With the release of Firmware version 1.13 for the PMW-F5 and F55, Sony enabled the use of EI gain when shooting using the Cine EI mode. It’s a little unusual in that the EI gain only goes below the default/native gain and not above.
So what is EI gain and how do you use it?
EI or Exposure Index gain is a way of changing the gain of the monitoring outputs, but not the gain of the primary recordings. Currently with the F5 and F55 the signal seen on the HDSDI, HDMI output and viewfinder outputs are the same signals as recorded to the internal SxS cards, so EI gain is also added to the internal recordings. However EI gain is not added to the raw recordings recorded by the R5, these are always done at the native ISO of the camera, 1250ISO for the F55 and 2000ISO for the F5. The EI Gain is selected in the camera menu, it cannot be selected from the hot keys around the info LCD.
The F55 supports Exposure Index values (400EI/640EI/800EI/1250EI) in Cine-EI mode.
The F5 supports Exposure Index value (640EI/800EI/1250EI/2000EI) in Cine-EI mode.
So as we can see from above we can reduce the EI gain, not increase it (the F3 and most other cameras allow an increase in gain). What happens when we do this? Lets say we decrease the EI gain on an F5 to 800EI or 640EI on an F55, just a little over a one stop reduction. The image seen on the monitor or viewfinder screen gets darker. If your using a light meter to expose you would set your light meter to the new EI ISO.
If exposing according to that light meter, it would tell you to open the iris by a little over 1 stop. This would then return the brightness on the monitor or viewfinder back to where it was before you selected the new EI ISO. Now remember that the actual raw recording ISO does not change. So as you open up the iris (as that’s what the light meter or monitor image tells you to do) the raw recordings become brighter. With most conventional cameras this could be a problem as it might result in an overexposed image. But the F5 and F55 have so much dynamic range that for most typical scenes the exposure will still be OK. My experience with the F5 and F55 is that when shooting at the native ISO you normally have plenty of headroom. Raising the recording levels like this will allow more shadow and dark detail to be seen. So what you are doing is in effect shifting the cameras latitude down by just over one stop. Assuming your correctly exposed at the cameras native ISO you get +7 stops and -7stops, but now with the EI ISO set 1.25 stops lower, the dynamic range becomes +5.75 stops -8.25 stops. In addition as you will be bringing levels down in post production (compared to footage shot at the native ISO) you will end up this less noise in the final results.
So the current EI ISO settings on the F5 and F55 allows you to shift your exposure range down to produce a less noisy image with greater shadow latitude, but decreased highlight latitude. Interestingly this is the opposite to the way EI ISO on the F3 works, the difference being due to the F5 and F55’s much higher dynamic ranges but more importantly the linear raw recording allowing for better use of the top end of the cameras dynamic range. A camera like the F3 that uses log recording cannot be significantly overexposed as the log compression of the highlights would cause problems with the brighter parts of the scene becoming excessively compressed. I would imagine that Sony will include EI ISO settings that are higher than the native ISO in future firmware releases, but for the moment we can only go down.
1.) i.LINK DVCAM Input
Stream signal recording via i.LINK DVCAM input is available.
2.) Adjusting the Brightness of the EVF Screen The default value of EVF has changed from ±0 to +30.
3.) Selecting AUDIO IN Reference Input Level The default value of Audio Input has changed from -40 dB to -60 dB.
4.) External Device Connector A message will be displayed if a USB device which is not supported by this unit has been connected.
5.) Auto Focus Auto focus performance is improved. Focus instability near the focal point and with a high-luminance background has been reduced.
6.) Auto Knee Auto Knee performance is improved. Emphasis of white for an object with a high-luminance background has been reduced.
7.) ATW ATW performance is improved. When ATW mode is set to Natural, automatic adjustment for wider brightness range when shooting outdoors is available.
8.) Matrix Adjustment Can be put on OFF in PICTURE PROFILE SET.
Sony released a further firmware update for the F5 and F55 on Friday. Version 1.13 adds EI exposure index modes to the Cine EI mode and the ability to trigger the record start and stop from the Hirose 4 pin power ports. In addition LED 4 can be assigned as a record indicator and there are some improvements to the cameras noise reduction in standard modes.
I’m running a one day intensive workshop for anyone looking at or working with large sensor cameras. If you are new to large sensor cameras or just want to know more about how to use them then this intensive one day workshop is for you. With a mixture of theory, practical demonstrations and hands-on sessions, I will guide you through every aspect of film making from lighting techniques to depth of field, gamma curves to RAW recording.
