Hi guys,

I've just been going over Abelcine's ISO ratings for the F3 natively:


-3 db. = 500 ISO
0 db = 800 ISO
+3 db = 1000 ISO
+6 db = 1600 ISO
+9 db = 2000 ISO
+12 db = 3200 ISO
+18 db = 6400 ISO

And I'm wondering if anyone has confirmed ratings for the camera in SLOG? Is it simply +1 stop to all of the ratings (I.e. 0 DB = 1600 ISO, 18 DB = 12800 ISO)?

Cheers

Those sensitivity ratings are for interlace only, in progressive the sensitivity is halved.

Base S-Log ISO is 800 when progressive and 1600 when interlaced. You really don't want to use any other gain/ISO settings with S-Log as you will not have the full dynamic range.

Abelcine also rated the F3 in non-slog 0db as 800ISO (as did I with a gray card and meter) - so there is not change in ISO between slog and non-slog in progressive frame mode?

Yeah, watching Andy's video again, I noticed that he mentioned that although the exposure is higher with SLOG, once you 'de-log' the footage and bring the levels back within normal ranges - your exposure ends up the same anyway.

The F3 when in progressive and not in S-log is ISO 400, not 800, it's only 800 in interlace.

The sensitivity of the camera increases by 1 stop (6db) when you switch to S-Log mode. You must remember that middle grey with S-Log should be between 34-38% while for non s-log with rev-709 it would be around 46%.

When you de-log the footage your middle grey will become brighter. If you wanted to make middle grey brighter without changing anything else you would have to increase the camera ISO, not decrease it, so it is obvious that the camera is clearly more sensitive in S-Log than standard modes.

Thanks Alister. How do you usually set exposure with the camera? Point the 'marker' at a grey card and adjust lighting/aperture until it hits 38%?

Yep, that's pretty much it, quite simple really. I also check that white from a good quality matt white card is so somewhere around 68%. You normally end up about 1.5 stops under conventional exposures.

The F3 when in progressive and not in S-log is ISO 400, not 800, it's only 800 in interlace.

The sensitivity of the camera increases by 1 stop (6db) when you switch to S-Log mode. You must remember that middle grey with S-Log should be between 34-38% while for non s-log with rev-709 it would be around 46%.

Speaking strictly non-slog I don't agree that progressive is 400. I rated the camera as 800ISO in progressive 24p 0db, non-slog mode using a grey card at middle grey @ 45%-50%. It took about 42 footcandles with taking lens at f4 to light reach exposure for middle grey. My light meter at f4 read 800ISO.

Andy also did this test and measured f4 at ISO400 but the camera lens was at f5.6 with middle grey at 50%. So really at f4 it would be ISO800. Interlace would double it as you've suggested.

Yep, that's pretty much it, quite simple really. I also check that white from a good quality matt white card is so somewhere around 68%. You normally end up about 1.5 stops under conventional exposures.

I agree with getting mid-tones low around 40% but the suggestion of putting white at 68% (only 3 stops over middle grey) makes for very strict light control and just isn't practical in many situations.

I agree with getting mid-tones low around 40% but the suggestion of putting white at 68% (only 3 stops over middle grey) makes for very strict light control and just isn't practical in many situations.

Attached is a snipit from the Sony's white paper on S-LOG. The values for 18% and 100% are shown in the image.

Obviously we all have to work in the real world so we aren't always able to adhere to design criteria but it is good to know what it is so we can understand the risks we take when we exceed it.

You can reference the complete white paper here - http://www.sony.co.uk/res/attachment/file/66/1237476953066.pdf

Speaking strictly non-slog I don't agree that progressive is 400. I rated the camera as 800ISO in progressive 24p 0db, non-slog mode using a grey card at middle grey @ 45%-50%.

It depends a lot on your picture profile. Using my DSC Chroma Dumonde chart and Leader waveform monitor, I have rated my camera at ISO 250 in progressive, non-S-LOG. Exposing so no grading will be necessary . . . if I get it right.

Yes true, since different profile will shift your grey around.

Me and Andy Shipsides were both using Rec 709 where middle grey is around 50%. We measured 800ISO.

