Is there a flicker issue using magnetic-ballast HMIs with the RED?

[Stephen Pruitt]

Hi all. . .

I see some HMIs on Ebay just now, but they are older Arrisun units with magnetic ballasts. Is there a flicker issue with these units when shooting 23.9 or 24fps? Are there any other issues with the magnetics that I should be aware of?

Thanks very much.

Stephen

[Barry Green]

Yes, you don't want to use magnetic HMIs with a rolling-shutter CMOS camera unless you're extremely careful about your shutter speeds. You won't be doing frame ramping with a magnetic HMI.

It doesn't "flicker" like a film camera does, instead the CMOS camera will get multiple rolling orange bands scrolling through the image. That doesn't happen with a high-frequency electronic ballast.

If you insist on trying it, don't use 1/48 shutter. That's just plain not allowed. You will probably be safe at 1/40 or 1/60 though.

[Stephen Pruitt]

Well, that's very good to know, Barry. . .

Geez, these HMIs are OBSCENELY expensive! Like $5500 EACH for a 1200W Arri from BHPhoto. Are there any options that are a lot less expensive???

Thank you very much.

Stephen

[M Taylor]

http://www.coollights.biz/hardlight-series-c-28.html

1200W HMI for $3200. The website says this light will be available from June 2nd, if you can wait that long. Richard Andrewski has been active on this forum in the past.

[Alexis Hanawalt]

Barry, where have you been all my life?

I ran into this on set this weekend, my first big Red shoot - as I posted here: http://www.reduser.net/forum/showthread.php?t=12213&highlight=pulses

Here is the sample footage I posted, the "rolling pulses" we now get to play with in post: http://www.rent-a-red.com/WavesSampleH264.mov

[Shawn Nelson]

I just shot a ton this weekend with a mag HMI, I have no idea what my Red was at (I was purely directing, yay!). Is there a way for me to find out shutter just from the R3D or the QTs?

[Barry Green]

Barry, where have you been all my life?
Over on DVXUser... :)


Here is the sample footage I posted, the "rolling pulses" we now get to play with in post: http://www.rent-a-red.com/WavesSampleH264.mov

Yep. It'll happen with any rolling-shutter camera. People have to educate themselves to the fact that the rolling shutter works differently from conventional video and film cameras, and you have to know what it will and won't do to your footage. For most circumstances it's fine, but for some it just plain isn't, and the sooner you find out about it and work around it the better off you'll be.

[Barry Green]

Geez, these HMIs are OBSCENELY expensive! Like $5500 EACH for a 1200W Arri from BHPhoto. Are there any options that are a lot less expensive???
Renting. New electronic-ballast HMIs are not inexpensive!

Cool Lights is making a series of inexpensive HMI/metal halide lights which work well. They're not Arri-class gear, they're inexpensive Chinese imports, but they're less expensive. I have one of the 150w units, cost me $390 and tested it with the Red and it worked fine unless you really cranked the shutter speed up to 1/1000 or so.

[Shawn Nelson]

Weird, so I just got word from my DP that my footage was shot 23.98 and 1/48 and yet my footage looks great, no rolling stuff.

[Alex Carr]

Whatever frame rate you shoot you have to adjust your shutter speed, this is not only in Red, it is with Varicam, HVX-200, and many, many more...

I have magnetic Ballasts with my 1200's and they work fine when I shoot certain shutter speeds, which does limit me...

[Brian Ferguson]

Weird, so I just got word from my DP that my footage was shot 23.98 and 1/48 and yet my footage looks great, no rolling stuff.

In film... There are "windows" with HMI's where you are in phase with the pulse of (North America) 60HZ and 24 frame-ish and 180 degree is sort of on the edge - 24 fps and 144 degree shutter with magnetic ballast is safer. They are not hard on/off windows they are sweet spots in the windows. 24 fps and 1/48th (180 degree) shutter are on the edge but usually safe. I would think that even with the rolling shutter it would be safe. Consult an ASC manual for this type of info, at least for a film based opinion. The good thing about RED is that it is WYSIWYG you can play back the reference QT movie and see what you have before you move on to the next scene.

