More Content / Content Suggestion?

More content like this is available on Ryan's Website.

Is there a video that you would like to see published?
Email your suggestion to: mail @ ryanewalters.com (No Spaces) and Ryan will take it into consideration.

Monday, April 16, 2012

Zacuto Shoot Out 2012: A Master Cinematographer's Experience


Back in February of this year Zacuto & Kessler asked me to join them for their Revenge 2012 camera shootout as the Red Epic Master Cinematographer. While I was honored to be a part of an event where I would be working alongside such greats as Bruce Logan, ASC, Rodney Charters, ASC, Nancy Schreiber, ASC, Michael Negrin, ASC & Curtis Clark, ASC, I accepted the offer with a bit of trepidation. After years of running various tests, and making those results public, I have become acutely aware that no matter how stringent a test is run, inevitably people are going to find fault, and there is no way to make everyone happy. Add to that the fact that critique from the Reduser community is even more stringent, and it is a recipe for some harsh criticism and open flaming of one's character. Even though I knew I was walking into dangerous ground by being the Red Epic Master Cinematographer, this was an opportunity I could not pass up.

More Perspective On The Test Results Here.

(Left to Right: Rodney Charters, ASCBruce Logan, ASCDen Lennie)
The Test Parameters
This years test was overseen and administered by Bruce Logan, ASC, and conducted on a stage at Tribeca Flashpoint Academy in Chicago. Featuring nine of the latest cameras, and consisting of three shots within one scene, this test would show off each camera's strengths and weaknesses in a real production environment. The master shot consisted of a wide range of exposure values. It was full of color, texture, detail, and had a fair bit of movement from both the talent and the camera. The next two shots were bits of coverage that highlighted specific elements within the scene like over exposure, and potential banding in transitions from dark to light on even colored surfaces. The test was well thought out and implemented - which is to be expected from a professional like Bruce Logan, ASC.

There were two portions of the test: the empirical and what was called the "Creative Shot." For the empirical portion of the test, the Master Cinematographer could tweak any camera settings he or she wanted, add any ND filtration desired, and set the exposure to wherever he or she felt would be appropriate for the setup. The three shots of the scene were filmed in traditional order - starting with the master and moving in for coverage. It should also be noted that during this portion each camera was able to shoot a "light" version, and a "dark" version of each shot. The light version was what we thought would be slightly over-exposed, and the dark version was what we thought was properly exposed, or just slightly underexposed. Then in post, we were able to choose the better of the two exposures for grading.

The second part of the test was back on the master shot and deemed the "Creative Shot" It was at this point that the Master Cinematographer could do whatever he or she wanted to do with the lighting to make the scene look the best to them using their camera system. There were two limitations to be worked within: 1) The exterior "sun" could not be altered in levels by dimming the lamps, or turning any of them off. We could ND or net the window if we wanted - but the lamps had to remain how they were, and 2) We could not relight the scene - the direction and position of the existing lights could not be changed. In the interior, fill light could be added, lights could be scrimmed, etc. It should also be noted that we could add any filtration to the camera that we wanted. That was our sandbox to play within for the test and I'll go into detail about the choices I made for the Red Epic a little later.

The LCD Screen From The Epic Shooting The Dynamic Range Chart - WITHOUT HDRx
The Dynamic Range Chart
While this shootout is not about shooting charts, there was one chart that was shot - a back lit dynamic range chart produced by DSC Labs. For this chart, exposure was set so that the very first chip was placed at clipping. This was verified by watching the external waveform flatten out on the first chip, and more importantly, by using the exposure check mode in the camera. At T8 1/3, only the first chip was clipping. 15 seconds of this shot was rolled off before implementing HDRx. For the HDRx exposure the lens was opened up to T2.8 1/3 (+3 EV*), shutter was set to 360 degrees (+1 EV), and then the frame rate was set to 6 fps (+2 EV) for a grand total of an increase of 6 stops. HDRx was then turned on and set to 6 to compensate accordingly. While I'm going to refrain from speaking definitively about how many stops I was able to see, as I was just viewing it off of the 5.6" EPIC LCD and then on an iMac using RedCine-X Pro, what I will say is that there were an astonishing number of visible bars - more than I had ever seen before with any other system.

