Tag Archives: log

Noise, ISO, Gain, S-Log2 v S-Log3 and exposure.

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.

Advertisements

ProRes Raw Over Exposure Magic Tricks – It’s all smoke and mirrors!

There are a lot of videos circulating on the web right now showing what appears to be some kind of magic trick where someone has shot over exposed, recorded the over exposed images using ProRes Raw and then as if by magic made some adjustments to the footage and it goes from being almost nothing but a white out of over exposure to a perfectly exposed image.

This isn’t magic, this isn’t raw suddenly giving you more over exposure range than you have with log, this is nothing more than a quirk of the way FCP-X handles ProRes Raw material.

Before going any further – this isn’t a put-down of raw or ProRes raw. It’s really great to be able to take raw sensor data and record that with only minimal processing. There are a lot of benefits to shooting with raw (see my earlier post showing all the extra data that 12 bit raw can give). But a magic ability to let you over expose by seemingly crazy amounts isn’t something raw does any better than log.

Currently to work with ProRes Raw you have to go through FCP-X. FCP-X applies a default sequence of transforms to the Raw footage to get it from raw data to a viewable image. These all expect the footage to be exposed exactly as per the camera manufacturers recommendations, with no leeway. Inside FCP-X it’s either exposed exactly right, or it isn’t.

The default decode settings include a heavy highlight roll-off. Apple call it “Tone Mapping”. Fancy words used to make it sound special but it’s really no different to a LUT or the transforms and processes that take place in other raw decoders. Like a LUT it maps very specific values in the raw data  to very specific output brightness values. So if you shoot just a bit bright – as you would often do with log to improve the signal to noise ratio – The ProRes raw appears to be heavily over exposed. This is because anything bright ends up crushed into nothing but flat white by the default highlight roll off that is applied by default.

In reality the material is probably only marginally over exposed, maybe just one to 2 stops which is something we have become used to doing with log. When you view brightly exposed log, the log itself doesn’t look over exposed, but if you apply a narrow high contrast 709 LUT to it, it then the footage looks over exposed until you grade it or add an exposure compensated LUT.  This is what is happening by default inside FCP-X, a transform is being applied that makes brightly exposed footage look very bright and possibly over exposed – because thats the way it was shot!

This is why in FCP-X  it is typical to change the color library to WCG (Wide Color Gamut) as this changes the way FCP-X processes the raw, changing the Tone Mapping and most importantly getting rid of the highlight roll off. With no roll-off, highlights and any even slight over exposure will still blow out as you can’t show 14 stops on a conventional 6 stop TV or monitor. Anything beyond the first 6 stops will be lost, the image will look over exposed until you grade or adjust the material to control the brighter parts of the image and bring them back into a viewable range. When you are in WCG mode in FCP-X the there is no longer a highlight roll off crushing the highlights and now because they are not crushed they can be recovered, but there isn’t any more highlight range than you would have if you shot with log on the same camera!

None of this is some kind of Raw over exposure magic trick as is often portrayed. It’s simply not really understanding how the workflow works and appreciating that if you shoot bright – well it’s going to look bright – until you normalise it in post. We do this all the time with log via LUT’s and grading too! It can be a little more straight forward to recover highlights from Linear Raw footage as comes form an FS5 or FS7 compared to log. That’s because of the way log maintains a constant data level in each highlight stop and often normal grading and colour correction tools don’t deal with this correctly. The highlight range is there, but it can be tricky to normalise the log without log grading tools such as the log controls in DaVinci Resolve.

Another problem is the common use of LUT’s on log footage. The vast majority of LUT’s add a highlight roll off, if you try to grade the highlights after adding a LUT with a highlight roll off it’s going to be next to impossible to recover the highlights. You must do the highlight recovery before the LUT is added or use a LUT that has compensation for any over exposure. All of these things can give the impression that log has less highlight range than the raw from the same camera. This is not normally the case, both will be the same as it’s the sensor that limits the range.

The difference in the highlight behaviour is in the workflows and very often both log and raw workflows are miss-understood. This can lead to owners and users of these cameras thinking that one process has more than the other, when in reality there is no difference, it’s appears to be different because the workflow works in a different way.

