Variable Shutters and Flickers

Discuss the Art, Science and Craft of Cinematography here
Post Reply
Aravindan.G.P.
Newbie
Newbie
Posts: 29
Joined: Wed Aug 26, 2009 2:52 pm

Variable Shutters and Flickers

Post by Aravindan.G.P. »

Recently I had the opportunity to write something on Shutter angles and Hertz cycle of HMI. I thought I would post it here for others to bounce their thoughts on it.

I think there should not be a need to elaborate the role that a shutter plays in the film camera. Shutters, linked with the intermittent motion of the pull down claw act as a mechanism to expose the film and at the same time make it possible for the image to be viewed through the eye piece.

The shutter spins 24 times per second exposing and blocking the film plane. This means that the time at which the film plane would be exposed for 24 fps would be 1/48th of a second.

The rate at which the variable shutter spins is usually fixed. It is a constant. What can be changed is the angle of the shutter ( 90 to 11.2 degrees) and the running time of the film frames (48, 64 fps etc).

Any decrease in the shutter angle or increase in the fps would warrant increased time needed for exposure, which can usually be achieved by opening the T-stop or changing the emulsion to higher speed etc.

Assuming that the camera operates on a crystal controlled motor, we had been advised to keep the shutter at 172.8 degrees whenever we had used HMI lights. What is the logic behind this?


We all know that we operate our electrical appliances at 220 Volts AC at 50Hz cycle. This means that the current reverses and alternates its polarity 50 times on one side and 50 times on the other, in one second. Hence, when a HMI light is connected to the supply, the light pulsates 100 times (50Hz x 2) in one second on either side.

So, for every shutter rotation, the light actually pulsates 4.17 times. That is 100/24 = 4.17. Extending that, we can conclude that when the shutter is open and film plane is exposed, there is 2.09 pulsation of the HMI light that reaches it and 2.09 when the shutter is closed.


Taking this as the criteria, certain optimum shutter angles are prescribed for different settings, they are


24 fps / 60 Hz 144 degree shutter
24fps / 50 Hz 172.8 degree shutter
25 fps/50 Hz 180 degree shutter


The exact formula by which one arrives at the shutter angle so that it just has that much time to allow two pulses of light from HMI to reach the film plane is,

24 x 360 x 1/ time fraction.

Time fraction is the amount of time the film plane is exposed.

For Eg:

24x 360 = 8640 x 1/50 = 172.8
24x 360 = 8640 x 1/48 = 180
24x 360 = 8640 x 1/60 = 144

Note that when the exposure is one fiftieth of a second with the shutter angle 172.8 degrees it perfectly falls in sync with the 50 Hz cycle in which the HMI light would be discharging 2 pulses of light for the exposure. Similarly, the last case where the shutter angle if 144 degrees and the Hz cycle is 60. The trick is that the shutter should be open so that the two pulses of light can travel and hit the film plane. If there is a delay then there is a flicker. Hope this helps in understanding what we already knew. Kindly share your thoughts..
Karthik Ganesh
Newbie
Newbie
Posts: 11
Joined: Tue Sep 29, 2009 3:21 pm

Re: Variable Shutters and Flickers

Post by Karthik Ganesh »

Nice article ara. Just that the calculation of 4.17 cycles was a little complicated for me. I have always thought of a different logic. Basically 50hz is 50 complete cycles of the HMI light, i.e., 50 cycles of alternating current and if we dont get a complete cycle we get a flicker.

So the need is to synchronise these 50 cycles to the time the shutter is open so that it can reach the film plane. So it boils down to the shutter speed of the single frame - can be 24 fps or 25 fps at 50 hz or 60 hz.

Shutter speed = shutter angle /( 360 * fps)

At 24 fps at 50 hz we have to get 1/50 shutter speed

1/50 = shutter angle /(360 * 24)

shutter angle = 360 * 24/50 = 8640/50 = 172.8

for 24 fps at 60 hz we need to get a shutter speed of 1/60

shutter angle = 360 * 24 / 60 = 144

Actually its the same calculation but i didnt get the 4.17 part. So thought will post it.
Aravindan.G.P.
Newbie
Newbie
Posts: 29
Joined: Wed Aug 26, 2009 2:52 pm

Re: Variable Shutters and Flickers

Post by Aravindan.G.P. »

Thanks machi...

Correct me if I am wrong....

You have mentioned that "at 24fps at 50 hz, we will have to get 1/50 shutter speed".....isn't it 1/48?, just wondering....however at 25 fps at 50hz we do get 1/50 shutter speed.

