The key benefit to shooting in V-Log is to have the ability to extend the camera’s dynamic range, and thus have greater latitude in post-production. In order to make the best possible use of shooting in V-Log, you have to expose properly otherwise there is a real chance of gaining unintentional digital noise in your image.
As those that have shot in Log would know, the video captured is not viewable as it appears, and instead appears to be washed and flat. This is because the camera sensor is preserving its maximum colour information possible, and thus the most malleable image when in post-production. This is the result of using the cameras’ logarithmic transfer curve (hence Log).
Whilst this curve has the advantage of gaining as much detail within the image, it has not been optimised for our vision, and is therefore much more demanding in terms of exposure. The impact of these factors is that it leads to significantly more work in post-production, however it benefits the end-result in providing an image that is both detailed and refined.
Most gamma curves found in cameras – known as scene files, such as ‘natural’, ‘standard’ and so on – have been designed to show what our vision perceives on-screen. The benefit of using scene files is less time spent on post-production. The major disadvantage of using these however is that the dynamic range of the image is reduced, with the gamma curve only picking up the most visible information, such as skin tones for example.
As a result, darker images and those that are highly exposed come off badly. Blacks are often crushed, and if you try to correct them in post-production then you will notice the appearance of ‘banding’ within the image. The advantage of shooting in V-Log therefore is that the ‘bit’ distribution between light and dark pixels is much more balanced, and that you are working with a conventional gamma curve that will look much more like an ‘S’.
V-Log therefore is a suitable mode for recording, but not necessarily for viewing. This poses a problem – how do you properly expose and complete a good white balance on an image that is flat?
LUTs – a brief overview
LUTs are used to modify between two images – the original and the displayed image, thanks to a mathematical calculation. LUTs are often used on reference monitors or external recorders when capturing content in Log, and as a result have become extremely common, with many editing or calibration softwares also capable of using them. Having a good LUT file in your viewing monitor will give you a good idea of what you are actually filming will look like once calibrated.
A LUT will verify the information related to the colour and gamma from the source and transform it, so that the display in which the LUT is loaded onto will correspond to the will of the operator. For each given input value there is only one output value that never changes.
So far so good . It seems pretty simple, right? We often forget that each camera model will assign specific colourimetry values to what its sensor will have spotted, therefore the same LUT will handle information from different cameras differently.
For example, if you are shooting with a DVX200 and want to get a realistic image to the best of the camera sensor’s capabilities, you would have to use a LUT that has been specifically developed for the DVX200 and not for another camera. In concrete terms, this means that even if the DVX200, VariCam and the GH5 all use V-Log, their sensors are all different and therefore behave differently, so the LUT applied to these three models may not necessarily produce exactly the same results.
That being said, you can still use the LUT of your choice because in the end the result must correspond to your taste, and if a LUT dedicated to another camera gives more satisfying or more creative results, then full steam ahead!
Using LUTs within V-Log
The use of LUTs provides a way of interpreting the image when on a shoot, and therefore acts as a first-step in post-production. Shooting in Log took a long time for camera operators to get used to when it was first introduced, because at the time there was no way to visualise the application of a LUT in real-time, and translate the washed-out image into something visible. Technology advances however have meant that LUTs can now be loaded in-camera – often via a SD card – and that their interpretation can be either enabled or disabled when on location.
The downside to using LUTs is that it automatically applies a given gamma curve, the result of which may not correspond to the way that you would have worked your images manually. As a result, LUTs can be seen as providing a preview that can be described as ‘reasonable’, but should not be totally relied upon since different calibration choices could have been made.
Lighting for V-Log
When using V-Log, it is essential to use the in-camera tools that you have available to adjust the lighting levels – namely the zebras, the oscilloscope, and the histogram. As we have already discussed, all gamma curves are different and so must be adjusted accordingly. For example, the levels that are often familiar for most to use (e.g. 70 IRE for flesh tones, 100 IRE for highlights) are simply not adaptable when using V-Log because the curve does not follow the same values for exposure as other Logs.
We have put together a simple example with the GH5 and use of V-Log Lite, which has a dynamic range of 12 stops whereas V-Log has a dynamic range of 14+ stops. Highlights will start to appear in the zebras and on the oscillograph at around 80 IRE. Nothing will be visible beyond 81 IRE, even if you open up the iris to its maximum and point it at the sun. This is because V-Log was of course originally designed for the VariCam and its dynamic range of 14+ stops, but with V-Log Lite’s dynamic range of 12 stops, it will always be limited to 80 IRE.
The right-hand exposure technique
We have suggested a number of techniques to support exposing when in Log. The first is the right-hand exposure technique, which involves the use of the histogram as a tool to demonstrate the distribution of brightness within an image. As a rule, the more the image tends towards the highlights, the more the curve of the histogram will move to the right. Those that use this technique argue that the noisiest areas of an image are in the darkest tones and shadows, and that by taking an image out of the dark area, it is possible to rework it in post-production.
In other words, the more we expose towards to right of the histogram, the more we will be able to obtain a detailed image. Remember our discussion around bit distribution: highlights and midtones are allocated the same number of bits, whereas dark areas have less bits available, therefore the fewer dark areas an image has then the more detailed its content will be. The drawback to this technique is the amount of time it takes in post-production to re-balance all the images, particularly with reducing the digital noise found within the images.
The 18% grey technique
18% grey – or middle grey as it is also known – is a standard in both video and photography, that reflects 18% of the light the image receives. It is very often used on test charts and is very easy to obtain in specialised stores. A reflection of 18% represents the average reflection of an average scene, hence the fact that automatic exposure of a camera is often designed to correctly expose areas of the image whose reflection is about 18%. Using this exposure technique, you will find that the oscillography and zebras are much more useful than the histogram.
As with the V-Log Lite example, unlike a conventional gamma curve, it doesn’t expose the mid-grays until about 50-55 IRE. It will actually expose at 42 IRE, so with a medium grey, you should see it appear on your oscillograph at about 42 IRE for a clean exposure.
For information, here are the reflection levels of the V-Log Lite mode:0% reflection (black) : 7.3 IRE
18% reflection (medium grey) : 42 IRE
90% reflection (white) : 61
IRE absolute white : 80 IRE
Exposing for a logarithmic gamma mode is not always as simple as setting your light levels at 18% and exposing at 42 IRE. From this, it leaves us wondering whether a fine balance between the right-hand exposure technique and the medium-grey exposure technique, would be the best solution despite it potentially proving complicated.
As a result, all decisions made on exposure will always offer a range of pros and cons. When exposing in V-Log we recommend the use of a basic LUT that you will use in post-production, the oscillograph to visualise your values, but also the zebra to have an on-screen alert for when you burn the information. False colours are another representation that gives colours relative to the exposure according to a scale, as shown below. The use of all of these tools provide you with a greater understanding of which IRE zone the colour information will be processed for each camera.