Log and Raw Don’t have highlight a highlight roll off.

This just keeps coming up over and over. Almost all log gamma curves and the majority of raw recording formats don’t have a highlight roll-off. Any roll off that you might see is probably in the LUT’s that you are using.

The whole point of log and raw is to capture as much information about the scene that you are shooing as you can. Log normally achieves this by recording every stop above middle grey with a constant amount of data, so even the very brightest stop has the same amount of recording data as the ones below it – there is no roll off.

In conventional limited range recordings such as Rec-709, hypergamma, cinegamma etc, highlight roll-offs work by reducing the contrast in the highlights to make the amount of data needed to record the very brightest stops much smaller than used for the rest of the image. This allows 2 or 3 stops to be squeezed into a very small recording range, keeping most of the recording data available for a nice bright high contrast image. The reduction in contrast in the extreme highlights helps hide any highlight handling problems and makes it appear as though the sensors clipping point is reach in a more pleasing soft manner.

But you don’t want this in a log or raw recording as it makes grading much harder as the footage will contain different contrast ranges, each needing it’s own grading adjustments. Also by reducing contrast in the highlights you are reducing the data. It would be very difficult to un-pick a highlight roll off and if you did want to expand the data back out you will get issues such as banding.

S-log-levels Log and Raw Don't have highlight a highlight roll off.
Chart showing S-Log2 and S-Log3 plotted against f-stops and code values. Note how little data there is for each of the darker stops, the best data is above middle grey and there is no highlight roll-off. Note that current sensor only go to +6 stops over middle grey so S-Log2 and S-Log record to different peak levels.

S-Log2 and S-Log3 like almost all log gammas have no highlight roll-off. The only roll off is from middle grey and down. So if you underexpose you will start to roll away the data in your scenes mid range and that’s not good. Expose for the mid range, this is the most important part of any image. If your highlights are a bit clipped don’t worry about this too much. In post production you can add a roll off in the grade that will make any clipped highlights roll away gently. Adding a bit of highlight diffusion in post will also nicely mask any clipped highlights and make them look natural.

Shooting Flat – No it’s not!

I know that many of my readers like to shoot log. One of the most common terms used around shooting log is “shooting flat”. Lets take a look at that term and think about what it actually means.

One description of a flat image might be – “An image with low contrast”. Certainly an image with low contrast can be considered flat.

Once upon a time shooting flat meant lighting a scene so that there was very little contrast. The background in an interview might be quite well  lit. You would avoid deep shadows or strong highlights. This was done because cameras had very limited dynamic ranges. These flat images of low contrast scenes could then have the contrast boosted in post production to make them look better.

8 years ago, with the advent of DSLR cameras that could shoot with film like depths of field it became fashionable to shoot flat because digital film cameras  when shooting using log produced an image that looks flat when viewed on a conventional TV or monitor.

But lets think about that for a moment. A typical digital cinema camera can capture 14 stops of dynamic range. A scene with 14 stops of dynamic range contains a huge contrast range, perhaps a brilliant bright sky and deep shadows, you can possibly describe the capture a scene with 14 stops of dynamic range as “flat”?

The answer is you can’t – or at least you shouldn’t because the recording  isn’t flat. The dynamic range that most digital cinema cameras can capture is not flat, not at all.

The problem is that a normal TV or video monitor can’t show a very big dynamic range. A conventional TV can only show around 6 stops. If you take a log video signal with a 14 stop image and try to show that on a 6 stop screen you will be squashing the highlights and shadows closer together, so the highlight that was at +14 stops in the scene and is recorded at 100%, gets pushed closer to the deepest shadow in the scene that is recorded at 1%.

On a normal 6 stop TV the 100% recording level is shown at +6 stops while the deepest shadow will be at 1%, so now the 14 stop recording is being shown with only 6 stops between the deepest black and the brightest highlight. Instead of the highlight being dazzlingly bright it’s now just a bright white and not all that much brighter than the shadows. As a result the image on the screen looks all wrong, nothing like what you recorded and it appears to be “flat”.

