Topics

High light output for 2000 FPS - advice required

DPmatlosz
 

First off:   think for a second, how would I light this same shot at 24fps?...  Now,   use bigger sources.  it's that simple.  So in this case, your talent it too close to the BG, You only front lit your subject,  there is no back light (all good lighting starts toward the lens)  I have lit bigger scenes with smaller and less lights and got great results. https://vimeo.com/channels/highspeedcinema   
 Too often camera folks throw out everything they have ever learned or known when they light for High speed.  I have heard camera folk profess, that you need to blast the talent with massive amounts of front light. Now unless that is a style you are trying to achieve, not sure why one would approach anything that way. In the end and again, light like you light 24fps, and you'll be fine.  

Best: 
Jimmy Matlosz
DP, LA, CA, ID

Simon Bishop
 

If there was ever a video that proved that ‘the sound is half of the picture’ then this video is it ;-)

 

It is crying out for an appropriate sound effect to make the picture ‘sing’. I have many friends in audio post production who could help to make this video so much more engaging ;-)

 

Regards to all, Simon Bishop, Sound recordist and led lighting junkie, currently in Manchester…

 

From: cml-lighting@... [mailto:cml-lighting@...] On Behalf Of steadirob@...

 

I managed to find that footage again, there is now a short video on Vimeo, password HmiArc


https://vimeo.com/269428012

steadirob@...
 
Edited

I also have some other footage that shows that the arc wander is often more visible on 25 to 100 fps footage, as it shows up to the eye as a real flicker. On high speed the fluctuations are spread out and less noticeable. These test were made after complains of flicker at normal shooting speeds up to 100-200 fps. 
On the reflector of the ArriMax is clearly visible that the intensity and the color of the light is varying.
However, since this footage clearly shows some brand names, I am hesitant to make it visible... The company agreed to replace these lamps and while the problem comes back now and then, it was apparently a bad batch.....

Edit: forgot to sign:
Rob Erik van Gelder
Head of maintenance
Lighthouse Film Service
Bangkok, Thailand.

steadirob@...
 

It was shot a few years ago, so I do not have all details ready anymore. I believe the slowest and most close-up arc was at 7500 fps (windowed to a small area on the Phantom chip to reach that frame rate)  Most other shots are on 1000 or 2000 fps fps. (also windowed)

Phil Badger
 

Thanks for posting that!  Very cool to actually see it like that.  What was the camera frame rate again?? 

Phil Badger
Gaffer Los Angeles

steadirob@...
 

I managed to find that footage again, there is now a short video on Vimeo, password HmiArc


https://vimeo.com/269428012

AJ
 

Rob will you
PLEASE post this footage- would be amazing to see the phenomena you are speaking about!

Arc-wondering,
Alan Jacobsen
DP -NYC

On May 10, 2018, at 10:53 PM, steadirob@... wrote:

Phantom 1000 fps footage of the arc moving all over the place between the electrodes. (1000 fps, small shutter angle, long zoom lens and 2x ND 1.5, so 10 stops of ND

Marty Oppenheimer
 

I believe the answer is that an HMI is NOT an arc light.  It is a discharge light.  Unlike a carbon arc, DC, constant arc, an HMI is discharging at (if I'm correct) twice the hertz rate.  To the eye, it is a constant light.  To rotating shutters  or short exposure cameras, it is on off on off on off.

Marty Oppenheimer
Oppenheimer Camera Products
Seattle


On 5/10/2018 7:39 PM, Jonathon Sendall wrote:
Which is why with HMI’s at high frame rate you can get arc wander. It doesn’t happen all the time though. Does anybody know the reason for this? Age? Type of arc?

Jonathon Sendall
DP, London
On Fri, 11 May 2018 at 03:27, Glenn Lee Dicus <glenndicus@...> wrote:
I would also add, HMI’s are Arc lights, there is no filament. The light has a light emitting arc across the anode and cathode, or it doesn’t.

Glenn Dicus
La based Loader

Mark Weingartner, ASC
 


On 10May, 2018, at 19:53 10, steadirob@... wrote:

We traced it down to Arc-wander, which is X times amplified by all the tiny mirrors from the parabolic reflector. These are all small flat shiny surfaces and any small movement of the ark in whatever direction is immediately projected/mirrored in the light output. And there is almost never a stable arc. 

Rob brought up a great point and an extreme example…  the segmented reflector offers really high efficiency… especially when shooting “normal” frame rates when the arc wander is likely to even out over the 1/50th of a second exposure time…   some older, less efficient reflector/lens combinations are  less prone to this.  If you are ganging up multiple lights and using a bit of diffusion you can average out the arc wander.