*EV = Exposure Value. Think of it as being equivalent to a stop, 1 EV = 1 Stop. A change of 1 EV can be arrived at through many different parameters. The following examples are all changes of +1 EV: 180 degree shutter to 360 degree shutter, 24 fps to 12 fps, T4 to T2.8.

Filmworkers Club Chicago's Baselight
Photo Courtesy of Lan & Vu Bui
The Grade
After I completed my portion of the test, the footage was sent over to Filmworkers Club Chicago for grading by talented colorists on their Baselight system. Each Master Cinematographer then got to sit with Bruce Logan, ASC, and the colorist and choose the best exposure to use from the empirical portion of the test. After making these selections, the clip was then graded to match the grade of the baseline camera - the Alexa. After matching all three clips, I then had the chance to really play with my own clip. Again, I was able to grade the clip I saw fit to get the look I wanted using any of the tools on hand.

Hanging out during the grade with Den Lennie & Mick Jones
Photo Courtesy of Lan & Vu Bui 
My Experiences On Set And In The Grade
As a working cinematographer, it is a rare opportunity to be able to watch a fellow cinematographer work, let alone watch some of the greats at work. I stood by and quietly watched as they effortlessly managed the crew with poise and confidence, drawing from years of experience. I made sure I was close enough to hear what was going on, keeping my ears pointed in their direction, but far enough back to be out of the way. It is tough to put into words, but the experience on set was better than anything I have experienced in any other learning environment to date.

Dialing in the Epic on set.
Photo Courtesy of Lan & Vu Bui 
My experience on set was further enhanced by the opportunity to follow along with several other cinematographers through to the grade. It is here where I got the opportunity to see why they made the choices that they did in their lighting setup and how that impacted the choices they made in the grade. It reaffirmed to me the importance of knowing your tools from production to post and the importance of using all the tools at your disposal to get the results you intend. How much you get in camera, and how much you leave to post will be determined by your restraints on set, your budget for post, and your own personal preferences. (My tendency and preference is to get as much of the look as possible in camera, and then finesse the image in post).

