Here are two sets of LUT’s for use in post production with the PXW-FS7, PMW-F5 and PMW-F55.
These LUT’s are based around the Sony 709(800) LUT and the Sony LC-709TypeA LUT (Arri Alexa look). But in addition to the base LUT designed for when you shoot at the native ISO there are LUTs for when you shoot at a lower or higher EI.
When you shoot at a high or low EI the resulting footage will be either under or over exposed when you add the standard LUT. These LUT’s include compensation for the under or overexposure giving the best possible translation from SGamut3.cine/S-log3 to rec-709 or the Alexa look and result in pleasing skin tones and a nice mid range with minimal additional grading effort.
If you find these LUT’s useful please consider buying me a coffee or beer.
It’s very easy to create your own User3D LUT for the Sony PXW-FS7 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).
Simply shoot a correctly exposed Slog3 clip at the native ISO. You must use the same Gamut as you intend to use in any other productions that will use this LUT, I recommend SGamut3.cine.
Import the clip in to Resolve and grade it as you wish the final image to look. A very good way to do this is to include a MacBeth chart or DSC Labs OneShot or CDM chart within the test shot. Then use Resolves excellent chart matching tool to get create a starting point for the grade. The big benefit of this approach is that Resolve will provide a very good gamma correction moving your footage from Slog3 to Rec-709. Once you have used the chart to move you into the correct gamma range just tweak and fine tune the image to get your desired look. Then once your happy with your look, right click on the clip in the timeline and “Export LUT”. Resolve will then create 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.
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.
NOTE: If you use a Mac to copy the LUT to the SD card the Mac will place some junk files on the SD card with almost exactly the same name as the LUT. The camera will see this junk file and display it in the list of LUT files. The junk file will have a “._” before the LUT name. Don’t try to load these junk files by mistake, they won’t work. Make sure you scroll down the LUT list to the real LUT file when you try to load the LUT’s in to the camera.
A very useful feature not well documented on the FS7, F5 and F55 cameras (and the F65 too) is the High/Low Key feature.
The High/Low Key function works by changing the brightness range of the image displayed in the viewfinder, this is very useful when shooting in the Cine-EI Mode and using a LUT to help judge your exposure.
My preferred LUT for exposure assessment is the 709(800) LUT. As this LUT is compatible with the gamma curve used in most TV’s and monitors it provides a nice contrasty image with what I would call “normal” brightness levels (middle grey at 42%, white at 90%, skin tones around 60-70%). So if you expose via the 709(800) LUT so that the pictures look right on the screen or in the viewfinder then your S-Log recordings will also be correctly exposed.
But the 709(800) LUT, like most LUT’s cannot show the full 14 stop capture range of the the S-Log recordings. So sometimes you might see an image via the LUT that looks correctly exposed but the highlights might look clipped or blown out as they are beyond the range of what the LUT can show as in the image below where the sky looks blown out. This is where the High/Low Key function comes in to play.
To access the function you have to assign High/Low Key to one of the cameras assignable buttons. Once assigned to a button on the first press of the button the viewfinder or monitor image will show the High Key parts of the shot. To do this the VF or monitor picture is made darker so that you can “see” into the full highlight capture range. “High Key” will be displayed in the top left hand corner of the viewfinder. As you can see in the image below we can now see that the sky is not blown out, so we know the S-log recording will be OK.
The second press of the button shows the Low Key (darker) parts of the scene. This is done by making the image much brighter so you can “see” into the shadows better and the entire under exposure range of what is being recorded is shown. “Low Key” is displayed in the top left of the viewfinder screen.
The third press of the button returns the image to the normal range that the LUT can show.
So by using the High / Low Key function you can see the entire range that the camera is capturing, check for over exposure or under exposure issues without having to turn the LUT on or off. This is a really useful function that I recommend you take advantage of when shooting with CineEI and LUT’s. However do remember to make sure you are back to the standard view range when setting your exposure level.
I have created two new sets of film look LUT’s for the FS7, F5 and F55. These LUT’s should be used in conjunction with S-Log3 and SGamut3.cine. They will give you a high contrast film like look that mimics Kodak film stock with slightly blue dark and mid tones and warm skin tones. The are 3D cube LUT’s and the two sets include LUT’s with compensation for under and over exposure.