As frequency is measure in Hertz and one hertz would equal to one complete sinusoidal waveform for an AC current, I said it releases 2 pulses of light for one Hertz cycle..
Aravindan.G.P.
Newbie
Newbie
Posts: 29
Joined: Wed Aug 26, 2009 2:52 pm

Re: Variable Shutters and Flickers

Post by Aravindan.G.P. »

Forgot to add that the 4.17 is 100 pulses divided by 24 fps...I think it is mentioned already. Acutally it comes around to 4.1666667 and rounded off to 4.17. Thanks.
Karthik Ganesh
Newbie
Newbie
Posts: 11
Joined: Tue Sep 29, 2009 3:21 pm

Re: Variable Shutters and Flickers

Post by Karthik Ganesh »

Yeah. To avoid flickers we need a shutter speed equivalent of 1/50 seconds. And since at 24 fps at 180 degrees shutter angle, the shutter speed is 1/48, we have to use shutter angle of 172.8, to get a 1/50 seconds shutter speed when we use normal HMI to avoid flickers.
Nikhil Mulay
Jr. Member
Jr. Member
Posts: 52
Joined: Wed Dec 13, 2006 8:43 am

Re: Variable Shutters and Flickers

Post by Nikhil Mulay »

Just one doubt about this... while all these calculations go on to assume that we want EXACTLY two pulses of light from the HMI to fall onto the film.. but that will only be possible when the camera shutter starts exactly at the zero crossover point of the ac cycle. The probability of this happening is extremely small. So in this case two cycles of light in totality would fall on the film, but exactly two cycles would be a remote chance. e.g. say the camera starts up at the peak of the ac cycle, then half of the first cycle, one complete second cycle and half of the third cycle.... and so on. So would changing the shutter angle actually guarantee that only two cycles of light would fall on the film plane? Sorry to be so anal about this... Just a small thing that i could not understand!!
Aravindan.G.P.
Newbie
Newbie
Posts: 29
Joined: Wed Aug 26, 2009 2:52 pm

Re: Variable Shutters and Flickers

Post by Aravindan.G.P. »

Hi Nikhil,
I was not able to post the reply without getting some numbers jumbled up due to the formatting in this site. I had attached the reply as 'Hi Nikhil'. Please refer to that. I think you will find your answer there.
Attachments
Hi Nikhil.doc
Reply reg. Variable shutters and Flickers
(46.5 KiB) Downloaded 840 times
Aravindan.G.P.
Newbie
Newbie
Posts: 29
Joined: Wed Aug 26, 2009 2:52 pm

Re: Variable Shutters and Flickers

Post by Aravindan.G.P. »

Hi Param, if you can extract the word document in the above attachment and put it in ftiipeople as such, it would be great. Thanks.
Nikhil Mulay
Jr. Member
Jr. Member
Posts: 52
Joined: Wed Dec 13, 2006 8:43 am

Re: Variable Shutters and Flickers

Post by Nikhil Mulay »

Was able to copy n paste successfully!! Thanks for the explanation!! got it now!!
Hi Nikhil,
I saw your post pretty late, as we were busy in a shoot. Ok, whatever you had asked is very relevant…I will try to answer from whatever I had read and understood. Before that, please don’t apologise for being anal. In life, its much better to be anal about something, than just be ‘banal’.
Ok, we are not making any herculean task to ‘exactly’ have two pulses of light fall onto the film plane. We are very much restricted by physics and economics that we get only those two pulses of light (It is due to the fact that we decided to produce and distribute electricity as 50Hz AC). The two pulses of light is a Rosetta stone only at 24 fps, otherwise it can range from 1 pulse of light to 100 pulses of light on a 50Hz AC current.
Flickers appear whenever there is an uneven exposure in the frames. Exposure depends on the amount of light and the amount of time. If the amount of light which falls on the film plane is consistent over an amount of time, then there will be no flicker. When it is not so, flickers will arise.
As mentioned already, the light pulses at twice the mains frequency ie in a 50Hz cycle, the light pulses 100 times. This means that at every half cycle of the of current the light pulses once. If we correlate it with the normal frame rate of 24 fps, then,
At 1/24th of a second, the light pulses 4 times and
At 1/48th of a second , which is the amount of time the film plane will remain exposed, there will be two light pulses.
Now, as long as the light pulse rate remains as an integer multiple of the frame rate, then there will be no flicker. That is as long as you chose one of the below frame rate, irrespective of the shutter, there should not be flicker. In this case what one is trying to do is to match up with the mains frequency. There are some caveats when you try to do this.
1. The camera should run on crystal controlled motor
2. The generator should preferably crystal controlled. (They did have some crystal controlled gennies, then when the electronic ballast for HMI arrived, they dispensed with them)
3. Extreme vigilance should be exercised to have both the mains frequency and the frame rate in synch, though they have a non-synch latitude upto 4% where there may not be appreciable flicker.




