Color Rendition problem with stage shows at DVinfo.net
DV Info Net

Go Back   DV Info Net > The Tools of DV and HD Production > Open DV Discussion

Open DV Discussion
For topics which don't fit into any of the other categories.

Reply
 
Thread Tools Search this Thread
Old December 14th, 2008, 07:39 PM   #1
Major Player
 
Join Date: Oct 2006
Location: Burbank CA
Posts: 466
Color Rendition problem with stage shows

I've posted a problem like this before but without the pictures. I consistently run into this problem whenever recording theatre shows at the local schools and theaters. The issue occurs in the red/blue spectrum. I've tested the camera at home and proven that there is nothing wrong with it. (See the chair shot). It faithfully resolves several shades of red and magenta accurately. I've taken video and photo stills with the camera.

the attached pics show the stage setup and the color test photo. The color test photo was taken with the camera and should show 4 distinct shades of red. A dark purple pillow, deep red Santa hat, magenta blanket and finally, a bright red Xmas bow.

The back wall in the stage shot is supposed to look like the red bow. The wall is supposed to have much more of an orange tint.

I consistently run into this. Shots that have highly saturated red tend to have blue phasing.

I took some time on the second night to play with the custom presets. No matter what changes I made to Red Gain, Blue Gain or Red/Blue matrixing, I could not match the color in the LCD or viewfinder that I was seeing with my eyes on stage. The same off color image was recorded on tape. So, the LCD, viewfinder and tape all produced an off-red image. It doesn't seem to matter whether I set a preset 3200k white balance or try a custom WB on the camera. Other shades reproduce accurately.

Hopefully the attached pics show the effect. Any input would be appreciated.

Canon XH-A!
Attached Thumbnails
Color Rendition problem with stage shows-color-still.jpg   Color Rendition problem with stage shows-img_0417.jpg  

David Morgan is offline   Reply With Quote
Old December 15th, 2008, 07:39 AM   #2
Trustee
 
Join Date: Jan 2008
Location: North Conway, NH
Posts: 1,745
I'm groping around here but lighting is the big variable here. If you opened the iris or slowed the shutter at the show it would lighten the back wall red color. It looks like there wasn't a lot of light at the show.
Tripp Woelfel is offline   Reply With Quote
Old December 18th, 2008, 01:00 AM   #3
Trustee
 
Join Date: Jan 2004
Location: Scottsdale, AZ 85260
Posts: 1,538
Where to start?

The patio shot has very modest contrast. (light vs dark values)
The stage shot pushes range of contrast right to the boundries.

The light reaching the lens from the patio shot is diffuse.
The light reaching the lens from the stage is harshly bright and dark.

The light reaching the lens on the patio shot travels a short distance. (relatively)
The light reaching the lens on the stage travels a proportionally longer distance.

Let's just look at the last element. If the stage is 60 feet from scene to camera - and the patio scene is 6 feet from scene to camera, the light intensity is 100 times LESS hitting the lens from the stage, relative to that from lens to the back yard scene.

Using our friend the inverse square principal (light, like audio being wave phenomena)

I suspect that's having a pretty profound impact on color rendering capability.

Lots of things going on here. I'll let others bring up more.

But the bottom line is that a typical theater setting is just about the WORSE situation you'll likely find yourself in for achieving good quality video.
Bill Davis is offline   Reply With Quote
Old December 20th, 2008, 01:12 PM   #4
Major Player
 
Join Date: Oct 2006
Location: Burbank CA
Posts: 466
Bill,
I appreciate your input on this perplexing subject. You can tell by the lack of responses, it's not an easy issue to solve. The interesting thing here is that the red spectrum is usually the only area where I can't get accurate reproduction. Greens etc... are good.
However, the distance vs intensity proposition doesn't always hold up because the lighting intensity on the stage varies. Even with full up brightness on stage (not shown here) and plenty of on stage light, the camera has a lot of trouble getting the fully saturated red as in the xmas bow. Seems that your stating that high contrast lighting combined with lack of intensity could be causing this issue?
I've seen other posts on the forum that dance around this same issue.
David Morgan is offline   Reply With Quote
Old December 20th, 2008, 03:44 PM   #5
Inner Circle
 
Join Date: Nov 2006
Location: Tallahassee, FL
Posts: 4,100
Questions:

1. What color is the backdrop under daylight?
2. What kind of lights are on that stage, and are the gelled?
3. Is that a gobo makking a pattern on the backdrop?
4. How did you white balance the camera?
5. Take a friend along with a still camera, or different model video cam. Do they get similar results?

