IRE guidlines

[Matt Dennie]

I was wondering if anyone had a table of IRE values for exposure of people in various conditions like night int. with ambient room light and ext. with moonlight , daytime. Side question: should "night" footage be exposed normally and brought down in post, or underexposed in camera?

[David Mullen ASC]

I was wondering if anyone had a table of IRE values for exposure of people in various conditions like night int. with ambient room light and ext. with moonlight , daytime. Side question: should "night" footage be exposed normally and brought down in post, or underexposed in camera?

 

That's pretty vague or generic because I can imagine a very bright night scene and I can imagine a very dark daytime scene. But in general, I think it's better to go halfway when it comes to making things darker, underexpose halfway to the darkness you like and then color-correct / print down to the final level of darkness. This gives you some leeway in case you change your mind as to the level of darkness.

[Ronald Gerald Smith]

Matt,

 

No such table exists but a fun way to study IRE values is to take screenshots of interesting scenes and pull them up in a editing program like FCP and then looking at the waveform display on the editing program. Also, if you have photoshop you can look at the HSV (HSB) values - V (or B) being value or brightness and they are numbered from 0-100 and represent IRE. I like to take screenshots of just the face or maybe just a cheek or the forehead just to see how they exposed that part of the face. Once you start seeing these still frames through waveform you start to see patterns emerge.

 

In a brighter outdoor settings I find that the main light is around 70 IRE for caucasian skin and lighter skinned asian and latino, maybe around 60 IRE for darker skinned asian and latino and caucasian and lighter skinned african american and about 50 IRE for darker skinned african-american. I find that moonlight the moon hits at around 40-50 IRE and tends to look a bit odd when over 50 IRE. Shadow areas of the face (fill side) on more contrasty scenes are usually around 20-30 IRE, and in some really stylized and over contrasty films it can even be from 0-10 IRE. In some people with very pale skin the main light can even be 90 IRE maybe next to a window or in direct sunlight. In a lot of fashion lighting the main light can even expose skin to 100 IRE and it looks natural and nice but you need a camera that can handle those highlights - I am pretty sure that the original photos were probably not shot at 100 IRE but that it was brought up in post to make it pop more in a magazine or on the internet fashion 'e-zines'. Rim lights and kickers that are 90 IRE are very common because a lot of people like to see a hot bright edge especially in music video settings. \

 

In 'darker' settings like candle lit dinner I would expose whiter skin at around 50 IRE and make shadow areas around 10 IRE. The same applies for lamp lit scenes during the evening.

 

In moonlight lit scenes I might expose whiter skin around 40-50 IRE, shadows areas around 10 IRE.

 

In daylight scenes where there is direct hard sunlight I might expose for 90 IRE, and shadow areas anywhere from 20 IRE-50 IRE depending on how contrasty you want it to look.

 

I daylight scenes where it is overcast I might expose for 70 IRE and shadow areas would fall somewhere around 50 IRE because there is not much contrast during overcast days but I might add some negative fill using some floppies and bring the shadow side down as low as I can.

 

In daylight interiors where the actor is near a window and there is soft bounced light coming in and the room itself is not lit - I might go contrasty like 90 IRE on the bright side and 30 IRE in the dark side or make it more subtle like 70 IRE key, 30-40 IRE fill.

 

Again these are just broad estimates and it gets boring if you mechanize your exposures but the above should serve as a rough guideline as to how people expose different situations.

[John Sprung]

It's interesting how some prehistoric things linger.... Guess what "IRE" stands for -- the Institute of Radio Engineers. Back in the 1930's SMPTE was still SMPE, and the radio guys were working on TV. Their "IRE" units were hundredths of a volt in the picture part of the signal which was zero to one volt. There was also a zero to -0.4 volt side used for sync pulses.

 

Analog TV cameras and monitors were nonlinear, IIRC, something like a 0.45 power function for the camera, and 2.2 power for the CRT monitors. That multiplies out real close to unity, so things look reasonable. But the real IRE scale is nonlinear.

 

So, on modern digital cameras with loads more dynamic range than they had at the dawn of NTSC, I wonder what they mean by IRE. The term clearly has long outlived its actual definition.

 

 

 

 

 

-- J.S.

[Ronald Gerald Smith]

It's interesting how some prehistoric things linger.... Guess what "IRE" stands for -- the Institute of Radio Engineers. Back in the 1930's SMPTE was still SMPE, and the radio guys were working on TV. Their "IRE" units were hundredths of a volt in the picture part of the signal which was zero to one volt. There was also a zero to -0.4 volt side used for sync pulses.

 

Analog TV cameras and monitors were nonlinear, IIRC, something like a 0.45 power function for the camera, and 2.2 power for the CRT monitors. That multiplies out real close to unity, so things look reasonable. But the real IRE scale is nonlinear.