Weingartner
LA



steadirob@...
 

Aside from all the information provide before I have one more flicker related issue that now comes up more frequently .

Several DP's want to shoot high-speed with the new Arri-Max type lamps because of their high output. We experienced several complaints about "flicker" , even with the 300 Hz HS-ballasts for the 18K Arrimax. 
Changing lamps,, ballasts, head-feeder cable etc did not help.
We traced it down to Arc-wander, which is X times amplified by all the tiny mirrors from the parabolic reflector. These are all small flat shiny surfaces and any small movement of the ark in whatever direction is immediately projected/mirrored in the light output. And there is almost never a stable arc. 
We had a particularly arc-wander-prone batch of 12K lamps from a respected manufacturer who agreed to take them all back after I showed Phantom 1000 fps footage of the arc moving all over the place between the electrodes. (1000 fps, small shutter angle, long zoom lens and 2x ND 1.5, so 10 stops of ND).

It was very interesting to see the glowing plasma jumping from one electrode to the other, changing color when the glow fades. But what was also visible is a  larger glowing gas-mass surrounding the actual plasma bridge and this has a shape of a candle flame. Due to turbulence inside the gas this "flame" is constantly moving, waving side to side or in some cases even turning around the plasma bridge. Then due to this moving a-symmetric body of glowing gas, the light output and its color constantly fluctuates.

So we advise to use diffusion when using HS-12/18K Arrimax  (or any other Arrimax-type), never use the naked lamp.
Many DP's do not like to hear this.... 

Rob Erik van Gelder
Head of Maintenance
Lighthouse Film Service
Bangkok, Thailand

Mark Weingartner, ASC
 


On 10May, 2018, at 19:39 41, Jonathon Sendall <jpsendall@...> wrote:

Which is why with HMI’s at high frame rate you can get arc wander. It doesn’t happen all the time though. Does anybody know the reason for this? Age? Type of arc? 

You can get arc wander with both AC arcs such as HMIs and DC arcs such as short-arc Xenons as found in projection lamphouses, follow spots,  and the parabolic reflector xenon lights without which music videos and TV commercials would have looked very different in the late eighties and early nineties.

As the electrodes heat up and wear, you can get varying resistance at different points so the “anchor” of the arc can move a little bit, but you can also get the arc between the two electrodes moving a little bit due to differential heat moving the gas around in the globe.
Longer arcs have more room to wander than shorter arcs but how much the wander affects light output has as much to do with reflector/lens design as it does the globe itself.   With a very specific polished parabolic reflector if the arc is moving in and out of the focus of the parabola the effectiveness of the reflector to deliver a parallelish beam can vary a lot.

With a fresnel lens with the typical spherical reflector as seen in older design HMI fresnels the same amount of arc wander is likely to make less of a difference in the field brightness.

Sometimes the overall brightness will change - sometimes the amount of light will stay relatively constant but different parts of the field will alternately get brighter and dimmer as the arc moves relative to the reflector.   Even fairly light diffusion can go a long way to evening out arc wander and if you are bouncing the light or pushing through heavy diffusion you will probably eliminate its effect entirely.

As there are a number of variables described above (as well as other variables discussed and not discussed, as always YOUR MILEAGE MAY VARY

Mark Weingartner, ASC
DP vfx etc 
still on the road
back on the computer instead of doing my work
Shhhh… don’t tell anyone

Jonathon Sendall
 

Which is why with HMI’s at high frame rate you can get arc wander. It doesn’t happen all the time though. Does anybody know the reason for this? Age? Type of arc?

Jonathon Sendall
DP, London

On Fri, 11 May 2018 at 03:27, Glenn Lee Dicus <glenndicus@...> wrote:
I would also add, HMI’s are Arc lights, there is no filament. The light has a light emitting arc across the anode and cathode, or it doesn’t.

Glenn Dicus
La based Loader

Glenn Lee Dicus
 

I would also add, HMI’s are Arc lights, there is no filament. The light has a light emitting arc across the anode and cathode, or it doesn’t.

Glenn Dicus
La based Loader

Mark Weingartner, ASC
 

That doesn't work on tech scouts.

It also doesn't work if you have no waveform, or a low res waveform in the monitor.

Some types of flicker aren't visible on low latency on-set displays.
Sometimes I lose jobs because I am not willing to do them without the tools that I need to be comfortable.
Someone else will take the job - sometimes they get lucky and sometimes they fail.
Their tolerance for failure may be higher than mine… Other than riding a motorcycle, climbing, filming off of the tops of high buildings and structures, and being freelance for more decades than I would like to admit, I’m actually pretty damned risk-averse.