Base Lighting Setup
The Red Epic Test & DP's Choice: Why I did what I did ...
For the empirical portion of the test I wanted to be at a shooting stop of T2.8, (A creative choice I felt fit the scene- and it also matched the stop of the Alexa) so after taking my readings with my meter* I added the appropriate amount of filtration Schneider Platinum IRND's in either .3 or .6 strengths depending on if we were shooting the light or dark version. While it is true that at these levels of ND the additional IR filtration is not absolutely needed, I have chosen to use it as my default ND filtration when using the Epic based on the results of my previous ND Testing. Next I used the exposure check mode in the camera to see if there was still any clipping in the scene. This revealed a couple of places that had some small amounts of specular highlight clipping. I then decided to roll the iris down until that clipping went away - which was either 1 or 2 stops, again depending on the level of ND in the lens. I opened the lens back up to T2.8 and invoked HDRx setting it to either 1 or 2 to match the number of stops I needed to retain all of the highlight information. When I got to post with the empirical test, I had clips that retained the entire dynamic range of the scene allowing them to be graded as needed. Because the clips would be graded in a Baselight system, I would have the flexibility to work natively with the RAW files, manipulating them in a completely RAW environment with floating 32-bit precision until they went out for delivery.
Ryan's changes to the setup.
For the "Creative Shot" portion of the test, what I was most concerned about was the shadow areas of the set, specifically: the woman by the door in the background, frame right; as the gentleman walks to the couch, and then the details in his hair as he sits down in the foreground. What I appreciate about working with a RAW image, and a camera like the Epic, is the ability to have multiple creative options available to me to tackle any situation that I am presented. In this instance, I had four very different options available to me which would all result in the same outcome: more detail in the shadows.
  • Option 01: Change nothing. Use the range of the chip to dig out the details in the areas that I wanted it.
    • The fastest and easiest choice would have been to do nothing and then use my post tools to get the image where I wanted it. Unless this is the look you have decided on getting during preproduction, it is not how I prefer to work (AKA "Fix it in post"). Additionally, even though the Epic with the MX chip is very clean, I am still very noise adverse, and I don't like to dig into the shadows to bring out detail as there is still a chance that it will introduce noise into the picture. I prefer to fix it in camera, and finesse it in post, rather than just relying on the "fix it in post" mentality. This is a personal preference, and since I wasn't backed into a corner due to schedules or other factors, I passed on this option.
  • Option 02: Open up 1 stop, and use HDRx to protect the highlights.
    • I was at an exposure of T2.8, and the lens went to a T2. Opening up a stop would have been the second quickest and easiest choice to make. If I were pressed on time, and I had no other option, this is the choice I would have had to make. Personally, I do not like shooting at wide apertures like T2 or T1.4 as I feel it becomes too distracting to the viewer, it makes life more difficult for my 1st AC, and with everything so blurry and out of focus in the background the environment of the characters gets lost. I was fine on time, so I crossed this option off my list.
  • Option 03: Lose the ND, and use HDRx to protect the highlights.
    • Taking the ND out of the lens and compensating with HDRx would have been the next fastest choice. It would only be slightly longer than option 02 as it would have involved the removal of the filter and its return to the filter case. This would have kept my shooting stop the same (T2.8) and kept us moving if we were pressed for time - which we were not. I passed on this option as I knew that in the grade for this shot I was not going to use HDRx. (More on HDRx later).
  • Option 04: Bring up the fill level in the areas where I wanted more detail.
    • This is the most involved and time consuming option I have to choose from, and because we were fine on time, I chose it. Using this option yields the best looking final images if a complete post process with a professional colorist is going to be implemented. By bringing up the shadow areas through the addition of fill light, I was creating a "fat negative." I allowed myself the flexibility in post to choose how quickly the shadow detail rolled off. This is where the finessing of the image comes into play, and with the proper post tools and talent behind the wheel the benefits of working in RAW really pays off.

In addition to the lighting change to the set, I also changed the filtration in front of the lens. Along with the IRND I also added a Schneider 1/4 Hollywood Black Magic. Working at 5k Full Frame has a lot of added benefits when doing visual effects work in post or when shooting a scene that has a lot of fine detail in it that you want to capture. This is not always so flattering when it comes to capturing all of the detail in the talents' skin. This kind of resolution can really show off the imperfections of the talent, and reveal the use of makeup. (Even when working with young talent who has smooth skin). I want the talent I work with to look the best they can, and this can mean the use of filtration to help hide any blemishes that may appear on screen. What I like about the Hollywood Black Magic filter is that it combines the classic soft filter with the black frost filter. Practically, what this means is that it helps to hide blemishes in the skin (classic soft) as well as it helps with the roll off into the highlights (black frost) while still retaining crisp blacks (black frost). I prefer to not overdo it, so I typically stick to lighter strengths of diffusion. What I like most about this particular filter is that it does not immediately give itself away. If you know what to look for and evaluate the image carefully you can find it, but it doesn't jump off the screen - it is subtle.

To use HDRx or not to use HDRx, that is the question. For my setup, I did shoot with HDRx turned on, however, it was my plan to not use it when it came time to grade my clip. The only reason I chose to leave it turned on was to cover my bases. Had this been an actual commercial or feature production, I would have left it turned off knowing that I would not use it in post. With HDRx turned on I can confidently say that every highlight detail was captured. However, for this particular scene, my instincts told me that I did not want to see everything - most notably the translite out the window. Due to the space limitations of the stage, it had to be placed about 6 feet from the window. And the resulting image felt like a Disneyland backdrop. Retaining all of the detail in the scene was not the direction I wanted to head in aesetically. I wanted the exterior to feel bright and overexposed so that it felt more real, and less like we were on a stage.

