Jeff Coatney
Well-known member
A week ago I sat in a DI suite and compared footage shot outdoors in bright sun of backlit subjects like people, buildings, the ocean, sand, you name it. Matt Steinauer shot most of the footage and assembled a test reel of shots with exposures that would challenge any camera, much less a digital one. We intentionally tried to put our Red in some compromising positions: backlit subjects, walking in and out of sun and shadow, blown-out backgrounds, clipped highlights, heavy shadow detail juxtaposed against hot highlights, contrasting textures in motion, black and white clothing, in short we set out to fail the camera with every shot.
Matt's a very experienced DP and he put the Red through every situation he and I could think of that would stretch the camera to the breaking point. Would under-exposing the highlights crush the blacks in the shadows, or add too much noise? Let's shoot it and find out. Would opening up the shadows cause the highlights, in this case bright sunshine on stainless steel or a front-lit white t-shirt to clip? Let's shoot it and find out. For three days we shot. In one interior set-up, we lit the subject with a 1 watt LED. A 1 watt LED was the only light source in the room and we were wide open on the RED 18-50mm Zoom. In fact, we used the 18-50 and the 300mm RED lenses for the entire experiment. Would we get too much noise? Let's shoot it and find out.
When we sat in the DI suite to grade the shots, we expected un-recoverable exposures in at least half the shots. We wanted to find un-recoverable exposures. We want to know the limits of the gear, to find its' real operating envelope. Where is this camera's envelope? Where is the sweet-spot? Where does it fail? We have to know the limitations in order to say to a producer or a DP that is skeptical about relying on it, under what conditions it will perform with confidence. And, once that is proven, how does it look when you output it to film? We expected failure. We expected large areas of clipped highlights that were un-recoverable. We expected those characteristics because Matt, intentionally, did not use a lightmeter. We exposed using the build 15 histogram exclusively. Shot by shot we went through it, each time we investigated the areas of the frame that appeared to be clipped, we found detail. In some cases up to two more stops before it really clipped. And when it clipped, it did so beautifully. No in-organic artifacting, no bizarre color-fringeing, and certainly no halation. The shadows were equally as impressive. When we opened them up we found levels of detail up to two stops and surprisingly(!) little noise. I was shocked. We intentionally set-out to fail the camera. (Which I'm sure is something Jim and Red Team may not want to hear, so my apologies for the seeming lack of respect) But the camera performed beautifully. In the end, of all the shots in the test, all of them but one could be graded and brought within film range without the need for a secondary. The shot that needed the secondary was of the back-lit ocean, with a hot, nearly clipped highlight running through it and a canyon in shadow in the foreground. (I'll post some stills later) That one took a little extra time but the Colorist brought it right in line with the others.
So much for the grading. How's it gonna look on film? We found out last tuesday. We sat in the DI suite and they showed us the graded 4K DPX's that we did the previous thursday, but projected digitally at 2K. Then they rolled the 4K film. We switched back and forth. There was nothing about the film that indicated a digital camera. Despite repeated viewings, the film appeared to have been shot with a film camera (with very expensive glass, no less, much to the surprise of our Colorist) . We looked for failure and only found one small portion of one shot that had a bit of discernable noise. Let me define "discernable" here in context. After seeing the print projected a half-dozen times and noting that the grain in the film's emulsion masked any noise in the blacks, the colorist pointed out a small area of shadow that appeared to jitter ever so slightly for about 12 frames. Even after he pointed it out, it was difficult to catch on the next pass. Acceptable? IMO yes. Would a 2K film out take care of it? Probably.
We had always intended to do a film out test when we took delivery of the Red. Now that we've done it with build 15 material, we can't wait to get our hands on build 16. The results blew us away. I knew this camera was fantastic. I knew after seeing the 4K projection at NAB that this was a stunning piece of equipment. It was only after seeing the footage we shot on film in a reference screening environment that I finally turned to Matt sitting next to me and said "I feel like we just bought a Panavision camera and got all of Kodak, Fuji and Agfa for free."
Then almost simultaneously we both said "I can't wait for Red Ray to get here."
Although we were very happy with the results of the film out (ecstatic) , at the end of the day it is still just film. When you see 2K projected digitally against a 4K film print, digital projection can't get here fast enough.
For those of you that would like to see our test, I will gladly arrange a screening of it. PM me for details and if you have a project about to start and you are trying to make the decision to go RED or go Retro, I'll try to get you in sooner. For me there is no difference anymore. The footage sells itself. Make up your own mind after seeing it.