I shot a film a while back with a Red at 23.98 and a Phantom at up 500 frames a second and beyond and there is no way to get even a square wave (electronic) ballast not to flutter at those frame rates. We were using 18k HMI's to get maximum light for high speed shooting. A talented editor had a plug-in that removed the strobing effects. The promo looked great no strobing after post. It is a matter of knowing your tools.

briferg

[Shawn Nelson]

Ooh, I must find the no-strobe plugin!

[Cüneyt Kaya]

We were using 18k HMI's to get maximum light for high speed shooting. A talented editor had a plug-in that removed the strobing effects. The promo looked great no strobing after post. It is a matter of knowing your tools.

briferg

which plug in....? for which software application?
could be useful.

[Richard Andrewski]

Magnetic or Electronic?

Head / ballast noise with electronic solutions is one major issue to think about. Can you live with a bit of noise because on some electronic ignitor / ballast solutions there is a bit more noise due to the high frequency the bulb is being driven at. I would categorize it as an electronic "whine". The nice thing about this higher frequency drive is the lack of flicker at most commonly used shutter speeds. Not all electronic ballast HMIs make "head noise", but some do. In most cases, if the head is at a normal position from the subjects and from where the microphones are, you'll never record or hear the sound.

Magnetic ballasts do not have this problem as they drive the bulb at line frequency. But therein lies the issue being discussed here. They have less noise but flicker is definitely an issue at some shutter speeds. In addition, in my experience, magnetic ballasts do not give the same bulb lumen output that a good and efficient electronic ballast does. Put equivalent wattage magnetic and electronic driven systems side by side and compare brightness of output to see what I mean.

So, each one has its own benefits and inconveniences that dictate which one is best for you. For instance, if the HMI, the actor and microphone are all right next to each other, you may decide to use a magnetic ballast for that scene, otherwise, electronic ones will do fine for you with the good separation of the microphone / subject and HMI head.

I didn't notice if anyone else posted the HMI flicker formula but I'll show it here anyway. Apologies if its a repeat. This applies to anything with a 50 / 60 hz frequency such as a magnetic fluorescent or HMI ballast.

Safe camera speeds with 60hz lighting :

Camera speed in fps = shutter angle / 3 / LPEP

Where LPEP = number of light peaks that you want to capture per exposure
period (typically 2 per cycle, or hz).

Safe shutter angles with 60hz lighting :

Shutter angle = speed in fps * 3 * LPEP

For 50hz power replace 3 with 3.6 in both formulas.

What all this means is that if you don't capture at least two cycles of light output during your exposure then you start to get flicker.

So for example in 50hz situations and assuming you want 30fps it would be as follows to figure out shutter adjustment:

30*3.6*2 = 216

Many may think its better to run at a faster shutter speed but it's the opposite. Think about it like this: Each shutter moment is a moment where we're capturing light. If you're light is flickering that means that its not refreshing at a very high rate. 50 to 60 times per second is not that much (a good fluorescent electronic ballast is refreshing at 20,000 to 40,000 times per second and a good HMI electronic ballast at up to about 150 times per second). You are practically guaranteed to see flickering if you're shutter is open at a moment the light is refreshing and not fully lit. The faster your shutter the more likely you are to be open at such a moment. Many feel that you need at least two refreshes of your light per shutter moment to be safe and not see flicker.

[Stephen Pruitt]

Thank you very much, Richard! Very informative. . .

. . . and also depressing.

:-(

Stephen

[Gunleik Groven]

Thanks.

If I am understanding this right, these are degrees, right (video background here...)

[Stephen Pruitt]

Yep. . . they are degrees. . . or, put another way, at 24fps, 144 degrees is the equivalent of a 1/60th exposure, whereas 216 degrees is 1/40th.

Now, exactly how much the look will deviate from using 144 or 216 relative to 180 degrees is something that I cannot address. I'm not a DP. Is there a noticeable difference?

