Michael Cioni
Well-known member
- Joined
- Mar 18, 2008
- Messages
- 341
- Reaction score
- 0
- Points
- 0
- Location
- 6381 DeLongpre | LA | CA | 90028
- Website
- www.lightiron.com
Working with our friends at Alternative Rentals, we recently did some ISO comparison testing and bracketing exposure tests between the RED One Mysterium X and the Sony F35. This test was meant to examine the behavior of the sensors in the exact same lighting conditions and explore their limitations in higher ISOs.
The obvious difficulty in doing comparisons like this is that the cameras to not have the same exact ISO points (though they are close), they do not record at the same resolution, or with similar compression schemes. To combat this, the evaluation was processed with absolute minimal transcode steps in both arenas. Plus, that RED was scaled down to 1080p for better evaluation alongside the F35 and the F35 was scaled up to 4K for better evaluation alongside the RED.
I will let the results speak for themselves, but there are some points I think warrant some discussion on this thread. First, I cannot stress enough the importance of not only evaluating images with a full debayer, but working with them, too. It is a misnomer to think that if you are finishing in 2K or 1080, then all you need is a 1/2 debayer because you "can't see it in 2K." This is an incorrect assumption - much like in a film transfer to 1080p or 2K, you wouldn't scan 1/2 the negative just because it is going from 4 perf to HD. Furthermore, if you are working on a 2K project that is going to have film distribution, then that film, regardless of what it is shot on, will be scaled up over 200% to fill the 35mm negative frame. From this perspective, half-debayered RED images going to film would look worse than F35 images undergoing the same scale. Full-debayered 2K RED images would perform almost the same as F35 and full-debayered 4K RED images would perform the best. This test demonstrates the importance of the debayer and the power of resolution. There is easily a dozen more advantages yielded by the full debayer, which can be discussed later, but hopefully this is clear:
If you or someone you love has seen RED footage and felt that it's resolution impact was unimpressive, then the first point of failure is likely in the debayering pipeline. Film, by it's very nature, is a soft format. So to make my RED projects look as best as they can, a full debayer must happen upstream of the scale (scaling may appropriate, that's totally fine) before moving into the DI.
The second element I'd like to point out is the ISO rating of the MX sensor. These are wondering thoughts, so keep that in mind... I have discovered a lot about this new sensor in recent weeks and I've racked my brain with a lot of people much smarter than me (both inside and outside of RED) and I'm reluctant to state exactly what ISO I think this sensor really is. However, a conclusion that we did come to today (thanks Deanan, Ted and Rob) is that ISO rating might not be the ideal or perhaps at least the only unit of measurement to best expose your MX footage. Above all, I am not a DP and do not pretend to be one. I only shoot with the camera on a regular basis because I feel it helps me better manipulate footage in post. But what I do believe is that this sensor behaves not only very differently from the RED One Mysterium, but even more differently from other digital cinema camera sensors. All that is to say it is my opinion that unlike the RED One Mysterium, the best images from the MX require the constant adjustment of ISO rating and relative T-stop settings. This means it is about to become more important that production and post both have a perfect handshake between which meta-data fields to maintain and which to reset. In other words, which is better, 320 or 1000? The answer is "yes" and "yes." For those of you who shoot a lot of RED, this should sound as new as it is exciting. For those of you who have a great post workflow, prepare to adjust. For those of you who have trouble exposing or getting the RED One Mysterium to look great, you need to be REDucated. I will pray for you.
As always, please be respectful of this shared content. It has always been my belief that sharing information with the community helps us all. I love collaborating with every RED user because I whole heartedly believe that there is a fight going on, but the fight is not amongst each other. I have never felt like I compete with other RED enthusiasts. It's cannibalistic and only slows our collective progress. The fight is between a small band of growing progressive filmmakers and a very large army of antiquated ideologies in desperate need of some full res debayering.
michael
http://www.lightirondigital.com/LightIRON_Digital/LiDigital__MX_4K_STLLS.html
The obvious difficulty in doing comparisons like this is that the cameras to not have the same exact ISO points (though they are close), they do not record at the same resolution, or with similar compression schemes. To combat this, the evaluation was processed with absolute minimal transcode steps in both arenas. Plus, that RED was scaled down to 1080p for better evaluation alongside the F35 and the F35 was scaled up to 4K for better evaluation alongside the RED.
I will let the results speak for themselves, but there are some points I think warrant some discussion on this thread. First, I cannot stress enough the importance of not only evaluating images with a full debayer, but working with them, too. It is a misnomer to think that if you are finishing in 2K or 1080, then all you need is a 1/2 debayer because you "can't see it in 2K." This is an incorrect assumption - much like in a film transfer to 1080p or 2K, you wouldn't scan 1/2 the negative just because it is going from 4 perf to HD. Furthermore, if you are working on a 2K project that is going to have film distribution, then that film, regardless of what it is shot on, will be scaled up over 200% to fill the 35mm negative frame. From this perspective, half-debayered RED images going to film would look worse than F35 images undergoing the same scale. Full-debayered 2K RED images would perform almost the same as F35 and full-debayered 4K RED images would perform the best. This test demonstrates the importance of the debayer and the power of resolution. There is easily a dozen more advantages yielded by the full debayer, which can be discussed later, but hopefully this is clear:
If you or someone you love has seen RED footage and felt that it's resolution impact was unimpressive, then the first point of failure is likely in the debayering pipeline. Film, by it's very nature, is a soft format. So to make my RED projects look as best as they can, a full debayer must happen upstream of the scale (scaling may appropriate, that's totally fine) before moving into the DI.
The second element I'd like to point out is the ISO rating of the MX sensor. These are wondering thoughts, so keep that in mind... I have discovered a lot about this new sensor in recent weeks and I've racked my brain with a lot of people much smarter than me (both inside and outside of RED) and I'm reluctant to state exactly what ISO I think this sensor really is. However, a conclusion that we did come to today (thanks Deanan, Ted and Rob) is that ISO rating might not be the ideal or perhaps at least the only unit of measurement to best expose your MX footage. Above all, I am not a DP and do not pretend to be one. I only shoot with the camera on a regular basis because I feel it helps me better manipulate footage in post. But what I do believe is that this sensor behaves not only very differently from the RED One Mysterium, but even more differently from other digital cinema camera sensors. All that is to say it is my opinion that unlike the RED One Mysterium, the best images from the MX require the constant adjustment of ISO rating and relative T-stop settings. This means it is about to become more important that production and post both have a perfect handshake between which meta-data fields to maintain and which to reset. In other words, which is better, 320 or 1000? The answer is "yes" and "yes." For those of you who shoot a lot of RED, this should sound as new as it is exciting. For those of you who have a great post workflow, prepare to adjust. For those of you who have trouble exposing or getting the RED One Mysterium to look great, you need to be REDucated. I will pray for you.
As always, please be respectful of this shared content. It has always been my belief that sharing information with the community helps us all. I love collaborating with every RED user because I whole heartedly believe that there is a fight going on, but the fight is not amongst each other. I have never felt like I compete with other RED enthusiasts. It's cannibalistic and only slows our collective progress. The fight is between a small band of growing progressive filmmakers and a very large army of antiquated ideologies in desperate need of some full res debayering.
michael
http://www.lightirondigital.com/LightIRON_Digital/LiDigital__MX_4K_STLLS.html
