Welcome to our community

Be a part of something great, join today!

Log in
  • Hey all, just changed over the backend after 15 years I figured time to give it a bit of an update, its probably gonna be a bit weird for most of you and i am sure there is a few bugs to work out but it should kinda work the same as before... hopefully :)

HMI - safe speeds

D

david_winters

Well-known member
Joined
Jan 27, 2007
Messages
512
Reaction score
0
Points
0
Location
Orange County, CA
On the Red One is it possible to shoot 23.98fps 1/48th with conventional magnetic ballasts (USA 60hz) and not have flicker problems? Based on the math this is not a safe speed as 120 does not evenly divide into 48. 1/40th or 1/60 shutter would be the proper procedure in this scenario, correct? Is anyone shooting 1/48th and not having problems?

Thanks
 
See here:
http://www.panavision.com.au/Information/Flickerfree60Hz.htm

Though 144 degrees is optimal for 24 fps crystal, you're safe at most shutter angles in terms of flicker. 144 degrees gets you two pulses per frame, 180 gets you something like 2 1/2 I think.

But "flicker" is the result of subtly different exposures per frame and you should be safe at 24 fps with most shutter angles.

Now 23.98 fps is a bit "off" from that safe 24 fps speed, but the amount of flicker (which can be measured on a waveform monitor) is so minor as to be more or less invisible.

However, below 72 degrees, you are capturing less than a full sine wave of an AC pulse, so depending on when you trigger the camera, you may be in sync with the peak or valley of the wave, creating some chance of underexposure, but not flicker.
 
As I recall, 48 fps will give you pretty severe flicker at 23.98 and 180 shutter. This from a previous post:

Page 42, THE 16MM CAMERA BOOK, says the formula is

safe fps = 4xLxA divided by 360xI

Where

fps is frames per second
L is line frequency, e.g. 50 or 60
A is shutter angle
I is any integer greater than zero

So let's say you've got a camera with 172.8 degree shutter angle in Europe and you want to generate a list of safe speeds, so

4 times 50Hz times 172.8 is 34,560, divided by 360 times 1 = 96 fps safe speed.

Another safe speed would be 34,560 divided by 360 times 4 = 24 fps.

Where this could fail is that the line frequency is not precisely controlled. Some countries have imprecise line frequency or you might be using a generator that has a varying frequency. I remember Cinetronics or Cinematography Electronics sold a frequency meter that you could use to test lights and monitors.

Of course now, this situation is not nearly so bad with a RED (as opposed to a film camera) because you can certainly shoot a take and watch the monitor for flicker. You can even turn the joystick knob when in the fps or shutter angle select menu and watch for flicker.
 
Here is what 23.98fps 1/48th (180 deg) with one magnetic 1200w LTM par looks like. Best viewed by scrubbing through clip.

http://www.innovatemedia.com/clips/winters/tests/hmi_flicker/23-98_1-48.mov

23.98 1/60th (180 deg)
http://www.innovatemedia.com/clips/winters/tests/hmi_flicker/23-98_1-60.mov

24 1/48th - The rolling bar is much more subtle and I could live with it on specific projects.
http://www.innovatemedia.com/clips/winters/tests/hmi_flicker/24_1-48.mov

I tested multiple frame rates with multiple heads & ballasts on mains power. Bottom line don't attempt to shoot 23.98 or 24 1/48 with magnetic ballasts. As the gentleman above have instructed, this is not a safe speed.

I also shot tests with electronic FF ballasts and experienced no flicker issues.

Thanks for the charts, links, and math.
 
Hey Dave thanks for posting the clips. I've been telling people about this since the day you showed me in your office. In addition to your last post, checking power is also a good thing to do. Like I told you, we had an issue with dirty power causing a flicker also. Clean power, clean footage.
 
Hey Weber,

So, we had the same rolling bar issue (flicker) on stage power and at the G&E rental house, so I know for sure power was not the issue. The ballasts and lamps went up to LTM for service and they were approved as in spec.

Bottom line it looks like you can "get away" with shooting 24fps 1/48th, as the rolling bar is very slow and subtle. However 23.98 1/48th is just enough out of sync that the rolling bar becomes more visible.

Thanks again man, spread the word. It's been years since I've made a flicker /FPS/shutter mistake. The last time was 1997 on 16mm, I hope this is the last one for me.
 
david_winters said:
Hey Weber,

So, we had the same rolling bar issue (flicker) on stage power and at the G&E rental house, so I know for sure power was not the issue. The ballasts and lamps went up to LTM for service and they were approved as in spec.

Bottom line it looks like you can "get away" with shooting 24fps 1/48th, as the rolling bar is very slow and subtle. However 23.98 1/48th is just enough out of sync that the rolling bar becomes more visible.

Thanks again man, spread the word. It's been years since I've made a flicker /FPS/shutter mistake. The last time was 1997 on 16mm, I hope this is the last one for me.
Click to expand...

Well you know, with newer technology you learn as you go. Can we make a "RED flicker free" chart?

What shutter angles did we figure at 23.98? Or did we NOT get to that? :hissyfit:
 
The charts referenced above are still relevant. I think it's as simple as taking the pulse rate of the ballast (60hz = 120) and divide by the shutter speed. If you get a full number then it's a safe speed. 120/48 = 2.5 not safe. 120/40 = 3 safe. 120/60= 2 safe. 120/30 = 4 safe.