To use the LUT’s in camera on the PXW-FS7, PMW-F5 or PMW-F55 then the LUT’s need to be placed in the PMWF55_F5 folder on an SD card. The correct folder structure should look like this:
I was lucky enough to get a chance to go out and shoot with a pre-production PXW-FS7 in Amsterdam during IBC. Guess what? It makes some very nice pictures!
In case you’ve had your head in the sand the last couple of weeks the PXW-FS7 is a new super35mm camcorder from Sony. It uses the same sensor as the Sony PMW-F5 and a lot of the camera is, I am sure, shared with the F5. Even the menu’s are almost exactly the same. It can record 4K internally on XQD cards using Sony’s XAVC codec. When the cameras start shipping next month you will be able to record 3840×2160 UHD/QuadHD as well as HD. Next year there will be an update to add 4096×2160 at up to 60fps.
Want to shoot slow motion? That’s no problem as the camera can go up to 180fps internally in HD and if you add an external raw recorder you can stretch that out to 240fps.
The XAVC codec options are great. You can choose between I frame for easy editing or long GoP which gives a smaller file size but needs more processing power to decode. The 10 bit 422 image quality is very similar in both cases, so choose which to use based on how much recording media you have and how powerful your edit machine is. If you still need the legacy HD XDCAM Mpeg codec then you have that too.
By adding the optional extension box to the rear of the camera you can even record ProRes HQ to the XQD cards (after a firmware update early next year). The extension box also adds the raw output needed to record raw to an external recorder such as The Odyssey 7Q or Sony R5 recorder. On top of that you also gain Timecode in and out plus genlock. To power all of this (and the camera) the extension box has a V-Mount battery plate on it’s rear. When not using the extension box the camera runs off BP-U type batteries, the same 12V batteries as used by an EX1 or PMW200 etc.
The FS7 has two different shooting modes. In custom mode the camera behaves pretty much like any other conventional camera where what you see in the viewfinder is what’s recorded on the cards. You can alter the cameras gamma curve, matrix and other settings, but basically what you see is what you get. The other mode is the CineEI mode (just like an F5 or F55) and in this mode the camera records using SGamut3.cine and S-Log3. The aim being to capture the maximum possible dynamic range and in this mode the cameras sensitivity is locked to it’s native ISO of 2000. As S-Log3 results in a very flat picture (that’s great for grading and post work) the camera includes the ability to add a range of Look Up Tables (LUT’s) to the viewfinder or HDSDI output. LUT’s help you better judge exposure and give a more pleasing image prior to grading. You can even generate your own LUT’s in software such as Resolve and load them in to the camera. For exposure assistance the camera has a range of tools including a waveform, vectorscope or histogram display as well as zebras.
Ergonomically the camera is very interesting. It has Sony’s E-Mount lens mount so you can use just about any lens you want simply by adding a lens adapter. Using a metabones or Commlite adapter you can use Canon EF lenses with ease. Likewise PL or Nikon lenses with the appropriate adapters.
Designed to sit on the front of your shoulder and supplied with a handgrip on an adjustable arm (attached via a standard Arri type rosette) the camera is easy to use. There are a couple of assignable buttons on the hand grip as well as a small joystick for navigating through the cameras menu system. A large zoom rocker will control any E-Mount zoom lenses used such as the new 28-135mm f4 lens and a further assignable dial wheel can be used to control the lenses aperture or other functions. The hand grip uses the LanC protcol so it should be possible to use other LanC devices with this camera.
The camera is a little front heavy as it sits on the front of your shoulder. When you add the extension box and a V-mount battery the balance is much better as the weight is now set much further back. With a 3rd party shoulder mount such as the new Vocas one or the dedicated Sony VCT-FS7 mount the camera can be turned into a true shoulder mount camera.
The LCD viewfinder is mounted on a thin arm that gives it forwards and backwards adjustment as well as up and down adjustment, but there is no left right adjustment.
Overall I think the viewfinder is the weakest part of this camera. The images in the VF are quite reasonable (its 940×560 resolution) but the mounting mechanism and loupe are not the best. Maybe this will be improved before the camera ships. I made a lot of use of one of the hand grip assignable buttons to provide focus magnification while shooting to ensure focus was spot on and it’s nice to have the focus mag function so easily accessible.
