Scroll down to where it says “Stunning Cinematic Colour” and there you will find a video called “Orlaith” that shows both LUT’s applied to the same footage.
Orlaith is a gaelic name and it is pronounced “orla”. It is the name of a mythical golden princess. The short film was shot on a teeny-tiny budget in a single evening with an FX3 and FX6 using S-Log3 and SGamut3.cine. Then the LUTs were applied directly to the footage with no further grading.
I’ve covered this before, but as this came up again in an online discussion I thought I would write about it again. For decades when I was doing a lot of corporate video work we shot greenscreen and chroma key with analoge or 8 bit, limited dynamic range, standard definition cameras and generally got great results (it was very common to use a bluescreen as blue spill doesn’t look as bad on skin tones as green). So now when we have cameras with much greater dynamic ranges and 10 bit recording is it better to shoot for greenscreen using S-Log3 (or any other log curve for that matter) or perhaps Rec-709?
Before going further I will say that there is no yes-no, right-wrong, answer to this question. I will also add that Rec-709 gets a bad rap because people don’t really understand how gamma curves/transfer functions actually work and how modern grading software is able to re-map the aquisition transfer function to almost any other transfer function. If you use a colour managed workflow in DaVinci Resolve it is very easy to take a Rec-709 recording and map it to S-Log3 so that you can apply the same grades to the 709 as you would to material originated using S-Log3. Of course the 709 recording may not have as much dynamic range as an S-Log3 recording, but it will “look” more or less the same.
Comming back to shooting greenscreen and chromakey:
S-Log3: ? Shoot using 10 bit S-log3 and you have 791 code values available (95-886) to record 14/15 stops of dynamic range. so on average across the entire curve each stop has around 55 code values. Between Middle Grey and +2 stops there are approx 155 code values – this region is important as this is where the majority of skin tones and the key background are likely to fall.
Rec-709: ? Shoot using vanilla Rec-709 and you are using 929 code values (90-1019) to record 6/7 stops so each stop has on average across the entire curve has around 125 code values. Between Middle Grey and +2 stops there are going to be around 340 code values. ? That is not an insignificant difference, it’s not far off the difference between shooting with 10 bit or 12 bit. ? If you were to ask someone whether it is better to shoot using 10 bit or 12 bit I am quite sure the automatic answer would be 12 bit because the general concensus is – more bits is always better. ? A further consideration is that the Sony cameras operate at a lower ISO when shooting with standard gammas and as a result you will have an improved signal to noise ratio using 709 than when using S-log3 and this can also make it easier to achieve a good, clean, key. ? However you do also need to think about what it is you are shooting and how it will be used. If you are shooting greenscreen in a studio then you should have full control over your lighting and in most cases 6 or 7 stops is all you need, so Rec-709 should be able to capture everything comfortably well. If you are shooting outside with less control over the light perhaps Rec-709 won’t have sufficient range. ? If the background plates have been shot using S-Log3 then some people don’t like keying 709 into S-Log3. However a colour managed workflow can deal with this very easily. We should consider that 709 and S-Log3 in a workflow where grading is a big part should not be though of as “looks” but simply as transfer functions or maps of what brightness/saturation seen by the camera is recorded at what code value. Handle these transfer functions correctly via a colour managed workflow and both will “look” the same and both will grade the same within their respective capture limits. ? For an easy workflow you might chose to shoot the greenscreen elements using log with the same settings as the plates. There is nothing wrong with this, it works, it is a very commonly used workflow but it isn’t necessarily always going to be optimum. A lot of people will put a lot of emphasis on using raw or greater bit depths to maximise the quality of their keying, but overlook gamma choice altogether, simply because “Rec-709” is almost a dirty word these days. ? If you have more control, and want absolutely the best possible key, you might be better off using Rec-709. As you will have more data per stop which makes it easier for the keying software to identify edges and less noise. If using Rec-709 you want to chose a version of Rec-709 where you can turn off the camera’s knee as this will prevent the 709 curve from crushing the highlights which can make them difficult to grade. In a studio situation you shouldn’t need to use a heavy knee.
I suggest you experiment and test for yourself and not every situation will be the same, sometimes S-Log3 will be the right choice, other times Rec-709. ?
This is a question that comes up a lot. Especially from those migrating to a camera with a CineEI mode from a camera without one. It perhaps isn’t obvious why you would want to use a shooting mode that has no way of adding gain to the recordings.
If using the CineEI mode shooting S-log3 at the base ISO, with no offsets or anything else then there is very little difference between what you record in Custom mode at the base ISO and CineEI at the base EI.
But we have to think about what the CineEI mode is all about. It’s all about image quality. You would normally chose to shoot S-Log3 when you want to get the highest possible quality image and CineEI is all about quality.
The CineEI mode allows you to view via your footage via a LUT so that you can get an appreciation of how the footage will look after grading. Also when monitoring and exposing via the LUT because the dynamic range of the LUT is narrower, your exposure will be more accurate and consistent because bad exposure looks more obviously bad. This makes grading easier. One of the keys to easy grading is consistent footage, footage where the exposure is shifting or the colours changing (don’t use ATW with Log!!) can be very hard to grade.
Then once you are comfortable exposing via a LUT you can start to think about using EI offsets to make the LUT brighter or darker. When the LUT is darker you open the aperture or reduce the ND to return the LUT to a normal looking image and vice versa with a brighter LUT. This then changes the brightness of the S-log3 recordings and you use this offsetting process to shift the highlight/shadow range as well as noise levels to suit the types of scenes you are shooting. Using a low EI (which makes the LUT darker) plus correct LUT exposure (the darker LUT will make you open the aperture to compensate) will result in a brighter recording which will improve the shadow details and textures that are recorded and thus can be seen in the shadow areas. At the same time however that brighter exposure will reduce the highlight range by a similar amount to the increase in the shadow range. And no matter what the offset, you always record at the cameras full dynamic range.