It should double in SLOG (I haven't confirmed that) and it should double in interlaced (I know it gets brighter but I never bothered measuring it).

Good stuff here..

Remember were dealing with stops which are ratios here, not finite values.

When your dealing with a good quality mid grey card (18% reflectance) and a white card (90% reflectance) in S-Log if you put middle grey at 38%, then white will be 68%, if it isn't then you middle grey is incorrect.

The contrast ratio between white and middle grey never changes, no matter how bright or dark the scene the middle grey will reflect 18% of the light and the white 90%, so the brightness ratio between the mid grey and white always remains constant, white will always be 500% brighter than mid grey. So it doesn't matter how bright your scene, with the S-Log curve, if middle grey is 38% then white will be 68%, 500% or 2.5 stops brighter than middle grey, always is, it's a reflectance ratio and the difference between mid grey and white will always be 2.5 stops which on the S-Log curve puts white at 68% when mid grey is 38.

You don't want to over expose with S-Log. It is highly compressed, if you over expose faces etc you risk putting them in the highly compressed part of the log curve and this makes it very hard to recover a decent looking image. You really need to keep faces and skin tones at least a stop under white, so faces need to be down at around 45-50%.

Do consider that when you look at the Sony S-Log graph linked to in this thread that the horizontal scale is stops, so it is a Log scale which is why the plot looks like a straight line. If the horizontal scale was linear the compressed nature of the curve would be obvious.

This has been a tremendously helpful and informative thread, thanks so much for your input everyone.

Thanks Alister. How do you usually set exposure with the camera? Point the 'marker' at a grey card and adjust lighting/aperture until it hits 38%?

One thing I've found is a potential "gotcha" would be making adjustments to the LUT through EV or even just turning the LUT on will start to bump up your grey card reading from 38% in the middle exposure box % readout of the LCD. What I do is to set the S-Log exposure at 38%, then tape the iris ring so it doesn't move. Then I can turn on LUT's and do EV increments without being fooled that I've actually opened the iris which you might think has happened if you look at the 38% starting to jump up in value on the LCD. Having a separate S-Log monitor like the Samurai screen is good to make sure you don't move the exposure on this accidentally.

Faces do tend to fall into the 40 - 50% area at MG 38%, but I like to stop down 1/4 stop or so to get them as close to 40% as possible since there is plenty of room to open them up in post.

I just shot Angelina Jolie on Thursday with S-Log on an F3 and I was very happy to have convinced the producers to use S-Log because she showed up in a pure white satin dress and I was able to easily hold the exposure on her well-lit face and the dress with S-Log.

Bruce, do you recall where the white dress fell on either F3 spot meter or ext. WF if you had one?

Thanks,

S-Log held that white at just under 70%, the LUT was displaying it right about 100%, but that was where I set the LUT at for HG40 + EV.

ok bruce, thanks.

Bruce why and how do you use the LUT.
I haven't used SLog much but I found I like it better to just set the monitor to higher contrast & sat. That way my monitors waveform is accurate & I can yank it back to preset and see what I'm really carrying in the highlights. I figure if it can look good in the monitor I'm plenty safe as long as I set my levels correctly. That's assuming a pretty loose working environment though. For clients it mighty bounce around though. Do you turn up the monitor saturation when you use the LUT?

I found on this shoot last week that the LUT was not looking even as good as the uncorrected S-Log feed so in the end I turned it off because I'd done a good enough job with the producer so that he was comfortable knowing that the desaturated but perfectly exposed S-Log would be an easy grade for him.

I'm pretty much at the point that I'm going to either have to do live grading on set using the S-Log feed, Live Grade, HDLink, and a MBPro - or start making custom LOOKs for the F3 which might be able to address the desaturation and exposure issues with camera LUT generation. Leaning towards the first, however, CVP File Editor is now available on Mac so making 3D LOOKs for camera might be a good starting point.

CVP file editor doesn't allow you raise saturation, so it doesn't really make a true viewing LUT to my mind. Also since the existing LUT's throw away the added headroom you are gaining in sLog then if you are running that feed through a waveform you're getting inaccurate waveform info. If I could make a really good LUT that also held my headroom , that would be great, but without it I'm sticking with the tweaking the monitor for now.

