Sony's S-log For Dummies – Matt Allard explains how it works with his PMW-F3

By Matt Allard

Sony's S-log upgrade comes on a 32GB SXS card

Like me you probably hadn’t heard of Sony S-log or even had the faintest idea about what it is or what it does.

S-log was originally only available on super high end Cine cameras like the Sony F35. Now that it’s available for the Sony F3 I thought I’d do some research and try and explain it in as simple language as possible. Now before I do I’m not some super tekkie guy or an engineer so i’m not even going to try and explain why it does what it does from a technical perspective. I’m just going to explain what it does to an image in the real world. I have researched by reading various articles and I’ve tried to not make this article too technical. I am also happy to be corrected on anything I have gotten wrong.

S-log For Dummies from Matthew Allard on Vimeo.

The short video I have made is just some test shots to try and show you what S-log does and looks like. It’s not a scientific test, it’s just me using the camera. It hasn’t been recorded on an external recorder – its all been recorded straight onto the internal SxS cards. All comparison shots were done using the exact same lenses and settings unless otherwise noted.  I’ve shot comparisons between S-log/non S-log, graded S-log/ungraded S-log and graded non S-log/graded S-log. There are also comparisons between a Canon 7D using the Technicolor Cinestyle profile and the F3 with and without S-log.

In action with the Sony PMW-F3

In a nutshell, S-log is supposed to offer you exceptional tonal reproduction in both the high-lights and low-lights. In other words it can help expose the brightest part of your scene while still keeping the detail in the shadows. We have all shot something when, to expose for the brightest part of a scene, we have had to stop the camera down. This usually means losing all the detail in the shadows. S-log helps you to keep some of that information you would normally lose.

The human eye can see around a 14 stop range of light but it works in the reverse way of a camera sensor. We see only small changes in dark conditions but very big changes in bright ones. With a cameras sensor it sees big changes in dark conditions but only small changes in bright ones. What this means is our eyes see most things that are dark as being dark, but when it comes to things that are bright like say the sky, we can make out very gradual changes in brightness. Thats why a camera struggles when you try and shoot a bright sky while trying to expose the foreground correctly – it can’t easily recognize the gradual steps in bright scenes like your eye can.

Straight from a camera’s sensor an image appears very dark if it doesn’t have a gamma curve added to it. Almost all HD video cameras process video with a gamma curve so when it is viewed on your TV it  looks correct . This viewing method is referred to as REC709. What REC709 can’t do though is display the entire range of what your cameras sensor can capture. The F3 has a 16 bit sensor but without S-log you can’t capture all that information into a 10 bit signal. Here’s where S-log earns its money. It bends the range of the sensor to fit all the information into the allowable 10 bit space. It loses some information in this process but it now allows the camera to mimick the changes in brightness that are visable to our eyes. So remember earlier when I talked about why a camera struggles with replicating bright scenes? Well this is what S-log is helping correct.

The conventional way to combat the bright skies would be with a ND grad filter in front of the lens, but I found when using Sony F3 these filters dont tend to look as good as on say my Sony PMW-350 XDCAM. With the F3 you can see a noticeable graininess where my filters are graduated and I imagine that unless you are using the highest possible quality filters you will still be able to see the change in quality. S-log benefits you greatly by allowing you to do without filters in some cases because it handles the highlights so much better.

The S-log menu

S-log does have to be color corrected in post though. A non color corrected S-log image looks washed out, lacks color and contrast. You don’t have to record S-log to a 4.4.4 external recorder like a Convergent Designs Gemini or even to a 4.2.2 recorder like a AJA KiPro Mini. It can be recorded straight to SxS cards but this not the best way to use it as S-log benefits in the color correction process from recording to a much higher bitrate codec and broader color space. When shooting S-log you can only choose either 5600k or 3200k and you can only shoot in progressive and not interlaced (I couldn’t find an explanation as to why).

Are there any negatives to S-log? From my experiences so far you don’t need to use it for everything and it’s not a quick fix miracle that will magically make all your material look wonderful. It does add increased noise to the image so you have to be careful when using it.

The positives? It is fantastic when you have to shoot scenes that has a high level of contrast. I have been able to shoot a lot of high contrast scenes I would never even attempt without S-log. It is also quite remarkable in low light. Turning on S-log is like turning on extra lights in a room. It also provides an amazing reproduction of detail that you normally never see. At 18db with the S-log enabled the F3 can literally see in the dark. My eyes can’t see what this camera can with a fast lens.

My over all impression of S-log is “wow!” It is however a $3800US upgrade and is not for everyone. I do however thoroughly recommend it as it really does take your F3 and put it on steroids.

About Matthew Allard, Aljazeera Senior Field Cameraman, Kuala Lumpur:
Matt has been a Camera/Editor in TV news for more 20 years, previously working for both Channel 9 and Channel 10 in Australia. Twice Network Ten Australia’s cameraman of the year as well as being a Walkley Finalist for outstanding camerawork in 2006 (for coverage of the Cronulla Race Riots) and a Logie Finalist for outstanding news coverage 2006 (Bali 9). He is a multiple ACS (Australian Cinematographers Society) award winner. His Sword Maker story that was shot on a 7D won the prestigious Neil Davis International News Golden Tripod at the 2011 ACS Awards. He has covered news events in more than 35 countries, from major sporting events to terrorist bombings. Based out of the Kuala Lumpur broadcast centre in Malaysia he is an avid user and follower of new technology, shooting stories on HD broadcast cameras, the Sony F3 as well as new Canon DSLRs.