One issue I did find with the arm for the hand grip was that unless you fold it up out of the way you can’t slide the camera on and off a tripod. If you are using a base plate this is less of a problem but with a bare camera it’s a bit of a pain.
I found the operation of the camera almost identical to the PMW-F5. There are some differences however. The FS7 does not have a 2K center scan mode for the sensor. This is used on the F5/F55 to eliminate aliasing problems when shooting above 60fps where the 4K sensor is read out as a 2K sensor. On the F5/F55 if you don’t want to use the 2K center scan mode you can fit a special 2K low pass optical filter to eliminate aliasing above 60fps, but again this is not possible on the FS7.
Another thing the FS7 doesn’t have is the large side display of the F5 and F55. For conventional shooting this is not really a big deal. But if you are using the CineEI mode where you may be using LUT’s on different outputs not having this information clearly displayed is a bit of a nuisance. In fact during the shoot with the FS7 at one point I though I was shooting with a LUT when in fact I was not. The only way to be sure of how everything is set is to go into the cameras menu system.
But what about the image quality? Frankly it’s amazing! For the money the images this camera produces are remarkable. It is using the F5’s sensor and it does have 14 stops of dynamic range. S-log3 is a great gamma curve and the camera is very low noise, even at it’s native 2000 ISO. It was hard to tell as most of the shooting took place at night, but initially it doesn’t look like there is any difference between the quality of the footage from the FS7 and the PMW-F5. Great colours, low noise, high dynamic range with very pleasing roll off what more can you want? One area where there will be a difference is with raw. The PMW-F5 takes the Sony R5 directly docked on it’s back. The raw form the F5 is 16 bit while the raw from the FS7 is going to be recorded on an external recorder at only 12 bits. 12 bit linear raw is really pushing the limits of what is needed for linear raw. However we do already know that the 12 bit raw from Sony’s FS700 works well, so this should be no different.
Where this camera will be really good is when combined with the new 28-135mm f4 servo zoom lens. Typically par-focal lenses with this kind aperture and zoom ratio cost in excess of $30K. This lens will be around $2.7K. Being able to zoom in and out on a large sensor camera smoothly really increases the cameras flexibility making it much easier to use in run and gun type situations. The lens is never going to be an incredible performer at this price and when wide open I did find it a little soft, but for shear ease of use it’s really remarkable. The FS7 combined with this lens will be a killer combination and that’s why I have ordered one. It’s NOT replacing my F5, I love my F5 and I think that the F5 is a much better camera for drama or studio type shoots. But the FS7 will be very handy for fast and fluid productions. In addition, for the money this camera is an absolute bargain.
First of all. You can use either, LUT’s or Looks. But there is a quite marked difference in the way they behave, especially if you use EI gain.
At the native ISO there is little to choose between them. But just to confirm my earlier suspicions about the way the 3D LOOK’s behave I ran a quick test.
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 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.
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 LUT’s don’t cover the full range of the Slog curves so you will never see all of your dynamic range 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 LUT’s range is less than Slog’s range, but the LUT itself does not clip, instead highlights just go up above 109% and this is in my opinion 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.
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. 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.
In addition the higher contrast makes focus easier. I will often switch in and out of the LUT to look at how the Log is coping with any over exposure. This is my personal preference, but I do also use other LUT’s and Looks in particular the 709TypeA from time to time.
Well I have set myself quite a challenge here as this is a tough one to describe and explain. Not so much perhaps because it’s difficult, but just because it’s hard to visualise, as you will see.
First of all the dictionary definition of Gamut is “The complete range or scope of something”.
In video terms what it means is normally the full range of colours and brightness that can be either captured or displayed.
I’m sure you have probably heard of the specification REC-709 before. Well REC-709, short for ITU-R Recommendation, Broadcast Television, number 709. This recommendation sets out the display of colours and brightness that a television set or monitor should be able to display. Note that it is a recommendation for display devices, not for cameras, it is a “display reference” and you might hear me talking about when things are “display referenced” ie meeting these display standards or “scene referenced” which would me shooting the light and colours in a scene as they really are, rather than what they will look like on a display.
Anyway…. Perhaps you have seen a chart or diagram that looks like the one below before.