Just to clarify things a little further, if I rate the camera at 800 ISO (at 0dB in SLOG) on a lightmeter, that will give me Middle Grey at 38%?

Apparently ISO doubles in SLOG mode to 1600 from 800 but.........

I would strongly suggest you take a quick moment to light your grey card, fill your frame with it, then turn on your spot meter function and display the % IRE reading in camera, and adjust your lens iris until you get 38%

Now you can take a light meter at the grey card and dial up or down your ISO value until your f-stops match between the light meter and lens.

If you want to see a video by Abelcine where Andy Shipsides performs a similar test, here is a link: AbelCine Tests the Sony F3 ISO Rating | CineTechnica (http://blog.abelcine.com/2011/03/11/abelcine-tests-the-sony-f3-iso-rating/)

Sony rates S-Log 0 db at EI 800.
In my exposure test changing actual f-stops, shows 800 to be the middle between over and under exposure of the full S-Log DR.
I don't believe that you could determine actual usefulness of under or over exposure with a chart set at a given exposure without actually over and under exposing.
My 2¢

Sony PMW F3 S-Log Usable F-Stops Test on Vimeo

I have used SLOG with the camera rated at 800 ISO, but "The Film and Digital Times" article (http://pro.sony.com/bbsccms/assets/files/mkt/cinema/articles/36FDT-SonyF3-3-5-Med-Rez-RGB.pdf) however indicates on page 4 that "The ballpark sensitivity rating is ISO 800 (1600 in S-log), with an exposure range greater than 13 stop".

The best thing to do is shoot your own grey card and use spot meter function in the F3 to rate it yourself.

Rating your camera's ISO lower (say at 800 instead of 1600) will cause you to over expose by one stop, so you might place skin at 1 stop under middle grey using a camera rating of 800 but in reality you're placing skin right at middle grey.

Thanks Dennis, Douglas.

Dennis, having to wrangle the camera around to spot meter a grey card up close is what I'm hoping to be able to avoid (if I can). It definitely works for getting perfect exposure, I'm just wondering if I can make my life easier by picking up something like a Sekonic L758Cine (which is rather easier to wrangle).

Douglas, thanks for the video. Rated at ISO 800, the camera certainly still appears to have enough overexposure latitude in SLOG to turn in good results. Though with those lampshades in the background only a stop away from clipping at ISO800, I'm starting to think that rating the camera at ISO 1600 is probably the way to go (for the extra stop of overexposure latitude it provides), graded up it certainly still seems to hold all of the shadows and mid-tones well.

Cheers,

Mark

The confusion over the ISO rating comes from the way the F3's ISO doubles when you switch from progressive to interlace.
In interlace the base ISO is 800, in progressive 400. In S-Log mode that becomes 1600 in interlace and 800 in progressive.

Mark sorry if I wasn't clear, once you do your own verification of your ISO using your own grey card, light meter, camera and lenses, you won't need to use the grey card on your shoots or between shots, just use your light meter but the point is you did your own tests and calibration of equipment.

I say never trust any elses settings or observations and always do it yourself first, and in advance of your shoot. I just read a thread where a guy downloaded someones light meter profile for various cameras and went ahead and did a shoot without conducting any prior tests or validation. His footage was under exposed and he didn't know why. 1.) He used a profile for the wrong RED camera 2.) He ended up using the VF to judge exposure instead of "trusting" the light meter 3.) he did not test it.

Camera firmware updates can influence sensor sensitivity. Meters can be calibrated in 1/3 increments to give different readings, plus offer many modes of operation - often all of this resulting more often than not in the incorrect use of light meters - and then people hate them or say how useless they are. Everytime I shoot I use a light meter, I get the exact amount of light I need all day long as setups and camera angles change, and that consistency shows in my work.

More about how I use lightmeters and the F3 can found in my blog hingsberg's blog | Director & Cinematographer (http://www.hingsberg.com)

The confusion over the ISO rating comes from the way the F3's ISO doubles when you switch from progressive to interlace.
In interlace the base ISO is 800, in progressive 400. In S-Log mode that becomes 1600 in interlace and 800 in progressive.