Some jobs the cliuents are willing to try and see what works - for most of mine they want assurances that everything will work and for that I have to demand the tools I need to be able to make that assurance.




If you have the tools to analyze the problem you can.
If you don’t have the tools you can’t.

Personally I don’t think you should be using a Phantom or similar digital camera at 2000fps without a proper waveform monitor…

FYI if you play back at 24fps or 30 fps even the crappiest waveforms will show you the bounce… as long as it has enough resolution if the pulse is very confined in frame

If you can afford the lights and the Phantom you can afford the wfm. Some of our job involves educating our clients as to the
downside risks of their cost-management strategies.

One of the things we do to control flicker, ripple etc with Phantom cameras is choose different shutter angles (exposure durations) so as to avoid beat frequencies…
That you can’t really do definitively until you have the camera.

If it were easy all our relatives would be doing it too;-)

Weingartner
LA

Art Adams
 

Now that we have electronic cameras and waveform monitors it is easy to play back high speed footage and watch for a bounce on the WFM.
 
That doesn't work on tech scouts.

It also doesn't work if you have no waveform, or a low res waveform in the monitor.

Some types of flicker aren't visible on low latency on-set displays.

--
Art Adams
Director of Photography
San Francisco Bay Area

Bruce Aleksander
 

Great explanation from Mark W. regarding flicker from incandescent fixtures, and why the Maxi Brutes were safe at high frame rates. The only thing I’ll add is that the phenomena that keeps them from flickering is called thermal inertia. Their light is incandescent, and thermal inertial prevents the light from dipping perceptibly during the break in between the frequency cycles. The more robust and larger the filament, the greater their thermal inertia. Low wattage lamps cool down more between cycles than higher wattage lamps. There’s still a measurable dip, but it’s nearly imperceptible, except to sensitive flicker metering. 
I hope that helps add to the understanding. 

Bruce Aleksander 
Lighting Director
Seattle, Washington

Mark Weingartner, ASC
 

On 10May, 2018, at 10:20 23, John Roche <nycgaffer@...> wrote:

Mark when you say source 4 spot lights which do you mean and was this an HPL750 bulb?
regular ellipsoildals and the earlier 575 globes believe it or not
WEINGARTNER




TEST TEST TEST YOUR MILEAGE MAY DIFFER

John Roche
 

Hey guys how much is the new version... I’m going to guess more than any sane gaffer would spend. Still hurting from Sekonic 700

John Roche
C: 917-566-6256
F: 212-580-7237

On May 10, 2018, at 1:29 PM, Art Adams <art.cml.only@...> wrote:

Mark, I own the older version of that meter as well. What do you find to be tolerable limits for flicker when using it?

Thanks- Art

--
Art Adams
Director of Photography
San Francisco Bay Area

John Roche
 

Mark when you say source 4 spot lights which do you mean and was this an HPL750 bulb?

John Roche
C: 917-566-6256
F: 212-580-7237

On May 10, 2018, at 1:04 PM, Mark Weingartner, ASC <vfxmark@...> wrote:

On 10May, 2018, at 09:15 17, Art Adams <art.cml.only@...> wrote:

Mark suggested not bothering with DC when using Maxibrutes because, for some reason, they don't flicker.

Flicker happens when the filament cools between AC cycles. His theory is that, as they are sealed beam lamps, the heat is contained within the bulb so the filament doesn't have a chance to significantly cool between cycles. He said he's never had a problem with Maxis flickering at high speeds.

My gaffer went with Maxis and AC power on his shoot and had no issues at 1000fps. 


Coming in late to the game as usual

Just a bit of a disclaimer from a quoted so-called “expert” (me)

I had always been careful when using Raybeams (1kPar x30)  Dino Lights1kPAR x 24) or Maxi Brutes to use three phase power and to use sources from (ideally) all three phases from each direction.

In the case of RAY lights and Dino Lights they  were often built to work from 3 phase power so a single source was automatically using all three phases.  I found myself as a Phantom Tech on a shoot with single phase maxi-brutes in the shot and low and behold they did not flicker even when dimmed a little bit.  I then tried a few single PARs at about 600fps and was surprised that I saw no exposure ripple at all.

This was in conflict with the accepted wisdom in the 110volt world that anything under 5k (or some say 2k) was going to flicker.

I subsequently did some very high frame rate work and tried  using AC powered ETC Source 4 spotlights.  Lo and behold!  No exposure ripple!