I think what people misunderstand about CineEI is that it’s there to allow you to get the best possible, highly controlled images from the camera. Getting the best out of any camera requires appropriate and sufficient light levels. CineEI is not designed or intended to be a replacement for adding gain or shooting at high recording ISOs where the images will be already compromised by noise and lowered dynamic range.
CineEI exists so that when you have enough light to really make the camera perform well you can make those decisions over noise v highlights v shadows to get the absolute best “negative” with consistent and accurate exposure to take into post production. It is also the only possible way you can shoot when using raw as raw recordings are straight from the sensor and never have extra gain added in camera.
Getting that noise/shadow/highlight balance exactly right, along with good exposure is far more important than the use of external recorders or fatter codecs. You will only ever really benefit fully from higher quality codecs if what you are recording is as good as it can be to start with. The limits as to what you can do in post production are tied to image noise no matter what codec or recording format you use. So get that bit right and everything else gets much easier and the end result much better. And that’s what CineEI gives you great control over.
When using CineEI or S-Log3 in general you need to stop thinking “video camera – slap in a load if gain if its dark” and think “film camera – if its too dark I need more light”. The whole point of using log is to get the best possible image quality, not shooting with insufficient light and a load of gain and noise. It requires a different approach and completely different way of thinking, much more in line with the way someone shooting on film would work.
What surprises me is the eagerness to adopt shutter angles and ISO ratings for electronic video cameras because they sound cool but less desire to adopt a film style approach to exposure based on getting the very best from the sensor. In reality a video sensor is the equivalent of a single sensitivity film stock. When a camera has dual ISO then it is like having a camera that takes two different film stocks. Adding gain or raising the ISO away from the base sensitivity in custom mode is a big compromise that can never be undone. It adds noise and decreases the dynamic range. Sometimes it is necessary, but don’t confuse that necessity with getting the very best that you can from the camera.
Arri have a little trick in their cameras when shooting log to ProRes that the Sony Log cameras don’t have. When you change the Exposure Index in an Arri camera they modify the position of the exposure mid point and the shape of the Log-C gamma curve. There is actually a different Log-C curve for each EI. When you take this into post it has the benefit that the brightness at each EI will appear similar. But as the curve changes for each EI a different LUT is needed for each exposure if you want something shot at say 800EI to look the same as something shot at 200EI.
With a Sony camera the same S-Log curve is used for each Exposure Index and the LUT brightness is changed so that you end up altering the mid point of the recording as well as the highlight and shadow range. In post each EI will appear to be a different brightness. You can use the same LUT for each EI provided you do an exposure correction prior to adding the LUT or you can use dedicated offset LUT’s for each exposure.
But what you need to remember is that you are always working within a restricted recording range with either system. You can’t go darker than the black recording level or brighter than the highest value the codec can record.
If you do it in camera, as Arri do and change the log curve, at a low EI you seriously constrict the recording range (at 200 EI peak only reaches around 78IRE). This happens because at a low EI you put more light on to the sensor. So to keep the mid range looking a normal brightness in post it must be recorded at at a level that is offset downwards compared to normal. So with all the levels now offset downwards to compensate for the brighter exposure you end up recording your entire capture range with a reduced or compressed recording range. In addition to avoid clipping the blacks at a low EI the shadows are rolled off so you lose some detail and textures in the shadows. You can see the different Log-C curves in this Arri White paper.
Most people choose a low EI for 2 reasons, better signal to noise ratio and improved shadow range. The Arri method gives you the better SNR but while the dynamic range is preserved it’s recorded using less data and in particular the shadow data decreases compared to shooting at the base ISO.
Shoot at a high EI, you put less light on to the sensor. So to maintain similar looking mids in post everything has to be recorded at a higher level. Now you have a problem because the highlights will extend beyond the upper limits of the recording range so Arri have to add a highlight roll off at the top of the Log-C curve. This can present some grading challenges as the curve is now very different to regular Log-C. In addition the highlights are compressed.
Most people choose to shoot at a high EI to extend the highlight range or to work in lower light levels.
The latter is a bit of a pointless exercise with any log camera as the camera sensitivity isn’t actually any different, you are only fooling yourself into thinking it’s is more sensitive and this can result in noisy footage. If you using a high EI to extend the highlight range then really the last thing you want is the extra highlight roll off that Arri have to add at 3200 EI to fit everything in.
One thing here in Arri’s favour is that they can record 12 bit ProRes 444. 12 bits helps mitigate the compressed recording range of low EI’s provided the post workflow is managed correctly.
The beauty of the Sony method is the recording range never changes, so low EI’s and brighter recordings deliver better shadow ranges with more data in the shadows and mids and high EI’s with darker recordings deliver better highlight ranges with no additional data restrictions or additional roll-offs giving the cinematographer more control to choose the exposure mid point without compromise to the data at either end.
But it does mean that post need to be awake and that the shooter needs to communicate with post regarding the brighter/darker looking images. But to be honest if post don’t understand this and recognise what you have done either buy just looking at the footage or checking the metadata what chance is there of post actually doing a decent job of grading your content? This should be fundamental and basic stuff for a colourist/grader. For a colourist/grader to not understand this and how to work with this is like hiring a camera operator that doesn’t know what an ND filter is.