BUT THE DATA IN THE FILE IS NOT FLAT – that recording contains a high contrast, 14 stop image – it’s the inability of the TV or monitor to show it correctly that makes it look wrong, not that you have shot flat.

In the early days of DSLR shooting many DSLR shooters decided to mimic the way the image from a digital cinema camera looks flat on a normal TV, perhaps in the miss-guided belief that a flat image must always have a greater dynamic range. This definitely isn’t always the case. I can take any regular dynamic range image and make it look flat by reducing the contrast, raising the blacks a bit, shifting the gamma perhaps, that’s easy. But that doesn’t increase the dynamic range that is captured. Changing the capture range of a camera typically requires fundamental changes to the way it operates rather than simple tweaks to the basic picture settings.

So we went through a period where shooting a flat looking image with a DSLR was the trendy way to shoot because on a normal TV or monitor the image recorded is reminiscent of the image from a true digital cinema camera shooting log, even though in practice the “flat look”  was often damaging the image rather than improving it.

Now there are many digital cinema cameras that can capture a very big dynamic range using log encoding and these images look washed out and flat on a normal monitor or TV because of the miss-match between the camera and the monitor, not because the captured scene is flat. But we still call this shooting flat (wrong)!

Why? In many cases people like to leave the image this way as they like this “incorrect” look. Flat is trendy, it’s fashionable, at least to those inside the TV and Video production world. I’m not sure that the wider general audience really understands why their pictures look washed out.

If you have a monitor with high dynamic range display capabilities such as a Atomos Shogun Flame or Inferno, that can show a large dynamic range then you’ll know that if you feed it log and set the display range to HDR and choose the right gamma curve, the picture on the screen is no longer flat, it’s bright and contrasty. This isn’t a LUT or any other cheat. The monitor is simply showing the image with a range much closer to the capture range and now it looks right again.

storm-PQ-14stop-1024x577 Shooting Flat - No it's not!
This is a high dynamic range image. View it on an HDR TV set to HDR10 and it will be brilliantly bright, highly colorfull and full of contrast. On a regular TV or monitor it looks flat and washed out because the regular TV can’t show it properly.

So next time you use the term “shooting Flat” think very carefully about what it actually means and whether you are really shooting flat or whether it’s simply a case of using the wrong monitor. Using words or terms like this incorrectly causes all kinds of problems. For example most people think that log footage is flat and that that’s how it’s supposed to look. But it isn’t flat and it’s not supposed to look flat, we are just using the wrong monitors!

Should I shoot 8 bit UHD or 10 bit HD?

This comes up so many times, probably because the answer is rarely clear cut.

First lets look at exactly what the difference between an 8 bit and a 10 bit recording is.
Both will have the same dynamic range. Both will have the same contrast. Both will have the same color range. One does not  necessarily have more color or contrast than the other. The only thing you can be sure of is the difference in the number of code values. An 8 bit video recording has a maximum of 235 code values per channel giving 13 million possible tonal values. 10 bit recording has up to 970 code values per channel giving up to 912 million tonal values.
 
There is a lot of talk of 8 bit recordings resulting in banding because there are only 235 luma shades. This is a bit of a half truth. It is true that if you have a monochrome image there would only be 235 steps. But we are normally making colour images so we are typically dealing with 13 million tonal values, not simply 235 luma shades. In addition it is worth remembering that the bulk of our current video distribution and display technologies are 8 bit – 8 bit H264, 8 bit screens etc. There are more and more 10 bit codecs coming along as well as more 10 bit screens, but the vast majority are still 8 bit.
Compression artefacts cause far more banding problems than too few steps in the recording codec. Most codecs use some form of noise reduction to help reduce the amount of data that needs to be encoded and this can result in banding. Many codecs divide the image data into blocks and  the edges of these small blocks can lead to banding and stepping.
 
Of course 10 bit can give you more shades. But then 4K gives you more shades too. So an 8 bit UHD recording can sometimes have more shades than a 10 bit HD recording. How is this possible? If you think about it, in UHD each color object in the scene is sampled with twice as many pixels. Imagine a gradient that spans 4 pixels. In 4K you will have 4 samples and 4 steps. In HD you will only have 2 samples and 2 steps, so the HD image might show a single big step while the 4K may have 4 smaller steps. It all depends on how steep the gradient is and how it falls relative to the pixels. It then also depends on how you will handle the footage in post production.
 