I know everyone calls it flicker but the “flicker” we got from HMIs with bad capacitors is different from the pulsing we get from HMI’s set at the wrong frame rate/shutter angle  and both of those are different from the exposure ripple that we get when tungsten filaments heat and cool.

So what’s going on?

My theory is that in the case of both 1k or 1200w PAR 64 lamps  the CC-typed filament (coiled-coiled… a coiled helix)  does not radiate heat readily, and that the reflector coupled with the glass front lens of the lamp stops the inside quartz glass envelope from cooling fast enough to make enough heat differential between the envelope and the filament for the filament to cool down enough for us to see it.    I have not done extensive testing because it seems to work in practice.  It appears that the very very compact filament in tungsten Source 4 spotlight surrounded by all that glass and metal of the fixture acts the same way.

By the way, you can test this without having a Phantom camera if you have can find someone with an oscilloscope.   If you take a silicon blue photocell (available at any hobby electronics kit sort of place)   and hook it up to the oscilloscope you should be able to see the 120 Hz ripple  (100Hz in 50 cycle countries)


WARNING!!!
We who are working in 120v 60Hz countries  are safer from ripple than you who are working in 240v 50Hz countries.

Why?   Two reasons:
1. 60Hz vs 50Hz the ripples are closer together, so the filament has less time to cool between max voltage peaks
2. 120v vs 240v (or 220 or 240 or whatever)    FOR A GIVEN WATTAGE, your lamp is passing half the amperage… the filament is either thinner or twice as long and can cool faster by radiating heat from more surface area relative to overall mass of the filament.

Conclusions?
If you are using 115v PAR 64 lamps in fixtures that run them in series as many do in 240v countries you will have the same net cooling effect as we do in our 115v world BUT your voltage ripples about 17% slower so the filament has longer time to cool off.

If you are using lamps (bulbs, globes)  designed for 240volt then they are also possibly prone to cooling faster due to  more surface area or length for the same wattage (designed to pass half the current for the same wattage)      I have not done a side-by-side test  - this is only an hypothesis at this point but I have gotten away without flicker with some 2kw fresnels in the USA that I bet would cause flicker in a 240v 50Hz situation

So - be careful what advice you follow on this sort of thing… test whenever you can.
I own and use an older version of the B&S frequency meter described in this link

By looking at the percentage ripple of the light you  can get a decent prediction as to whether you will have a problem without having the Phantom camera available.

NOTE:  Even if you are shooting 1000fps  you only need to know the ripple of your light source which will be either 100Hz or 120Hz  (two times the mains frequency) when looking at this issue for high speed photography with tungsten sources.  This meter can also help you diagnose an HMI with a ballast problem if the ripple is excessive (especially compared to other HMIs in your set-up.

I have no connection with the company that makes the meter - I paid retail for mine, and I have their older model but I am too cheap to buy the new one.

I hope this long rambling complicated post serves to point out that there are several variables to pay attention to, not just wattage.

And keep in mind the issue of whether a given HMI will flicker at a high speed frame rate is different from whether a tungsten filament light will ripple at a high frame rate…
AC arcs like HMIs  completely turn off twice per power cycle… the capacitors and coils in the ballast or other fancy electronica try to minimize the time the light is off but it does literally produce no light at all twice per cycle whereas  the tungsten filament pulses more or less light depending on where it is in the power cycle.

If you can find a DC generator with tungsten lights the problem goes away…  
If you can find a DC rectified Xenon light the problem also goes away.
Companies like Luminys have made lights specifically designed for high speed work for years and years. (I like those guys but they didnt pay me to say that either)

Mark Weingartner, ASC
LA-based DP & vfx supervisor
out of town again but not for long
+1 818 970 6833










Mark Weingartner, ASC
 

On 10May, 2018, at 10:29 44, Art Adams <art.cml.only@...> wrote:

Mark, I own the older version of that meter as well. What do you find to be tolerable limits for flicker when using it?I don’t have a really good answer… depends on how prominent the source and the contrast overall… I mostly use it (and rarely now) to disqualify faulty lights more than to qualify ripple percentages.
Now that we have electronic cameras and waveform monitors it is easy to play back high speed footage and watch for a bounce on the WFM.

Best
Weingartner
on the road
gotta get off the computer

Art Adams
 

Mark, I own the older version of that meter as well. What do you find to be tolerable limits for flicker when using it?

Thanks- Art

--
Art Adams
Director of Photography
San Francisco Bay Area

Mark Weingartner, ASC
 

On 10May, 2018, at 09:15 17, Art Adams <art.cml.only@...> wrote:

Mark suggested not bothering with DC when using Maxibrutes because, for some reason, they don't flicker.