The Sony FS7/FX9/F5/F55/Venice cameras can do something similar to an Arri camera by baking in the S-Log3 LUT. Then in post the exposure will look the same at every EI. BUT you will lose some highlight range at a low EI’s and some shadow range at a high EI’s without gaining any extra range at the opposite end. As a result the total dynamic range does reduce as you move away from the base ISO.
In addition on the Venice, FS7/F5/F55 (and I suspect in a future update the FX9) you can bake in a user LUT to the SxS recordings. If you create a set of S-Log3 to S-Log3 LUT’s with EI offsets included in the LUT you could replicate what Arri do by having an offset and tweaked S-Log3 User LUT for each EI that you want to shoot at. You would not use the cameras EI control you would leave the camera st the base ISO. The LUT’s themselves will include the exposure offset. These will maintain the full dynamic range but just like Arri they will need to roll off the shadows or highlights within the LUT.
But monitoring will be tricky as you won’t have the benefit of a 709 type LUT for monitoring so you you may need to use an external monitor or viewfinder that can apply a LUT to it’s image. The good news is the same LUT would be used in the monitor for every version on the offset S-Log3 LUT that you are baking in as the exposure brightness levels will be the same for each offset.
So here you are a set of 4 S-Log3/S-Gamut3.cine offset LUT’s for those Sony cameras that will take a user LUT. I have named the LUT’s – 2S Down SL3C, 1S Down SL3C, 1S UP SL3C, 2S UP SL3C.
The name means (Number of Stops) (Down or Up) (Slog3.Cine).
So if the cameras base ISO is 2000 (F5/FS7 etc) and you want to shoot at the equivalent of 1000EI, which is 1 stop down from base you would use “1S Down SL3C”.
As always (to date at least) I offer these as a free download available by clicking on the links below. But I always appreciate a contribution if you find them useful and make use of them. I will let you pay what you feel is fair, all contributions are greatly appreciated and it really does help keep this website up and running. If you can’t afford to pay, then just download the LUT’s and enjoy using them. If in the future you should choose to use them on a paying project, please remember where you got them and come back and make a contribution. More contributions means more LUT offerings in the future.
Please feel free to share a link to this page if you wish to share these LUT’s with anyone else or anywhere else. But it’s not OK to to share or host these on other web sites etc.
I have written about this many times before, but I’ll try to be a bit more concise here.
So – You have recorded S-Log2 or S-Log3 on your Sony camera and at the same time recorded on an external ProRes Recorder such as an Atomos, Blackmagic or other ProRes recorder. But the pictures look different and they don’t grade in the same way. It’s a common problem. Often the external recording will look more contrasty and when you add a LUT the blacks and shadow areas come out very differently.
Video signals can be recorded using a several different data ranges. S-Log2 and S-Log3 signals are always Data Range. When you record in the camera the cameras adds information to the recording called metadata that tells your editing or grading software that the material is Data Range. This way the edit and grading software knows how to correctly handle the footage and how to apply any LUT’s.
However when you record to an external recorder the external recorder doesn’t have this extra metadata. So the recorder will record the Data Range signal that comes from the camera but it doesn’t add the metadata. The ProRes codec is normally used for Legal Range video and by default, unless there is metadata that says otherwise, edit and grading software will assume any ProRes recordings to be Legal Range.
So what happens is that your edit software takes the file, assumes it’s Legal Range and handles it as a Legal Range file when in fact the data in the file is Data Range. This results in the recording levels being transposed into incorrect levels for processing. So when you add a LUT it will look wrong, perhaps with very dark shadows or very bright over exposed looking highlights. It can also limit how much you can grade the footage.
What Can We Do About It?Premiere CC.
You don’t need to do anything in Premiere for the internal .mp4 or MXF recordings. They are handled correctly but Premiere isn’t handling the ProRes files correctly.
My approach for this has always been to use the legacy fast color corrector filter to transform the input range to the required output range. If you apply the fast color corrector filter to a clip you can use the input and output level sliders to set the input and output range. In this case we need to set the output black level to CV16 (as that is legal range black) and we need to set output white to CV235 to match legal range white. If you do this you will then see that the external recording appears to have almost exactly the same values as the internal recording. However there is some non-linearity in the transform, it’s not quite perfect.
Now when you apply a LUT the picture and the levels are more or less what you would expect and almost identical to the internal recordings. I say almost because there is a slight hue shift. I don’t know where the hue shift comes from. In Resolve the internal and external recordings look pretty much identical and there is no hue shift. In Premiere they are not quite the same. The hue is slightly different and I don’t know why. My recommendation – use Resolve, it’s so much better for anything that needs any form of grading or color correction.
It’s very easy to tell Resolve to treat the clips as Data Range recordings. In the media bin, right click on the clip and under “clip attributes” change the input range from “auto” to “full”. If you don’t do this DaVinci Resolve will assume the ProRes file to be legal range and it will scale the clip incorrectly in the same way as Premiere does. But if you tell Resolve the clip is full range then it is handled correctly.
This comes up again and again, hence why I am writing about it once again.
Raw should never be converted to log before recording if you want any benefit from the raw. You may as well just record the 10 bit log that most cameras are capable of internally. Or take log and output it via the cameras 10 bit output (if it has one) and record that directly on the ProRes recorder. It doesn’t matter how you do it but if you convert between different recording types you will always reduce the image quality and this is as bad a way to do it as you can get. This mainly relates to cameras like the PXW-FS7. The FS5 is different because it’s internal UHD recordings are only 8 bit, so even though the raw is still compromised by converting it to ProRes log, this can still be better than the internal 8 bit log.