Light Pulse Frame Rate Pulse rate
1 x 100 100
2 x 50 100
3 x 33.33 100
4 x 25 100
5 x 20 100
:
:
:
100 x 1 100


Your question becomes relevant here, when the camera should be in PHASE with the AC power. Like you said, it becomes important that the shutter opens at the same time the peaking occurs in the AC cycle. It does become significant at which point of the half wave the exposure starts. In my opinion, there is a small rider to it. I feel that the question of when the shutter should open vis-à-vis the AC cycle is more relevant when we are dealing with higher fps, irrespective of the shutter angle. For example, if we set the camera at 100fps in the above condition, we get only one pulse of light at 1/200th of a second to fall on the film plane. It is then pertinent to have the opening of the shutter and the Ac cycle in perfect synch. In this case, if both of them are perfectly synched, then you may not see the image you are shooting on your viewfinder.

What we had seen above is independent of the shutter angle. The other way to avoid a flicker is to choose the shutter angle along with the combination of frame rate. Many cinematographic books and website recommend the following combos.






50% 48% 40% 33.33%
180 172.8 144 120
1 50 48 40 33.33 100
1/2 25 24 20 16.66 50
1/3 16.66 16 13.33 11.11 33.33
1/4 12.50 12 10 8.33 25
1/5 10 9.60 8 6.66 20

What you should look at is the flexibility of the frame rates, theoretically., Consider the columns mentioning the shutter angles. When the shutter angle is 50% opened, ie 50% of 360, then the acceptable frame rates to avoid flicker would be 50, 25, 16.66, 12.50 and 10 fps. These frames rates gets reduced at ½, 1/3rd , ¼th and 1/5th of 50 fps. It is the same for shutter angles 172.8 and 144 and 120. What happens when we reduce the shutter angle is that it gets closed (like at 172.8 degrees, the area of the shutter that is opened is reduced by 2% as compared to 180 degrees, at 144 -10%, at 120 around 16%) thereby increasing the duty cycle of the shutter and to maintain the Hertz cycle, we correspondingly reduce the frame rates so that pulse rate of the light is an integer multiple of the frame rate. What essentially happens is that with the combo of shutter angle and the fps we restrict the hertz cycle of the shutter to one of the harmonics of the Alternating Current. In this case, it does not matter when the exposure cycle starts, as all half sine waves are identical. As long as we adjust the shutter angle and the fps to capture one or more half cycles, we can avoid flicker.

Now, whatever which was said above applies to the a magnetic ballast in a HMI light. Nowadays we usually have Electronic Ballast. In an electronic ballast, the 50 Hz is upped between 65 to 80 Hz so that it produces square waves instead of sine waves. When compared to the pulsation of the magnetic ballast, the pulsation of a electronic ballast will ‘appear’ continuous. It is also theoretically possible to connect 3 HMIs on each phase of a 3-phase AC current to produce a pulsation of 360 half cycles in, so that there is no delay for a flicker to be felt or recorded. There are also many studies conducted to study the flicker below 10 fps.

I should admit that I had not tried all of the above combinations when I was in the Institute. In retrospect, I should have written a script for each of them and tried it.

It took some time for me to write as I was referring some books and websites. I am sure there is somebody who is more adept with these technical matters and put the same in simple manner. I stand corrected if there are any mistakes.








I am sure there will be others who would be adept in answering the same doubt in a much simpler
User avatar
Paramvir Singh
Hero Member
Hero Member
Posts: 677
Joined: Thu Nov 23, 2006 12:54 pm
Location: Mumbai
Contact:

Re: Variable Shutters and Flickers

Post by Paramvir Singh »

thanks Nikhil, Aravindan and Karthik. Am recovering from a bad illness so havent read the thread carefully but will definitely do so and reply. in the meanwhile, i am posting the document with its tables in place.

note, you can also view the document in full screen mode

joyfulshankar
Newbie
Newbie
Posts: 1
Joined: Mon Feb 28, 2011 10:12 pm

Re: Variable Shutters and Flickers

Post by joyfulshankar »

will HMI flicker noticed if we are using video???? since video runs at 25fps and it will be in sync with the supply say 50hz.....wat are the safe running speeds for 50hz????
User avatar
Paramvir Singh
Hero Member
Hero Member
Posts: 677
Joined: Thu Nov 23, 2006 12:54 pm
Location: Mumbai
Contact:

Re: Variable Shutters and Flickers

Post by Paramvir Singh »

Only if your HMIs are flicker free and the generator is crystal controlled. Joyfulshanker, your account is being suspended for not being from FTII. Sorry but this is a closed group.
Post Reply