You can follow him on Twitter @mattaljazeera

daniel-chung-web-ad-640x120px 640x100-5 daniel-chung-web-ad-640x120px daniel-chung-web-ad-640x120px daniel-chung-web-ad-640x120px

Posted on September 22nd, 2011 by Matthew Allard | Category: DSLR video news, Sony F3 |

7 responses to "Sony's S-log For Dummies – Matt Allard explains how it works with his PMW-F3"

  1. Sam_Morgan_Moore Says:
    September 22nd, 2011 at 11:51 pm

    S Log for dummies

    OK Im a dummy on video post production

    SLog shoots flat

    Surely when you add a curve in post to create an image with proper contrast you will need to crush one end or the other

    Negating the wide DR gain becuae you just crushed it away

    Now with raw still we appreciate that wide DR because we have the post tools that can use the full sensor data – eg Lightroom

    With Lightroom we can selectively manipulate ares of an image (for exampl pull back window detail fomr images that are exposed for an interior)

    I can see that a (motion) post house with time and money to get the full range of the image from SLog could benefit by using computer heavy post software

    Im a little confused how Slog would benefit the solo shooter on a deadline beyond giving them a little more wiggle room in thier initial exposure when that shooter does not have access to the higher end post software ?

    I can see it give a litte more wiggle room (space for mis exposure) and also may control baked in looks which may be a little too saturatedd and video like – beyond that – im lost ?

  2. Matthew Allard Says:
    September 23rd, 2011 at 12:28 am

    Hi Sam,

    You don’t need expensive color correction software to manipulate S-log. I usually use Apples Color to do mine. For these examples I just did a super basic grade within FCP. S-log lets you record a much greater dynamic range. If you record something without having that information present you can’t recover it. It’s up to you in post what you do with the image. Color or Resolve just to name a few, are able to easily manipulate the full sensor data. Color is free, Resolve light is free so here is no money to spend.

    As with anything you shot it is up to you to use what tools you have available within the constraints of time. I grade all my material in Color and grading S-log is not much more difficult than grading my regular stuff. I don’t recommend S-log for everyone, but it is an amazing feature.

    The biggest gain with S-log is its ability to hold in detail in the highlights. This is something you can never recover without S-log. Yes you do have to crush blacks as any gamma curve like this will give you elevated black levels.

    S-log is basically a sort of RAW mode (it’s not actually a RAW image) for video or film. Just like in Lightroom you can manipulate the picture. The higher the color space and codec you record to the more you can push the image.

    I hope this helps.

  3. Sam_Morgan_Moore Says:
    September 23rd, 2011 at 12:38 am


    Im questioning (and im a vid post newb) the fundamental difference between stills and motion post

    Sure, with Slog you can bend the file a little more

    In stills we would use the raw lattitude to work on specific areas of an image

    eg to ‘burn in’ the sky

    When you can work on specific areas of an image, say the sky you can take the full depth from a file, be it raw or psuedo raw (ie a flat profile)

    If you are unable to work on a specific are of the image you cant do this as far as I can see ie burning the sky will crush the blacks in your shadows

    Now the sky is a bit of a simple example because im sure thats easy to do

    But to get the reall ‘stills post’ effect, where we would be all over an image, tweaking colour and contrast (maybe on a face) you need a post solution that can track motion

    An AKFAIK that is neither cheap nor fast ?


  4. Matthew Allard Says:
    September 23rd, 2011 at 1:41 am

    Hi Sam,

    I’m not any expert on any of this and I’m definitely not an expert in photography.I’l see if Dan can help yo out with this question as he is a professional photographer and can probably answer your question a little bit better than me.

    Hope we can come up with a better answer to your question.

  5. Sam_Morgan_Moore Says:
    September 23rd, 2011 at 1:52 am

    yeah you need to see the crimes that we commit on stills photos – thats lattitude !

    : )

  6. federico gnoli Says:
    September 23rd, 2011 at 12:42 pm

    Hi mat! I was wondering if you could help me, i would like to know what tripoddo you use with the F3, i primary use HDSLR, and travel a lot and I don’t know wich one I can buy.
    The cine dslr could be used also with an F3?

  7. Matthew Allard Says:
    September 23rd, 2011 at 9:47 pm

    Hi Federico,

    I usually use a Sachtler 20p head and 18p legs. This is a broadcast tripod so it’s not light nor cheap (around 15k US). I also have a Manfrotto 509 HD which I think is an excellent tripod (around $1500US). If you are after light but good quality and your mainly using HDSLR then i would have a look at the new Sachtler Ace tripod

    Hope this helps.

Leave a comment

You must be logged in to post a comment.

Website by Kevin Woo Designs