera, one needs to use a stable light source, and simply using a dimer will not provide that stability. Hence, the LED's with physical modifiers to achieve desired output for a given or desired EV output. Interesting.

I got side tracked reading all this info... I came back to this thread to comment that I installed a little program onto my computer that allows brightness adjustment of my computer's LED monitor... Although with note pad in full screen mode and brightness at 100%, I get an EV from my DSLR of EV10, and as I drop brightness in 10% increments, I see a full stop of change in shutter speeds, indicating EV droping 1EV/10% reduction. I may be able to simply use my computer's monitor to check my exposure meters? And above EV10, just use daylight?

Bill Burk · Mar 30, 2013

The issue with daylight is that it fluctuates, as you said, it's not easily controlled. A 4700K Incandescent bulb, on the other hand, might be something you can control with a feedback circuit or voltage regulator. If I were setting up a standard box, I'd make it one brightness, a standard, and then modify its brightness outside the box on the bench. You could use the old GE meter to monitor the brightness inside the box.

noacronym · Mar 30, 2013

I have found in my self-taught meter calibration endeavors that the best place to calibrate a meter is to concentrate on finding middle gray on the film curve, and finding middle brightness of actual picture-taking conditions. I do this, presuming my meter cell has lost some of its linearity. Were it not so I wouldn't be calibrating the meter in the first place. If you set your meter up for "sunny f/16", then it stands to reason the meter will be accurate from cloudy to sunny. But what about 8:45pm EDT in the south in July? That's where your lost meter cell linearity will get you in trouble. What about 9:15 EDT in the same location? By then it's dark, and even though the meter might be accurate to 1 minute exposure when it was new, your lost linearity will have you several stops off.
The point is, a lot of meters with weakened cells can be useable under normal hand-held situations. You're just not going to be able to use it on the bright side of Mercury to the dark side of Pluto like when it was new. Id still go on with a calibration if I was only 1 stop off in the middle of a cloudy-bright day. Any more than about a stop, and your cell is just too weak.

xo-whiplock · Mar 30, 2013

Bill Burk said:
The issue with daylight is that it fluctuates, as you said, it's not easily controlled. A 4700K Incandescent bulb, on the other hand, might be something you can control with a feedback circuit or voltage regulator. If I were setting up a standard box, I'd make it one brightness, a standard, and then modify its brightness outside the box on the bench. You could use the old GE meter to monitor the brightness inside the box.

I like standards, without them we all would be lost... LOL Thank you. This makes the most sense to me as well. It would be much easier to use physical filters upon a standard light source, then trying to manage a variable light source accurately for a DIY project and limited resources. I'm thinking a set of ND filters to peg say 3 different points along the exposure trajectory and the final point being the standardized light source without a filter for the brightest point. My camera's service manual says to use a brightness box set to two different EV's, and use a 50% filter to do one of the adjustments/checks. These checks/adjustments would be for high brightness and low brightness adjustments. There are other checks that can be done as well, but without such a box, it seems useless to even open up a camera to try and fix or adjust its exposure meter. This is why I asked for suggestion/help in making one. So far, I'll have to make due with my computer monitor, but I may well look for a used/broken LED monitor to use as my base for creating the box I need later....

Chan Tran · Mar 30, 2013

Dear XO! The color temperature of daylight changes but we want to calibrated light source to have a constant color temperature. Dichroic color head generally has 3200K lamp in there. With some Cyan and very little Magenta I can make it becomes a 4700K light source and it would stay that way constantly. I used the spyder2 with the HFHC software to measure color temperature and recheck by shooting the light source with a Nikon Coolpix with the color temp set to 4700K and see if the resulting image is gray.
The filtrations setting which is monitor by the 3 sensors and displayed on the unit digital displays are very accurate. I make a note of the readings for this condition and later all I have to dial in the amount of filter that gives the same reading then the light is the same even if the bulb has aged and changed color and intensity.
And then dial in equal amount of filter on all 3 channel would give it neutral density. Monitoring the actual brightness with an accurate spot meter as reference. (Well you must have an instrument that you can trust to be correct) so I would know how bright the light is.
It works very well except that I only have 5 stops range. The unit has 180cc adjustment and that's 6 stop but I waisted some of the filtration to increase the color temperature.

By the way I used it to check out the light meter on my Minolta SRT-101 and after checking it I decided not to use it. Even I supply the camera a constant 1.35v source via a regulated power supply. I can adjust the calibration resistor to make it read higher or lower but if it's correct at EV14 it's wrong at EV9 and vice versa so I decided to use the camera as a meterless camera rather than trying to get expensive battery for it.

noacronym · Mar 30, 2013

Chan Tran said:
Dear XO! The color temperature of daylight changes but we want to calibrated light source to have a constant color temperature. Dichroic color head generally has 3200K lamp in there. With some Cyan and very little Magenta I can make it becomes a 4700K light source and it would stay that way constantly. I used the spyder2 with the HFHC software to measure color temperature and recheck by shooting the light source with a Nikon Coolpix with the color temp set to 4700K and see if the resulting image is gray.
The filtrations setting which is monitor by the 3 sensors and displayed on the unit digital displays are very accurate. I make a note of the readings for this condition and later all I have to dial in the amount of filter that gives the same reading then the light is the same even if the bulb has aged and changed color and intensity.
And then dial in equal amount of filter on all 3 channel would give it neutral density. Monitoring the actual brightness with an accurate spot meter as reference. (Well you must have an instrument that you can trust to be correct) so I would know how bright the light is.
It works very well except that I only have 5 stops range. The unit has 180cc adjustment and that's 6 stop but I waisted some of the filtration to increase the color temperature.