Flicker happens when the filament cools between AC cycles. His theory is that, as they are sealed beam lamps, the heat is contained within the bulb so the filament doesn't have a chance to significantly cool between cycles. He said he's never had a problem with Maxis flickering at high speeds.

My gaffer went with Maxis and AC power on his shoot and had no issues at 1000fps. 


Coming in late to the game as usual

Just a bit of a disclaimer from a quoted so-called “expert” (me)

I had always been careful when using Raybeams (1kPar x30)  Dino Lights1kPAR x 24) or Maxi Brutes to use three phase power and to use sources from (ideally) all three phases from each direction.

In the case of RAY lights and Dino Lights they  were often built to work from 3 phase power so a single source was automatically using all three phases.  I found myself as a Phantom Tech on a shoot with single phase maxi-brutes in the shot and low and behold they did not flicker even when dimmed a little bit.  I then tried a few single PARs at about 600fps and was surprised that I saw no exposure ripple at all.

This was in conflict with the accepted wisdom in the 110volt world that anything under 5k (or some say 2k) was going to flicker.

I subsequently did some very high frame rate work and tried  using AC powered ETC Source 4 spotlights.  Lo and behold!  No exposure ripple!



I know everyone calls it flicker but the “flicker” we got from HMIs with bad capacitors is different from the pulsing we get from HMI’s set at the wrong frame rate/shutter angle  and both of those are different from the exposure ripple that we get when tungsten filaments heat and cool.

So what’s going on?

My theory is that in the case of both 1k or 1200w PAR 64 lamps  the CC-typed filament (coiled-coiled… a coiled helix)  does not radiate heat readily, and that the reflector coupled with the glass front lens of the lamp stops the inside quartz glass envelope from cooling fast enough to make enough heat differential between the envelope and the filament for the filament to cool down enough for us to see it.    I have not done extensive testing because it seems to work in practice.  It appears that the very very compact filament in tungsten Source 4 spotlight surrounded by all that glass and metal of the fixture acts the same way.

By the way, you can test this without having a Phantom camera if you have can find someone with an oscilloscope.   If you take a silicon blue photocell (available at any hobby electronics kit sort of place)   and hook it up to the oscilloscope you should be able to see the 120 Hz ripple  (100Hz in 50 cycle countries)


WARNING!!!
We who are working in 120v 60Hz countries  are safer from ripple than you who are working in 240v 50Hz countries.

Why?   Two reasons:
1. 60Hz vs 50Hz the ripples are closer together, so the filament has less time to cool between max voltage peaks
2. 120v vs 240v (or 220 or 240 or whatever)    FOR A GIVEN WATTAGE, your lamp is passing half the amperage… the filament is either thinner or twice as long and can cool faster by radiating heat from more surface area relative to overall mass of the filament.

Conclusions?
If you are using 115v PAR 64 lamps in fixtures that run them in series as many do in 240v countries you will have the same net cooling effect as we do in our 115v world BUT your voltage ripples about 17% slower so the filament has longer time to cool off.

If you are using lamps (bulbs, globes)  designed for 240volt then they are also possibly prone to cooling faster due to  more surface area or length for the same wattage (designed to pass half the current for the same wattage)      I have not done a side-by-side test  - this is only an hypothesis at this point but I have gotten away without flicker with some 2kw fresnels in the USA that I bet would cause flicker in a 240v 50Hz situation

So - be careful what advice you follow on this sort of thing… test whenever you can.
I own and use an older version of the B&S frequency meter described in this link

By looking at the percentage ripple of the light you  can get a decent prediction as to whether you will have a problem without having the Phantom camera available.

NOTE:  Even if you are shooting 1000fps  you only need to know the ripple of your light source which will be either 100Hz or 120Hz  (two times the mains frequency) when looking at this issue for high speed photography with tungsten sources.  This meter can also help you diagnose an HMI with a ballast problem if the ripple is excessive (especially compared to other HMIs in your set-up.

I have no connection with the company that makes the meter - I paid retail for mine, and I have their older model but I am too cheap to buy the new one.

I hope this long rambling complicated post serves to point out that there are several variables to pay attention to, not just wattage.

And keep in mind the issue of whether a given HMI will flicker at a high speed frame rate is different from whether a tungsten filament light will ripple at a high frame rate…
AC arcs like HMIs  completely turn off twice per power cycle… the capacitors and coils in the ballast or other fancy electronica try to minimize the time the light is off but it does literally produce no light at all twice per cycle whereas  the tungsten filament pulses more or less light depending on where it is in the power cycle.