Stephen

[Larry Gebhardt]

Hi Guys...I'm the lister on ebay with the HMI 1200 pars. I thought they were getting way above what they were worth...more than I paid for them used a few years ago. A bidder asked a question if they were "flicker less" and i responded by saying that they were NOT, as described in the listings. So, I began to worry that maybe people were not reading the listings, and did a google for HMI's and Flicker, and up came this very timely thread (I didn't happen to be on Reduser yesterday ot today...unusual for me)

So, this cleared it up a bit for me with regard to the Red camera and others...Thanks for all the information on this and as usual, the rest of the site. This is a really great forum to share and learn from.

The listings I have on ebay are Magnetic HMI's...so you need to shoot with a shutter spead that works...just like film cameras...in case anyone from here is bidding on them.

Thanks a lot

Larry

[Nick Gardner]

I have used magnetic ballast HMIs with the red with no problems. Can you randomly shoot whatever frame rates and shutter angles you want with no problems? NO. You can't do that with ANY camera. There are pros and cons to both Magnetic and electronic ballasts. Mag ballast are way less finicky than electronic ones. Electronic ones are smaller, lighter and you can do speed ramps. As long as you play by the rules, there should be no problems with magnetic ballasts.

Nick

[Stephen Pruitt]

Hi all. . .

There must be some reason (except effort and "aesthetics") why this isn't done:

Say we take a 4x4 piece of plywood. Say we mount 100 "bare bulb" sockets to the plywood (10 strips of 10 sockets). Say we put a switch on each strip. Say we put 100 14W compact fluorescent bulbs in the sockets. Bulbs are available in several different temperatures.

What have we created? A 100 bulb, 1400W fluorescent "fixture" with the light output of 6000 watts of tungsten? Or a worthless, unusable piece of experimental junk?

What say we?

Thanks much.

Stephen

[Alexis Hanawalt]

You'd be surprised how many people have built things like what you just described. No reason not to, other than the obvious compromises.

I built a set of knock-off Kino's for $250 a piece... ballasts are cheap and flickerless these days, and you can by high CRI florescent bulbs cheaper than Kino bulbs... they worked fine.

[Larry Gebhardt]

Well the two HMI 1200 Pars just sold...for a total of over $5000.00. That is twice what I thought they would bring.

Is it the same as with Zeiss primes and such, the numbers of Red cameras out there needing all sorts of production equipment, etc. driving prices up?

Just wondering (wish I had more 1200 Pars...although I will hold onto my Zeiss collection)

[Alexis Hanawalt]

I bought a lens a year and a half ago - it's worth 3x what I paid for it now.

[10s]

Demand is good, while it drives the prices up intially, the supply should catch up to meet demand. That's why there is an un-naturally high population ratio of chickens & cows to say, owls & mountain lions. An un-naturally high population of film related gear such as HMIs wouldn't hurt my feelings.

About the board with the 100 bulbs: It's about "throw" distance. FAYs are similar in that they took 9 PARS and put them together and they work great. They're been around since the 50s. FAYs have throw.

The HMI / shutter posting by Cool lights needs to be a sticky.

[Richard Andrewski]

Hi all. . .

There must be some reason (except effort and "aesthetics") why this isn't done:

Say we take a 4x4 piece of plywood. Say we mount 100 "bare bulb" sockets to the plywood (10 strips of 10 sockets). Say we put a switch on each strip. Say we put 100 14W compact fluorescent bulbs in the sockets. Bulbs are available in several different temperatures.

What have we created? A 100 bulb, 1400W fluorescent "fixture" with the light output of 6000 watts of tungsten? Or a worthless, unusable piece of experimental junk?

What say we?

Thanks much.