Like I said always do your own calibration of light meter to your camera and setup. Myself and some others have rated the F3 in progressive mode 1/48, 0db mode at 800 ISO, not 400.

Regardless, the main thing when using a light meter is to know when you are hitting that middle grey with a specific volume of light.

I would say that the camera has a wide latitude that in most cases it really doesn't matter as long as you are within 400 and 1600 EI for your flesh tones. It is only when you go into the extreme ends of f-stops 6.5 under and over exposure.
For example: Assuming that you have a person exposed correctly under sunlight. Your background is a dark brown wall with an EV of 6.5 f-stops over the person, you will still get a nice image no matter if you rate it at 800 or 1600. If instead the wall was white even with the amount of light you may want to under expose the person in order to get details on the white (EI 1600).
It is mostly a problem when you have a third object that the EV is 6.5 f-stops under the person. If this object is white no problem, but if it is dark you will provable get no details.
There are not too many situations were you have 13 f-stops spread. If you do and you understand how the sensor reacts, you can play with it to get a decent image. You may need to decide if you can live with blown highlights or detail less shadows.
It really doesn't matter mush if you rate it at 400, 800, 1200 or 1600. At the mid point it will look good. At the extremes it could make the difference.

And remember that mid in LOG isn't 50%, its 38%. ;)

And remember that mid in LOG isn't 50%, its 38%. ;)

Yep SLOG middle grey is 38% and definitely remember that your middle grey points shift depending on the gamma curve you will use.

Here is a PDF with all the cine gammas curves and their middle grey points for reference which you can easily store on your iPhone or other smart phone:

http://www.starcentral.ca/Sony%20PMW.pdf

Yep SLOG middle grey is 38% and definitely remember that your middle grey points shift depending on the gamma curve you will use.

Here is a PDF with all the cine gammas curves and their middle grey points for reference which you can easily store on your iPhone or other smart phone:

http://www.starcentral.ca/Sony%20PMW.pdf

I would like to add something that I've not read in this or other discussions regarding exposure.

If anyone looks at those gamma curves it is important to understand one of the things those curves are telling you. Where you see more vertical distance between the blocks it is telling you that there is more digital luma steps being recorded there. This translates into less compression being applied which equals more SUBTLE detail being recorded.

Lets step out of the camera for a second and talk about our own visual perception. We are acutely aware of how people are supposed to look when we observe each other with our own eyes. With that in mind it should become instantly apparent where skintones should fall on the exposure curve right? They should fall into the LEAST compressed part of those curves meaning the parts where the spaces between the stops are the greatest. When you need to bias skintone exposure you want to have highlights compressed less than shadows. Our visual system is less forgiving when it comes to highlights and how the subtle translucence of skin appears to us and more forgiving when looking towards the shadows. This assumes you are looking to capture the most accurate representation of a person which may not be your goal from a creative standpoint.

This is where it can be tricky to use a light meter alone to set exposure with a video camera that uses a gamma curve different from film. This is why I choose to use a waveform monitor for judging exposure whenever possible. It allows me to more accurately place the most important parts of the image in the range where the least compression is being applied to those elements. I acknowledge that may cause the overall exposure to not be as equal from shot to shot. My priority is to give the most information to the most important elements in the frame and the final luminance range for presentation will be set during finishing.

Obviously I'm not shooting ENG. :)

Anyway, if you understand what those curves are telling you it becomes possible to make better decisions on your exposure. I've offered my thoughts only as a single point of reference and not as an indication of the pinnacle of exposure theory.

Chris is makes complete sense what you have described, there is much more data information in the mid and even low-end tones, and highlights are compressed heavily similar to how "gentle" they will appear when shooting on film.

I've therefore been debating on other forums some of the potential benefits of shooting linear Raw based formats like RED, and now 4k Raw found on the new F5/F55.