What you see with your eye is a rough guide at best as to what the camera will see. Verify your instruments and trust them instead.
__________________
DVX100, PMW-EX1, Canon 550D, FigRig, Dell Octocore, Avid MC4/5, MB Looks, RedCineX, Matrox MX02 mini, GTech RAID, Edirol R-4, Senn. G2 Evo, Countryman, Moles and Lowels.
Perrone Ford is offline   Reply With Quote
Old December 20th, 2008, 07:06 PM   #6
Major Player
 
Join Date: Oct 2006
Location: Burbank CA
Posts: 466
Quote:
Originally Posted by Perrone Ford View Post
Questions:

1. What color is the backdrop under daylight?
2. What kind of lights are on that stage, and are the gelled?
3. Is that a gobo makking a pattern on the backdrop?
4. How did you white balance the camera?
5. Take a friend along with a still camera, or different model video cam. Do they get similar results?

What you see with your eye is a rough guide at best as to what the camera will see. Verify your instruments and trust them instead.
#1-I don't know. My guess is the backdrop is being colored by a gelled light. maybe a clue?
#3- Pattern, if I remember, was rotating so maybe some kind of effects generator instead of gobo?
#4- In this case, WB was internal camera preset at "indoor" 3200k. I haven't had great results trying to WB custom because nobody knows what the real color temp of the spotlights or other lights are. I have used a white reflector and a spotlight before to manual balance but I got the same sort of results with the reds and at times, worse true resolution of other colors. I've gone to using the WB preset.
#5-would have been nice. At least with my still shots of the chair (next day) I proved that the camera CAN resolve the mysterious shade of red i can never seem to get on stage. As noted in my original post, since I've had this problem before, I remember scenes with red in them at the time of shooting and intentionally look at them as soon as I can to compare what's on tape vs what I saw.

Maybe an ignorant question here but are u saying that the camera can produce more color shades than the human eye or visa-versa? Sort of like comparing chromaticity charts relating to RGB vs video spectrum.
However, I know as far as contrast goes, the human eye can handle gobs more contrast ratio than any instrument can.
David Morgan is offline   Reply With Quote
Old December 20th, 2008, 07:41 PM   #7
Inner Circle
 
Join Date: Nov 2006
Location: Tallahassee, FL
Posts: 4,100
The human eye will see far more colors than your camera; however, there are certain colors that the camera cannot reproduce unless it has white light to do it. So my guess is that the shade of red on that backdrop simply won't photograph with whatever color is coming off the stage lights.

It's like shooting under mercury vapor lamps. There is not full spectrum light so no matter WHAT you do, you are simply not going to get certain colors to register in the camera. My *guess* is that you have non-white light causing you to not be able to get the accurate colors you desire. And a white balance isn't going to replace the fact you aren't getting full spectrum light. I suspect this because you ARE getting accurate colors in other parts of the spectrum.

Hopefully, someone with better lighting knowledge will come along and give a definitive answer, but that is where I'm leaning with this.
__________________
DVX100, PMW-EX1, Canon 550D, FigRig, Dell Octocore, Avid MC4/5, MB Looks, RedCineX, Matrox MX02 mini, GTech RAID, Edirol R-4, Senn. G2 Evo, Countryman, Moles and Lowels.
Perrone Ford is offline   Reply With Quote
Old December 21st, 2008, 12:16 PM   #8
Major Player
 
Join Date: Oct 2006
Location: Burbank CA
Posts: 466
yes I get what ur saying.
I still have to deal with the "practical" problem. People look at the finished product and say:
Hey, that's not the color of the dress I was wearing!
David Morgan is offline   Reply With Quote
Old December 21st, 2008, 12:23 PM   #9
Inner Circle
 
Join Date: Nov 2006
Location: Tallahassee, FL
Posts: 4,100
Quote:
Originally Posted by David Morgan View Post
People look at the finished product and say:
Hey, that's not the color of the dress I was wearing!
...And that's not the same face you were wearing either.