So it is not as clear cut as often made out. For some shots with lots of textures 4K 8 bit might actually give more data for grading than 10 bit HD. In other scenes 10 bit HD might be better.
 
Anyone that is getting “muddy” results in 4K compared to HD is doing something wrong. Going from 8 bit 4K to 10 bit HD should not change the image contrast, brightness or color range. The images shouldn’t really look significantly different. Sure the 10 bit HD recording might show some subtle textures a little better, but then the 8 bit 4K might have more texture resolution.
 
My experience is that both work and both have pro’s and con’s. I started shooting 8 bit S-log when the Sony PMW-F3 was introduced 7 years ago and have always been able to get great results provided you expose well. 10 bit UHD would be preferable, I’m not suggesting otherwise (at least 10 GOOD bits are always preferable), but 8 bit works too. 

Using LUT’s for exposure – choosing the right LUT.

If using a LUT to judge the exposure of a camera shooting log or raw it’s really important that you fully understand how that LUT works.

When a LUT is created it will expect a specific input range and convert that input range to a very specific output range. If you change the input range then the output will range will be different and it may not be correct. As an example a LUT designed and created for use with S-Log2 should not be used with S-Log3 material as the the higher middle grey level used by S-Log3 would mean that the mid range of the LUT’s output would be much brighter than it should be.

Another consideration comes when you start offsetting your exposure levels, perhaps to achieve a brighter log exposure so that after grading the footage will have less noise.

Lets look at a version of Sony’s 709(800) LUT designed to be used with S-Log3 for a moment. This LUT expects middle grey to come in at 41% and it will output middle grey at 43%. It will expect a white card to be at 61% and it will output that same shade of white at a little over 85%. Anything on the S-Log3 side brighter than 61% (white) is considered a highlight and the LUT will compress the highlight range (almost 4 stops) into the output range between 85% and 109% resulting in flat looking highlights. This is all perfectly fine if you expose at the levels suggested by Sony. But what happens if you do expose brighter and try to use the same LUT either in camera or in post production?

Well if you expose 1.5 stops brighter on the log side middle grey becomes around 54% and white becomes around 74%. Skin tones which sit half way between middle grey and white will be around 64% on the LUT’s input. That’s going to cause a problem! The LUT considers anything brighter than 61% on it’s input to be a highlight and it will compresses anything brighter than 61%. As a result on the output of your LUT your skin tones will not only be bright, but they will be compressed and flat looking. This makes them hard to grade. This is why if you are shooting a bit brighter it is much, much easier to grade your footage if your LUT’s have offsets to allow for this over exposure.

If the camera has an EI mode (like the FS7, F5, F55 etc) the EI mode offsets the LUT’s input so you don’t see this problem in camera but there are other problems you can encounter if you are not careful like unintentional over exposure when using the Sony LC709 series of LUTs.

Sony’s  709(800) LUT closely matches the gamma of most normal monitors and viewfinders, so 709(800) will deliver the correct contrast ie. contrast that matches the scene you are shooting plus it will give conventional TV brightness levels when viewed on standard monitors or viewfinders.

If you use any of the LC709 LUT’s you will have a miss-match between the LUT’s gamma and the monitors gamma so the images will show lower contrast and the levels will be lower than conventional TV levels when exposed correctly. LC709 stands for low contrast gamma with 709 color primaries, it is not 709 gamma!

Sony’s LC709 Type A LUT is very popular as it mimics the way an Arri Alexa might look. That’s fine but you also need to be aware that the correct exposure levels for this non-standard LC gamma are middle grey at around 41% and white at 70%.

An easy trap to fall into is to set the camera to a low EI to gain a brighter log exposure and then to use one of the LC709 LUT’s and try to eyeball the exposure. Because the LC709 LUT’s are darker and flatter it’s harder to eyeball the exposure and often people will expose them as you would regular 709. This then results in a double over exposure. Bright because of the intentional use of the lower EI but even brighter because the LUT has been exposed at or close to conventional 709 brightness. If you were to mistakenly expose the LC709TypeA LUT with skin tones at 70%, white at 90% etc then that will add almost 2 stops to the log exposure on top of any EI offset.