S-Log like any other log is a compromise recording format. Log was developed to squash a big dynamic range into the same sized recording bucket as would normally be used for conventional low dynamic range gammas. It does this by discarding a lot of tonal and textural information from everything brighter than 1 stop above middle grey, instead of the amount of data doubling for each stop up you go in exposure, it’s held at a constant amount. Normally this is largely transparent as human vision is less acute in the highlight range, but it is still a compromise.
The idea behind Linear raw is that it should give nothing away, each stop SHOULD contain double the data as the one below. But if you only have 12 bit data that would only allow you to record 11 stops of dynamic range as you would quickly run out of code values. So Sony have to use floating point math or something very similar to reduce the size of each stop by diving down the number of code values each stop has. This has almost no impact on highlights where you start off with 100’s or 1000’s values but in the shadows where a stop may only have 8 or 16 values dividing by 4 means you now only have 2 or 4 tonal levels. So once again this is a compromise recording format. To record a big dynamic range using linear what you really need is 16 bit data.
In summary so far:
S-Log reduces the number of highlight tonal values to fit it a big DR in a normal sized bucket.
Sony’s FSRaw, 12 Bit Linear reduces the number of tonal Values across the entire range to fit it in a compact 12 bit recording bucket, but the assumption is that the recording will be at least 12 bit. The greatest impact of the reduction is in the shadows.
Convert 12 bit linear to 10 bit S-Log and now you are compromising both the highlight range and the shadow range. You have the worst of both, you have 10 bit S-Log but with much less shadow data than the S-log straight from the camera. It’s really not a good thing to do and the internally generated S-Log won’t have shadows compromised in the same way.
If you have even the tiniest bit of under exposure or you attempt to lift the shadows in any way this will accentuate the reduced shadow data and banding is highly likely as the values become stretched even further apart as you bring them up the output gamma range.
If you expose brightly and then reduce the shadows this has the effect of compressing the values closer together or pushing them further down the output curve, closing them together as they go down the output gamma range, this reduces banding. This is one of the reasons why exposing more brightly can often help both log and raw recordings. So a bit of over exposure might help, but any under exposure is really, really going to hurt. Again, you would probably be better off using the internally generated S-Log.
To make matters worse there is also often an issue with S-Log in a ProRes file.
If all that is not enough there is also a big problem in the way ProRes files record S-Log. S-Log should always be recorded as full range data. When you record an internal XAVC file the metadata in the clips tells the edit or grading software that the file is full range. Then when you apply a LUT or do your grading the correct transforms occur and all shadow textures are preserved. But ProRes files are by default treated as legal range files. So when you record full range S-Log inside a ProRes file there is a high likelihood that your edit or grading software will handle the data in the clip incorrectly and this too can lead to problems in the shadows including truncated data, clipping and banding, even though the actual recorded data may be OK. This is purely a metadata issue, grading software such as DaVinci resolve can be forced to treat the ProRes files as full range.
Having shot quite a bit of S-Log3 content on the new Sony PXW-FX9 I thought I would comment on my exposure preferences. When shooting with an FS5, FS7 or F5, which all use the same earlier generation 4K sensor I find that to get the best results I need to expose between 1 and 2 stops brighter than the 41% for middle grey that Sony recommend. This is because I find my footage to be noisier than I would like if I don’t expose brighter. So when using CineEI on these cameras I use 800EI instead of the base 2000EI
However the FX9 uses a newer state of the art back illuminated sensor. This more sensitive sensor produces less noise so with the FX9 I no longer feel it is necessary to expose more brightly than the base exposure – at either of the base ISO’s. So if I am shooting using CineEI and 800 base, I use 800EI. When shooting at 4000 base, I use 4000 EI.
This makes life so much easier. It also means that if you are shooting in a mode where LUT’s are not available (such as 120fps HD) then you can use the included viewfinder gamma assist function instead. Viewfinder gamma assist adds the same 709(800) look to the viewfinder as you would get from using the cameras built in 709(800) LUT. You can use the VF gamma assist to help judge your exposure just as you would with a LUT. Basically, if it looks right in the viewfinder, it almost certainly is right.
Testing various FX9’s against my Sekonic light meter the cameras CineEI ISO ratings seem to be spot on. So I would have no concerns if using a light meter to expose. The camera also has a waveform scope and zebras to help guide your exposure.
VF Gamma assist is available in all modes on the FX9, including playback. Just be careful that you don’t have both a LUT on and gamma assist at the same time.
Sony had often been criticized for having a default look to their cameras that wasn’t “film like”. This was no accident as Sony have been a leading producer of TV cameras for decades and a key thing for a broadcaster is that both old and new cameras should match. So for a very long time all of Sony’s cameras were designed to look pretty much like any other TV camera.
But this TV look wasn’t helping Sony to sell their film style cameras. So when they developed the image processing for the Venice camera a lot of research was done into what makes a pretty picture. Then over a period of about 18 months a new LUT was created for the Venice camera to take advantage of that sensors improved image quality and to turn the output into a beautiful looking image. This LUT was designed to still leave a little room to grade so is slightly flat. But it does include a big highlight roll off to help reserve a lot of the cameras dynamic range.
This LUT is called s709 (I think it simply stands for “Sony 709) and it’s a large part of the reason why, out of the box, the Venice camera looks the way it does. Of course a skilled colourist might only rarely use this LUT and may make the output from a Venice look very different, but a Venice with s709 is regarded as the default “Venice look”, and it’s a look that a lot of people really, really like. It’s what comes out of the SDI ports, is what’s seen in the viewfinder and can be recorded to the SxS cards unless you select the legacy 709(800) LUT. s709 is the LUT applied by default to X-OCN from Venice by default.