Now this shows several things. The big outer oval shape is what is considered to be the equivalent to what we can see with our own eyes. Within that range are triangles that represent the boundaries of different colour gamuts or colour ranges. The grey coloured triangle for example is REC-709.
Something useful to know is that the 3 corners of each of the triangles are whats referred to as the “primaries”. You will hear this term a lot when people talk about colour spaces because if you know where the primaries (corners) are, by joining them together you can find the size of the colour space or Gamut and what the colour response will be.
Look closely at the chart. Look at the shades of red, green or blue shown at the primaries for the REC-709 triangle. Now compare these with the shades shown at the primaries for the much larger F65 and F55 primaries. Is there much difference? Well no, not really. Can you figure out why there is so little difference?
Think about it for a moment, what type of display device are you looking at this chart on? It’s most likely a computer display of some kind and the Gamut of most computer displays is the same size as that of REC-709. So given that the display device your looking at the chart on can’t actually show any of the extended colours outside of the grey triangle anyway, is it really any surprise that you can’t see much of a difference between the 709 primaries and the F65 and F55 primaries. That’s the problems with charts like this, they don’t really tell you everything that’s going on. It does however tell us some things. Lets have a look at another chart:
This chart is similar to the first one we looked at, but without the pretty colours. Blue is bottom left, Red is to the right and green top left.
What we are interested in here is the relationship between the different colour space triangles. Using the REC-709 triangle as our reference (as that’s the type of display most TV and video productions will be shown on) look at how S-Gamut and S-Gamut3 is much larger than 709. So S-Gamut will be able to record deeper, richer colours than 709 can ever hope to show. In addition, also note how S-Gamut isn’t just a bigger triangle, but it’s also twisted and distorted relative to 709. This is really important.
You may also want to refer to the top diagram as well as I do my best to explain this. The center of the overall gamut is white. As you draw a line out from the center towards the colour spaces primary the colour becomes more saturated (vivid). The position of the primary determines the exact hue or tone represented. Lets just consider green for the moment and lets pretend we are shooting a shot with 3 green apples. These apples have different amounts of green. The most vivid of the 3 apples has 8/10ths of what we can possibly see, the middle one 6/10ths and the least colourful one 4/10ths. The image below represents what the apples would look like to us if we saw them with our eyes.
If we were shooting with a camera designed to match the 709 display specification, which is often a good idea as we want the colours to look right on the TV, the the greenest, deepest green we can capture is the 709 green primary. lets consider the 709 green primary to be 6/10ths with 10/10ths being the greenest thing a human being can see. 6/10ths green will be recorded at our peak green recording level so that when we play back on a 709 TV it will display the greenest the most intense green that the display panel is capable of. So if we shoot the apples with a 709 compatible camera, 6/10ths green will be recorded at 100% as this is the richest green we can record (these are not real levels, I’m just using them to illustrate the principles involved) and this below is what the apples would look like on the TV screen.
So that’s rec-709, our 6/10ths green apple recorded at 100%. Everything above 6/10 will also be 100% so the 8/10th and 6/10ths green apples will look more or less the same.
What happens then if we record with a bigger Gamut. Lets say that the green primary for S-Gamut is 8/10ths of visible green. Now when recording this more vibrant 8/10ths green in S-Gamut it will be recorded at 100% because this is the most vibrant green that S-Gamut can record and everything less than 8/10 will be recorded at a lower percentage.
But what happens if we play back S-Gamut on a 709 display? Well when the 709 display sees that 100% signal it will show 6/10ths green, a paler less vibrant shade of green than the 8/10ths shade the camera captured because 6/10ths is the most vibrant green the display is capable of. All of our colours will be paler and less rich than they should be.
So that’s the first issue when shooting with a larger colour Gamut than the Gamut of the display device, the saturation will be incorrect, a dark green apple will be pale green. OK, that doesn’t sound like too big a problem, why don’t we just boost the saturation of the image in post production? Well if the display is already showing our 100% green S-Gamut signal at the maximum it can show (6/10ths for Rec-709) then boosting the saturation won’t help colours that are already at the limit of what the display can show simply because it isn’t capable of showing them any greener than they already look. Boosting the saturation will make those colours not at the limit of the display technology richer, but those already at the limit won’t get any more colourful. So as we boost the saturation any pale green apples become greener while the deep green apples stay the same so we loose colour contrast between the pale and deep green apples. The end result is an image that doesn’t really look any different that it would have done if shot in Rec-709.