If you can find a DC generator with tungsten lights the problem goes away…  
If you can find a DC rectified Xenon light the problem also goes away.
Companies like Luminys have made lights specifically designed for high speed work for years and years. (I like those guys but they didnt pay me to say that either)

Mark Weingartner, ASC
LA-based DP & vfx supervisor
out of town again but not for long
+1 818 970 6833










Art Adams
 

>I was under the assumption that using an AC powered light under 5K would lead to flicker on the Phantom at that speed

A year or two ago my regular gaffer had a question about this, and I asked our own Mark Weingartner if the two of them could chat. They did, and Mark suggested not bothering with DC when using Maxibrutes because, for some reason, they don't flicker.

Flicker happens when the filament cools between AC cycles. His theory is that, as they are sealed beam lamps, the heat is contained within the bulb so the filament doesn't have a chance to significantly cool between cycles. He said he's never had a problem with Maxis flickering at high speeds.

My gaffer went with Maxis and AC power on his shoot and had no issues at 1000fps. That was a learning experience for me too, as I'd always used rectified DC for Maxis in the past when shooting high speed, and I was surprised to hear this. (I wasn't on that shoot.)

Small fresnels are different, as are open face units. Sealed beam units, though, don't seem to have this issue.

--
Art Adams
Director of Photography
San Francisco Bay Area

Jignesh Jhaveri, Photographer + DOP, Mumbai, India.
 

Thank you everyone for your suggestions and advice.

@ Ted - was limited by set size :-( and initially they wanted a wall shadow.

@ Cleverson - sunlight is a great idea! 

@ Pablo - good to know, we needed 4K for a digital zoom in / out in post

@ Seba and Mark - thanks - we went with Opal to get crisper light.

Jignesh Jhaveri, Photographer + DOP, Mumbai, India.
 

Thank you very much Art, you make some very good points :-)

This is a watch commercial, and the reason for high speed was a Zero Gravity concept. Always #OnTrend with Titan - May 2018

We were following the lighting reference of the Press / Still Image campaign, which I had lit with just one hard source and a fill. Unfortunately since this was being shot high speed, I needed to add more lights. 

I like the idea of using MaxiBrutes - but I was under the assumption that using an AC powered light under 5K would lead to flicker on the Phantom at that speed, and we don't seem to have DC powered ones here.

I wanted 5.6, as the talent was jumping, and a bit all over the place.

Kind Regards,
Jignesh Jhaveri
Photographer + DOP
Mumbai, India

Ted Langdell
 

On May 4, 2018, at 1:11 AM, Jignesh Jhaveri, Photographer + DOP, Mumbai, India. <jignesh.jhaveri@...> wrote:

I was disappointed with my setup for 2 reasons: 

1: Multiple shadows

Could you have pulled the talent far enough away from the wall to have the shadows fall only on the floor?

Ted

Ted Langdell
(530)301-2931

Dictated into and Sent from my iPhone, which is solely responsible for any weird stuff I didn't catch.


cleverson cassanelli
 

a 2000 fps with phanton in a job who it seens to be for fashion. on gray back, gray wall.

i will think like a fashion photographer, not as a cinematographer and....
will choose to shoot this 2000 fps outside, with sun light

offcourse, i dont know if it was possible for you. but that will be my first choice.

here is 3 frames in 240fps, sunlight with light grid butterfly or a big screen fabric to get down 2 stops of contrast
and bounce




here there is no option for sunlight, was a night shoot. 240fps too
one Maxibrut with butterfly for difusion and one T12 fresnel open focus for back light.
with a camera in a dolly


all shoots for the same film



Cleverson Cassanelli
DP/Sao Paulo/BR


2018-05-04 12:57 GMT-03:00 Pablo Vallejo <pablo@...>:

Hello Jignesh:

another tip
for velocities greater than 1000 fps the phantom flex 2k is better than the 4K


regards

Pablo Vallejo
Cinematographer

El 4 may 2018, a las 11:20, Mark Kenfield <mark@...> escribió:

To eliminate multiple shadows you need a denser diffusion, Opal's pretty light. I'd suggest Lite Grid Cloth (they called it "Half GC" when I was shooting Dehli) at a minimum. Lite Grid is great because it comes very close to the diffusion characteristics of Full Grid, but with much less light loss.

And the only way to close down your aperture is to throw more light at it (or shoot at a higher ISO).  