Throughout the day you will explore the benefits as well as some of the drawbacks of shooting with cameras that use APS-C and super35mm sized sensors, both from a practical point of view as well as a technical one. You will learn about how to manage your depth of field (DoF), whether you need a shallow DoF for a filmic look or a deeper DoF for easier shooting. There will be a session on lenses and lens adapters looking at the pro’s and con’s of using either DSLR lenses or PL mount lenses.
Then we will look at lighting, exploring the different types of lights available to the modern film maker and how to use them. We will spend time learning about contrast ratios and how to light and expose for a filmic look. In addition we will explain about the different types of gamma curves, log curves and raw recording methods used by these cameras. By combining the right lighting with the right gamma curve we will see how to maximise the performance available from these large sensor cameras. The day will end with a hands on practical session where we will put everything learnt into practice, setting up, lighting and shooting a test scene and then grading the footage shot.
Important
If you have your own camera then bring it along to use. If not, then, it doesn’t really matter as there will be cameras for you to try (minimum PMW-F5, PMW-F3, NEX-FS-700, Canon 550D, Sony NEX 5N). You will also be able to compare the images from both dedicated video cameras and DSLR’s.
This one-day workshop is a must for anyone interested in large sensor cameras and improving their film making techniques.
You will leave this session with:
Pros and cons of using cameras with large sensors
How to light and expose for a filmic look
Different types of gamma curves, log curves and raw recording methods
How to manage your depth of field
Techniques to maximise the performance of large sensor cameras
The pro’s and con’s of using lens adapters, DSLR and PL mount lenses
and much more…
Who should attend:
This Masterclass is suitable for anyone who is interested in learning how to get the best from a large sensor cameras or improve their film making techniques in general. Cinematographers, DoP’s, camera operators, freelancers, focus pullers.
What attendees say..
“It was a great day, Alister certainly knows his stuff and is a good presenter as well” Bruna Fionda – Lighting Camerawoman
“The rightful thanks entirely belong to Visual Impact for this superb opportunity to learn from the genius Aistair Chapman.” Dhruv Gadhvi – Cameraman
“Just a quick note to say thanks for the Sony PMW-200 Masterclass…. It was not only great to learn more about the PMW200, but also to meet the other attendees.” Steve Vousden – News Craft Coordinator, BBC Wales
“Thanks for organising the event at Teddington. It was enjoyable and interesting and helpful. A big thank you to all your staff who participated.” Chris Morphet – Cameraman
“I really enjoyed the day found it so useful. All the staff we so friendly and informative.” Tony Jones – Director- Lighting Cameraman/DoP/Editor/PD
Well I’ve just returned home from NAB and a week of Tornado Chasing in the USA. For the Tornado chasing I was shooting in 4K using my Sony F5. I’ve shot run and gun with my F3 and FS700 in the past when shooting air-shows and similar events. But this was very different. Tornado chasing is potentially dangerous. You often only have seconds to grab a shot which involves leaping out of a car, quickly setting up a tripod and camera and then framing and exposing the shot. You often only have time for one 30 second shot before you have to jump back into the car and move on out ahead of the storm. All of this my be happening in very strong winds and rain. The storms I chased last week had inflow winds rushing into them at 50+ MPH.
The key to shooting any thing fast moving, like this, is having whatever camera kit your using well configured. You need to be able to find the crucial controls for exposure and focus quickly and easily. You need to have a way of measuring and judging exposure and focus accurately. In addition you need a zoom lens that will allow you to get the kinds of shots you need, there’s no time to swap lenses!
For my storm chasing shoot I used the Sony F5 with R5 recorder. This was fitted with a Micron bridge plate as well as a Micron top cheese plate and “Manhandle”. Instead of the Sony viewfinder I used an Alphatron viewfinder as this has a waveform display for exposure. My general purpose lens was a Sigma 18-200mm f3.5-f6.5 stabilised lens with a Canon mount. To control the iris I used a MTF Effect iris control box. For weather protection a CamRade F5/F55 Wetsuit. The tripod I used for this shoot was a Miller 15 head with a set of Carbon Fibre Solo legs.