But, it’s even worse that just a difference to the saturation. Look at the triangles again and compare 709 with S-Gamut. Look at how much more green there is within the S-Gamut colour soace than the 709 colour space compared to red or blue. So what do you think will happen if we try to take that S-Gamut range and squeeze it in to the 709 range? Well there will be a distinct colour shift towards green as we have a greater percentage of green in S-Gamut than we should have in Rec-709 and that will generate a noticeable colour shift and the skewing of colours.
This is where Sony have been very clever with S-Gamut3. If you do take S-Gamut and squeeze it in to 709 then you will see a colour shift (as well as the saturation shift discussed earlier). But with S-Gamut3 Sony have altered the colour sampling within the colour space so that there is a better match between 709 and S-Gamut3. This means that when you squeeze S-Gamut3 into 709 there is virtually no colour shift. However S-Gamut3 is still a very big colour space so to correctly use it in a 709 environment you really need to use a Look Up Table (LUT) to re-map it into the smaller space without an appreciable saturation loss, mapping the colours in such a way that a dark green apple will still look darker green than a light green apple but keeping within the boundaries of what a 709 display can show.
Taking this one step further, realising that there are very few, if any display devices that can actually show a gamut as large as S-Gamut or S-Gamut3, Sony have developed a smaller Gamut known as S-Gamut3.cine that is a subset of S-Gamut3.
The benefit of this smaller gamut is that the red green and blue ratios are very close to 709. If you look at the triangles you can see that S-Gamut3.cine is more or less just a larger version of the 709 triangle. This means that colours shifts are almost totally eliminated making this gammut much easier to work with in post production. It’s still a large gamut, bigger than the DCI-P3 specification for digital cinema, so it still has a bigger colour range than we can ever normally hope to see, but as it is better aligned to both P3 and rec-709 colourists will find it much easier to work with. For productions that will end up as DCI-P3 a slight saturation boost is all that will be needed in many cases.
So as you can see, having a huge Gamut may not always be beneficial as often we don’t have any way to show it and simply adding more saturation to a seemingly de-saturated big gamut image may actually reduce the colour contrast as our already fully saturated objects, limited by what a 709 display can show, can’t get any more saturated. In addition a gamut such as S-Gamut that has a very different ratio of R, G and B to that of 709 will introduce colour shifts if it isn’t correctly re-mapped. This is why Sony developed S-Gamut3.cine, a big but not excessively large colour space that lines up well with both DCI-P3 and Rec-709 and is thus easier to handle in post production.
It’s been brought to my attention that there is a lot of concern about the apparent noise levels when using Sony’s new Slog3 gamma curve. The problem being that when you view the ungraded Slog3 it appears to have more noise in the shadows than Slog2. Many are concerned that this “extra” noise will end up making the final pictures nosier. The reality is that this is not the case, you won’t get any extra noise using Slog3 over Slog2. Because S-Log3 is closer to the log gamma curves used in other cameras many people find that Slog3 is generally easier to grade and work with in post production.
So what’s going on?
Slog3 mimics the Cineon Log curve, a curve that was originally designed, back in the 1980’s to match the density of film stocks. As a result the shadow and low key parts of the scene are shown and recorded at a brighter level than Slog2. S-Log2 was designed from the outset to work with electronic sensors and is optimised for the way an electronic sensor works rather than film. Because the S-Log3 shadow range has more gain than S-log2, the shadows end up a bit brighter than it perhaps they really needs to be and because of the extra gain the noise in the shadows appears to be worse. The noise level might be a bit higher but the important thing, the ratio between wanted picture information and un wanted noise is exactly the same whether in Slog2 or Slog3.
Let me explain:
The signal to noise ratio of a camera is determined predominantly by the sensor itself and how the sensor is read. This is NOT changing between gamma curves.
The other thing that effects the signal to noise ratio is the exposure level, or to be more precise the aperture and how much light falls on the sensor. This should be same for Slog2 and Slog3. So again no change there.