Cheers,

Mark Kenfield
Cinematographer

www.dreamsmiths.com.au
mark@...
0400 044 500

On 4 May 2018 at 18:11, Jignesh Jhaveri, Photographer + DOP, Mumbai, India. <jignesh.jhaveri@photolink.co.in> wrote:
Hello,

I recently had a shoot that required us to shoot at 2000 FPS and I used the following light:

2 x T24s, 2 x T12s which gave me an effective output of 60,000 watts as the bulbs inside were of the 20,000 and 10,000 variety.

I softened the light using an Opal diffuser and it was placed 20' (6 meters) away.

I was disappointed with my setup for 2 reasons: 

1: Multiple shadows
2: Not enough output 

<dummyfile.0.part>

<dummyfile.1.part>

Questions:

1: Is there a way to unify the light to reduce the multiple shadows? 

2: Whats the best way to have more quantum of light, I would have been happier shooting at a smaller stop - 5.6

Your experience and advice would be greatly appreciated.

Kind Regards,
Jignesh Jhaveri
Photographer + DOP
Mumbai, India





Pablo Vallejo
+34609112660
+34911733447
pablo@...
calle La Granja 28
Pol.Ind de Alcobendas, Madrid
www.el28studio.com






--
Cleverson Cassanelli

diretor de fotografia
11 96635 0917
São Paulo

Pablo Vallejo
 

Hello Jignesh:

another tip
for velocities greater than 1000 fps the phantom flex 2k is better than the 4K


regards

Pablo Vallejo
Cinematographer

El 4 may 2018, a las 11:20, Mark Kenfield <mark@...> escribió:

To eliminate multiple shadows you need a denser diffusion, Opal's pretty light. I'd suggest Lite Grid Cloth (they called it "Half GC" when I was shooting Dehli) at a minimum. Lite Grid is great because it comes very close to the diffusion characteristics of Full Grid, but with much less light loss.

And the only way to close down your aperture is to throw more light at it (or shoot at a higher ISO).  

Cheers,

Mark Kenfield
Cinematographer

www.dreamsmiths.com.au
mark@...
0400 044 500

On 4 May 2018 at 18:11, Jignesh Jhaveri, Photographer + DOP, Mumbai, India. <jignesh.jhaveri@...> wrote:
Hello,

I recently had a shoot that required us to shoot at 2000 FPS and I used the following light:

2 x T24s, 2 x T12s which gave me an effective output of 60,000 watts as the bulbs inside were of the 20,000 and 10,000 variety.

I softened the light using an Opal diffuser and it was placed 20' (6 meters) away.

I was disappointed with my setup for 2 reasons: 

1: Multiple shadows
2: Not enough output 

<dummyfile.0.part>

<dummyfile.1.part>

Questions:

1: Is there a way to unify the light to reduce the multiple shadows? 

2: Whats the best way to have more quantum of light, I would have been happier shooting at a smaller stop - 5.6

Your experience and advice would be greatly appreciated.

Kind Regards,
Jignesh Jhaveri
Photographer + DOP
Mumbai, India





Pablo Vallejo
+34609112660
+34911733447
pablo@...
calle La Granja 28
Pol.Ind de Alcobendas, Madrid
www.el28studio.com



karelbata@...
 

Nice informative post Art!

Last time I did this I used 3/4 backlighting from three 5k fresnels and added sheets of poly for fill. (In 3D too - which is why the only CU I can find here is a red/cyan anaglyph) I wanted a hard-edged night-time look (and it's meant to be a bit stylised), but it would have been easy enough to juggle it around to get more bounced fill for a more daytime look.

Shot at 500 - 1000 fps. You shot at 2000? That must have been really slow! Go to 1000 and you gain a stop!

A neutral gray wall, as you had, would perhaps need a little extra light thown on it. But did you want it so flat? I would have used a gobo or something to break it up, or perhaps have a diagonal slash of light? Anyway, something.

Stills are a bit contrasty. The original was more natural. (And she did have blue hair! Looked cool.) A blog on the shoot here





Seba Mendelberg
 

Mayte bring the whole light pack from one side.
Unify the pack with a butterfly with hilite or half soft frost 12x12ft.
Use 12x12 butterfly with ultrabounce for the other side of the triangle.

Best,
--

Sebastián Mendelberg
Gaffer,
Argentina.
+54 11 5600 2627
http://m.imdb.com/name/nm1875287

Mark Kenfield
 

To eliminate multiple shadows you need a denser diffusion, Opal's pretty light. I'd suggest Lite Grid Cloth (they called it "Half GC" when I was shooting Dehli) at a minimum. Lite Grid is great because it comes very close to the diffusion characteristics of Full Grid, but with much less light loss.