The second question that has to be asked when using HDRx, is how does the motion rendition look aesethically? When shooting HDR imagery, it is my opinion that as of this writing, the implementation of HDRx and the use of the Magic Motion tool in post is the best implementation of HDR motion shooting to date. If I need to shoot HDR images, then this is the tool I'm going to use. However, with that said, aesethically, I do not care for the motion rendition of HDRx. It is a personal preference, a creative choice that each person has to make. And while I think it is the best implementation out there, it is still not for me. So for the projects that I work on, I treat the camera as if it doesn't have the HDRx mode.
One last look at the Setup with Bruce Logan, ASC
When it came time for the grade of my clip, I was happy with the choices that I made on set. Increasing the fill allowed me to work with the colorist to really dial in the levels of the shadows making sure that the detail was kept where I wanted it, but fell off where I did not want it. The same was true with the highlights. I pushed them up to overexposure to make the feeling of the exterior more believable. I had more than enough detail in the highlights to start with so that they rolled off nicely into overexposure. The Hollywood Black Magic filter performed as I expected it to - keeping all of the important details in the image, but removing the imperfections in the skin, and helping the highlights to roll off in a pleasant way while retaining crisp blacks.

*Yes, I use a meter when shooting digitally. While I know that people like Rodney Charters, ASC, & Nancy Schreiber, ASC, do not use a meter when shooting digitally, I am under no illusion that I have their same level of experience. They have trained their eyes to the point where they can light by eye. And I'm still in that process of training my eye. If you want to learn more about using a meter, check out my training video here.

My Takeaways
As I have stated before, what matters most is the person behind the lens, not the camera or any of the tech. The camera is a tool- a means to an end. If the story, acting, production design, lighting, and sound are terrible, the camera is not going to save the production. This point was hammered home to me as I sat in on some of the other grading sessions. I was continually impressed with how well some of the smaller cameras and "less professional" codecs held up in the grade. This proved to me that if a tool is used well you can walk away with some amazing imagery.

The camera that impressed me most was the Sony FS100*. According to the specs, this little camera should not produce images that rival the Alexa or the Epic. Even using the footage off of the SD card compressed at 4:2:0, the colorist was able to match the properly exposed FS100 to the Alexa no problem. It was the people behind the camera that made that happen. Does that mean that the Alexa should be ditched for the FS100? Of course not - that is absurd to think or say. There are a number of applications where the high quality imagery coming off an Alexa is going to be needed and preferred. Instead, what it does mean, is that if you have one of the "lesser" cameras, and you are pining over getting something bigger and better thinking that it will improve your images, you are going to be sadly mistaken. All it is going to do for you at this point is increase your overhead - which can be a very foolish thing to do for the beginning filmmaker.

Before you get your panties all in a bunch, the numbers do have some significance. They do point towards the potential quality, or lack thereof, that you can get out of a camera system. But they are only pointers - not definitives. You need to test the tool out for your specific application to see if it will deliver the results you need. Todays sensors, codecs, and algorithms for recording imagery have advanced so much that I think applying what we learned & experienced with the 4:2:0 / 8-bit imagery from 2003 to todays 4:2:0 / 8-bit imagery is short-sighted. Technology has not stood still since its first introduction. If you're not happy with the images you are producing, take a good, long, hard look at the person behind the lens before you jump to the conclusion that it's the cameras fault.