The other thing to really note is that with Raw you are getting 16-bit resolution, or 65,000 steps of data quantization divided by your camera dynamic range in stops. With 10-bit you only get 1000 steps. I see value in both linear and higher bit depth for high end projects but its not easy to argue or convince others of this, especially when feature films go through post in 10-bit dpx format - but that's AFTER it was shot in 12-bit Raw and converted.

+1 for what Chris said. It is vital with log that your mid range is in the correct part of the curve and in fact the lower down the curve you are the less compressed the image is. This is the biggest mistake made by some users of S-Log and similar gammas. They forget that this is a log compressed system where each stop of additional exposure is in effect allocated only half of the data of the previous stop. If you place mid grey at 38% as designed, even simple 90% reflectance white (which should be hitting about 68%) is being compressed 9 times more than mid grey. Your highlights are 42x more compressed than mid grey. Over exposing skin tones and mid tones makes it very difficult for the colourist to extract a meaningful amount of data and thus reconstruct the subtle tones and textures that make faces etc look pleasing to the viewer. That's why the EI values only go up and not down, so if exposing correctly you will only ever under expose with EI which pushes mids down the curve. This is of course where linear raw excels as there is no log compression so you can place your mid tones wherever you want within the cameras dynamic range.

Yes Alister I've done that on a few projects and learned the hard way. It was for that reason I felt that shooting SLOG needs a little more attention when shooting than that of linear RAW.

With LOG your highlghts are way more "baked in" as opposed to linear where you can shift your gamma without compromising data throughout the tonal range. I'm not saying its some fail safe, clipped highlights will still be clipped but linear will allow you more highlight "flexibility" if you will.

Would you sort of agree to that?

You guys all have it totally inside out regarding S-Log.

The entire point of an S-Log gamma curve is to effectively linearize the amount of bits per stop. That is why it looks flat as the brightness values are fairly evenly spread. That is why it's so easily gradable and highlights are easily recoverable if not clipped, giving it a more film like highlight response.

Regular non log gamma curves become more compressed with half the bits per stop, not log.

All you have to do is plot the code values against stops to see this is true. There is even a nice graph out there somewhere Sony released in the S-Log whitepaper.

You guys all have it totally inside out regarding S-Log.

The entire point of an S-Log gamma curve is to effectively linearize the amount of bits per stop. That is why it looks flat as the brightness values are fairly evenly spread. That is why it's so easily gradable and highlights are easily recoverable if not clipped, giving it a more film like highlight response.

Regular non log gamma curves become more compressed with half the bits per stop, not log.

All you have to do is plot the code values against stops to see this is true. There is even a nice graph out there somewhere Sony released in the S-Log whitepaper.

You mean this one?

http://i1119.photobucket.com/albums/k636/hingsberg/Curves_4.png

That's a marketing slide that might be vaguely representative of the dynamic of S-Log range squashed into a rec709 display after grading. It's pretty meaningless.

This is from the Sony whitepaper. It's the F35, not F3 curve, but the concept is the same.

Stops are not a liner measurement so even if you are allocating equal luma steps to each stop you are actually compressing the info as the exposure goes up.

I've said nothing about stops being linear.

Various people have been asserting S-Log compresses highlights greater than regular gamma curves.

It's actually the reverse. S-Log gives you more room through relatively even numbers of bits allocated rather than reducing bits per stop as you near clipping.

Yes, I understand that. It is correct that relative to other gamma curves s-log is a smoother roll off in the highlights.

I should have been more descriptive. I was talking about the difference between the chart Dennis posted and the one you posted. They are both fundamentally correct. The difference is one is showing the info in stops and the other is showing the info as linear.

While that slide is roughly analogous to S-Log displayed on a REC709 monitor, it's totally unrepresentative of the data actually recorded is S-Log, and easily misinterpreted, as shown in this thread.

Exposure isn't more critical in S-Log than regular gammas, it's actually more forgiving, and it doesn't compress highlights more than regular gammas, it compresses significantly less.

Regular gammas have a fat part of the bandwidth where skin tones work best, simply because skin has naturally very fine colour and luminance graduations. The fatter part of the gamma data curve gives a more realistic representation of skin. It doesn't look flat when exposed there, hence the old skin tone exposure rules.