As these two key factors do not change when you switch between Slog2 and slog3, there is no change in the signal to noise ratio between Slog2 and Slog3. It is the ratio between wanted picture information and noise that is important. Not the noise level, but the ratio. What people see when they look at ungraded SLog3 is a higher noise level simply because ALL the signal levels are also higher, both noise and desirable image information. So the ratio between the wanted signal and the noise is actually no different for both Slog2 and Slog3.
Gamma is just gain, nothing more, nothing less, just applied by variable amounts at different levels. In the case of log, the amount of gain decreases as you go further up the curve.
Increasing or decreasing gain does NOT significantly change the signal to noise ratio of a digital camera (or any other digital system). It might make noise more visible if you are amplifying the image more than normal in an underexposure situation where you are using that extra gain to compensate for not enough light. But the ratio between the dark object and the noise does not change, it’s just that as you have made the dark object brighter by adding gain, you have also made the noise brighter by the same amount, so the noise also becomes brighter and thus more obvious.
Lets take a look at some Math. I’ll keep it very simple, I promise!
Just for a moment to keep things simple, lets say some camera has a signal to noise ratio of 3:1 (SNR is normally measured in db, but I’m going to keep things really simple here).
So, from the sensor if my picture signal is 3 then my noise will be 1.
If I apply Gamma Curve “A” which has 2x gain then my picture becomes 6 and my noise becomes 2. The SNR is 6:2 = 3:1
If I apply Gamma Curve “B” which has 3x gain then my picture becomes 9 and my noise becomes 3. The SNR is 9:3 = 3:1 so no change to the ratio, but the noise is now 3 with gamma B compared to Gamma A where it is 2, so the gamma B image will appear at first glance to be noisier.
Now we take those imaginary clips in to post production:
In post we want to grade the shots so that we end up with the same brightness of image, so lets say our target level after grading is 12.
For the gamma “A” signal we need to add 3x gain to take 6 to 18. As a result the noise now becomes 6 (3 x 2 = 6).
For the gamma “B” signal (our noisy looking one) we need to use less gain in post, only 2x gain, to take 9 to 18. When we apply 2x gain our noise for gamma B becomes 6 (2 x 3 = 6).
Notice anything? In both cases the noise in the final image is exactly the same, in both cases the final image level is 18 and the final noise level is 6, even though the two recordings started at different levels with one appearing noisier than the other.
OK, so that’s the theory, what about in practice?
Take a look at the images below. These are 400% crops from larger frames. Identical exposure, workflow and processing for each. You will see the original Slog2 and SLog3 plus the Slog 2 and Slog 3 after applying the LC-709 LUT to each in Sony’s raw viewer. Nothing else has been done to the clips. You can “see” more noise in the raised shadows in the untouched SLog3, but after applying the LUTs the noise levels are the same. This is because the Signal to Noise ratio of both curves is the same and after adding the LUT’s the total gain applied (camera gain + LUT gain) to get the same output levels is the same.
It’s interesting to note in these frame grabs that you can actually see that in fact the S-Log3 final image looks if anything a touch less noisy. The bobbles and the edge of the picture frame look better in the Slog3 in my opinion. This is probably because the S-Log3 recording uses very slightly higher levels in the shadow areas and this helps reduce compression artefacts.
The best way to alter the SNR of a typical video system (other than through electronic noise reduction) is by changing the exposure, which is why EI (Exposure Index) and exposure offsets are so important and so effective.
Slog3 has a near straight line curve above middle grey. This means that in post production it’s easier to grade as adjustments to one part of the image will have a similar effect to other parts of the image. It’s also very, very close to Cineon and to Arri Log C and in many cases LUT and grades designed for these gammas will also work pretty well with SLog3.
The down side to Slog3?
Very few really. Fewer data points are recorded for each stop in the brighter parts of the picture and highlight range compared to Slog2. This doesn’t change the dynamic range but if you are using a less than ideal 8 bit codec you may find S-Log2 less prone to banding in the sky or other gradients compared to S-Log3. With a 10 bit recording, in a decent workflow, it makes very little difference.
When an engineer designs a gamma curve for a camera he/she will be looking to achieve certain things. With Sony’s Hypergammas and Cinegammas one of the key aims is to capture a greater dynamic range than is possible with normal gamma curves as well as providing a pleasing highlight roll off that looks less electronic and more natural or film like.
