Household LED Lighting Mimics Shutter Failure

[Golem]

I was shooting at the wall, checking miscelaneous

lens hoods as to their fit on various lenses, as to

vignetting, etc ..... so I was shooting the same

image over and over. But the resulting exposures

appeared to be varied over a range of about 0 to

1+ EV, like some kinda erratic bracketing.

 

Then I noticed that some were brighter in their

upper half, or lower half, or shadowed along one

of the long-side edges, which would mean problems

with the timing between first and second curtains

of the shutter. So I slowed the shutter way down to

speeds where such timing is almost irrelevent and

my fears were confirmed. Those exposures did not

exhibit the problems seen in the higher speeds.

 

Anywho, I had recently fitted that room with LED

lighting, which I didn't realized has even worse

pulsing effects than compact fluorescents. All the

lens shade testing occured at night. Next day when

it occured to me to possibly blame the LED lights,

I ran shutter speeds tests by daylight. No problem

to be seen. I retested in the LED-lit room and the

problem was still there. It's definitely the LED

lighting causing the problem.

 

`

[Al Pha]

Same experience here. I regularly shoot in a Hare Krishna temple that has LED lights on the altar and in the main room. Shots taken at the same settings (aperture, shutter speed, ISO) look very different because of the LED lights.

[Golem]

As this type of general purpose lighting proliferates,

an old quandary is somewhat transmogrified.

 

For those of us typically shooting at marginal handheld

speeds indoors, I had puzzled over whether IBIS was the

answer [for obvious reasons] or was it rather more of a

benefit to go for improved High ISO quality, to allow

raising the shutter speed.

 

The latter option would not only reduce the effect of a

less than steady hand holding the camera, but also help

with reducing blur from subject motion. Acoarst, such a

quandary has two differing solutions depending upon the

degree of subject motion ... if any at all.

 

And now ... AND NOW ... we have strobing LED lighting

that may mean avoiding speeds over 1/60, or possible

even avoiding speeds over 1/25. There's a minor thing

about whether you have 50 or 60 cycle AC mains current.

A more significant factor is the max synch speed of your

shutter. Synch speed is the highest speed at which the

mechanical shutter uncovers the entire sensor. However

at max synch speed, that full uncovering does not dwell

over the sensor but for a brief moment, yet the entire

travel time of the shutter is much more than that brief

moment, so the LED could be dark at that very moment.

IOW it's like an improperly synched strobe shot.

 

 

===================================================

 

 

Kinda Techy Stuff:

 

As speeds slow further below max synch speed, the full

open "moment" get longer. The "wrong moments", those

that yield streaks, bars, shadows, and generally darker

exposures do NOT get longer, so they'll proportionally

contribute less to the overall time of exposure. Most

shutters today are a traveling "window" moving across

the sensor. The SPEED of travel never changes, but the

"window" can vary from a narrow slit, crosswise to the

direction of travel, to a rather large long rectangle

whose long side parallels the direction of travel. The

longer the rectangle, the more the LED pulsing is able

to complete its cycles during FULL uncovering period

[uncovering the sensor]. But the longer the rectangle

the SLOWER the actual shutter speed. Obviously, in the

"traveling slit" mode [faster shutter speeds] the light

source must be absolutely steady as the slit traverses

the full size the sensor. IOW, it's a reeeeeally fast,

very small, mechanical scanner. You wouldn't want the

lamp in your flatbed scanner pulsing during its travel

across the image, right ?

 

 

====================================================

 

 

For the Geekishly Curious Only:

 

The actual "windows" do not exist physically. If they

were physical, they could be larger than the space in

the camera where the shutter lives. The windows exist

in time, not in space .... sort of. Yes and no. They

existed on a roll of cloth [with rectangular windows in

it] many decades ago, when cameras were larger. Today,

the slit-sized opennings are physically there, but the

rectangular opennings longer than the dimensions of the

sensor are compressed, mechanically, not digitally. If

you know that there are 1st and 2nd curtains in the

shutter [remember "Electronic First Curtain"? well, for

the moment, forget it]. To understand the mechanical

compression idea, you have to realize that "1st and 2nd

curtains" is archaic jargon, far more acurately stated

as "Leading and trailing blades". The travel speed of

the blades never changes, but the shutter timing circuit

determines how much lead time [head start] to allow the

leading blade before the second blade is allow to go

chase after it. That delay time determines the effective

shape of the slit or rectangular "window". A rather long

delay time effects a rectangular "window" that is longer

[can be MUCH longer] than the physical size of the sensor

and the physical space allowed for the shutter mechanism.

 

`

[pilsburypie]

Just trying to get my head round this from a video perspective. Sure on a still photo, you might get the "wrong moment" of the frequency of the LED light, but how noticeable would this appear on video shot with the same shutter speed? Video in its simplest term is a flick book of still images. I suppose it depends on how often it takes a "duff" frame....

[Golem]

Just trying to get my head round this from a video

perspective. Sure on a still photo, you might get

the "wrong moment" of the frequency of the LED

light, but how noticeable would this appear on video

shot with the same shutter speed? Video in its

simplest term is a flick book of still images. I

suppose it depends on how often it takes a "duff" frame....

Hard to wrap my head around the idea that you've

never observed that result. It's a sort of rolling, cyclic,

and continuing "passing shadow" effect. There was

even a rather similar problem in analog [pre-digital]

video. And that was before LED lighting. "Legacy"

fluorescents would provoke it ... so it tended to affect

mostly "Spot News" and some documentary footage.

 

As to seeing the effect of LED lighting in digital video,

at the traditional cinematic shutter speed of about 1/25

[+/-] the effect would be minimized but still be present.

Current technology is already pushing video shutter

speeds to 3-digit speeds, which increases the problem.

 

`

[pilsburypie]

I'm behind the times and only just got LEDs in my kitchen. Not taken any video in there and can't remember seeing it elsewhere. Why does this happen in rooms with LED lighting and not with LED video lights that now seem the norm? Is it due to mains lights in your home working off AC but video light LEDs having a DC converter in them thus removing frequency?

[Golem]

I'm behind the times and only just got LEDs in my kitchen. Not taken any video

in there and can't remember seeing it elsewhere. Why does this happen in rooms

with LED lighting and not with LED video lights that now seem the norm? Is it due

to mains lights in your home working off AC but video light LEDs having a DC

converter in them thus removing frequency?

  

Yes .... sortalike that.

  

Dedicated LED video lights run on clean pure DC. 

  

Household LED lamps run on a cheap built-in AC

to DC converter that has 60 "dead instants" per 

second in its DC delivery. IOW it pulses at 60hz,

unless it's 50hz .... depends upon where you live.

[JCC]

Most of my shooting is in performance venues.  Increasingly they have gone to LED stage lighting. 

 

Not only is there the color banding you mention, but RGB white is very peaky, and not true white.  The shadows can be very unflattering. 

 

JCC