S-Log has relatively even bandwidth spread over all stops, so skin can look natural almost anywhere in the spectrum. Now you can expose scenes as they look naturally to your eye if that's your aesthetic, and you don't end up with a flat dull grey faces in the dark or bright monotone under hard light.

So this entire SLOG video presented on behalf of Sony is incorrect?

VideON | F3 S-Log Stage Presentation | Home Page Featured Category (http://pro.sony.com/bbsc/video/related-home_page_featured/video-F3_S_Log_Stage_Presentation/)

Haven't watched it. Pick a point to discuss.

It talks about compression of the highlights, and that you have less bits of data for it, but as I understand it you are saying the opposite is true.

Look at the slide I posted, directly from Sony's engineers S-Log whitepaper. It clearly shows the code values incrementing in a linear fashion.

If the video you linked says the opposite about S-Log, I'd believe the engineers. I'd also believe my own eyes when grading F3 S-Log in Resolve. There is A LOT of information in the highlights you can recover.

In my experience, the Sony marketing departments and Sony engineering departments don't seem to communicate very well. The marketing department tends to make technical errors as they are not engineers and do not appear to run things past people who know often enough.

I'd believe the engineers. I'd also believe my own eyes when grading F3 S-Log in Resolve. There is A LOT of information in the highlights you can recover.
+1 You can only recover on very few color grading programs like Resolve, Color, etc,.

Edit: I shoot 10 bit 422. I assume 444 is better and 420 can't be as good.

I think your forgetting some key points David.

When looking at the Sony S-Log white paper, the relationship between the data bits used and f-stops appears linear, because it is. When looking at the graphs in the Sony white paper you need to understand that the horizontal scale is a log scale, not a linear one, so the line on the graph appears straight, because the scale is log. F-Stops are logarithmic. Replace the scale with a linear scale and you would see the familiar log curve. Because the scale is log the line appears straight, but what you are seeing is a non linear relationship between illumination and recording data.

In the video, during my presentation I clearly state that log works by using roughly the same amount of data bits per stop. Exactly as per the graph you are referring to. BUT cameras and monitors are not log devices, they are linear. The scale in the chart is a log scale which is why the line looks straight. If you replaced the log scale with a % scale or lux scale you would see a very highly curved, very non-linear response curve.

S-Log does compress highlights more aggressively than standard gammas and cinegammas. It has to, that's how you squeeze a greater dynamic range into the same sized bit bucket. There is no free lunch, if you want to fit something bigger into the same box, your going to have to squash it more. In order to squeeze more dynamic range into the same bit bucket with S-log the onset of compression is much earlier than standard gammas and cinegammas and the earlier onset of this compression means that accurate exposure is more important and is also part of the reason why middle grey is shifted lower down the exposure range, to keep the important stuff away from the more compressed parts of the curve. Adding and extra stop and a half of dynamic range requires 200% more compression over the cinegammas and hypergammas gammas. Yes there is a lot of information in the highlights of an S-Log image that can be extracted, but if it's anything like skin tones or natural textures, things the human visual system is good a noticing problems with, you will definitely pick up on the compression.

First, camera sensors and monitors are linear devices, but the f-stop scale is a log scale and finally our visual system is essentially a log system.

In a perfect world we would simply take the full linear output from the camera and feed it in to an equally linear monitor and bingo, we would have a very accurate representation of the scene being filmed that to our eyes would look completely true to life.

Sadly though it's not as simple as that as our own visual system is not linear, it is logarithmic, like the f-stop scale. So we don't perceive a significant brightness change in an image unless you DOUBLE the brightness. With linear devices like a video monitor or TV, to make the image look brighter you must double the output, which requires double the data. As we work in a world where there is a finite limit on how much bandwidth or data that is available this is actually very difficult to do, especially if you want a really large brightness range.

IMPORTANT: The numbers given in the examples below are for illustration only, they are not accurate and do not represent actual data or exposure levels, thet are just there to illustrate the principle behind log.