What is Color Science
Colour Science is the new fancy term that Red have turned into a catch-all for anything to do with colour and it’s now much abused. Every color video camera ever made uses color science to determine the way the image looks. it’s nothing new. All colour science is, is how all the different elements of a camera and it’s workflow work to produce the final colour image. But in the last couple of years it seems to have become to mean “color magic” or “special sauce”.
If we are to be totally accurate the only camera with Venice colour science is Venice. No other camera has exactly the same combination of optical filters, sensor, processing, codecs and workflow. No other camera will replicate exactly the way Venice responds to light and turns it into a color image. You might be able to make the output of another camera appear similar to a Venice, but even then it won’t be the same colour science. What it would be is perhaps the “Venice look”.
The FS5 II and it’s new default look.
So when Sony released the FS5 II they were very careful to describe the default mode as providing a Venice “like” image, tuned to provide softer, alluring skin tones using insight and expertise gained during the development of Venice. Because that’s what it is, it looks more like Venice than previous generations of Sony cameras because it has been tuned to output a image that looks similar. But it isn’t really Venice color science, it’s a Venice look-a-like or at least as similar as you can get, even though it’s a very different sensor, but with a touch of extra contrast added to make it more suitable for an out of the box look that won’t necessarily be graded.
And the PXW-FX9 and s-Cinetone?
The FX9 has new colour filters, a new sensor, new processing. But it is not a Venice. In Custom mode it has what Sony are now calling “S-Cinetone” which is set to become their new default look for their film style cameras. This once again is based on the Venice look and shares many similarities to the Venice colour science, but it will never be the full Venice colour science because it can’t be, it’s different hardware. S-Cinetone is a combination of a gamma curve called “original” and a matrix called “S-Cinetone” in the FX9. When used together S-Cinetone gives similar colours to Venice but has increased contrast suitable for direct-to-air applications where the material won’t be graded (s709 in comparison is flatter). S-Cinetone has a very gentle highlight roll off and produces a film like look that is tailored for video productions rather than the flatter s709 look which is designed for on set monitoring on film style shoots. If you want you can mix different gamma curves with the S-Cinetone matrix to have the Venice like colours but with different contrast ranges to suit the scene that you are shooting. If you need a broadcast safe image you can use Hypergamma1 with the S-Cinetone matrix.
Is the Venice look always there?
Previous generations of Sony cameras used a common default 709 gamma often denoted as STD5 combined with a 709 colour matrix. This is what most of us probably called the “Sony look”. The exact colour science in each camera with this look would have been quite different as there were many combinations of filters, sensors and processing, but those variations in processing were designed such that the final output of generations of Sony TV cameras all looked almost exactly the same. This too still exists in the FX9 and when set to STD5 the FX9 will produce an image very, very close to earlier generations of Sony camera. But from this new sensor with the latest filters etc. you can still have the old look. This just demonstrates how the broad brush use of the term colour science is so confusing as the FX9 is a new camera with new colour science, but it can still look just like the older cameras.
What about when I shoot S-Log3?
When shooting S-Log3 with the FX9, then you are shooting S-Log3. And S-Log3/S-Gamut3 )or S-gamut3.cine) is a set standard where certain numerical values represent certain real world colours and brightnesses. So the S-Log3 from an FX9 will look very similar to the S-Log3 from a Venice, which is similar to the S-Log3 from a F55 which is similar to the S-Log3 from an FS7.
But compared to an FS7 at least, the different, improved sensor in the FX9 will mean that it will be able to capture a bigger dynamic range, it will have less noise and the sensors response to colour is much improved. BUT it will still be recorded in the same manner using the same gamma curve and colour space with the same numerical values representing the same brightness levels and colours. However the fact that the sensor is different will mean there will be subtle differences within the image. One obvious one being the extra dynamic range, but also things like better colour separation and more true to life color response at the sensor level.
Then you apply the s709 LUT, the very same LUT as used for Venice. So those very same numerical values are turned into the same expected colours and brightness levels. But because it’s a different sensor some values may have been better captured, some worse, so while overall the image will look very, very similar there will be subtle differences, it’s the subtle differences that make one look more natural or more pleasing than the other. For example the FX9 image will have less noise and greater DR than the image from and FS7. In addition the FX9 images will have more pleasing looking skin tones because from what I have seen the sensor responds better to the tones that make up a face etc.
Why not use the same name for s709 and S-cinetone?
S-Cinetone is different to s709. One is a gamma curve plus colour matrix designed to be recorded as is for television and video applications. You can’t change middle grey or white, you can’t alter the highlight or shadow ranges, other than by using alternate gammas with the S-Cinetone matrix. The default “original” gamma curve has more contrast than the S-Log3 + s709 LUT and the colours although similar are slightly different.
s709 is a LUT applied to S-Log3 material designed to provide a film like look for on set monitoring. Both S-Cinetone and s709 will look similar, but they are two different things that require two very different workflows, to call them the same thing would be confusing. You get a call from the producer “I want you to shoot S-Cinetone”…. Which one? The log one or the S-Cinetone one?
Because the FX9’s optical low pass filter, ND filter, sensor colour filters, pixels, sensor output circuits and initial processing of the image are all the same whether in S-Cinetone or S-Log3, then those aspects of the colour science are common for both. But when shooting s-Log3 you have a huge range of options in post, not just s709.
So in reality the FX9 has several different color sciences. One that mimics a default Venice camera without needing to shoot log and grade. One that mimics earlier generations of sony TV cameras. Another that mimics a Sony Venice when shooting S-Log3 and using the s709 LUT.