Stephen

Not worthless but something that has its own uses and its own strengths and weaknesses. Not the equivalent of just 6000w tungsten but the equivalent of (maybe 6000w depending upon how efficient the small ballasts are in the base of the bulbs) tungsten behind diffusion which limits its throw and is emanating from a large surface area so therefore very soft. You can't make the fluorescent bulbs hard no matter what you do. If you had a big enough one, you'd have something similar to a Fisher light but a cool version. Great for car commercials for instance where you want smooth and even lighting. No need for a fan either like with a fisher light.

[Richard Andrewski]

Thanks.

If I am understanding this right, these are degrees, right (video background here...)

Yes sorry degrees.

[Brian Ferguson]

which plug in....? for which software application?
could be useful.

I just sent an email to find out, I am pretty sure all of the editing was done in Final Cut.

[Cüneyt Kaya]

I just sent an email to find out, I am pretty sure all of the editing was done in Final Cut.

thanks

[M Taylor]

which plug in....? for which software application?
could be useful.


I just sent an email to find out, I am pretty sure all of the editing was done in Final Cut.

Not aware of a deflicker plugin for Final Cut Pro, but...

http://www.granitebaysoftware.com/Product_gbdeflicker.aspx

...this is an After Effects PC only plugin that is used for removing flicker in timelapse sequences. I haven't used it personally but, judging from the video tutorials on the site, it should work on HMI flicker.

If the Granite Bay plugin being PC only is a deal breaker, The Foundry's Tinderbox 1 DeFlicker plugin work on OSX, but you have to buy the whole bundle of 20 plugins - not cheap!

[Steve Freebairn]

Not aware of a deflicker plugin for Final Cut Pro, but...

http://www.granitebaysoftware.com/Product_gbdeflicker.aspx

...this is an After Effects PC only plugin that is used for removing flicker in timelapse sequences. I haven't used it personally but, judging from the video tutorials on the site, it should work on HMI flicker.

If the Granite Bay plugin being PC only is a deal breaker, The Foundry's Tinderbox 1 DeFlicker plugin work on OSX, but you have to buy the whole bundle of 20 plugins - not cheap!

That plugin is sweet! Should be built right into Ae.

[Nelson Goforth]

Say we take a 4x4 piece of plywood. Say we mount 100 "bare bulb" sockets to the plywood (10 strips of 10 sockets).


I'm sure it has been done. But then you've got to build a case for it, etc. That's the great thing about Kinos (and some others) - they are self-contained and to some degree self-encased. I could go out and build something similar, but it would be very expensive to try to custom build what Kino and others have done, and if it wasn't expensive it probably wouldn't last very long, or be heavy as hell.

"Throw" isn't the only criteria. What a lamp like this does is provide a very broad source that will fit in a very small space, without putting out a lot of heat. On Dumb and Dumber we used Jay Yowler's custom "Wall of Light" (24 tubes I think - a predecessor to his BFL) when we were backed up against the antique paneling in the bar at the Stanley Hotel. Talent could pass by a couple feet away and be well lit, and we didn't fry the finish on the paneling. The light array took a while to put together (lots of tubes with little space between them), but it could be carried and managed pretty easily.

[Dj Joofa]

That plugin is sweet! Should be built right into Ae.

I think that plugin is designed to remove the flicker effects *between* different frames, and perhaps not *within* the same frame. Methodology, does exist however, that can even take care of within-frame flicker effects.

[Obin Olson]

[QUOTE=Stephen Pruitt;208458]Hi all. . .

There must be some reason (except effort and "aesthetics") why this isn't done:

Say we take a 4x4 piece of plywood. Say we mount 100 "bare bulb" sockets to the plywood (10 strips of 10 sockets). Say we put a switch on each strip. Say we put 100 14W compact fluorescent bulbs in the sockets. Bulbs are available in several different temperatures.

What have we created? A 100 bulb, 1400W fluorescent "fixture" with the light output of 6000 watts of tungsten? Or a worthless, unusable piece of experimental junk?

What say we?



watch out, I have done it, they MUST be HIGH CRI bulbs or you will get muddy images!

my 2 cents worth just sayin

[Mike Prevette]

"worthless, unusable piece of experimental junk"

The green spikes in the CFL's are awful. I have not personally found any that are worth a damn. Not saying they don't exist, just saying I haven't found any (and I buy a lot of them).