I got burned because we didn't notice the rolling bar on set, and my last few RED w HMI jobs were at 29.97fps, so I missed it.

Thanks
Dave
 
I don't know if it's that big of an issue if you aren't speeding up the footage in post. I have shot hundreds of hours of footage with Mag HMIs at 23.98 and 180 degree shutter, and have never once noticed any flickering. Not with my eye, not on a waveform, never. Neither have any of the numerous editiors, of VFX guys working with the footage.

Obviously there is something going on from the test you did, but I don't know if it's of concern practically.

Thanks for the clips,

Nick
 
23.98 1/30, 1/40th, 1/60 are safe.

Nick, your response is what I was hoping to hear when I posted the first message. When I saw the first edit on my job, with the flicker issue, I immediately sent an email out to the stage, camera rental, grip electric rental, that we had a flicker issue and I wanted to round up all the gear and re-create the problem. I also called a few other DPs I know and trust to get their real world experience shooting magnetic with 23.98 180 deg shutter.

Their responses were consistent with yours, "I've been doing it for years, no issues." Then "the generator must have been off sync, the house mains was bad, the ballasts were out of spec, old hmi globes." After those items were ruled out I started to suspect the RED rolling shutter being more sensitive to phase shifts vs mechanical film shutters and global shutters on other 24p CCD based systems.

On this project we had a mix of magnetic and electronic so, some setups are more visible vs others. The only time it boned us was on a few multiple plate passes, where we had to roto and selective color correct various layers with blended masks to make it all line up. It wasn't really that big of a deal, but if I didn't have control of post on this job it would have been a different experience.

I hope this is a good heads up for others. Be careful shooting 23.98 1/48th (180 deg) as it's not a safe speed. It's hard to see on set, but much more visible in the bay, and when you CC dropping the midtones. It only caused problems for us on the shots with layering. From my tests 24.00 fps 1/48th is much less problematic.
 
I did two RED features last summer at 23.98 with 180 degree shutters and standard magnetic ballasts on overhead flos, etc. -- no flicker problems.

Like I said, the flicker is measurable on a waveform but generally considered too mild to be noticeable.

This came up years ago on my very first feature because we had a lot of 24-frame monitor playback, which is actually 23.976 fps, and we had a lot of magnetic ballasts on fluorescents in the sets, so I shot tests... and didn't see any flicker problems.

I have had flicker problems over the years though with bad ballasts, bad heads, bad globes, bad gennies, etc. -- I don't think you ever can completely eliminate the risk, but at least with digital, you should see problems live on the set monitors.

Now you say that 23.98 + 1/48th (180 degrees) is bad... but which is it, because 23.98 at 180 degrees isn't exactly 1/48th and 23.98 at 1/48th isn't exactly 180 degrees -- if 24 at 180 degrees is "good", which is 1/48th, then what is the goal at 23.98 -- to get to exactly 1/48th or to exactly 180 degrees?
 
David Mullen ASC said:
Now you say that 23.98 + 1/48th (180 degrees) is bad... but which is it, because 23.98 at 180 degrees isn't exactly 1/48th and 23.98 at 1/48th isn't exactly 180 degrees -- if 24 at 180 degrees is "good", which is 1/48th, then what is the goal at 23.98 -- to get to exactly 1/48th or to exactly 180 degrees?
Click to expand...

Wow ... Pretty Deep :)

For me, I would say 180 degrees ...... so what's the answer?
At 23.98 what is the safe shutter angle?
 
David Mullen ASC said:
Now you say that 23.98 + 1/48th (180 degrees) is bad... but which is it, because 23.98 at 180 degrees isn't exactly 1/48th and 23.98 at 1/48th isn't exactly 180 degrees -- if 24 at 180 degrees is "good", which is 1/48th, then what is the goal at 23.98 -- to get to exactly 1/48th or to exactly 180 degrees?
Click to expand...


Hi David,

Thanks for taking the time to reply. From the tests I shot, links to some of it posted in this thread, the "rolling bar" is more noticeable in the 23.976 fps 1/48th footage vs 24.000 1/48th. No rolling bar at 23.976 1/60th.

24.000 1/48th the rolling bar travels at a slower rate and there is slightly less exposure variation across the frame compared to 23.976 1/48th.

In the Red menu does 1/48th = 180 deg when recording at 23.976 or 24.000? I should probably drop the "degrees" from my descriptions as that sort of label is really only correct for a mecnical shutter, no?
 
Ok, re-reading this I better understand your question. My labels and slate are what was displayed in the RED menu. I don't know which is truly 1/48th sec.
 
I have found all of the CMOS cameras are very sensitive to flicker and 23.98 or 24 at 180 degrees is NOT safe. I always see flicker when I prep a Red at a rental house that has fluoros.

144 degrees is safe (1/60th second).

An easy way to test is to aim the camera at your light (or something brightly lit exclusively by it) and stop way down until the light looks dim in the monitor. If it is flickering you can see it better this way.
 
I've found this to be a good guide for the formulas behind the calculations:

http://www.psps.com/tips/pdf/psps_19.pdf

Someone should make that a sticky as it comes up often around here. According to calcs it should be 144 degrees but many say they don't find an issue with 180 degrees so what you would have to guess is that something is off in the ballast or bulb in some cases that makes it problematic.
 
You must log in or register to reply here.
Back
Top