Unless you shoot all of this under balanced white light, colors are going to get changed. It's the same on a Hollywood set or a Broadway play. No way around it.
__________________
DVX100, PMW-EX1, Canon 550D, FigRig, Dell Octocore, Avid MC4/5, MB Looks, RedCineX, Matrox MX02 mini, GTech RAID, Edirol R-4, Senn. G2 Evo, Countryman, Moles and Lowels.
Perrone Ford is offline   Reply With Quote
Old December 21st, 2008, 12:30 PM   #10
Major Player
 
Join Date: Jan 2007
Location: New York City
Posts: 523
Quote:
Originally Posted by David Morgan View Post
The interesting thing here is that the red spectrum is usually the only area where I can't get accurate reproduction. Greens etc... are good.
This is reaching way back into unused parts of my memory so take it as such:

IIRC, the green CCD (or, when I learned this, tube) accounts for 60% of the picture. And, thinking of our friend ROY G BIV, that makes sense--green is in the middle. And red and blue are on the ends.

I also seem to recall that red gets the short end of the stick compared to blue--something about imager effiencies I think. In good light situations, the camera is working harder to produce a good red picture than green or blue. In low light, it just may not have enough to work with.

What can you do about this? Increase the light coming into the camera!

ie. open iris, leave the shutter open longer, increase the gain...

Or get a camera with larger sensors and a faster lens.

Physics can be a bitch!
__________________
Andy Tejral
Railroad Videographer
Andy Tejral is offline   Reply With Quote
Old December 21st, 2008, 12:33 PM   #11
Inner Circle
 
Join Date: Nov 2006
Location: Tallahassee, FL
Posts: 4,100
Actually blue has the least luminance, green the most.
__________________
DVX100, PMW-EX1, Canon 550D, FigRig, Dell Octocore, Avid MC4/5, MB Looks, RedCineX, Matrox MX02 mini, GTech RAID, Edirol R-4, Senn. G2 Evo, Countryman, Moles and Lowels.
Perrone Ford is offline   Reply With Quote
Old December 21st, 2008, 12:58 PM   #12
Wrangler
 
Join Date: Dec 2002
Location: Mays Landing, NJ
Posts: 11,797
Some good responses here! I have been shooting archive video of all our operas for a number of years now, and have tried every trick I can think of as well (currently using a Sony Z1, previously PDX-10 and VX-2000). After awhile I just had to accept that you simply cannot get all the colors "right" when shooting live performance where no attempt has been made to adjust the lighting for video. Since I also design theatrical scenery and lighting, this was a huge frustration to me for the first couple years. Finally I just came to accept it.

I play with the color corrector when I edit and try to bring things into a reasonably similar range as what I saw on stage, or at least something which I find pleasing. I have very rarely had anybody say "hey my costume is the wrong color" actually, but they often are! It's not unusual for our operas to contain a mix of halogen conventional fixtures, HMI moving lights, Xenon follow spots and high powered video projectors (Xenon, HMI or ???). Forget it... there's just no way you can make all of those things look right together. In post you can mess around and try to find the best compromise, but that's about it.

I designed the lights for an opera last spring and used a mix of Halogen and HMI, much of which was not gelled in order to give the harsh kind of look the director wanted. We were both happy with the lighting, so after shooting the video I was anxious to see how it turned out. After watching several segments of the raw tapes I was so unhappy that I put them away and still haven't edited them! Not only are there color issues (which should be fixable), but the lighting was very contrasty and used severe angles from the sides and top. This approach is very effective for an audience member sitting in the theatre, but just about the worst of all worlds for a video camera!
Boyd Ostroff is offline   Reply With Quote
Old December 22nd, 2008, 02:04 AM   #13
Trustee
 
Join Date: Jan 2004
Location: Scottsdale, AZ 85260
Posts: 1,538
Quote:
Originally Posted by David Morgan View Post
Bill,
I appreciate your input on this perplexing subject. You can tell by the lack of responses, it's not an easy issue to solve. The interesting thing here is that the red spectrum is usually the only area where I can't get accurate reproduction. Greens etc... are good.
However, the distance vs intensity proposition doesn't always hold up because the lighting intensity on the stage varies. Even with full up brightness on stage (not shown here) and plenty of on stage light, the camera has a lot of trouble getting the fully saturated red as in the xmas bow. Seems that your stating that high contrast lighting combined with lack of intensity could be causing this issue?
I've seen other posts on the forum that dance around this same issue.

We're touching on some of the issues - and I'll freely admit that I'm not a physicist and don't fully understand this myself. But it's pretty clear that since the days of color TV, RED has been the MOST difficult part of the visible spectrum to record and reproduce.

Note that most cameras are sensitive, not just to visible red, but to the near infared spectrum just adjacent to visible. More sensitive than the human eye, which can easily be demonstrated by videotaping the business end of a typical infared remote control and noting how the camera turns what's invisible to the human eye, visible.

