I’m doing some 45 min presentations on the fundamentals of working with S-Log at NAB on the Sony Booth. The presentation covers: What is S-log, how does it work, how do I expose, record and grade it. I’m on the 35mm stage at 4.30 on Monday and 9.30 on Tuesday and Wednesday. You can also watch excerpts from my “Night of Wonders” video of the Northern Light in the Sony 4K theater.

This is a question I was asked by email, actually it’s a rather good question as for a long time as a cameraman, I wasn’t sure exactly what it was and how it worked. I did know (from reading the manual) that DCC was a supposedly magic way to get better highlight handling, so I did tend, in the past, to have it switched on most of the time. These days however, DCC is something that I tend not to use.

So what is DCC? well DCC is basically an automatic knee function. Going back to basics, we have to consider that standard camera gammas have a very limited dynamic range, typically 6 to 7 stops. This is mainly for historical reasons to ensure that grandma with her 20 year old TV, which may only have a dynamic range of 5 or 6 stops still gets a sensible picture, even if shot with the latest and greatest technology. However, most modern cameras are capable of capturing a much greater dynamic range than 6 stops. Fortunately the human visual system is tuned to analyse the mid ranges (faces, plants etc) in far more detail than highlights. This means that we can compress highlights and most viewers won’t notice, and that’s exactly what the knee circuit in a camera does. It takes anything above a certain brightness level (the knee point) and compresses it. As an example if we consider that our standard gammas allows us to have a 7 stop range, then if we take everything brighter than stop 6 and compress it by 3:1 that gives us a 9 stop range squeezed into a 7 stop recording and the average viewer probably won’t notice the compression. As a camera operator your life is much easier because the sky doesn’t blow out and over expose quite so easily.
So most pro cameras will have a knee circuit operating at a fixed point that helps us deal with highlights. That’s great and very useful, but it is introducing compression and there is a fixed point at which the compassion starts. This makes the compression quite ugly due to it’s “on” or “off” nature (look at hypergammas or cinegammas for something in between). A typical example of ugly knee compression would be a face where the cheeks or forehead are shiny and quite bright, as a result the knee kicks in on these areas and gives them a very plastic and un-natural look.

This is where DCC is supposed to come in. It stands for Dynamic Contrast Control, which sounds very fancy indeed. In reality it is nothing more than an automatic knee. With a conventional knee, the knee point (onset of compression) and knee slope (amount of compression) are fixed. But with DCC the knee point and knee slope adjusts automatically depending on how bright the highlights are. That’s great if your lighting is constant and the scene doesn’t change as the knee will adjust for optimum contrast handling, but not so clever if the highlights start changing mid shot. For example if you have DCC on and you pan across a bright window you may see the brightness of the window go up and down as you pan past. You may also see clouds or the sky changing in apparent brightness as the sun comes in and out, while the rest of the scene remains constant. Another way of thinking about auto knee or DCC is as automatic gain just for highlights.
I would never normally use auto gain for the rest of my image, so why have auto gain just for the highlights. It looks really nasty when you see it operating and in very bright scenes can make judging exposure harder as you may be fighting the auto knee. Open the iris, knee compresses more, close the iris, knee compresses less and so on. So these days I choose not to use DCC, instead I have a couple of different picture profiles with different fixed knee settings for different lighting situations. Or I use cinegammas, hypergammas or S-log. Cinegammas and hypergammas have much more pleasing highlight handling than conventional gammas + knee, but you may need to add a little black press to restore some contrast if your not going to grade (negative black gamma).

Santa Fe Workshops are holding an intensive 3 day, hands on 3D workshop in Hong Kong  on the 28-29-30th of March. This is a full hands on workshop featuring a wide range of cameras and rigs, including beam splitters from a couple of manufacturers, remotely controlled rigs and one piece cameras. There will be sessions on all aspects of 3D production from budgeting and planning all the way through post production to delivery. Whether you are working on a tight budget or producing a full length feature film there will be expert advice and guidance to help you complete your 3D project. More details can be found here: http://www.hdworkshops.com/sfhd/

Here’s a trailer/teaser for “A Diamond In The Mind”, the Duran Duran video that will be released on Blu-Ray later in the year. I was technical director on the shoot at the end of last year. We used 12 PMW-F3’s with some very exotic lenses and a special picture profile I created to match the lighting and feel of the concert. The end result is an incredible work of art. The film was produced by Hangman Studios, with Gavin Elder directing, James Tonkin producing and Den Lennie as DoP.

The engineers at Sony’s Shinegawa factory have been busy. Announced today is the new Sony FS-700 which at first look appears to be a re-vamped FS-100, but under look closely and it’s actually a very different camera.

