Fluorescent tube intensity/colour cycle captured using rolling shutter

[Guillermo Luijk]

Electronic rolling shutter is usually of a similar order of magnitude as the domestic electric AC voltage (50Hz/60Hz). By capturing a vertically oriented fluorescent tube using electronic shutter, and analysing the resulting linear RAW values along the tube, we can obtain the time profile of this kind of light very accurately thanks to sensor linearity (lens vignetting was not corrected):



We can see that the tube never completely stops emitting light, and in the minimums low frequencies (warm spectrum) become dominant. However in general high frequencies (cool light) are highly dominant in the peaks, so the averaged colour perceived by human eye tends to be cold.

Regards

[Jack Hogan]

Nice!  Good work Guillermo.

Jack

[Bart_van_der_Wolf]

Hi Guillermo,

Nice demonstration.

The cooler UV+blue/green peaks combined with the warmer fluorescent afterglow, as seen through the Bayer CFA filter bands, are clearly shown. Also, the stroboscopic effect of Fluorescent tubes is obvious, and that can cause issues with light metering, and with shorter exposure times. Color accuracy is non-existent in practical photography, and such a peaked and variable illumination spectrum is almost impossible to color-balance.

Cheers,
Bart

[bjanes]

Hi Guillermo,

Nice demonstration.

The cooler UV+blue/green peaks combined with the warmer fluorescent afterglow, as seen through the Bayer CFA filter bands, are clearly shown. Also, the stroboscopic effect of Fluorescent tubes is obvious, and that can cause issues with light metering, and with shorter exposure times. Color accuracy is non-existent in practical photography, and such a peaked and variable illumination spectrum is almost impossible to color-balance.

Bart,

A nice demonstration, indeed. However, the stroboscopic effect that Guillermo demonstrated can be greatly mitigated by replacing a magnetic ballast which operates at 60 Hz (USA) or 50 Hz (Europe) by an electronic ballast which usually operates at 20,000 Hz or higher (see here).

I formerly used a fluorescent light box (formerly used to sort Kodachromes) to illuminate Stouffer wedges and other targets for photographic tests. The old box worked well for low shutter speeds, but produced inconsistent results at higher shutter speeds. The problem was resolved by replacing the old magnetic ballast with an electronic device.

Regards,

Bill

[Bart_van_der_Wolf]

Bart,

A nice demonstration, indeed. However, the stroboscopic effect that Guillermo demonstrated can be greatly mitigated by replacing a magnetic ballast which operates at 60 Hz (USA) or 50 Hz (Europe) by an electronic ballast which usually operates at 20,000 Hz or higher (see here).

Hi Bill,

Yes, replacing the starter/ballast can make a difference, but the specifications of the fluorescent coating (slow or fast decay) also make a difference. In the end, there will still be some fluctuation (intensity and color), so something to be aware of.

I'm sensitive to the 50Hz cycle we have here, but LED lighting is taking over very rapidly due to better efficiency. Also a spiky spectrum, but less fluctuation.

Cheers,
Bart

[bjanes]

I'm sensitive to the 50Hz cycle we have here, but LED lighting is taking over very rapidly due to better efficiency. Also a spiky spectrum, but less fluctuation.

Bart,

LEDs are being rapidly implemented here in the USA also, but they are too susceptible to flicker. Many operate at twice the mains frequency as documented here. I have been switching over to LEDs in my own residence and have not been bothered by flicker but have no means of measuring flicker and it is not documented on the packaging of the lamps. This flicker is of concern when LEDs are used in photographic applications where a direct current driver would be desirable.

In 2009 the U.S. Environmental Protection Agency (EPA) recommended that the operational frequency of LEDs be raised to 150 Hz, but has since relented (see here). What is the situation in Europe?

Regards,

Bill

[Bart_van_der_Wolf]

Bart,

LEDs are being rapidly implemented here in the USA also, but they are too susceptible to flicker. Many operate at twice the mains frequency as documented here. I have been switching over to LEDs in my own residence and have not been bothered by flicker but have no means of measuring flicker and it is not documented on the packaging of the lamps. This flicker is of concern when LEDs are used in photographic applications where a direct current driver would be desirable.

Yes, LEDs can also flicker, potentially even more than fluorescent tubes, but it depends on their design. I think the color may be less variable during the cycling of AC power, but that also depends on their design.

Apparently, there is no uniform way of quantifying flicker, and that's why it's not mentioned in the specs. Also, the combination with dimmers may exacerbate issues. Here is a decent summary of the various aspects involved.

In 2009 the U.S. Environmental Protection Agency (EPA) recommended that the operational frequency of LEDs be raised to 150 Hz, but has since relented (see here). What is the situation in Europe?

Probably somewhat similar to your situation in my country, but I do not know how other countries are doing. Incandescent bulbs above a certain wattage are no longer allowed to be sold. I've replaced just about all lights by LEDs over the last 2 years (since light color has become more agreeable), and the electricity bill has dropped quite a bit as a result. The last 2 infrequently used FL-tubes will be replaced when they die by LED strips, and a few halogen spots in the hall are only used briefly during the dark season when entering. The products sold here are not bothering me that much, so apparently they have sufficient phase switching capability that I hardly see a stroboscopic effect when waving my hand in front of a dark background.

Cheers,
Bart