Also, back when engineers were staring to encode video to storage systems, it always seemed to be the RED information that was most prone to smearing and time base errors.

The RED information also occupiess the longest wavelengths of the visible spectrum - around 700nm - but I can't speak on how that might affect it's clarity over distance, other than to note that in audio waves, low frequency (long wavelength) information travels distance much more effectively than do short wavelengths. (See, or rather HEAR, teenage operated car stereos in the night for confirmation) and that MIGHT mean that at distance, the RED light being reflected over say 100 feet to a camcorder in an auditorium might be more efficient than the shorter wavelengths and that might tend to overdrive the red spectrum - relative to the rest of the light - and cause some kind of oversaturation.

I also REALLY think the inverse square thing is critical here. Falloff over distance of waves (light or sound) is profound and inescapable.

I suspect all this and more might be contributed to making it difficult to record colors on video at distance. And a stage play or concert is one of the MOST common events where the video must be captured at a significant distance from the event without the benefit of specialized compensatory lighting like you encounter in sporting events.

As others have noted here. The reality of it being VERY hard to record a stage play that LOOKS great to the human eye - and successfully translate the effect of that lighting on the human eye via a camcorder - is well known and pretty universal.

I think the only way to REALLY do justice to a live stage performance on video is to both light it AND set up camera positioning and blocking specifically for video for a few performances and tweek everything for the camera rather than the eye.

Sorry it's not easier. But that's just the reality as I see it.
Bill Davis is offline   Reply With Quote
Old December 22nd, 2008, 03:26 AM   #14
Major Player
 
Join Date: Oct 2007
Location: Scotland, Ayr www.amour weddingvideos.co.uk
Posts: 305
Quote:
Originally Posted by Bill Davis View Post
I think the only way to REALLY do justice to a live stage performance on video is to both light it AND set up camera positioning and blocking specifically for video for a few performances and tweek everything for the camera rather than the eye.

Sorry it's not easier. But that's just the reality as I see it.
Bill, I wish dance schools would listen to this as most people here probably know we seam to be an after thought and how many times have i heard "no you cant film there, the parents want to see the show and i want to put seats there"
The last show I filmed the school had paid for a full stage and lighting company to come in (who were great to work with )but I was forced to the rear of the arena and my other camera had to be moved each night to accomodate more seating .Its tempting to turn round and walk away from shows like this but equipment needs to be paid for.
On the other hand the last company I worked for didnt have the budget for expensive lighting for their Ice skating show and wanted me to film as part of the show a solo skater in total darkness apart from a 1K spotlight following her(ouch!) The spotlight just didnt have the power to reach the length of the rink so we agreed to film the solo with lights up then again with lights down for the parents and I would include it at the end of the dvd for her parents.
Attached Thumbnails
Color Rendition problem with stage shows-magnum-solo-3.jpg   Color Rendition problem with stage shows-magnum-solo4.jpg  

John Estcourt is offline   Reply With Quote
Old December 22nd, 2008, 08:19 PM   #15
Major Player
 
Join Date: Oct 2006
Location: Burbank CA
Posts: 466
thx guys for the reply's.

Must obviously be something the eye can't see, created by the stage lighting.
One of the only concepts that keeps me grounded is a classic example:
Shooting a person at the podium who is conducting a powerpoint presentation on a nearby screen. Some projectors are set to 9300k WB for graphics. If u WB on the presenter and then pan to the screen, it is usually blue, which makes sense if your balanced at 3200k or even 5600k on the person. YET, the eye does not perceive the screen with a blue cast. Only the camera does. Our eye compensates, the camera doesn't.
David Morgan is offline   Reply
Reply

DV Info Net refers all where-to-buy and where-to-rent questions exclusively to these trusted full line dealers and rental houses...

B&H Photo Video
(866) 521-7381
New York, NY USA

Scan Computers Int. Ltd.
+44 0871-472-4747
Bolton, Lancashire UK


DV Info Net also encourages you to support local businesses and buy from an authorized dealer in your neighborhood.
  You are here: DV Info Net > The Tools of DV and HD Production > Open DV Discussion

Thread Tools Search this Thread
Search this Thread:

Advanced Search

 



All times are GMT -6. The time now is 11:03 AM.


DV Info Net -- Real Names, Real People, Real Info!
1998-2024 The Digital Video Information Network