Out of the box it will offer 1920×1080 recording at 24/25/30 fps plus a special burst mode that will allow you to shoot at 120 fps for 16 seconds or 240 fps for 8 seconds at full 1920×1080. This is a remarkable feat and while 8 seconds doesn’t sound long, do remember that shoot for 8 seconds at 240 fps will result in a clip almost a minute and a half long at 24 fps. At lower resolutions it’s even possible to shoot at up to 960 fps.

To achieve these high frame rates a new sensor is being used. This has 11.6 million pixels and was designed for video applications. The fact that it has 11.6 million pixels means that with a future firmware upgrade the camera will also be able to shoot at 4K. You won’t be able to record 4K in-board (nor 10920×1080 50/60P for that matter). To record these formats you will need an external recorder. To get the signal to the recorder the FS-700 is equipped with a single 3G HDSDi connector and there will be a special 4K recorder from Sony.

Addressing some of the complaints about the original FS-100 the FS-700 features built in ND filters while retaining the Sony E mount lens system. Some of the function buttons have been made larger to suit those users with big fingers or wearing gloves and additional commonly used controls, like image expand (for focussing) have been added to the hand grip. The addition of the 10 bit 3G HDSDi output is also very welcome and the camera can be switched between 25 and 30fps making it a world camera.

What’s less clear in the Sony press release is what will come out of the Sdi connector when shooting 1920 x 1080. Will it still be 4:2:0 or will this camera output 4:2:2 or even 4:4:4? As the internal recording media is still SD cards or the FMU it appears likely that the recording format will still be AVCHD, so that would mean 4:2:0, 8 bit internal recordings.

Even if the output remains as 4:2:0 the ability to shoot burst of 120 or 240 fps video certainly make this an interesting camera. We’ll have to wait and see how the sensor performs as it will have much smaller pixels than the FS100 or F3 so it may not be as sensitive and might be more noisy. It is after all still rare to get something for nothing, so there may be a price to pay for the ability to shoot at 4K.

US street price is pegged to be less than $10K USD and it should be available in June. Hopefully I’ll get to play with one at NAB in just over a weeks time.

News hot off the press from Convergent Design is that the tried and trusted NanoFlash is to be reduced in price by $1000. That’s a significant reduction bringing the street price down to around $1800 USD. The NanoFlash is still the recorder of choice for many of the productions I am involved in, simply because it has a well proven track record for reliability. In addition the Mpeg files it produces are compact so your data wrangling takes less time than with larger codecs. This is significant when you have several cameras shooting all day. The Nano3D is still one of the few 3D recorders on the market that really works, so all in all this is good news for anyone looking for a tried and tested external recording option.

I’ve just returned from a shoot for the Ducati Moto GP team at the Jerez race circuit in Spain. I was shooting in 3D using a pair of F3’s on a Hurricane Rig while other crews shot on Alexa’s, F3’s and C300’s.  It was really interesting to see all 3 cameras in use. Each showed strengths and weaknesses, but all produced good looking images. The Alexa is a fantastic camera and produces great images, but it’s big and heavy. Add a big optimo zoom or similar and it really becomes quite a beast to use. The camera operator shooting on the Alexa was nowhere near as mobile as the other crews. For him an assistant was essential to help with the suitably large tripod, large batteries and frequent media changes.  So, great pictures but fewer set-ups with the Alexa.

The C300’s and F3’s offered much better portability. There were quite a few gripes about the C300’s ergonomics (in particular the need to use the menus to change basic settings) and the usual problems with the F3’s EVF. The F3’s S-Log appeared to offers better dynamic range than the Canons C-Log, but the canon operators found it easier to expose the C-Log. I guess that’s because the C-Log mid range is much closer to a standard gamma and the extra chroma makes the image more pleasing prior to grading. The F3’s were being used with Nikon fit lenses and the C300’s with Canon DSLR lenses, so anyone swapping between cameras had to deal with opposing focus directions! The biggest gripe of the day was with the audio controls on the C300. There were a lot of interviews and some green screen work, shot of a head hight tripod. When the C300 is at eye level the monitor unit ends up just above eye level and the audio controls can only be seen if it’s mounted upright. So the mic ends up pointed at the roof and the LCD is so far up in the air it’s next to useless. The F3’s recorded onto Convergent Designs Gemini’s. This gave the data wranglers a bit of a nightmare as the amount of data uncompressed HD creates is huge, but the pictures do look incredible.

An interesting shoot and it was fun to see each camera operator looking at the other cameras with envious eye’s. Alexa op wanting the light weight of the other cams. C300 op’s wanting the F3’s audio controls and LCD, F3 op’s wanting the C300’s EVF. It just shows that there is no one camera perfect for every application.