By the way I used it to check out the light meter on my Minolta SRT-101 and after checking it I decided not to use it. Even I supply the camera a constant 1.35v source via a regulated power supply. I can adjust the calibration resistor to make it read higher or lower but if it's correct at EV14 it's wrong at EV9 and vice versa so I decided to use the camera as a meterless camera rather than trying to get expensive battery for it.

There you go--on that Minolta the meter cell has lost linearity. The CdS cell is worthless. Accurate at one light level, off at others. It's a shame.

Bill Burk · Mar 30, 2013

Chan Tran,

Were you adjusting both variable resistors and also mechanical linkage? Sounds like a calibration dance that requires a few recursions until you get it to work...

http://www.willegal.net/photo/photo-first_page.htm

Chan Tran · Mar 31, 2013

Bill Burk said:
Chan Tran,

Were you adjusting both variable resistors and also mechanical linkage? Sounds like a calibration dance that requires a few recursions until you get it to work...

http://www.willegal.net/photo/photo-first_page.htm

One of the resistors is for adjusting the battery check level so it's not important and can easily adjusted. The other is for the metering itself. I didn't do anything with the mechanical because unlike other camera the SRT-101 mechanical linkage is totally independent of the meter movement. It's quite an interesting design where I don't see in other camera. All user's adjustment like aperture, shutter speed, ISO would only move the mechanical pointer. The light intensity moves the meter movement.

xo-whiplock · Mar 31, 2013

I'm following but I got lost at setting up a color enlarger properly... never done it but I'm sure I could learn... Chan, you are like advanced class and I'm beginning basics class... LOL I get your point thought. You have a stable light source and it allowed you to determine that the meter in your SRT101 is not reliable, so you use it full manual... My XE-7 is kinds made to use the meter and it's such a nice camera, I wanted to make sure after having to part out my first one due to shutter at 1/1000th sec. not opening and causing unexposed frames for which I could not fix on my own, I use the slide resistor from it in another XE-7 that I broke the slide resistor on while doing a clean up and getting the slide resistor to read again. Once I put the replacement slide resistor in and tested and measured shutter speeds, I compared meter reading with my DSLR aimed at different EV's. Close enough, so I shot a test roll this morning and they all came out exposed properly. I'm a happy camper. Scanning negatives is what I'm working on now... I've ruled out VueScan due to output that is not as good as the Epson Scan software's output. Oh, and to not use digital ice... LOL. I have like 8 SRT's? 101's and a couple of 102's all with mirror lock up. My first GAS camera because they are so basic and good at being mechanical, and easy take the top off and unstick the meter needles from the foam... LOL. I'll be using one for micro stuff and maybe telescope... The XE-7 is my baby and I love winding and shooting it. I have a pretty good kit of Rokkor lenses, so I good to go. Just need to shoot more and practice. I don't think I'll be able to DIY a stable light source, other then the idea of using an LED monitor to do a basic not scientific check of the my camera's exposure meters. If something is way off, I'll have to send it in to keh.com or someplace like it to get it done right in a lab by a pro. I value my cameras, and I need to have the faith and trust in their operation before running film though them. Right now, I'm confident in my XE-7 to do me right, and my Canon QL17 GIII. These two will do me fine for now, and I'll work on checking the rest of my cameras as time permits. Thanks all for the feedback and assistance, info and education... Happy Easter Egg Day!

Chan Tran · Mar 31, 2013

I use the Minolta dual scan IV for my film scanning. I always use it in manual exposure mode especially if I want to check the exposure on film. In auto mode the scanner software automatically correct for exposure error as well as doing the color balance for the shot. Which is great for quick scan but you would not know if your exposure is right on or not. Using manual exposure setting on my scanner I can easily tell if my film exposure is off and by how much.
Your XE-7 is an automatic camera and it does call for using it in automode and thus the meter must work correctly. I used an SRT-101 for many years (The cameraless years) and didn't use its meter at all and had good exposures most of the time. I started having many bad exposures when I bought a spanking new Nikon F5 in 2002 and tried out it matrix metering system.

John Koehrer · Apr 1, 2013

Kyoritsu used EV's 9, 12 & 15 The camera/meter is set for ASA 100 as a standard value.
When changing levels the knob had a mechanical feel and sound to it more like changing ND filters not like rotating an electrical switch.
The tester also allowed changing the "K" factor to a value used by the camera maker.

While fiddling with the "K" knob there was a slight difference in brightness possibly in color temp.

Brightness Box?

Privacy & Transparency

Privacy & Transparency