And the only way to close down your aperture is to throw more light at it (or shoot at a higher ISO).  

Cheers,

Mark Kenfield
Cinematographer

0400 044 500

On 4 May 2018 at 18:11, Jignesh Jhaveri, Photographer + DOP, Mumbai, India. <jignesh.jhaveri@...> wrote:
Hello,

I recently had a shoot that required us to shoot at 2000 FPS and I used the following light:

2 x T24s, 2 x T12s which gave me an effective output of 60,000 watts as the bulbs inside were of the 20,000 and 10,000 variety.

I softened the light using an Opal diffuser and it was placed 20' (6 meters) away.

I was disappointed with my setup for 2 reasons: 

1: Multiple shadows
2: Not enough output 





Questions:

1: Is there a way to unify the light to reduce the multiple shadows? 

2: Whats the best way to have more quantum of light, I would have been happier shooting at a smaller stop - 5.6

Your experience and advice would be greatly appreciated.

Kind Regards,
Jignesh Jhaveri
Photographer + DOP
Mumbai, India

Art Adams
 

There are times when multiple shadows don't bother me so much, usually when I'm shooting in white rooms and I want to create some interest. I'll stagger the lights at different heights and subtly change the hues with variations of CTO and CTB to emulate the feel of bounce light coming through windows. This works best when the lights are all coming from roughly the same direction. Once you cross them, and get shadows going different directions (left and right, instead of all left or all right) it looks lit.

That's what you have here. You've done a traditional key/fill setup with multiple shadows, and honestly it just looks lit.

What comes to mind:

I don't know why you kept your lights so far back. I'd get them as close as possible, through diffusion. I'd probably add a diffusion "wall" at the end of your set, on the left side, using a 12'x8' frame of half grid (or full grid, if I could get away with it) that extended from the left corner of the set toward the lens. I'd stack the T24s vertically at the very left edge to create soft punchy light that scraped the wall and lit the subject from the side to give him a bit of modeling, and then use the T12s to fill the rest of the frame as soft key-side fill. I'd fill in the shadows with a bounce off frame right (8'x8' ultra bounce, maybe) and try to darken the wall a bit with a 4'x8' floppy on frame left.

I'd have the diffusion as close as I could to the subject while keeping the floppy just outside of frame as that does several things: it feels natural, like window light; shadows on the wall might still be noticeable but they'd have a much softer edge to them, which feels natural; and you'd get large specular highlights in the subject's skin. It would be very difficult to darken the top of the frame, as you've done here, but the light would feel real and less lit. If there were multiple shadows they'd all fall the same direction, which feels like window light, and if there were obvious distinct shadows you could stagger the heights of the lights to make them feel more random. You could darken the top of the frame in post or with a grad filter in front of the lens.

Honestly, though, I wouldn't have used fresnel lights at all for this, unless I used one T12 from as close to the set wall as I could, from the left side, to rake across the talent's body. A hint of hard light from the side brings clothes alive, and I assume that's what you're selling here. The hard light would have been barely brighter than the soft source, but it adds a bit of sparkle. I would still have lit the 4' of diffusion closest to the wall brighter than the rest, and then filled the frame overall as directional fill, but I'd have used MaxiBrutes. You don't need fresnels here: they're too inefficient and too hard. I'd use a wall of modulated soft light, and I'd do it with big, sloppy, efficient and bright lights through diffusion.

Oh, and I don't know why you'd want to shoot at T5.6. I'd shoot this at T2 if it gave me the quality of light that I wanted. It's a fairly flat shot, and I don't care whether the flat wall is in focus or not.

I don't have a lot of great pictures of the above technique in action, but here's the closet I have. It's not exact, but hopefully you get the idea.





--
Art Adams
Director of Photography
San Francisco Bay Area

Jignesh Jhaveri, Photographer + DOP, Mumbai, India.
 

Hello,

I recently had a shoot that required us to shoot at 2000 FPS and I used the following light:

2 x T24s, 2 x T12s which gave me an effective output of 60,000 watts as the bulbs inside were of the 20,000 and 10,000 variety.

I softened the light using an Opal diffuser and it was placed 20' (6 meters) away.

I was disappointed with my setup for 2 reasons: 

1: Multiple shadows
2: Not enough output 





Questions:

1: Is there a way to unify the light to reduce the multiple shadows? 

2: Whats the best way to have more quantum of light, I would have been happier shooting at a smaller stop - 5.6

Your experience and advice would be greatly appreciated.

Kind Regards,
Jignesh Jhaveri
Photographer + DOP
Mumbai, India