The board would have to be painted white, and you would have to use it a considerable distance from your subject to avoid multiple shadows. Also HEAVY! YOur better off spending the $400 to make it, and the 6 hours of your time buying a nice used light off ebay.

[Jim Hoffman]

It seems like a ready made light using common CFLs would be okay for use on a small green screen set up? green spikes... who cares?

I also want to get back to the original question about the k5600s - Cools seem like good price per punch but it also seems like the k5600s have flexibility in their head design that would allow you to use them in a few different ways...no? perhaps a combo of both?

There was discussion of head noise with an electronic ballast. Can anyone say one way or the other if it is an issue with Cools or k5600s?

[Eileen Ryan]

I see some HMIs on Ebay …, but they are older Arrisun units with magnetic ballasts. Is there a flicker issue with these units when shooting 23.9 or 24fps? Are there any other issues with the magnetics that I should be aware of?

I won’t address the issue of flicker and frame rate/shutter angles because it is well established elsewhere in this forum that there are safe windows that are flicker free as long as the power supply is stable.

What I would like to add is that, based on recent research done by a Boston Gaffer I work with by the name of Guy Holt, if you don’t have access to the newest Power Factor Corrected (PFC) electronic ballasts, you are better served by using the older magnetic ballasts on an inverter generator (like the Honda EU6500is) over non-PFC electronic ballasts on conventional AVR generators (like the Honda EX5500 or ES6500.) I know this is contrary to the conventional wisdom, so I will explain his findings in more detail.

When electronic square wave HMI ballasts came on the market, they were at first thought to be the solution to all the problems inherent in running HMI lights on small portable generators. By eliminating the flicker problem associated with magnetic ballasts, they also eliminated the need for the expensive and ultimately unreliable AC governors required for flicker free filming with magnetic HMI ballasts and portable gas generators. Since they are not frequency dependent, it was thought at first that electronic square wave ballasts would operate HMIs more reliably on small portable generators – even those without frequency governors.

For this reason, as soon as electronic square wave ballasts appeared on the market, many lighting rental houses replaced the more expensive crystal governed portable generators with less expensive non-synchronous portable generators. The theory was that an electronic square wave ballast would operate reliably on a non governed generator and allow filming at any frame rate, where as a magnetic HMI ballast operating on an unreliably AC governed generator allowed filming only at permitted frame rates.

In practice, electronic square wave ballasts turned out to be a mixed blessing. Part of the problem with operating electronic HMI ballasts on portable gas generators in the past has to do with the purity of the power waveform they generate. With an applied voltage waveform distortion of upwards of 19.5%, conventional AVR generators (like the Honda EX5500 or ES6500) do not interact well with the leading power factor (current leads voltage) of the capacitive reactance created by electronic square wave HMI ballasts. The net result is harmonic currents are thrown back into the power stream, which results in a further degradation of the voltage waveform and ultimately to equipment failure or damage.

The oscilloscope shots of the power waveforms below is from an article Guy wrote for his company newsletter on the use of portable generators in motion picture lighting and is typical of what results from the operation of a 1200W HMI with non-power factor corrected ballast on grid power (left), on a conventional AVR generator (Honda EX5500) (middle), and inverter generator (Honda EU6500is)(right.) The adverse effects of the harmonic noise generated by non-PFC electronic ballasts and exhibited here in the middle shot, can take the form of overheating and failing equipment, circuit breaker trips, excessive current on the neutral wire, and instability of the generator’s voltage and frequency. Harmonic noise of this magnitude can also damage HD digital cinema production equipment, create ground loops, and possibly create radio frequency (RF) interference (for more details use this link (http://www.screenlightandgrip.com/html/emailnewsletter_generators.html).)