 

So, on modern digital cameras with loads more dynamic range than they had at the dawn of NTSC, I wonder what they mean by IRE. The term clearly has long outlived its actual definition.

 

 

 

 

 

-- J.S.

 

 

Hi John, looking at your post it does seem to me to be a prehistoric term.... do you think that I used it incorrectly? I'm not an expert and I'm quite greener even more so than a novice. Maybe a more relevant term would be 'luminance level'. Any ideas?

[John Sprung]

Hi John, looking at your post it does seem to me to be a prehistoric term.... do you think that I used it incorrectly?

 

The waveform monitor function of an editing program may well be trying to stay consistent with the original IRE, which was defined for oscilloscope waveform monitors. CRT oscilloscopes often had a waveform setting built in, just because it was quite easy to do. As for Photoshop, I don't know. It would take some searching thru manuals or asking an expert to see if they also aimed to match the old IRE curve. Based on the numbers you give, I'd guess maybe not. Skin at 60 - 90 IRE would be quite hot. We used to knock title and credit text down from 90 to about 60 - 70 for a more subtle, less video look.

 

Bottom line, numbers need units, and units need standards. So, any time you see numbers coming at you, always find out what the units are, and what standard defines their meaning.

 

 

 

 

-- J.S.

[Ronald Gerald Smith]

The waveform monitor function of an editing program may well be trying to stay consistent with the original IRE, which was defined for oscilloscope waveform monitors. CRT oscilloscopes often had a waveform setting built in, just because it was quite easy to do. As for Photoshop, I don't know. It would take some searching thru manuals or asking an expert to see if they also aimed to match the old IRE curve. Based on the numbers you give, I'd guess maybe not. Skin at 60 - 90 IRE would be quite hot. We used to knock title and credit text down from 90 to about 60 - 70 for a more subtle, less video look.

 

Bottom line, numbers need units, and units need standards. So, any time you see numbers coming at you, always find out what the units are, and what standard defines their meaning.

 

 

 

 

-- J.S.

 

Very informative. Yes, I know it's contradicting my original post but yes I too believe that less hot lit images tend to be more filmic and a lot of the underexposed work by cinematographers such as Lance Acord looks so beautiful and cinematic to me they glow within themselves somehow.

 

You are very right photoshop seems to be less a useful tool for judging brightness (well looking at the HSV indicator at least) - mainly because a waveform is so much better to judge a larger area of exposure.

 

Yes I think that skin at 90 IRE seems to be hot but it is so common nowadays with things being on the web and needing to be bright and pop out of your screen. I think IRE that high is mostly used for otherwordly or fashion/beauty applications rather than any realistic narrative piece (most of the time). I think 90 IRE is akin to overexposing skin by 2 stops (from my limited knowledge I am assuming that it is), and I think that some cinematographers like to overexpose skin that much especially if the situation/story calls for it. Such as like an interrogation scene where the hot light is pointed at the suspect, or when someone is stuck in a hole for days and then the first sign of light is blinding and the skin almost burns because the light is so bright compared to the darkness inside the hole.

 

I remember reading years ago when I first started learning about videography that people recommend ideal caucasian skin exposure is 70 IRE, which is zone 6 if you were using the zone system.

Edited March 9, 2011 by Ronald Gerald Smith

[Ronald Gerald Smith]

Very informative. Yes, I know it's contradicting my original post but yes I too believe that less hot lit images tend to be more filmic and a lot of the underexposed work by cinematographers such as Lance Acord looks so beautiful and cinematic to me they glow within themselves somehow.

 

You are very right photoshop seems to be less a useful tool for judging brightness (well looking at the HSV indicator at least) - mainly because a waveform is so much better to judge a larger area of exposure.

 

Yes I think that skin at 90 IRE seems to be hot but it is so common nowadays with things being on the web and needing to be bright and pop out of your screen. I think IRE that high is mostly used for otherwordly or fashion/beauty applications rather than any realistic narrative piece (most of the time). I think 90 IRE is akin to overexposing skin by 2 stops (from my limited knowledge I am assuming that it is), and I think that some cinematographers like to overexpose skin that much especially if the situation/story calls for it. Such as like an interrogation scene where the hot light is pointed at the suspect, or when someone is stuck in a hole for days and then the first sign of light is blinding and the skin almost burns because the light is so bright compared to the darkness inside the hole.

 

I remember reading years ago when I first started learning about videography that people recommend ideal caucasian skin exposure is 70 IRE, which is zone 6 if you were using the zone system.

 

I want to add to the above:

I remember reading years ago when I first started learning about videography that people recommend ideal caucasian skin exposure is 70 IRE, which is zone 6 if you were using the zone system. It's probably not the best idea following to that advice because it limits people and limits creativity with exposure but it is just something that I thought was generally accepted mostly by videographers and news people who want to have a target/goal so they are not trying to make some kind of creative exposure decision seconds before it is time to roll a live broadcast.