As I have stated before, the camera is a tool, and it is only one small part of a much bigger picture. If you get hung up on that one small part of the bigger picture, you may be shooting yourself in the foot. It is always best to choose the appropriate tool for the job no matter what that tool happens to be. For example, I shot a commercial for Adidas and the correct tool for that job was 2 Epic's. I needed to be able to shoot 120 fps and I needed an ultra high resolution RAW image for the visual effects work that needed to be done. Knowing the project requirements, the choice was a no brainer - of course I'm going with the Epic. For another project that I shot for Comcast, the end deliverable was a small internet video, and they wanted a tape based format that could easily fit in to their existing post workflow, and the footage needed to be FedEx'ed as soon as the shoot was complete to meet their turnaround time. For this project, the right choice was the Sony Z1U shooting in SD 4:3. (Yes, there are still some projects shooting in SD 4:3, amazingly enough ...) As I did preproduction for a TV Pilot for a Cable TV Network we were leaning towards using a GH2 due to budget requirements and the speed at which a small camera can be used at. And finally to come full circle, I had another Adidas commercial where the best choice was the FS100. The camera is just a tool. Getting hung up on the spec sheet is foolish and short-sighted in my opinion. If you always need to have the latest and greatest in order to feel better about the images you create, by all means have at it. I'm going to continue to select the tool that best suits the needs of the production, and then suck every last ounce of quality I can out of that tool.

*The Canon C300 also really impressed me. The 8-bit codec held up really well in the grading suite. However, at $16,000 I expect more out of a camera system, so it was less surprising to me than a $5,000 camera.

A Note About The Tech
If you haven't picked up on it already, I'm a firm believer that it is the person behind the tech that matters most - not the tech itself. But that does not discount the tech completely. The quality level of the tech allows for greater flexibility during production. With more sophisticated equipment comes less hassle, and more solutions to possible problems. For example, while you can get great images out of the 7D, using that camera introduces additional problems on set, like the need for recording quality sound and figuring out a production friendly way to rig the camera. Whereas with more "professional" tools like the C300, these issues disappear as it allow for XLR's and quality internal recording of the audio. As I previously demonstrated when using a camera like the Epic you can have four different solutions for tackling how you want to expose a scene. More professional tools allow for greater flexibility on set, and in post.

So the tech does matter, but it doesn't matter in the way most people think it matters. Better tech should yield less hassle on set, and it should allow for more creative choices during production and post production. It is about allowing the "problems" to disappear so that you can focus on the creative pursuit of filmmaking. It is not going to substantially increase or change the quality level of the final imagery if what is in front of the lens is poor quality to begin with. While this may seem a bit crass - You can polish a turd all you want, but in the end it will still be a turd.

Check It Out For Yourself
If you are at NAB this year, I HIGHLY recommend that you stop by and watch the results of the test which you will find at Zacuto & Kessler's Booth: C9848. I think there is a good chance you'll miss several of the cameras. While there will be a cameras that will be easy to spot, I think you'll be surprised to see how similar many of the images are. The only way I know which is which is because of being a part of the grading session of several of the cameras and I know what to look for. If I had not been a part of it, I think I would have a much more difficult time telling them apart.

Until Next Time - Get Out There And Shoot,
Ryan E. Walters, Cinematographer 

6 comments:

  1. This is a great read. Thank you :-)

    ReplyDelete
    Replies
    1. Great to see you still catching up on cameras :)

      Delete
  2. Great write up about the test Ryan, it was great to work with you and do our little bit with the FS100- I concur that the on set experience with these great Cinematographers was a truly humbling end enhancing experience for s all- let alone share the billing with them. It was a very memorable experience

    ReplyDelete
    Replies
    1. Thanks. Likewise. I look forward to our paths crossing more in the future. :)

      Delete
  3. Excellent rundown, I have to say I liked the EPIC image the best of all, then Alexa and then GH2 with a close 4th the C300. You did a superb job.

    ReplyDelete
    Replies
    1. Thanks James. :) When I saw the blind test at Skywalker Ranch, I chose Alexa, Epic, FS100, and then the GH2. It all goes to show that personal preference plays a lot into the perception of the final image. It was a great opportunity to be involved with this production. :)

      Delete