There is a video on YouTube right now where the author claims that the Sony Alpha cameras don’t record correctly internally when shooting S-Log2 or S-Log3. The information contained in this video is highly miss-leading and the conclusion that the problem is with the way Sony record internally is incorrect. There really isn’t anything wrong with the way Sony do their recordings. Neither is there anything wrong with the HDMI output. While centered around the Alpha cameras the information below is also important for anyone that records S-Log2 or S-log3 externally with any other camera.
Some background: Within the video world there are 2 primary ranges that can be used to record a video signal.
Legal Range uses code value 16 for black and code value 235 for white (anything above CV235 is classed as a super-white and these can still be recorded but considered to be beyond 100%).
Full or Data Range uses code value 0 for black and code value 255 for white or 100%.
Most cameras and most video systems are based on legal range. ProRes recordings are almost always legal range. Most Sony cameras use legal range and do include super-whites for some of the curves such as Cinegammas or Hypergammas to gain a bit more dynamic range. The vast majority of video recordings use legal range. So most software defaults to legal range.
But very, very importantly – S-log2 and S-log is always full/data range.
Most of the time this doesn’t cause any issues. When you record internally in the camera the internal recordings have metadata that tells the playback, editing or grading software that the S-Log files have been recorded using full range. Because of this metadata the software will play the files back and process them at the correct levels. However if you record the S-Log with an external recorder the recorder doesn’t always know that what it is getting is full range and not legal range, it just records it, as it is, exactly as it comes out of the camera. That then causes a problem later on because the externally recorded file doesn’t have the right metadata to ensure that the full range S-Log material is handled correctly and most software will default to legal range if it knows no different.
Lets have a look at what happens when you import an internally recorded S-Log2 .mp4 file from a Sony A7S into Adobe Premiere:
A few things to note here. One is Adobe’s somewhat funky scopes where the 8 bit code values don’t line up with the normally used IRE values used for video productions. Normally 8 bit code value 235 would be 100IRE or 100%, but for some reason Adobe have code value 255 lined up with 100%. My suspicion is that the scope % scale is not video % or IRE but instead RGB%. This is really confusing. A further complication is that Adobe have code value 0 as black, again, I think, but am not sure that this is RGB code value 0. In the world of video Black should be code value 16. But the scopes appear to work such that 0 is black and that 100 is full scale video out. Anything above 100 and below 0 will be clipped in any file you render out.
Looking at the scopes in the screen grab above, the top step on the grey scale chart is around code value 252. That is the code value you would expect it to be, that lines up just nicely with where the peak of an S-Log2 recording should be. This all looks correct, nothing goes above 100 or below 0 so nothing will be clipped.
So now lets look at an external ProRes recording, recorded at exactly the same time as the internal recording and see what Premier does with that:
OK, so we can see straight away something isn’t quite right here. In an 8 bit recording it should be impossible to have a code value higher that 255, but the scopes are suggesting that the recording has a peak code value of something around CV275. That is impossible, so alarm bells should be ringing. Something is not quite right here. In addition the S-Log2 appears to be going above 100, so that means if I were to simply export this as a new file, the top of the recording will be clipped and it won’t match the original. This is very clearly not right.
Now lets take a look at what happens in Adobe Premiere when you apply Sony’s standard S-Log2 to Rec-709 LUT to a correctly exposed internal recording:
This all looks good and as expected. Blacks are sitting down just above the 0 line (which I think we can safely assume is black) and the whites of the picture are around code value 230 or 90, whatever that means. But they are certainly nice and bright and are not in the range that will be clipped. So I can believe this as being more or less correct and as expected.
So next I’m going to add the same standard LUT to the external recording to see what happens.
OK, this is clearly not right. Our blacks now go below the 0 line and they look clipped. The highlights don’t look totally out of place, but clearly there is something going very, very wrong when we this normal LUT to this correctly exposed external recording. There is no way our blacks should be going below zero and they look crushed/clipped. The internal recording didn’t behave like this. So what is going on with the external recording?
To try and figure this out lets take a look at the same files in DaVinci Resolve. For a start I trust the scopes in Resolve much more and it is a far better programme for managing different types of files. First we will look at the internal S-Log2 recording:
Once again the levels of the internal S-Log2 recordings look absolutely fine. Our peak is around code value 1010 which would be 252 in 8 bit. Right where the brightest bits of an S-log2 file should be. Now lets take a look at the external recording.
If you compare the two screen grabs above you can see that the levels are exactly the same. Our peak level is around CV1010/CV252, just where it should be and the blacks look the same also. The internal and external recordings have the same levels and look the same. There is no difference (other then perhaps less compression and fewer artefacts in the ProRes file). There is nothing wrong with either of these recordings and certainly nothing wrong with the way Sony record S-Log2 internally. This is absolutely what I expect to see.
BUT – I’ve been a little bit sneaky here. As I knew that the external recording was a full range recording I told DaVinci Resolve to treat it as a full range recording. In the media bin I right clicked on the clip and under “clip attributes” I changed the input range from “auto” to “full”. If you don’t do this DaVinci Resolve will assume the ProRes file to be legal range and it will scale the clip incorrectly in the same way as Premiere does. But if you tell Resolve the clip is full range then it is handled correctly.
This is what it looks like if you allow Resolve to guess at what range the S-Log2 full range clip is by leaving the input range setting to “auto”:
In the above image we can see how in Resolve the clip becomes clipped because in a legal range recording anything over CV235/CV940 would be an illegal super white. Resolve is scaling the clip and pushing anything in the original file that was above CV235/CV940 off the top of the scale. The scaling is incorrect because Resolve doesn’t know the clip is supposed to be full range and therefore not scaled. If we compare this to what Premiere did with the external recording it’s actually very similar. Premiere also scaled the clip, only Premiere will show all those “illegal” levels above it’s 100 line instead of clipping then as Resolve does. That’s why Premiere can have those “impossible” 8 bit code values going up to CV275.