http://www.screenlightandgrip.com/images/generators/wwaveform_elec_ballast.jpg

As is evident in the oscilloscope shots below of a 1200W magnetic HMI ballasts on grid power (left), on power generated by a conventional AVR generator (middle), and power generated by an inverter generator (right), the lagging power factor caused by the inductive reactance of magnetic ballasts has by comparison only a moderately adverse effect on the power waveform. Outside of causing a voltage spike in the inverter power, magnetic ballasts actually show a positive effect on the already distorted power waveform of the Honda EX5500 conventional generator. For this reason magnetic ballasts work better on conventional generators with frequency governors than do non-PFC electronic square wave HMI ballasts.


http://www.screenlightandgrip.com/images/generators/wwaveform_mag_ballast.jpg

These oscilloscope shots show that if you don’t have access to the newest PFC electronic ballasts, the older magnetic ballasts are in fact cleaner running on portable gas generators than non-PFC electronic ballasts. And, where inverter generators like the Honda EU6500is do not require crystal governors to run at precisely 60Hz, you can operate magnetic HMI ballasts reliably on them. In addition, the smaller magnetic ballasts (575-2500W) offer the distinct advantage of being less expensive and draw less power (once they have come up to speed) than the commonly available non-PFC electronic equivalents (13.5A versus 19A for a 1.2kw.)

Of course there are downsides to using magnetic ballasts. One down side is that you are restricted to using only the safe frame rates and shutter angles. Another downside to magnetic ballasts is that you can’t load the generator to full capacity because you must leave “head room” for their higher front end striking load. When choosing HMIs to run off portable generators, bear in mind that magnetic ballasts draw more current during the striking phase and then they “settle down” and require less power to maintain the HMI Arc. By contrast, an electronic ballasts “ramps up”. That is, its’ current draw gradually builds until it “tops off.”

For example, even though a 2.5 magnetic ballast draws approximately 26 amps you will not be able to run it reliably on the 30A/120V twist-lock receptacle of a 6500W generator. As mentioned above, magnetic ballasts have a high front end striking load. So, you must leave “head room” on the generator to accomodate the strike. And, even though the twist-lock receptacle is rated for 30 Amps, conventional 6500W generators are only capable of sustaining a peak load of 27.5 Amps per leg for a short period of time. Their continuous load capacity (more than 30 minutes) is 23 Amps per leg. And, if there is any line loss from a long cable run the draw of a 2.5 magnetic ballast will climb to upward of 30 Amps. To make matters worse, as we saw in the oscilloscope shots above, the lagging power factor caused by the inductive reactance of the magnetic ballast kicking harmonic currents back into the power stream causes a spike in the supply voltage that can cause erratic tripping of the breakers on the generator or ballast. (for a more detailed explanation of why that is I, again, suggest you my newsletter article.) In my experience the load of a 2.5kw magnetic ballast is too near the operating threshold of an unmodified 6500W generator for it to operate reliably.


http://www.screenlightandgrip.com/images/HD_PP_Demo_Transformer-Distro.jpg
PFC 2.5 & 1.2 HMI Pars, PFC 800w Joker HMI, Kino Flo Flat Head 80, 2 ParaBeam 400s, and a ParaBeam 200 powered by a modified Honda EU6500is through a 60A Full Power Transformer/Distro

The only sure way to power a 120V 2.5kw (or even a 4kw) HMI magnetic ballast on a portable gas generator is from its' 240V circuit through a 240v-to-120v step down transformer like the one Guy’s company, Screenlight & Grip, manufactures for their modified Honda EU6500is. Their 60A Full Power Transformer/Distro will step down the enhanced 7500W output of their modified Honda EU6500is generator to a single 60A 120V circuit that is capable of accommodating the high front end striking load, and even the voltage spikes, of either a 2.5kw or 4kw magnetic ballast at 120V. For details about the benefits of using step down transformers with portable generator, read his article. It is available at www.screenlightandgrip.com/html/emailnewsletter_generators.html (http://www.screenlightandgrip.com/html/emailnewsletter_generators.html).

Eileen Ryan, Gaffer, Boston