The F-stop scale is logarithmic. Each additional stop of exposure requires twice as much illumination as the previous stop and doubles the tonal range of the image. So if it takes 2 light bulbs to provide the first stop of illumination, then it would take 2 more light bulbs to increase the exposure by 1 stop. Four light bulbs for 2 stops, 8 lightbulbs for 3 stops, 16 light bulbs for 4 stops, 32 for 5 stops and so on. 10 stops would need 1024 bulbs!

Remember a camera is a linear device. To directly record the output from the sensor, if each lightbulb takes one bit of data to record, you would only need 512 bits of data to record 9 stops, but would need twice that, 1024 bits to record 10 stops and a whopping 4096 bits to record 12 stops. So as you can see for a camera, a linear device, to record ever greater exposure ranges requires massive amounts of extra data.

Lightbulbs/sensor data 2-----4-----8-----16-----32-----64-----128-----256-----512-----1024-----2048
F-stop-------------------1-----2-----3------4------5-------6-------7--------8-------9-------10---------11

You can see from this crude example that if you are using an 8 bit recording system with only 240 bits of data to play with, that recording anything beyond about 9 stops requires something different (the knee).

What you must remember is that as each additional stop doubles the tonal range of the image, if you only allocate the same amount of data to each stop, then as the image gets brighter then progressively less and less data is being used to record the linear output from the camera.

Lightbulbs/sensor data 2-----4-----8-----16-----32-----64-----128-----256-----512-----1024-----2048
F-stop-------------------1-----2-----3------4------5-------6-------7--------8-------9-------10---------11
S-Log data bits--------10----20----30----40-----50-----60------70------80------90------100------110

Taking the first stop, the sensors output uses 2 bits of data. With S-log we have 10 bits available to record it, no problem.
Taking the 10th stop, the sensors linear output uses up bits 512 to 1024, 512 bits of data, yet with log recording, allocating roughly the same amount of data to each stop, we squeeze that into only 10 bits of recorded data. So, while the first stop had an excess of data, the brighter stop can only be recorded by discarding sensor data. This is the key to the way log works, as you go up the exposure range, more and more sensor data is discarded, thus more and more tonal information is discarded in the brighter parts of the image.

Conventional gammas like 709 are nearly linear, so almost twice as much data is used to record subsequent stops. It must be remembered that video camera sensors, TV's and monitors are linear devices. For them to accurately portray the real world then you would ideally just take the full linear feed from the camera and simply send that to the monitor. This would however require a massive amount of bandwidth or data for anything beyond about 6 stops (because each extra stop of brightness would need DOUBLE the amount of data). That's why Rec-709 is only actually designed for 6 stops, because it is a linear gamma curve (although most camera manufacturers cheat and provide a Rec-709 compatible gamma that gives around 8 or 9 stops) Because us humans have a visual system that is essentially log we can cheat the system and record using highlight compression (reduced information in the highlights) and the viewer will not in most cases notice this and this is where cinegammas, hypergammas and log come in to play, discarding data in the highlights.

If your still not convinced take a look at the unprocessed linear raw output from a DSLR or camera like the BMCC or Epic. What you will see even with a correctly exposed image is a very, very dark image with perhaps a few highlights here and there. That's because the vast majority of the image data is getting used to record the high lights and very brightest parts of the scene so the bulk of the image (the mid range etc) ends up way way down the data range making the image look very very dark. With the new F5/F55 with 14 stops of DR and 16 bit linear raw your mid tones will have around 2,000 bits of data per stop while your highlights may use a whopping 30,000 bits.

422, 444, 420 make absolutely no difference to dynamic range. With a RGB 444 recording it will be slightly easier to make large post production colour shifts than 422 or 420. The difference between 422 and 420 in progressive is extremely small and makes very little difference to grading, it only really effects chroma resolution. The colour space is the same for 444, 422 and 420.

Who needs film school when you've got Alister?! Thanks again for clear, detailed and incredibly informative post mate.

I think your forgetting some key points David.

S-Log does not compress highlights more than regular gammas. Easily testable in Resolve. I use it to my advantage everyday.

Of course the scale is logarithmic, that's why it can allocate the bits evenly. Seriously, that is the only reason log gammas exist, to allow even bit distribution rather than reducing bit distribution in the highlights.