Just to be complete here, I did also test the internal .mp4 recordings in Resolve switching between “auto” and “full” range and in both cases the levels stayed exactly the same. This shows that Resolve is correctly handling the internally record full range S-Log as full range.
What about if you add a LUT? Well you MUST tell Resolve to treat the S-Log2 ProRes clip as a full range clip otherwise the LUT will not be right, if your footage is S-Log3 you also have to tell Resolve that it is full range:
Both the internal and external recordings are actually exactly the same. Both have the same levels, both use FULL range. There is absolutely nothing wrong with Sony’s internal recordings. The problem stems from the way most software will assume that the ProRes files are legal range. But if it’s an S-Log2 or S-Log3 recording it will in fact be full (data) range. Handling a full range clip as legal range means that highlights will be too high/bright or clipped and blacks will be crushed. So it’s really important that your software handles the footage correctly. If you are shooting using S-Log3 this problem is harder to spot as S-Log3 has a peak recording level that is well with the legal range, so you often won’t realise it’s being scaled incorrectly as it won’t necessarily look clip. If you use LUT’s and your ProRes clips look crushed or highlights look clipped you need to check that the input scaling is correct. It’s really important to get this right.
Why is there no difference between the levels when you shoot with a Cinegamma? Well when you shoot with a cinegamma the internal recordings are legal range so the internal recordings get treated as legal range and so do the external recordings, so they don’t appear to be different (In the YouTube video that led to this post the author discovers that if you record with a normal profile first and then switch to a log profile while recording the internal and external files will match. But this is because now the internal recording has the incorrect metadata, so it too gets scaled incorrectly, so both the internal and external files are now wrong – but the the same).
Once again: There is nothing wrong with the internal recordings. The problem is with the way the external recordings are being handled. The external recordings haven’t been recorded incorrectly, they have been recorded as they should be. The problem is the edit software is incorrectly interpreting the external recordings. The external recordings don’t have the necessary metadata to mark the files as full range because the recorder is external to the camera and doesn’t know what it’s being sent by the camera. This is a common problem when using external recorders.
What can we do in Premiere to make Premiere work right with these files?
You don’t need to do anything in Premiere for the internal .mp4 recordings. They are handled correctly but Premiere isn’t handling the full/data range ProRes files correctly.
My approach for this has always been to use the legacy fast color corrector filter to transform the input range to the required output range. If you apply the fast color corrector filter to a clip you can use the input and output level sliders to set the input and output range. In this case we need to set the output black level to CV16 (as that is legal range black) and we need to set output white to CV235 to match legal range white. If you do this you will then see that the external recording appears to have almost exactly the same values as the internal recording. However there is some non-linearity in the transform, it’s not quite perfect. So if anyone knows of a better way to do this do please let me know.
Now when you apply a LUT the picture and the levels are more or less what you would expect and almost identical to the internal recordings. I say almost because there is a slight hue shift. I don’t know where the hue shift comes from. In Resolve the internal and external recordings look pretty much identical and there is no hue shift. In Premiere they are not quite the same. The hue is slightly different and I don’t know why. My recommendation – use Resolve, it’s so much better for anything that needs any form of grading or color correction.
Even though I have written about these many times before the message still just doesn’t seem to be getting through to people.
Since the dawn of photography and video the only way to really change the signal to noise ratio and ultimately how noisy the pictures are is by changing how much light you put onto the sensor.
Gain, gamma, log, raw, etc etc only have a minimal effect on the signal to noise ratio. Modern cameras do admittedly employ a lot of noise reduction processes to help combat high noise levels, but these come at a price. Typically they soften the image or introduce artefacts such as banding, smear or edge tearing. So you always want to start off with the best possible image from the sensor with the least possible noise and the only way to achieve that is through good exposure – putting the optimum amount of light onto the sensor.
ISO is so confusing:
But just to confuse things the use of ISO to rate an electronic cameras sensitivity has become normal. But the problem is that most people have no clue about what this really means. On an electronic camera ISO is NOT a sensitivity measurement, it is nothing more than a number that you can put into an external light meter to allow you to use that light meter to obtain settings for the shutter speed and aperture that will give you the camera manufacturers suggest optimum exposure. That’s it – and that is very different to sensitivity.
Lets take Sony’s FS7 as an example (most other cameras behave in a very similar way).
If you set the FS7 up at 0dB gain, rec-709, it will have an exposure rating of 800 ISO. Use a light meter to expose with the meters ISO dial set to 800. Lets say the light meter says set the aperture to f8. When you do this the image is correctly exposed, looks good (well as good as 709 gets at least) and for most people has a perfectly acceptable amount of noise.
Now switch the camera to S-Log2 or S-Log3. With the camera still set to 0dB the ISO rating changes to 2000 which give the impression that the camera may have become more sensitive. But did we change the sensor? No. Have we added any more gain? No, we have not, the camera is still at 0dB. But if you now expose at the recommended levels, after you have done your grading and you grade to levels similar to 709 the pictures will look quite a lot noisier than pictures shot using Rec-709.
So what’s going on?
If you now go back to the light meter to expose the very same scene, you turn the ISO dial on the light meter from 800 to 2000 ISO and the light meter will tell you to now set the aperture to f13 (approx). So starting at the f8 you had for 800 ISO, you close the aperture on the camera by 1.3 stops to f13 and you will have the “correct” exposure.