RED TEAM, you folks build one hell of a camera. In the hands of a top-of-the-line Colorist in a state-of-the-art DI suite, the results can be stunning and in my opinion, indistinguishable from film acquisition. But don't take my word for it, see for yourself.:biggrin:
Matt's a very experienced DP and he put the Red through every situation he and I could think of that would stretch the camera to the breaking point. Would under-exposing the highlights crush the blacks in the shadows, or add too much noise? Let's shoot it and find out. Would opening up the shadows cause the highlights, in this case bright sunshine on stainless steel or a front-lit white t-shirt to clip? Let's shoot it and find out. For three days we shot. In one interior set-up, we lit the subject with a 1 watt LED. A 1 watt LED was the only light source in the room and we were wide open on the RED 18-50mm Zoom. In fact, we used the 18-50 and the 300mm RED lenses for the entire experiment. Would we get too much noise? Let's shoot it and find out.
When we sat in the DI suite to grade the shots, we expected un-recoverable exposures in at least half the shots. We wanted to find un-recoverable exposures. We want to know the limits of the gear, to find its' real operating envelope. Where is this camera's envelope? Where is the sweet-spot? Where does it fail? We have to know the limitations in order to say to a producer or a DP that is skeptical about relying on it, under what conditions it will perform with confidence. And, once that is proven, how does it look when you output it to film? We expected failure. We expected large areas of clipped highlights that were un-recoverable. We expected those characteristics because Matt, intentionally, did not use a lightmeter. We exposed using the build 15 histogram exclusively. Shot by shot we went through it, each time we investigated the areas of the frame that appeared to be clipped, we found detail. In some cases up to two more stops before it really clipped. And when it clipped, it did so beautifully. No in-organic artifacting, no bizarre color-fringeing, and certainly no halation. The shadows were equally as impressive. When we opened them up we found levels of detail up to two stops and surprisingly(!) little noise. I was shocked. We intentionally set-out to fail the camera. (Which I'm sure is something Jim and Red Team may not want to hear, so my apologies for the seeming lack of respect) But the camera performed beautifully. In the end, of all the shots in the test, all of them but one could be graded and brought within film range without the need for a secondary. The shot that needed the secondary was of the back-lit ocean, with a hot, nearly clipped highlight running through it and a canyon in shadow in the foreground. (I'll post some stills later) That one took a little extra time but the Colorist brought it right in line with the others.
So much for the grading. How's it gonna look on film? We found out last tuesday. We sat in the DI suite and they showed us the graded 4K DPX's that we did the previous thursday, but projected digitally at 2K. Then they rolled the 4K film. We switched back and forth. There was nothing about the film that indicated a digital camera. Despite repeated viewings, the film appeared to have been shot with a film camera (with very expensive glass, no less, much to the surprise of our Colorist) . We looked for failure and only found one small portion of one shot that had a bit of discernable noise. Let me define "discernable" here in context. After seeing the print projected a half-dozen times and noting that the grain in the film's emulsion masked any noise in the blacks, the colorist pointed out a small area of shadow that appeared to jitter ever so slightly for about 12 frames. Even after he pointed it out, it was difficult to catch on the next pass. Acceptable? IMO yes. Would a 2K film out take care of it? Probably.
We had always intended to do a film out test when we took delivery of the Red. Now that we've done it with build 15 material, we can't wait to get our hands on build 16. The results blew us away. I knew this camera was fantastic. I knew after seeing the 4K projection at NAB that this was a stunning piece of equipment. It was only after seeing the footage we shot on film in a reference screening environment that I finally turned to Matt sitting next to me and said "I feel like we just bought a Panavision camera and got all of Kodak, Fuji and Agfa for free."
Then almost simultaneously we both said "I can't wait for Red Ray to get here."
Although we were very happy with the results of the film out (ecstatic) , at the end of the day it is still just film. When you see 2K projected digitally against a 4K film print, digital projection can't get here fast enough.
For those of you that would like to see our test, I will gladly arrange a screening of it. PM me for details and if you have a project about to start and you are trying to make the decision to go RED or go Retro, I'll try to get you in sooner. For me there is no difference anymore. The footage sells itself. Make up your own mind after seeing it.
RED TEAM, you folks build one hell of a camera. In the hands of a top-of-the-line Colorist in a state-of-the-art DI suite, the results can be stunning and in my opinion, indistinguishable from film acquisition. But don't take my word for it, see for yourself.:biggrin:
