Sony have just released a firmware update for the PXW-FS7 and PXW-FS7 II cameras. This is a minor update with only one new feature being added which is the ability to alter the recorded ISO when shooting in the Cine-EI mode.
When the camera is set to CineEI, this new function is turned on and off in the menu under System – Base Settings – Rec/Out EI Applied.
What does it do?
When shooting normally using Cine EI, assuming that no LUT is applied to SDI1/Rec the camera always records at it’s base sensitivity (2000 ISO exposure rating) with no added gain. This is done to ensure that the cameras full dynamic range is always available and that the full recording range of either S-Log2 or S-Log3 is always available.
Then the EI system is used to apply a LUT just to the viewfinder or SDI 2 for monitoring. The gain of the LUT can then be changed to provide a brighter or darker viewfinder/monitor image. For example setting the EI to 1000 EI would make the viewfinder image darker than the base setting of 2000EI by 1 stop.
Because you are viewing this darker image you would then open the cameras aperture by 1 stop to compensate. Opening the aperture up results in a brighter recording. A brighter recording, achieved by putting more light onto the sensor will have less noise than a darker exposure, so the end result is brighter recorded images with less noise.
This process is often referred to as “rating” the camera and it is in many cases preferable to “rate” the FS7 around a stop slower (Viewfinder is darker, less sensitive, so that means you end up opening up the aperture) than the base 2000 ISO rating to gain a cleaner image that typically gives much greater flexibility in post production. So many users will set the EI on an FS7 to 1000 or 800 (It’s no co-incidence that I find get the sweet spot to be 800EI which happens to match the rating that Sony give the FS7 when shooting Rec-709).
Because in the normal EI mode there is no change to the cameras actual recording gain (the recordings take place at the equivalent of 2000 ISO) there is no change to the dynamic range. The camera will always capture 14 stops no matter what you set the EI to. However if you open the aperture by an extra stop (selecting 1000 EI, which results in a 1 stop darker viewfinder image, so to compensate you open up 1 stop) you move the mid point of the exposure up 1 stop. This means you will reduce the over exposure headroom by 1 stop but at the same time you gain one stop of under exposure range. You will see 1 stop further into the shadows, plus there will be less noise, so the shadow range becomes much more useable.
Conventional EI mode and Post Production.
In post production these brightly exposed images will need some degree of adjustment. If you are doing a virgin grade from scratch then you don’t really need to do anything extra or different, you will just grade it to taste.
If you are using a LUT you will need to either use an exposure compensated LUT (I always provide these in any of my free LUT sets) or you will need to correct the exposure before applying the LUT. If you apply a standard LUT and then try to correct the exposure the results will often not be satisfactory as the LUT determines many things such as where any highlight roll-off occurs. Correcting after this can result in washed out of flat looking skin tones. So really you need to make the exposure correction to the material before it is passed to the LUT.
No loss of dynamic range with conventional EI.
It’s worth noting that even though the levels are reduced to “normal” levels when applying an exposure compensated LUT or through grading this should not reduce the dynamic range. You do not just shift the range down (which would hard clip the blacks and cause a loss of DR at the low end). What you are typically doing is reducing the gain to bring the levels down and this allows the information in the new extended shadow range to be retained, so nothing is lost and your footage will still have 14 stops of DR along with nice clean shadows and mid tones.
Is this all too difficult?
However, some people find that the need to correct the exposure prior to adding the LUT difficult or time consuming (I don’t know why, they just do. It takes no longer to add a compensated LUT than a normal LUT). Or some people find it difficult to get a good looking image from brightly exposed footage (probably because they are grading after the LUT has been applied). For these reasons Sony have added the ability to bake the EI change directly into the recording by shifting the gain of the recordings to match the selected EI.
Rec/Out EI Applied:
So now if you enable Rec/Out EI Applied any change you make to the cameras EI settings will now also be applied as a gain change to the recordings. If you set the EI to 1000, then the recordings will take place at 1000 ISO and not 2000 ISO. This means that you do not have to make any exposure corrections in post production, just apply a standard LUT.
You will loose some of your dynamic range:
The down side to this is that you are now changing the gain of the camera. Changing the gain away from 0dB will reduce the dynamic range and affect the recording range. So, for example if you wish to shoot at 1000 EI have Rec/Out EI Applied you will be recording with -6dB gain and an effective ISO of 1000. You will have 1 stop less of shadow range as the cameras effective sensitivity is being reduced by 1 stop but the sensors clip/overload point remains the same. So when you open the aperture to compensate for the lower sensitivity you will have the same shadow range as base, but loose one stop off the top. The images will have less noise, but there will be no additional shadow information and a reduction in highlight range by 1 stop, the DR will be 13 stops.
Another side effect of this is that the peak recording level is also reduced. This is because the cameras clipping point is determined by the sensor. This sensor clip point is normally mapped to the peak recording level and the cameras noise floor is mapped to the black level (you can’t see things that are darker than the sensors noise floor no matter what level of gain you use as the noise will always be higher than the object brightness).
If you reduce the gain of the signal this level must decrease as a result. This means that S-Log3 which normally gets to around ~94% will now only reach ~85%, the change to S-Log2 is even greater (S-Log3’s peak recording level will reduce by 8.9% for every stop down you go, S-Log2 will reduce by 12% for each stop you go down ).
Raising the EI/ISO will also reduce the dynamic range as the gain is applied after the sensor. So the sensors clip point remains the same, so the brightest highlight it handles remains fixed. Adding gain after this simply means the recordings will clip earlier, but you will get a brighter mid range, brighter (but not more) shadows and a noisier picture.
What happens in post with Rec/Out EI Applied?
In post production these range and peak level changes mean that while a standard LUT will result in a correct looking mid range (because middle grey and skin tones will be at “normal” levels) there may be some problems with highlights never reaching 100% in the case of a low EI/ISO. Or being excessively clipped in the case of a high EI/ISO. Remember LUT’s are designed to work over very specific ranges. So if the input to the LUT doesn’t reach the peak level the LUT is expecting then the output from the LUT will also be reduced. So often there will still be the need to do some additional grading of highlights prior to the application of the LUT, or the need to use LUT’s designed specifically for each ISO rating (and the design of these LUT’s is more complex than a simple exposure offset).
Is it really easier?
So while this new feature will simplify the workflow for some situations where an alternate ISO/EI has been used – because exposure correction in post production won’t be needed. It may actually make things more difficult if you have bright highlights or need to be sure that your finished video meets expected standards where highlights are at 100%. You will still need to do some grading.
I don’t recommend that you use it.
Personally I do not recommend that you use this new feature. There are plenty of exposure compensated LUT’s available online (I have lots here). Tweaking the exposure of log footage in post production isn’t that difficult, especially if you use a color managed workflow. My guess is that this is aimed at FCP-X users where FCP-X applies a default LUT as standard. In this instance footage shoot with an offset exposure will look over/under exposed while footage shoot with the EI/ISO Applied will look normal (except for the highlights). So on the face of things the workflow may appear simpler. But you are loosing dynamic range and surely the primary reason for shooting with log is to maximise the dynamic range and gain the greatest possible post production flexibility. This new feature reduces dynamic range and as a result reduces post production flexibility.
Of course just because I don’t recommend it’s use, it doesn’t mean that you can’t or shouldn’t use it if it works for you, just make sure you fully understand what it is doing.
First: If there is something you wish to know about please try the search box to search this web site. You should find a search box on the left of every page when using a normal web browser. If you are using a mobile web browser then the search box will appear towards the bottom of each page, you may need to scroll down to find it, but it’s there.
A great function that allows you to check the extreme ends of your log exposure in the CineEI mode is the Hi/Low Key function. It has to be assigned to a button before you can use it, but it provides a very fast way to check what is really going on in your highlights and shadows. You’ll find all you need to know about High and Low key here.
Just a reminder that my guide to shooting with Cine EI for the PMW-F5 and F55 cameras is still just as valid today as it was when I wrote it back in 2013. There have been a few tweaks to the cameras menu here and there, but the principles and basic operation have not changed.
So if you are new to Cine-EI please take a look at the guide. It takes you through how to shoot with Cine EI, which LUT’s to use and how to expose them.
This is a question that comes up time and time again. I’ve been using the F5 and FS7 for almost 5 years. What I’ve discovered in that time is that the one thing that people notice more than anything from these cameras is noise if you get your exposure wrong. In addition it’s much harder to grade a noisy image than a clean one.
Lets take a look at a few key things about how we expose and how the F5/FS7 works (note the same principle applies to most log based cameras, the FS5 also benefits from being exposed brighter than the suggested base settings).
What in the image is important? What will your audience notice first? Mid-range, shadows or highlights?
I would suggest that most audiences first look at the mid range – faces, skin tones, building walls, plants etc. Next they will notice noise and grain or perhaps poor, muddy or murky shadows. The last thing they will notice is a few very brightly highlights such as specular reflections that might be clipped.
The old notion of protecting the highlights comes from traditional gamma curves with a knee or highlight roll off where everything brighter than a piece of white paper (90% white) is compressed into a very small recording range. As a result when shooting with conventional gamma curves ALL of the brighter parts of the image are compromised to some degree, typically showing a lack of contrast and texture, often showing some weird monotone colors. Log is not like that, there is no highlight roll off, so those brighter than white highlights are not compromised in the same way.
In the standard gammas at 0dB the PXW-FS7, like the PMW-F5 is rated at 800 ISO. This gives a good balance between noise and sensitivity. Footage shoot at 0dB/800ISO with the standard gammas or Hypergammas generally looks nice and clean with no obvious noise problems. However when we switch to log the native ISO rating of the cameras becomes 2000 ISO, so to expose “correctly” we need to stop the aperture down by 1.3 stops. This means that compared to 709 and HG1 to HG4, the sensor is being under exposed by 1.3 stops. Less light on the sensor will mean more noise in the final image. 1.3 stops is the equivalent of 9dB. Imagine how Rec709 looks if it is under exposed by 1.3 stops or has to have +9dB of gain added in. Well – thats what log at 2000 ISO will look like.
However log has lots of spare headroom and no highlight compression. So we can choose to expose brighter than the base ISO because pushing that white piece of paper brighter in exposure does not cause it to become compressed.
If you open the aperture back up by 1.3 stops you get back to where you would be with 709 in terms of noise and grain. This would be “rating” the camera at 800 ISO or using 800 EI. Rating the camera at 800EI you still have 4.7 stops of over exposure range, so the only things that will be clipped will in most cases be specular reflections or extreme highlights. There is no TV or monitor in existence that can show these properly, so no matter what you do, they will never be true to life. So don’t worry if you have some clipped highlights, ignore them. Bringing your exposure down to protect these is going to compromise the mid range and they will never look great anyway.
You should also be extremely cautious about ever using an EI higher that 2000. The camera is not becoming more sensitive, people are often misslead by high EI’s into thinking somehow they are capturing more than they really are. If you were to shoot at 4000 EI you will end up with footage 15dB noisier than if you were shooting the same scene using 709 at 800 ISO. That’s a lot of extra noise and you won’t necessarily appreciate just how noisy the footage will be while shooting looking at a small monitor or viewfinder.
I’ve been shooting with the F5 and then the FS7 for almost 5 years and I’ve never found a situation where I going to an EI higher than 800 would have resulted in a better end result. At the same time I’ve seen a lot of 2000 EI footage where noise in the mid range has been an issue, one particular example springs to mind of a high end car shoot where 2000 EI was used but the gloss and shine of the car bodywork is spoilt because it’s noisy, especially the darker coloured cars.
Of course this is just my opinion, based on my own experience, others may differ and the best thing you can do is test for yourself.
I have just completed a long overdue update to my guide to CineEI and S-Log2/S-Log3 on the PXW-FS7. I have made some changes to reflect the new expanded zebra range available in the latest firmware versions as well as the new waveform operation. I have also added in the two videos I have made on using CineEI. The first video covers setting up the Cine EI mode and how to expose S-Log2/3 and then the second video explains how to use the exposure index function to offset your exposure for less noise and grain.
So with the FS7 now shipping and the first units landing in peoples hands I have put together a comprehensive guide to using S-Log3 and CineEI on the PXW-FS7. Please follow this link to read or download the guide to CineEI on the PXW-FS7.
It’s important to note that S-Log3 has a peak recording level of 92IRE so never goes above this. Don’t be surprised to find that your overall levels are going to be much lower than you would normally use for conventional 709 shooting. In addition I can’t stress enough how important it is to learn how to use LUT’s (look up tables) in camera and in post production with this camera. It will make your life so much simpler and easier. LUT’s may sound complicated and difficult, but they are not. If you want to create your own LUT’s take a look at this guide here.
The FS7 is an incredibly powerful camera. But if you really want to get the most from the Cine-EI mode and S-Log then you need to adjust the way you shoot. You can’t just apply normal Rec-709 exposure levels to S-Log3, it’s not designed to work that way. However by using the 709(800) LUT on the viewfinder output you can expose based on the viewfinder image as you would normally, while the S-Log3 recordings will be at the correct levels. So do learn how to implement LUT’s correctly, it will make your life so much easier. Take a look at this video for an idea of how it works. The video features an F5 but the FS7 is the same.
While you’re at it you might also want to take a look at this article on the S-log3 gamma curve. Many people will look at the S-log and think that it looks noisy and be worried by this. You shouldn’t be. The shape of the log curve means that before grading and application of a LUT it can emphasise noise. However once you use a LUT to convert from S-log3 to 709 you will find that most of the noise will go away. Again, please use a LUT as simply trying to grade S-log3 in to 709 space is often not as effective as adding the right LUT. If you really know what you are doing, by using S-Curves and log grading tools it is possible to grade the native S-log3 in a 709 environment, but LUT’s do make it simpler. Another useful way to get from S-log3 to 709 is to use the new color chart tool in Resolve which recognises and corrects either a Macbeth chart or DSC One Shot chart to the correct levels automatically. When you set up this process in Resolve you will select the source gamma as S-Log3 so the correction compensates for the gamma curve as well as adding color correction. I’lll write this up in more depth in the next couple of weeks.
So enjoy your FS7 if you have one. As soon as mine arrives I will write up the correct way (or at least the designed way) to use the Cine-EI mode, in the mean time the F5/F55 Cine-EI guide can be used, the process is exactly the same on the FS7.
This guide to Cine-EI is based on my own experience with the Sony PMW-F5 and F55. There are other methods of using LUT’s and CineEI. The method I describe below, to the best of my knowledge, follows standard industry practice for working with a camera that uses EI gain and LUT’s.
If you find the guide useful, please consider buying me a beer or a coffee. It took quite a while to prepare this guide and writing can be thirsty work.
If you want you can download this guide as a PDF by clicking on this link: Ultimate Guide to CineEI on the PMW. I’d really appreciate a drink if your going to take away the PDF.
In this guide I hope to help you get the very best from the Cine EI mode on Sony’s PMW-F5 and PMW-F55 cameras. The cameras have two very distinct shooting mode, Cine EI and Custom Mode. In custom mode the camera behaves much like any other traditional video camera where what you see in the viewfinder is what’s recorded on the cards. In custom mode you can change many of the cameras settings such as gamma, matrix, sharpness to create the look you are after in-camera. “Baking-in” the look of your image in camera is great for content that will go direct to air or for fast turn around productions. But a baked-in look can be difficult to alter in post production. In addition it is very hard to squeeze every last drop of picture information in to the recordings in this mode.
The other mode, Cine-EI, is primarily designed to allow you to record as much information about the scene as possible. The footage from the camera becoming in effect a “digital negative” that can then be developed in post and the final highly polished look of the film or video created. In addition the Cine-EI mode mimics the way a film camera works giving the cinematographer the ability to rate the camera at different ISO’s to those specified by Sony. This can be used to alter the noise levels in the footage or to help deal with difficult lighting situations.
One further “non-standard” way to use Cine-EI is to use a LUT (Look Up Table) to create an in-camera look that can be baked in to the footage while you shoot. This offers an alternative to custom mode. Some users will find it easier to create a specific look for the camera using a LUT than they would by adjusting camera settings such as gamma and matrix.
MLUT’s, LUT’s and LOOK’s (all types of Look Up Tables) are only available in the Cine-EI mode.
THE SIMPLIFIED VERSION:
Before I go through all the “why’” and “hows” first of all let me just say that actually CineEI is easy. I’ve gone in to a lot of extra detail here so that you can full master the mode and the concepts behind it.
But in it’s simplest form, all you need to do is to turn on the MLUT’s. Choose the MLUT that you like the look of, or is closest to the final look you are after. Expose so that the picture in the viewfinder or on your monitor looks how you want and away you go.
Then in post production bring in your Slog footage. Apply the same LUT as you used when you shot and the footage will look as shot, only as the underlying footage is either raw or Slog you have a huge range of adjustment available to you in post.
THAT’S IT! If you want, it’s that simple.
If you want to get fancy you can create your own LUT and that’s really easy too (see the end of the document). If you want less noise in your pictures use a lower EI. I shoot using 800EI on my F5 and 640EI on the F55 almost all the time.
Got an issue with a very bright scene and strong highlights, shoot with a high EI.
Again, it’s really simple.
But anyway, lets learn all about it and why it works the way it works.
LATITUDE AND SENSITIVITY.
The latitude and sensitivity of the F5/F55, like most cameras is primarily governed by the latitude and sensitivity of the sensor. The latitude of the sensor in these cameras is around 14 stops. Adding different amounts of conventional camera gain or using different ISO’s does not alter the sensors actual sensitivity to light, only how much the signal from the sensor is amplified. Like turning up or down the volume on a radio, the sound level gets higher or lower, but the strength of the radio signal is just the same. Turn it up loud and not only does the music get louder but also any hiss or noise, the ratio of signal to noise does not change, they BOTH get louder. Turn it up too loud and it will distort. If you don’t turn it up loud enough, you can’t hear it, but the radio signal itself does not change. It’s the same with the cameras sensor. It always has the same sensitivity, but with a conventional camera we can add or take away gain (volume control?) to make the pictures brighter or darker (louder?).
Sony’s native ISO ratings for the cameras, 1250ISO for the F55 and 2000ISO for the F5 are values chosen by Sony that give a good trade off between sensitivity, noise and over/under exposure latitude. In general these native ISO’s will give excellent results. But there may be situations where you want or need different performance. For example you might prefer to trade off a little bit of over exposure headroom for a better signal to noise ratio, giving a cleaner picture. Or you might need a very large amount of over exposure headroom to deal with a a scene with lots of bright highlights.
The Cine EI mode allows you to change the effective ISO rating of the camera.
With film stocks the manufacturer will determine the sensitivity of the film and give it an Exposure Index which is normally the equivalent of the films measured ASA/ISO. It is possible for a skilled cinematographer to rate the film stock with a higher or lower ISO than the manufacturers rating to vary the look or compensate for filters and other factors. You then adjust the film developing and processing to give a correctly exposed looking image. This is a common tool used by cinematographers to modify the look of the film, but the film stock itself does not actually change it’s base sensitivity, it’s still the same film stock with the same base ASA/ISO.
Sony’s Cine EI and EI modes on Red and Alexa are very similar. While it has many similarities to adding conventional video camera gain, the outcome and effect can be quite different. If you have not used it before it can be a little confusing, but once you understand the way it works it is very useful and a great way to shoot. Again, remember that the actual sensitivity of the sensor itself never changes.
CONVENTIONAL VIDEO CAMERA GAIN.
Increasing conventional camera gain will reduce the cameras dynamic range as something that is recorded at maximum brightness (109%) at the native ISO would be pushed up above the peak recording level and into clipping if the conventional camera gain was increased because we can’t record a signal larger than 109%. But as the true sensitivity of the sensor does not change, the darkest object the camera can actually detect remains the same. Dark objects may appear a bit brighter, but there is still a finite limit to how dark an object the camera can actually see and this is governed by the sensors noise floor and signal to noise ratio (from the sensors own background noise). Any very dark picture information will be hidden in the sensors noise. Adding gain will bring up both the noise and darkest picture information, so anything hidden in the noise at the native ISO (or 0db) will still be hidden in the noise at a higher gain or ISO as both the noise and small signal are amplified by the same amount.
Using negative conventional gain or going lower than the native ISO may also reduce dynamic range as picture information very close to black may be shifted down below black when you subtract gain or lower the ISO. At the same time there is a limit to how much light the sensor can deal with before the sensor itself overloads. So even though reducing the ISO or gain may make the picture darker, the sensor clipping/overload point remains the same, so there is no change to the upper dynamic range, just a reduction in recording level.
As Sony’s Slog2 and Slog3 are tailored to capture the cameras full 14 stop range this means that when shooting with Slog2 or Slog3 the gamma curve will only work as designed and deliver the maximum dynamic range when the camera is at it’s native ISO. At any other recording ISO or gain level the dynamic range will be reduced. IE: If you were to use SLog2 or SLog3 with the camera in custom mode and not use the native ISO by adding gain or changing the ISO, you will not get the full 14 stop range that the camera is capable of delivering.
EXPOSURE LEVELS FOR DIFFERENT GAMMA CURVES AND CONTRAST RANGES.
It’s important to know (or better still understand) that different gamma curves with different contrast ranges will require different exposure levels. The TV system that we use today is currently based around a standard known as Rec-709. This standard specifies the contrast range that a TV set or monitor can show and which recording levels represent which display brightness levels. Most tradition TV cameras are also based on this standard. Rec-709 does have some serious restrictions, the brightness and contrast range is very limited as these standards are based around TV standards and technologies developed 50 years ago. To get around this issue most TV cameras use methods such as a “knee” to compress together some of the brighter part of the scene into a very small recording range.
As you can see in the illustration above only a very small part of the recording “bucket” is used to hold a moderately large compressed highlight range. In addition a typical TV camera can’t capture all of the range in many scenes anyway. The most important parts of the scene from black to white is captured more or less “as is”. This leaves just a tiny bit of space above white to squeeze in a few highly compressed highlights. The signal from the TV camera is then passed directly to the TV and as the shadows, mid range and skin tones etc are all at more or less the same level as captured the bulk of scene looks OK on the TV/Monitor. Highlights may look a little “electronic” due to the very large amount of compression used.
But what happens if we want to record more of the scenes range? As the size of the recording “bucket”, the codec etc, does not change, in order to capture a greater range and fit it in to the same space, we have to re-distribute how we record things.
Above you can see that instead of just compressing a small part of the highlights we are now capturing the full dynamic range of the scene. To do this we have altered the levels that everything is recorded at. Blacks and shadows are a little lower, greys and mids are a fair bit lower and white is a lot lower. By bringing all these levels down we make room for the highlights and the really bright stuff to be recorded without being excessively compressed.
The problem with this though is that when you output the picture to a monitor or TV it looks odd. It will lack contrast as the really bright stuff is displayed at the same brightness as the conventional 709 highlights. White is now only as bright as faces would be with a conventional TV camera and Faces are only a little bit above the middle grey level.
This is how Slog works. By re-distributing the recording levels we can squeeze a much bigger dynamic range into the same size recording bucket. But it won’t look quite right when viewed directly on a standard TV or monitor. It may look dark and certainly a bit washed out.
I hope you can also see from this that whenever the cameras gamma curve does not match that of the TV/Monitor the picture might not look quite right. Even when correctly exposed white may be at different levels, depending on the gamma being used, especially if the gamma curve has a greater range than the normal Rec-709 used in old school TV cameras.
THE CORRECT EXPOSURE LEVELS FOR SLOG-2 and SLOG-3.
Lets look at the correct exposure levels for SLog-2 and SLog-3. As these gamma curves have a very large dynamic range the recording levels that they use are different to the levels used by the normal 709 gamma curve used for conventional TV. As a result when correctly exposed, Slog looks dark and low contrast on a conventional monitor or in the viewfinder. The table below has the correct levels for middle grey and 90% reflectance white for the different types of Slog.
The white level in particular is a lot lower than we would normally use for TV gamma. This is done to make extra space above white to fit in the extended range that the camera is capable of capturing, all those bright highlights, bright sky and clouds and other things that cameras with a smaller dynamic range struggle to capture.
Let’s now take a look at how to set the correct starting point exposure for SLog-3. You can use a light meter if you wish, but if you do want to use a light meter I would first suggest you check the calibration of the light meter by using the grey card method below and comparing the what the light meter tells you with the results you get with a grey or white card.
The most accurate method is to use a good quality grey card and a waveform display. For the screen shots seen here I used a Lastolite EzyBalance Calibration Card. This is a pop up grey card/white card that fits in a pocket but expands to about 30cm/1ft across giving a decent sized target. It has middle grey on one side and 90% reflectance white on the other. With the MLUT’s off, set the exposure so that the grey card is exposed at the appropriate level (see table above).
Note: To get the waveform to display you must have BOTH the SDI SUB and Viewfinder MLUT’s OFF or BOTH ON. The waveform is turned on and off under the “VF” – “Display ON/OFF” – “Video Signal Monitor” settings of the main menu. Sadly the cameras built in waveform display is not the best so it may help to use an external monitor with a better waveform display.
If you don’t have access to a better waveform display you can use a 90% reflectance white card and zebras. By setting up the Zebras with a narrow aperture window of around 3% you can get a very accurate exposure assessment for various shades of white. For SLog-3 set the Zebras to 3% aperture and the level at 61%. Sadly the zebras don’t go below 60%. For Slog-2 using 60% will be accurate enough, a 1% error is not going to do any real harm. You can use exactly the same method for S-Log2 just by using the SLog-2 levels detailed in the chart above.
The image above shows the use of both the Zebras and Waveform to establish the correct exposure level for S-Log3 when using a 90% reflectance white card or similar target. Please note that very often a piece of white paper or a white car etc will be a little bit brighter than a calibrated 90% white card. If using typical bleached white printer paper I suggest you add around 4% to the white values in the above chart to prevent under exposure.
SO HOW DOES CINE-EI WORK?
Cine-EI (Exposure Index) works differently to conventional camera gain. It’s operation is similar in other cameras that use Cine-EI or EI gain such as the F5, F55, F3, F65, Red or Alexa. You enable Cine-EI mode in the camera menus Base Settings page. On the F5 and F55 it works in YPbPr, RGB and RAW modes.
IMPORTANT: In the Cine-EI mode the ISO of the recordings remains fixed at the cameras native ISO (unless baking in a LUT, more on that later). By always recording at the cameras native ISO you will always have 14 stops of dynamic range.
YOU NEED TO USE A LUT:
Important: For Cine-EI mode to work as expected you should monitor your pictures in the viewfinder or via the SDI/HDMI output through one of the cameras built in MLUT’s (Look Up Table), LOOK’s or User3D LUT’s. So make sure you have the MLUT’s turned on. If you don’t have a LUT then it won’t work as expected because the EI gain is applied to the cameras LUT’s. At this stage just set the MLUT’s to on for the Sub&HDMI output and the Viewfinder out.
EXPOSING VIA THE LUT/LOOK.
At the cameras native ISO (2000 on F5, 1250 on F55), when shooting via a LUT you should adjust your exposure so that the picture in the viewfinder looks correctly exposed. If the LUT is correctly exposed then so too will the S-log. As a point of reference, middle grey for Rec-709 and the 709(800) LUT should be at, or close to 42%.
This is really quite simple, generally speaking when using a LUT, if it looks right, it probably is right. However it is worth noting that different LUT’s may have slightly different optimum exposure levels. For example the 709(800) LUT is designed to be a very close match to the 709 gamma curve used in the majority of monitors, so this particular LUT is really simple to use because if the picture looks normal on the monitor then your exposure will also be normal.
The above images show the correct exposure levels for the 709(800) LUT. Middle grey should be 42% and 90% white is… well 90%. Very simple and you can easily use zebras to check the white level by setting them to 90%. As middle grey is where it normally is on a TV or monitor and white is also where you would expect to see it, when using the 709(800) LUT, if the picture looks right in the viewfinder then it generally is right. This means that the 709(800) LUT is particularly well suited to being used to set exposure as a correctly exposed scene will look “normal” on a 709 TV or monitor.
Many of the other LUT’s and Looks however capture a contrast range that far exceeds the Rec-709 standard used in most monitors. So you may need to adjust your exposure levels slightly to allow for this.
The LC709-TypeA Look is very popular as a LUT for the PMW-F5 and F55 as it closely mimics the images you get from an Arri Alexa (“type A” = type Arri).
The “LC” part of the Look’s name means Low Contrast and this also means – big dynamic range. Whenever you take a big dynamic range (lots of shades) and show it on a display with a limited dynamic range (limited shades) all the shades in the image get squeezed together to fit into the monitors limited range and as a result the contrast gets reduced. This also means that middle grey and white are also squeezed closer together. With conventional 709 middle grey would be 42% and white around 80-90%, but with a high dynamic range/low Contrast gamma curve white gets squeezed closer to grey to make room for the extra dynamic range. This means that middle grey will remain close to 42% but white reduces to around 72%. So for the LC709 Looks in the F5/F55 optimum exposure is to have middle grey at 42% and white at 72%. Don’t worry too much if you don’t hit those exact numbers, a little bit either way does little harm.
RECOMMENDED LUT EXPOSURE LEVELS.
Here are some white levels for some of the built in LUT’s. The G40 or G33 part of the HG LUT’s is the recommended value for middle grey. Use these levels for the zebras if you want to check the correct exposure of a 90% reflectance white card. I have also include an approximate zebra value for a piece of typical white printer paper.
709(800) = Middle Grey 42%. 90% Reflectance white 90%, white paper 92%.
HG8009(G40) = Middle Grey 40%. 90% Reflectance white 83%, white paper 86%.
HG8009(G33) = Middle Grey 33%. 90% Reflectance white 75%, white paper 80%.
The “LC709” LOOK’s = Middle Grey 42%. 90% Reflectance white 72%, white paper 77%.
DONT PANIC if you don’t get these precise levels! I’m giving them to you here so you have a good starting point. A little bit either way will not hurt. Again, generally speaking if it looks right in the viewfinder or on your monitor screen, it is probably close enough not to worry about it.
If you ever need to confirm the correct levels for any given LUT or Look it’s really easy. Put the camera in to CineEI. Turn OFF the MLUT’s (remember you can turn LUT’s on and off from the cameras side information screen by pressing the CAMERA menu button until you see the LUT options at the bottom of the display). With the MLUT’s OFF use a grey card or white card (or maybe your calibrated light meter) to set the exposure for the Slog curve you have chosen.
Once you have established the correct exposure for the SLog, without adjusting anything else, turn on the MLUT’s, ensure the camera is at the native ISO and choose the LUT or LOOK that you want to check. Now you can measure the grey and white point for the LUT/Look you have chosen and see on the monitor or in the viewfinder what the correct exposure looks like via the LUT. It probably won’t be vastly different from normal 709 in most cases, especially middle grey this tends to stay very close to 42%, but it’s useful to do this check if you are at all unsure.
If you can, use a LUT, not a LOOK. That’s my recommendation, not a hard and fast rule but if you’re new to CineEI, LUT’s and Looks please read on as to why I say this. Otherwise skip on to Baking-in the LUT/LOOK.
I recommend that for exposure evaluation you should normally try to use one of the cameras built in MLUT’s not the LOOK’s. Especially if you are new to LUT’s and Looks. This is because the LUT’s behave differently to the Looks when you use a high or low EI.
I found that when you lower the EI gain, below native, the output level of the LOOK lowers, so that depending on the EI, the clipping, peak level and middle grey values are different. For example on my PMW-F5 at 500 EI the LC709TypeA LUT has a peak output (clipping) level of just 90% while at 2000 ISO it’s 98%. This also means that middle grey of the LOOK will shift down slightly as you lower the EI. This means that for consistent exposure at different low EI’s you may need to offset your exposure very slightly (it is only very slight). It also means that at Native EI if the waveform shows peak levels at 90% you are not overexposed or clipped, but at low EI’s 90% will mean clipped Slog, so beware of this peak level offset with the LOOK’s.
When you raise the EI of the LOOKS, the input clipping point of the Look profile changes. For each stop of EI you add the LOOK will clip one stop earlier than the underlying Slog. For example set the LC709TypeA LUT to 8000 ISO (on my PMW-F5) and the LOOK itself hard clips 2 stops before the actual SLog3 clips. So your LOOK will make it appear that your Slog is clipped up to 2 stops before it actually is and the dynamic range and contrast range of the LOOK varies depending on the EI, so again beware.
So, the Looks may give the impression that the Slog is clipped if you use a high ISO and will give the impression that you are not using your full available range at a low ISO. I suspect this is a limitation of 3D LUT tables which only work over a fixed 0 to1 input and output range.
What about the 1D LUT’s? Well the built in LUT’s don’t cover the full range of the Slog curves so you will never see all of your dynamic range at all at once. However I feel their behaviour at low and high EI’s is a little bit more intuitive than the level shifts and early clipping of the LOOKs.
The 1D LUT’s will always go to 109%. So there are no middle grey shifts for the LUT, no need to compensate at any ISO. In addition if you see any clipping below 109% then it means your SLog is clipping, for example if you set the camera to 500 ISO (on an F5), when you see the 709(800) LUT clipping at 105% it’s because the Slog is also clipping.
At High ISO’s you won’t see the top end of the SLog’s exposure range anyway because the 1D LUT’s range is less than Slog’s range, but the LUT itself does not clip, instead highlights just go up above 109% where you can’t see them and this in my opinion is more intuitive behaviour than the clipped LOOK’s that don’t ever quite reach 100% and clip at lower than 100% even when the Slog itself isn’t clipped.
At the end of the day use the ones that work best for you, just be aware of the limitations of both and that the LUT’s and LOOKs behave very differently. I suggest you test and try both before making any firm decisions, but my recommendation is to use the LUT’s rather than the LOOK’s when judging exposure.
Using the built in MLUT’s?
There are 5 built in MLUT’s: P1 709(800), P2 HG8009G40, P3 HG8009G33, P4 Slog2, P5 Slog3. You can only select the Slog2 LUT when the camera is set to SLog2 in the base settings, the same for the SLog3 LUT. You can also create your own 1D LUT’s in Sony’s Raw Viewer software and user 3D LOOK’s in most grading suites but that’s a whole other subject that I’m not going to cover right here (see this article for user 3D Look creation or go to the additional LUT creation section at the bottom of this document), for now lets just consider the built in LUT’s.
All 3 of the other built in LUTs have an 800% exposure range. The camera itself has a 1300% exposure range when your shooting in CineEI, Raw, SLog2 or SLog3 (1300% more than standard gamma). So if you want to see your full exposure range then you should select SLog as your LUT or turn the LUT’s off. However the pictures will be flat looking and lack contrast, which makes accurate focussing harder and you must set your exposure using the SLog levels given above, so in addition Slog 2 will look dark. Note that if you press the “camera” button by the side LCD screen 2 times you can use the hot keys around the LCD to change LUT and turn the LUT’s on and off.
If you use MLUT’s P1, P2 orP3 then the viewfinder pictures will have near “normal” contrast. Your exposure levels will be more normal looking (although P3 should have middle grey at 33% so should look a touch darker than normal) but you won’t be seeing the full recorded range, only 800% out of the possible 1300% is displayed, so some things might look clipped in the viewfinder while the actual recording is not. I suggest switching the LUT’s off momentarily to check this, or connect a second monitor to the Main SDI to monitor the non LUT output. Do note that if using Slog2/Slog3 as a LUT and at a positive EI ISO you won’t see your full recording range. You will still see up to a stop more than the 800% LUT’s but at high EI’s the Slog2 or Slog3 LUT will clip slightly before the camera recordings. At low EI’s the Slog LUT’s will clip at the same time as the camera, but the level of the clipping point on any waveform display via the LUT output will be lower and this can be a little confusing. Unfortunately if you turn off the LUT’s you can’t get the cameras built in waveform display. This is where an external monitor with waveform becomes very handy to monitor the non LUT native ISO Slog output from the main HDSDI output.
Personally I prefer to use the 709(800) LUT for exposure as the restricted range matches that of most consumer TV’s etc so I feel this gives me a better idea of how the image may end up looking on a consumers TV. The slightly restricted range will help highlight any contrast issue that may cause problems in post. It’s often easier to solve these issues when shooting rather than leaving it to later. Also I find my Slog exposure more accurate as the LUT’s restricted range means you are more likely to expose within finer limits. In addition as noted above I fell the LUT’s behaviour is more predictable and intuitive at high and low EI’s than the LOOK’s.
BAKING IN THE LUT/LOOK.
When shooting using a high or low EI, the EI gain is added or subtracted from the LUT or LOOK, this makes the picture in the viewfinder or monitor fed via the LUT brighter or darker depending on the EI used. In Cine-EI mode you want the camera to always actually record the raw and S-Log2/S-log3 at the cameras native ISO (1250 ISO for F55 or 2000 ISO for F5). So normally you want to leave the LUT’s OFF for the internal recording. Just in case you missed that very important point: normally you want to leave the LUT’s OFF for the internal recording!
Just about the only exceptions to this might be when shooting raw or when you want to deliberately record with the LUT/Look baked in to your SxS recordings. By “baked-in” I mean with the gamma, contrast and color of the LUT/Look permanently recorded as part of the recording. You can’t remove this LUT/look later if it’s “baked-in”.
No matter what the LUT/Look settings, if you’re recording raw on the R5 raw recorder the raw is always recorded at the native ISO. But the internal SxS recordings are different. It is possible, if you choose, to apply a LUT/LOOK to the SxS recordings by setting the “Main&Internal” MLUT to ON. The gain of the recorded LUT/LOOK will be altered according to the CineEI gain settings. This might be useful to provide an easy to work with proxy file for editing, with the LUT/LOOK baked-in while shooting raw. Or as a way create an in-camera look or style for material that won’t be graded. Using a baked-in LUT/LOOK for a production that won’t be graded or only have minimal grading is an interesting alternative to using Custom Mode that should be considered for fast turn-around productions.
In most cases however you will probably not have a LUT applied to your primary recordings. If shooting in S-Log2 or S-Log3 you must set MLUT – OFF for “Main&Internal” See the image above. With “Main&Internal MLUT OFF” the internal recordings, without LUT, will be SLog2 or Slog3 and at the cameras native ISO.
You can tell what it is that the camera is actually recording by looking in the viewfinder. At the center right side of the display there is an indication of what is being recorded on the cards. Normally for Cine-EI this should say either SLog2 or Slog3. If it indicates something else, then you are baking the LUT in to the internal recordings.
CHANGING THE EI.
At the native ISO you have 6 stops of over exposure latitude. This is how much headroom your shot has. Your over exposure latitude is indicated on the cameras side LCD panel as highlight latitude.
REDUCING THE EI.
So what happens when you halve the EI gain to 1000EI? 1 stop of ISO will subtracted from the LUT. As a result the picture you see via the LUT becomes one stop darker (a good thing to know is that 1 stop of exposure is the same as 6db of gain or a doubling or halving of the ISO). So the picture in the viewfinder gets darker. But also remember that the camera will still be recording at the native ISO (unless baking-in the LUT).
As you can see from the side panel indication, the cameras highlight latitude decreases by 1 stop.
Why does this happen and whats happening to my pictures?
First of all lets take a look at the scene, as seen in the cameras viewfinder when we are at the native EI (This would be 1250 on the F55 and 2000ISO on an F5) and then with the EI changed one stop down so it becomes 500EI on F55 or 1000EI on the F5. The native ISO on the left, the one stop lower EI on the right.
So, in the viewfinder, when we lower our EI by one stop (halving the EI) the picture becomes darker by 1 stop. Note that if you were using the waveform display or histogram the indicated levels would also become lower. The waveform, histogram and zebras all measure the output from the LUTor the image seen in the viewfinder. So as this becomes one stop darker, they would also read 1 stop darker/lower. If using an external monitor with a waveform display connected to the SDI SUB out (SDI’s 3&4) or HDMI and the LUT is enabled for “Sub&HDMI” this too would get darker and the levels decrease by one stop.
What do you do when you have a dark picture? Well most people would normally compensate for a dark looking image by opening the iris to compensate. As we have gone one stop darker with the EI gain, to return the viewfinder image back to the same brightness as it was at the native EI you would open the iris by one stop.
If using a light meter you would start with the meter set at 2000/1250 ISO and set your exposure according to the what the meter tells you. Then you reduce your EI gain on the camera (the viewfinder image gets darker). Now you also change your ISO on the light meter to the new EI ISO. The light meter will then tell you to open the iris by one stop.
So now, after reducing the EI by one stop and then compensating by opening the iris by 1 stop, the viewfinder image is the same brightness as it was when we started.
But what’s happening to my recordings?
Remember the recordings, whether on the SxS card (assuming the Main&Internal LUT is OFF) or RAW always happens at the cameras native ISO (2000 on the F5 and 1250 on the F55), no matter what the EI is set to. As a result, because we opened the iris by 1 stop to compensate for the dark viewfinder or new light meter reading the recording will have become 1 stop brighter. Look at the image below to see what we see in the viewfinder alongside what is actually being recorded. The EI offset exposure as seen in the viewfinder (left hand side) looks normal, while the actual native ISO recording (right hand side) is 1 stop brighter.
How does this help us, what are the benefits?
When I take this brighter recorded image in to post production I will have to bring the levels back down to normal as part of the grading process. As I will be reducing my levels in post production by around 1 stop (6db) any noise in the picture will also be reduced by 6db. The end result is a picture with 6db less noise than if I had shot at the native ISO. Another benefit may be that as the scene was exposed brighter I will be able to see more shadow information.
Is there a down side to using a low EI?
Because the actual recorded exposure is brighter by one stop I have one stop less headroom. However the F5 and F55 have an abundance of headroom so the loss of one stop is often not going to cause a problem. I find that going between 1 and 1.5 EI stops down rarely results in highlight issues. But when shooting very high contrast scenes and using a low EI it is worth toggling the LUT on and off to check for clipping in the SLog image. This can be done from the side panel of the camera by pressing the “Camera” button until you see the MLUT controls on the side display and turning the MLUT on or off.
What is happening to my exposure range?
What you are doing is moving the mid point of your exposure range up in the case of a lower EI. This allows the camera to see deeper into the shadows but reduces the over exposure latitude. The reverse is also possible. If you use a higher EI you shift your mid point down. This gives you more headroom for dealing with very bright highlights, but you won’t see as far into the shadows and the final pictures will be a little noisier as in post production the overall levels will have to be brought up to compensate for the darker overall recordings.
Cine-EI allows us to shift our exposure mid point up and down. Lowering the EI gain gives you a darker VF image so you tend to overexpose the actual recording which reduces over exposure headroom but increases under exposure range (and improves the signal to noise ratio). Adding EI gain gives a brighter Viewfinder image which makes you underexpose the recordings, which gives you more headroom but with less underexposure range (and a worse signal to noise ratio).
When shooting raw information about the EI gain is stored in the clips metadata. The idea is that this metadata can be used by the grading or editing software to adjust the clips exposure level in the edit or grading application so that it looks correctly exposed (or at least exposed as you saw it in the viewfinder via the LUT). The metadata information is recorded alongside the XAVC or SSTP footage when shooting SLog2/3. However, currently few edit applications or grading applications use this metadata to offset the exposure, so S-Log2/3 material may look dark/bright when imported into your edit application and you may need to add a correction to return the exposure to a “normal” level. As the footage is log you should use log corrections to get the very best results. As an alternative you can use a correction LUT to move the exposure up and down like the ones for SLog2 available here, created by cameraman and DP Ben Turley. http://www.turley.tv. Sony’s Raw Viewer software does correctly read the Slog2/3 metadata and will automatically add any required offsets. If shooting raw the majority of grading and editing applications will correctly read the metadata in the raw footage and apply the correct exposure offset, so raw normally looks correctly exposed.
If shooting raw then you may choose to add the S-Log2/3 LUT to your internal SxS recordings. This will then add the EI-gain to the internal SxS recordings, so they will become brighter/darker as if applying actual gain while only the raw recordings remain at the native ISO. This may be useful if you wish to use the SxS recordings as a proxy file for the edit and would like the proxy files to look similar to the way the final footage will look after grading and correction for the EI offset.
WHAT IF YOU ARE SHOOTING USING HFR (High Frame Rate) AND LUT’S CANT BE USED.
In HFR you can either have LUT’s on for everything including internal recording, or all off, not LUT at all. This is not helpful if your primary recordings are internal SLog.
So if you can’t use the LUT’s you can use the VF High Contrast mode. Sadly this is only available in the viewfinder, but I find that it is much more obvious if your exposure is off when you use the VF High Contrast mode.
The VF High Contrast Mode acts as a 709(800) LUT for the viewfinder only. So expose at the native ISO, by eye, using normal 709 type levels and your Slog-3 should be pretty close to perfect.
The camera automatically turns this mode OFF when you power the camera down, so you must re-enable it when you power cycle the camera. This is probably a good thing as it means you shouldn’t accidentally have it turned on.
Sadly zebras etc either measure the LUT output or the Slog, they are NOT effected by the viewfinder HC mode, so in HFR they will be measuring the SLog. Also if the LUT’s are off then you can’t use different EI gains.
CineEI allows you to “rate” the camera at different ISO.
You MUST use a LUT for CineEI to work as designed.
A low EI number will result in a brighter exposure which will improve the signal to noise ratio giving a cleaner picture or allow you to see more shadow detail. However you will loose some over exposure headroom.
A high EI number will result in a darker exposure which will improve the over exposure headroom but decrease the under exposure range. The signal to noise ratio is worse so the final picture may end up with more noise.
A 1D LUT will not clip and appear to overexpose as readily as a 3D LOOK when using a low EI, so a 1D LUT may be preferable.
When viewing via a 709 LUT you expose using normal 709 exposure levels. Basically if it looks right in the viewfinder or on the monitor (via the 709 LUT) it almost certainly is right.
When I shoot with my F5 I normally rate the camera as 800EI. I find that 5 stops of over exposure range is plenty for most situations and I prefer the decrease in noise in the final pictures. I rate the F55 similarly at 640EI. But please, test and experiment for yourself.
QUICK GUIDE TO CREATING YOUR OWN LOOK’s (Using DaVinci Resolve).
It’s very easy to create your own 3D LUT for the Sony PMW-F5 or PMW-F55 using DaVinci Resolve or just about any grading software with LUT export capability. The LUT should be a 17x17x17 or 33x33x33 .cube LUT. This is what Resolve creates by default and .cube LUT’s are the most common types of LUT in use today.
First simply shoot some test Slog3 clips at the cameras native ISO. You must use Slog3 if you want to use User 3D LOOK’s in the camera. In addition you should also use the same color space for the test shot as you will when you want to use the LUT. I recommend shooting a variety of clips so that you can asses how the LUT will work in different lighting situations.
Import and grade the clips from the test shoot in Resolve creating the look that you are after for your production or as you wish your footage to appear in the viewfinder of the camera. Then once your happy with the look of the graded clip, right click on the clip in the timeline and “Export LUT”. Resolve will then create and save a .cube LUT.
Then place the .cube LUT file created by the grading software on an SD card in the PMWF55_F5 folder. You may need to create the following folder structure on the SD card. So first you have a PRIVATE folder, in that there is a SONY folder and so on.
PRIVATE : SONY : PRO : CAMERA : PMWF55_F5
Put the SD card in the camera, then go to the “File” menu and go to “Monitor 3D LUT” and select “Load SD Card”. The camera will offer you a 1 to 4 destination memory selection, choose 1,2,3 or 4, this is the location where the LUT will be saved. You should then be presented with a list of all the LUT’s on the SD card. Select your chosen LUT to save it from the SD card to the camera.
Once loaded in to the camera when you choose 3D User LUT’s you can select between user LUT memory 1,2,3 or 4. Your LUT will be in the memory you selected when you copied the LUT from the SD card to the camera.
Camera setup, reviews, tutorials and information for pro camcorder users from Alister Chapman.