Storm chasing with a PMW-F5
Overall I was pleased with the way this setup worked. The F5’s ergonomics really help as the logical layout makes it simple to use. The 18-200mm lens is OK. I wish it was faster for shooting in low light but for the daytime and dusk shots, f3.5 (at the wide end) is OK. The F5 is so sensitive that it copes well even with this slow lens. The CamRade wetsuit is excellent. Plenty of clear windows so you can see the camera controls and a well tailored yet loose fit that allows you to get easy access to the camera controls. I’ve used Miller Solo legs before and when you need portability they can’t be beaten. The are not quite as stable as twin tube legged tripods, but for this role they are an excellent fit. The Miller 15 head was also just right. Not too big and bulky, not too small. The fluid motion of the head is really smooth.
Storm Chasing in the USA with the PMW-F5
So what didn’t work? Well I used the Element Technica Micron bridge plate. I really like the Micron bridge plate as it allows you to re-balance the camera on the tripod very quickly. But it’s not really designed for quick release, it’s a little tricky to line up the bridge plate with the dovetail so I ended up removing and re-fitting the camera via the tripod plate which again is not ideal. The Micron Bridge plate is not really designed for this type of application, when I go back storm chasing in May I’ll be using a baseplate that locks into a VCT-14 quick release plate, not sure which one yet, so I have some investigating to do. The VCT-14 is not nearly as stable or as solid as the Micron, but for this application speed is of the essence and I’m prepared to sacrifice a little bit of stability. The Micron bridge plate is better suited to film style shooting and in that role is fantastic, it’s just not the right tool for this job.
Rainbow under a severe thunderstorm.
The MTF-Effect unit is needed to control the aperture of the Canon mount lens, it also powers the optical image stabiliser. But it’s a large square box. I had it mounted on the top of the camera, not in the best place. I need to look at where to mount the box. I’m actually considering re-housing the unit in a custom made hand grip so I can use it to hold the camera with my left hand and have iris control via a thumbwheel. I also want to power it from one of the camera’s auxiliary outputs rather than using the AA batteries internally. The other option is the more expensive Optitek lens mount which I’m hoping to try out soon. I’m also getting a different lens. The Sigma was fine, but I’m going to get a Sigma 18-250mm (15x) f3.5-f6.5 for a bit more telephoto reach. The other option I could have used is my MTF B4 adapter and a 2/3″ broadcast zoom, but for 4K the Tamron will have better resolution than an HD lens. If I was just shooting HD then the broadcast lens would probably be the best option. After dark I swapped to my Sigma 24-70mm f2.8 for general purpose shooting and this worked well in low light but with the loss of telephoto reach, I need to look into a fast long lens but these tend to be expensive. If you have deep enough pockets the lens to get would probably be the Fujinon Cabrio 19-90 T2.9, but sadly at the moment my budget is blown and my pockets are just not that deep. The Cabrio is very similar to an ENG broadcast lens in that it has a servo zoom, but it’s PL mount and very high resolution. Another lens option would be the Canon CN-E30-105mm T2.8, but overall there isn’t a great deal of choice when it comes down to getting a big zoom range and large aperture at the same time, in a hand-held package. If I was working with a full crew then I would consider using a much larger lens like the Arri Alura 18-80 or Angenieux Optimo 24-290, but then this is no longer what I would consider run and gun and would require an assistant to set up the tripod while I bring out the camera.
A Supercell thunderstorm looking like a flying saucer.
From an operating point of view one thing I had to do was to keep reminding myself to double check focus. If you think focus is critical in HD, then it’s super critical for 4K. Thunderstorms are horrid things to try and focus on as they are low contrast and soft looking. I had to use a lot of peaking as well as the 1:1 pixel function of the Alphatron viewfinder, one of the neat things about the Alphatron is that peaking continues to work even in the 1:1 zoom mode. As I was shooting raw and using the cameras Cine EI mode to make exposure simpler I turned on the Look Up Tables on the HDSDI outputs and used the P1 LUT. I then exposed using the waveform monitor keeping my highlights (for example the brighter clouds) at or lower than 100%. On checking the raw footage back this looks to have worked well. Quite a few shots needed grading down by 1 to 1.5 stops, but this is not an issue as there is so much dynamic range that the highlights are still fine and you get a cleaner, less noisy image. When shooting raw with the F5 and F55 cameras I’d rather grade down than up. These cameras behave much more like film cameras due to the massive dynamic range and raw recording, so a little bit of overexposure doesn’t hurt the images as it would when shooting with standard gammas or even log. Grading down (bringing levels down) results in lower noise and a cleaner image.