To achieve these things though, sometimes compromises have to be made. The problem being that our recording “bucket” where we store our picture information is the same size whether we are using a standard gamma or advanced gamma curve. If you want to squeeze more range into that same sized bucket then you need to use some form of compression. Compression almost always requires that you throw away some of your picture information and Hypergamma’s and Cinegamma’a are no different. To get the extra dynamic range, the highlights are compressed.
To get a greater dynamic range than normally provided by standard gammas the compression has to be more aggressive and start earlier. The earlier (less bright) point at which the highlight compression starts means you really need to watch your exposure. It’s ironic, but although you have a greater dynamic range i.e. the range between the darkest shadows and the brightest highlights that the camera can record is greater, your exposure latitude is actually smaller, getting your exposure just right with hypergamma’s and cinegamma’s is very important, especially with faces and skin tones. If you overexpose a face when using these advanced gammas (and S-log and S-log2 are the same) then you start to place those all important skin tone in the compressed part of the gamma curve. It might not be obvious in your footage, it might look OK. But it won’t look as good as it should and it might be hard to grade. It’s often not until you compare a correctly exposed sot with a slightly over shot that you see how the skin tones are becoming flattened out by the gamma compression.
But what exactly is the correct exposure level? Well I have always exposed Hypergammas and Cinegammas about a half to 1 stop under where I would expose with a conventional gamma curve. So if faces are sitting around 70% with a standard gamma, then with HG/CG I expose that same face at 60%. This has worked well for me although sometimes the footage might need a slight brightness or contrast tweak in post the get the very best results. On the Sony F5 and F55 cameras Sony present some extra information about the gamma curves. Hypergamma 3 is described as HG3 3259G40 and Hypergamma 4 is HG4 4609G33. What do these numbers mean? lets look at HG3 3259G40
The first 3 numbers 325 is the dynamic range in percent compared to a standard gamma curve, so in this case we have 325% more dynamic range, roughly 2.5 stops more dynamic range. The 4th number which is either a 0 or a 9 is the maximum recording level, 0 being 100% and 9 being 109%. By the way, 109% is fine for digital broadcasting and equates to bit 255 in an 8 bit codec. 100% may be necessary for some analog broadcasters. Finally the last bit, G40 is where middle grey is supposed to sit. With a standard gamma, if you point the camera at a grey card and expose correctly middle grey will be around 45%. So as you can see these Hypergammas are designed to be exposed a little darker. Why? Simple, to keep skin tones away from the compressed part of the curve.
Here are the numbers for the 4 primary Sony Hypergammas:
Cinegammas are designed to be graded. The shape of the curve with steadily increasing compression from around 65-70% upwards tends to lead to a flat looking image, but maximises the cameras latitude (although similar can be achieved with a standard gamma and careful knee setting). The beauty of the cinegammas is that the gentle onset of the highlight compression means that grading will be able to extract a more natural image from the highlights. Note than Cinegamma 2 is broadcast safe and has slightly reduced recording range than CG 1,3 and 4.
Standard gammas will give a more natural looking picture right up to the point where the knee kicks in. From there up the signal is heavily compressed, so trying to extract subtle textures from highlights in post is difficult. The issue with standard gammas and the knee is that the image is either heavily compressed or not, there’s no middle ground.
In a perfect world you would control your lighting (turning down the sun if necessary ;-o) so that you could use standard gamma 3 (ITU 709 standard HD gamma) with no knee. Everything would be linear and nothing blown out. This would equate to a roughly 7 stop range. This nice linear signal would grade very well and give you a fantastic result. Careful use of graduated filters or studio lighting might still allow you to do this, but the real world is rarely restricted to a 7 stop brightness range. So we must use the knee or Cinegamma to prevent our highlights from looking ugly.
If you are committed to a workflow that will include grading, then Cinegammas are best. If you use them be very careful with your exposure, you don’t want to overexpose, especially where faces are involved. getting the exposure just right with cinegammas is harder than with standard gammas. If anything err on the side of caution and come down 1/2 a stop.
If your workflow might not include grading then stick to the standard gammas. They are a little more tolerant of slight over exposure because skin and foliage won’t get compressed until it gets up to the 80% mark (depending on your knee setting). Plus the image looks nicer straight out of the camera as the cameras gamma should be a close match to the monitors gamma.