HMI to Tungsten Lumen Efficiency

[Mark Kenfield]

Hi fellas,

 

Guy just blew my mind a bit on a Reduser thread about the upcoming 10kw Tungsten equivalent LED fresnel that Mole Richardson have coming out.

 

I'd always worked under the assumption that an HMI fixture is going to give you roughly 4x the light output per watt of an equivalent tungsten fixture. i.e. 1.2kw HMI = 5kw of tungsten, 2.5kw of HMI = 10kw of tungsten etc.

 

Now if you compare an 1800w M18 to a 2000w Arrilite 2000+ (both of which use arrimax open-face reflectors), this does appear to ring true. At 30' full flood (63° beam compared to a 62° beam, so very close) you're looking at very close to 4x the output from the M18 compared to the 2000+.

 

However looking through the photometrics for fresnels (on the Arri calculator), the HMIs don't seem to fair much better than 2x the lumen efficiency of their tungsten counterparts.

 

Now I almost never get to work with bigger fresnels (excluding occasional 5ks in studios) as everything's open faced lights into frames most of the time, so I've never really encountered this discrepancy before.

 

So what I'm wondering is, why does HMI lose so much of its efficiency advantage compared to tungsten when going through a fresnel lens?

 

Cheers,

Mark

[Phil Rhodes]

I'm not sure. On paper the luminous efficacy (to use the correct term) benefit of most kinds of modern gas-discharge light source, including fluorescent and HMI, is about 4:1 over tungsten, and this should hold if they're placed in identical optical assemblies. Are you completely sure they're using the same lens and reflector setups?

 

HMI is actually not that brilliant a technology by modern standards - it's fundamentally a 1960s mercury vapour lamp with tweaks - and things like ceramic metal halide can do better, but I'm as surprised by your findings as you are.

 

P

[Aapo Lettinen]

I guess it has something to do with the fact that HMI is a round point source whereas larger tungsten bulbs usually have the filament arranged to a flat rectangular configuration. thus the HMI can waste more light to the sides whereas the tungsten filament is more of a directional source

[Adrian Sierkowski]

I think a lot of this comes more from lamp/lens designs than anything else.

[andrew ward]

The fresnels are bigger in tungsten.

And youre probably comparing a tungsten fresnel to an hmi par with fresnel lens in? The reflectors arent optimised for fresnels. Sunray small hmis feel punchy cause they have glass mirror reflectors.

[andrew ward]

Btw everything should be fresnels. Its total DP wank that theyre not.

[Phil Rhodes]

PARs are generally considerably more efficient than fresnels, which is not so much wank as an objective fact.

[andrew ward]

Says the DP.

[Phil Rhodes]

I notice your occupation is listed as "cinematographer", whereas mine is not. It is, regardless, true.

[Mark Kenfield]

Interestingly, at the lower end of the spectrum (1.2kw HMI fresnel) the 4:1 ratio persists, and it's as the wattages increase that the comparative efficacy drops back down to 2:1.

[Aapo Lettinen]

Btw everything should be fresnels. Its total DP wank that theyre not.

you're only using older fixtures and not for example arrisuns/maxes I assume

[andrew ward]

Phil. I just wanted people to like me more.

[andrew ward]

"Comparative efficacys" used to be my favorite band until their difficult second album.

[Guy Holt]

Interestingly, at the lower end of the spectrum (1.2kw HMI fresnel) the 4:1 ratio persists, and it's as the wattages increase that the comparative efficacy drops back down to 2:1.

 

There is factor not yet discussed that may attribute to this phenomena. The HMI spectrum consists of a large ultra-violet component that can be dangerous if exposed to for an extended period of time. For this reason HMI fixtures incorporate UV filtration. On the open face ARRIMAX fixtures and Par fixtures, it consists of a thin perfectly clear sheet of glass. On Fresnel fixtures the UV filtration is in the Fresnel lens itself. You can see this clearly if you hold up a Fresnel lens from an HMI next to a Fresnel lens from a quartz head – they are not the same glass. The spectrum generated by tungsten lamps, by comparison, don’t include dangerous wavelengths and so the lens doesn’t have to filter out part of the spectrum. For this reason, the tungsten lamp/lens combination is more efficient, since it doesn’t filter out part of the spectrum generated by the lamp.

 

This would also account for why, as Mark has noticed, the efficiency of HMI Fresnels get worse with the bigger lights. The bigger lights require bigger lens that also have to be thicker to maintain structural integrity under the higher heat generated by the bigger lamps.

 

Guy Holt, Gaffer, ScreenLight & Grip, Lighting Rental and Sales in Boston

[Mark Kenfield]

There is factor not yet discussed that may attribute to this phenomena. The HMI spectrum consists of a large ultra-violet component that can be dangerous if exposed to for an extended period of time. For this reason HMI fixtures incorporate UV filtration. On the open face ARRIMAX fixtures and Par fixtures, it consists of a thin perfectly clear sheet of glass. On Fresnel fixtures the UV filtration is in the Fresnel lens itself. You can see this clearly if you hold up a Fresnel lens from an HMI next to a Fresnel lens from a quartz head – they are not the same glass. The spectrum generated by tungsten lamps, by comparison, don’t include dangerous wavelengths and so the lens doesn’t have to filter out part of the spectrum. For this reason, the tungsten lamp/lens combination is more efficient, since it doesn’t filter out part of the spectrum generated by the lamp.

 

This would also account for why, as Mark has noticed, the efficiency of HMI Fresnels get worse with the bigger lights. The bigger lights require bigger lens that also have to be thicker to maintain structural integrity under the higher heat generated by the bigger lamps.

 

Guy Holt, Gaffer, ScreenLight & Grip, Lighting Rental and Sales in Boston

Eureka! You're brilliant Guy, I knew there must have been something obvious to cause the comparative loss of efficacy.

 

This is all very useful knowledge to have.