BUT: now you are putting 1.3 stops less light on to the sensor so the signal coming from the sensor is reduced by 9dB and as a result the sensor noise that is always there and never really changes is much more noticeable. As a result compared to 709 the graded S-Log looks noisy and it looks noisier by the equivalent of 9dB. This is not because you have changed the cameras sensitivity or changed because you have changed the amount of camera gain but because compared to when you shoot in 709 the sensor is being under exposed and as a result it is outputting a signal 9dB lower. So in post production when you grade or add a LUT you have to add 9dB of gain to get the same brightness as the original direct rec-709 recording and as well as making the desirable image brighter it also makes the noise 9dB higher (unless you do some very fancy noise reduction work in post).
So what do you do?
It’s common simply to open the aperture back up again, typically by 1.5 stops so that after post production grading the S-log looks no more noisy than the 709 from the FS7 – Because in reality the FS7’s sensor works best for most people when rated at the equivalent of 800 ISO rather than 2000 – probably because it’s real sensitivity is 800 ISO.
When you think about it, when you shoot with Rec-709 or some other gamma that won’t be graded it’s important that it looks good right out of the camera. So the camera manufacturer will ensure that the rec-709 noise and grain v sensitivity settings are optimum – so this is probably the optimum ISO rating for the camera in terms of noise, grain and sensitivity.
So don’t be fooled into thinking that the FS7 is more sensitive when shooting with log, because it isn’t. The only reason the ISO rating goes up as it does is so that if you were using a light meter it would make you put less light onto the sensor which then allows the sensor to handle a brighter highlight range. But of course if you put less light onto the sensor the sensor won’t be able to see so far into the shadows and the picture may be noisy which limits still further the use of any shadow information. So it’s a trade-off, more highlights but less shadows and more noise. But the sensitivity is actually the same. Its’s an exposure change not a sensitivity change.
So then we get into the S-Log2 or S-Log3 debate.
First of all lets just be absolutely clear that both have exactly the same highlight and shadow ranges. Both go to +6 stops and -8 stops, there is no difference in that regard. Period.
And lets also be very clear that both have exactly the same signal to noise ratios. S-log3 is NOT noisier than S-log2. S-log 3 records some of the mid range using higher code values than S-Log2 and before you grade it that can sometimes make it appear like it’s noisier, but the reality is, it is not noisier. Just like the differing ISO ratings for different gamma curves, this isn’t a sensitivity change, it’s just different code values being used. See this article if you want the hard proof: http://www.xdcam-user.com/2014/03/understanding-sonys-slog3-it-isnt-really-noisy/
Don’t forget when you shoot with log you will be grading the image. So you will be adjusting the brightness of the image. If you grade S-Log2 and S-Log3 to the same brightness levels the cumulative gain (the gain added in camera and the gain added in post) ends up the same. So it doesn’t matter which you use in low light the final image, assuming a like for like grade will have the same amount of noise.
For 8 bit records S-Log2 has different benefits.
S-Log2 was designed from the outset for recording 14 stops with an electronic video camera. So it makes use of the cameras full recording range. S-Log3 is based on an old film log curve (cineon) designed to transfer 16 stops or more to a digital intermediate. So when the camera only has a 14 stop sensor you waste a large part of the available recording range. On a 10 bit camera this doesn’t make much difference. But on a 8 bit camera where you are already very limited with the number of tonal values you can record it isn’t ideal and as a result S-Log2 is often a better choice.
But if I shoot raw it’s all going to be so much better – isn’t it?
Yes, no, maybe…. For a start there are lot’s of different types of raw. There is linear raw, log raw, 10 bit log raw, 12 bit linear, 16 bit linear and they are all quite different.
But they are all limited by what the sensor can see and how noisy the sensor is. So raw won’t give you less noise (it might give different looking noise). Raw won’t give you a bigger dynamic range so it won’t allow you to capture deeper or brighter highlights.
But what raw does normally is to give you more data and normally less compression than the cameras internal recordings. In the case of Sony’s FS5 the internal UHD recordings are 8 bit and highly compressed while the raw output is 12 bit, that’s a 4 fold increase in the amount of tonal values. You can record the 12bit raw using uncompressed cDNG or Apples new ProResRaw codec which doesn’t introduce any appreciable compression artefacts and as a result the footage is much more flexible in post production. Go up to the Sony Venice, F5 or F55 cameras and you have 16 bit raw and X-OCN (which behaves exactly like raw) which has an absolutely incredible range of tonal values and is a real pleasure to work with in post production. But even with the Venice camera the raw does not have more dynamic range than the log. However because there are far more tonal values in the raw and X-OCN you can do more with it and it will hold up much better to aggressive grading.
It’s all about how you expose.
At the end of the day with all of these camera and formats how you expose is the limiting factor. A badly exposed Sony Venice probably won’t end up looking anywhere near as good as a well exposed FS7. A badly exposed FS7 won’t look as good as a well exposed FS5. No camera looks good when it isn’t exposed well.
Exposure isn’t brightness. You can add gain to make a picture brighter, you can also change the gamma curve to change how bright it is. But these are not exposure changes. Exposure is all about putting the optimum amount of light onto the sensor. Enough light to produce a signal from the sensor that will overcome the sensors noise. But also not so much light that the sensor overloads. That’s what good exposure is. Fiddling around with gamma curves and gain settings will only every make a relatively small difference to noise levels compared to good exposure. There’s just no substitute for faster lenses, reflectors or actually adding light if you want clean images.
And don’t be fooled by ISO ratings. They don’t tell you how noisy the picture is going to be, they don’t tell you what the sensitivity is or even if it’s actually changing. All it tells you is what to set a light meter to.
Camera setup, reviews, tutorials and information for pro camcorder users from Alister Chapman.