Frame grab from the F5 of a Supercell storm with a grey funnel cloud beneath.
So you can run and gun in an intense fast moving environment with a large sensor camera. It’s not as easy as with a 2/3″ or 1/2″ camera. You have to take a little more time double checking your focus. The F5 is so sensitive that using a F3.5-F6.5 lens is not a huge problem. A typical 1/2″ camera (EX1, PMW-200) is rated at about 300 ISO and has an f1.8 lens. The F5 in Cine EI mode is 2000 ISO, almost 3 stops more sensitive. So when you put an f3.5 lens on, the F5 ends up performing better in low light, even at f6.5 it’s only effectively one stop less sensitive. For this kind of subject matter you don’t want to be at f1.8 – f2.8 with a super 35mm sensor anyway as the storm scenes and shots involved work better with a deep focus range rather than a shallow one.
Having watched the footage from the shoot back in HD on a large screen monitor I am delighted with the quality of the footage. Even in HD it has better clarity than I have seen in any of my previous storm footage. This is I believe down to the use of a 4K sensor and the very low noise levels. I’d love to see the 4K material on a 4K monitor. It certainly looks good on my Mac’s retina display. Hopefully I’ll get back out on the plains and prairies of Tornado Alley later in May for some more storm chasing. Anyone want to join me?
My facebook and twitter readers will already know that I’m in the USA doing a spot of storm chasing. The last few days have been a bit of a practice, chasing storms, jumping out of the car and getting the camera set up quickly (I’m shooting in 4K raw with the PMW-F5). I already have some really nice footage, but nothing spectacular yet, just some pretty scenic shots, some thunderstorms and a little bit of lightning. Today however looks to be a very interesting day. The Storm Prediction Center has issued a public advisory for severe weather including damaging winds, large hail and tornadoes. I’m right in the best place to catch some of the action and all set to go. However a tornado may only be 100m/100 yards across and the area where there is a 15% chance of tornadoes with 25 miles of a given point (a very high tornado probability) covers an area of about 50,000 square miles, that’s an area almost the size of the UK. So for the next couple of hours I will be looking at weather forecasting data and computer models trying to figure out exactly where to go.
There is a very serious risk of large and very violent tornadoes today across much of Oklahoma. The storms and tornadoes will be moving North East at about 25-35mph and possibly passing through some very heavily populated areas. In addition hail stones the size of tennis balls are expected along with straight line winds of up to 100 mph around some of the stronger storms. What’s going to cause all of this mayhem? Well the North of the US is currently sitting under some very cold winter air, it’s snowing about 300 miles north of here. Just to the south of me it’s hot, very hot and humid. High above strong jet stream winds are set to increase in strength today and dig down towards the south. As the jet stream digs south it will push that cold northern air south driving a cold front that will crash into the warm southern air. This cold front will push the warm air upwards and aided by the jet stream winds this creates atmospheric instability and lift that causes violent storms to break out along the front. The final ingredient is wind shear. Winds near the ground will be coming from the south towards the front while winds high above in the jet stream will be coming from the west. This will cause the stronger thunderstorms to start to spin and rotate which helps them to become much stronger turning them into “Supercells”. Supercell storms are responsible for almost all violent tornadoes.
So, wish me luck. Today will be a frantic and hectic day. Hopefully I’ll get some interesting 4K footage to share with you all very soon.
This is a really neat online tutorial (http://www.canonoutsideofauto.ca) on exposure, aperture and shutter and how they work together to create different looks to a photo. There is also a clever simulation of what you pictures would look like if taken with different settings. This is great for kids, photography beginners or those that want to take better pictures but are not used to using manual modes.
Here’s a link to my NAB presentation on the Sony PMW-F5 and F55. I give some insight into the cameras, sensors, shooting with Hypergammas or raw and a brief workflow demo.
Waveform and measurement options on the Convergent Design Odyssey7Q
Hot off the press from NAB is the announcement that the Convergent Design Odyssey 7Q will be able to work directly with the FS700 to record 4K and 2K with from the FS700 with 2K raw going up to 240fps. This is really great news. I’ll be posting a video blog about this later in the week. for now here is an extract from the official press release
LAS VEGAS, NAB Booth C11001, April 8, 2013 — Sony is announcing that its affordable 4K production camcorder, the NEX-FS700, will support a direct connection to Convergent Design’s new Odyssey7Q. The combination will allow recording of 2K RAW at up to 240 frames per second, with 4K video enabled via a single 3G connection.
The NEX-FS700 can achieve a high frame rate of up to 240 fps recording in 2K RAW while recording to the Odyssey 7Q. The 240 content fps is recording continuously, without windowing the imager or line doubling the signal. This assures full resolution at all times without windowing artifacts.
The direct connection to Convergent Design’s new Odyssey 7Q complements Sony’s own newly announced recording solution using the AXS-R5 RAW recorder and the new HXR-IFR5 interface unit. Now professional users have even more flexibility in choosing a workflow that meets their needs for today’s diverse client requirements.
“The addition of RAW and 4K recording unleashes the full power of the FS700’s state of the art 4K imager, vastly increasing the potential applications for the camcorder and resulting in tremendous flexibility in post-production,” said Peter Crithary, marketing manager for large sensor technology at Sony Electronics. “Now, interoperability with the exciting Odyssey 7Q in addition to our own recording technology gives users a wide range of cost effective choices when working with diverse workflows.”
I have to say that the more I use my F5 the more I’m coming to love the images it produces. However it has taken a little while to really find the best way to expose it when shooting raw.
In order to record raw the camera has to be in the Cine EI mode which means that the internal recordings use S-Log2. S-Log2 uses lower black values than S-log and as a result looks less flat, even though it has greater dynamic range, so judging exposure is a little easier but still tricky, especially if your not used to the way log looks.
One issue at the moment is the lack of any built in waveform, histogram or spot meter, these should come with a later firmware update. In the mean time there are several things you can do. You can use an external monitor with a waveform display (I use the Alphatron EVF which now has a waveform display). You can use a light meter or you can use one of the built in LUT’s and then use zebras as you would normally. If using the LUT’s please remember that at the moment when the LUT’s are ON not only is the LUT applied to the HDSDI and HDMI outputs, but it is also applied to the internal recordings, so you are no longer recording S-Log2. Sony define Middle grey for s-Log2 as 32% and when exposed like this the images do look quite under exposed but do grade very well. The issue is that while this is great for Log where you do need to try to keep skin tones and the mid ranges in the lower more linear part of the gamma curve it’s not ideal for raw. This is because the raw is linear raw. There is no highlight or mid range compression as with standard gammas or log, so where you put middle grey is much less important as skin tones and mid tones will grade equally well even if exposed very high.
By deliberately over exposing the raw you can minimise noise and the F5/F55 raw is incredibly tolerant of over exposure. So the camera behaves much like a film camera when shooting raw and IMHO often benefits from exposing brighter with raw than the S-Log optimum. So a waveform display that allows you to see where your highlights are helps judging raw exposure much better than just sticking mid grey at 32%. When you shoot with the majority of video cameras you are always conscious of protecting your highlights because over exposure looks really bad and makes grading a nightmare. With most cameras the limited dynamic range and the way traditional gamma curves compress the highlights means that many of us camera operators and DP’s will deliberately slightly under expose when shooting video.
But when you shoot linear raw and have 14 stops of dynamic range to play with it really is very different. There is so much over exposure headroom that you are hard pushed to over expose the camera anyway. Because each stop contains the full amount of data it really doesn’t matter where you place your exposure range. Provided the brightest parts of your scene are not actually clipping you can afford to push your exposure levels up. There really is no need to underexpose. Exposing brighter brings an added benefit and that is that after grading you will end up with less noise than a scene exposed darker. So for the F5 and F55 when shooting raw (just as with the F65) I favour exposing on the brighter side and this is how you would work with a film camera. Film doesn’t like under exposure, under expose film and it gets noisy and grainy, just like underexposing raw with an F5/F55.
When shooting raw I prefer not to use a LUT and use a waveform display to keep an eye on my highlights. I will typically expose S-Log2 so that my highlights sit just below 100%. Very often the log will look a little over exposed (but the raw will be fine). If use a grey card and set middle grey to the recommended 32%, very often this will result in a darker exposure than I will actually use for shooting raw. This does mean that the S-Log2 footage may not grade as well as it should being a bit over exposed, but the raw looks fantastic and I can minimise noise levels by shooting this way.
Cinematographer and film maker Alister Chapman's Personal Website