Above middle grey with 709(800) a 1 stop exposure change results in an a 20% change in brightness, with LC709TypeA the same exposure change only gives a just over 10% change, as a result over or under exposure is much less obvious and harder to measure or judge by eye with LC709. The cameras default zebra settings for example have a 10% window. So with LC709 you could easily be a whole stop out, while with 709(800) only half a stop.

Personally when shooting I don’t really care too much about how the image looks in terms of brightness and contrast. I’m more interested in using the built in LUT’s to ensure my exposure is where I want it to be. So for exposure assessment I prefer to use the LUT that is going to show the biggest change when my exposure is not where it should be. For the “look” I will feed a separate monitor and apply any stylised looks there. To understand how my highlights and shadows, above and below the LUT’s range are being captured I use the Hi/Low Key function.

If you are someone that creates your own LUT’s an important consideration is to ensure that if you are shooting test shots, then grading these test shots to produce a LUT it’s really, really important that the test shots are very accurately exposed.

You have 2 choices here. You can either expose at the levels recommended by Sony and then use EI to add any offsets or you can offset the exposure in camera and not use EI but instead rely on the offset that will end up in the LUT. What is never a good idea is to add an EI offset to a LUT that was also offset.

Why are Sony’s ISO’s different between standard gammas and log?

With Sony’s log capable cameras (and most other manufacturers) when you switch between the standard gamma curves and log gamma there is a change in the cameras ISO rating. For example the FS7 is rated at 800 ISO in rec709 but rated at 2000 ISO in log. Why does this change occur and how does it effect the pictures you shoot?

As 709 etc has a limited DR (between around 6 and 10 stops depending on the knee settings) while the sensor itself has a 14 stop range, you only need to take a small part of the sensors full range to produce that smaller range 709 or hypergamma image. That gives the camera manufacturer some freedom to pick the sweetest part of the sensors range. his also gives some leeway as to where you place the base ISO.

I suspect Sony chose 800 ISO for the FS7 and F5 etc as that’s the sensors sweet spot, I certainly don’t think it was an accidental choice.

What is ISO on an electronic camera? ISO is the equivalent sensitivity rating. It isn’t a measure of the cameras actual sensitivity, it is the ISO rating you need to enter into a light meter if you were using an external light meter to get the correct exposure settings. It is the equivalent sensitivity. Remember we can’t change the sensor in these cameras so we can’t actually change the cameras real sensitivity, all we can do is use different amounts of gain or signal amplification to make the pictures brighter or darker.

When you go switch the camera to log you have no choice other than to take everything the sensor offers. It’s a 14 stop sensor and if you want to record 14 stops, then you have to take 100% of the sensors output. The camera manufacturer then chooses what they believe is the best exposure mid point point where they feel there is an acceptable compromise between noise, highlight and lowlight response. From that the manufacture will get an ISO equivalent exposure rating.

If you have an F5, FS7 or other Sony log camera, look at what happens when you switch from rec709 to S-Log2 but you keep your exposure constant.

Middle grey stays more or less where it is, the highlights come down. White will drop from 90% to around 73%. But the ISO rating given by the camera increases from 800ISO to 2000ISO. This increased ISO number implies that the sensor became more sensitive – This is not the case and a little missleading. If you set the camera up to display gain in dB and switch between rec709 (std gamma) and S-Log the camera stays at 0dB, this should be telling you that there is no change to the cameras gain, no change to it’s sensitivity. Yet the ISO rating changes – why?

The only reason the ISO number increases is to force us to underexpose the sensor by 1.3 stops (relative to standard gammas such as rec709 and almost every other gamma) so we can squeeze a bit more out of the highlights. If you were using an external light meter to set your exposure if you change the ISO setting on the light meter from 800 ISO to 2000 ISO  the light meter will tell you to close the aperture by 1.3 stops. So that’s what we do on the camera, we close the aperture down a bit to gain some extra highlight range.

But all this comes at the expense of the shadows and mid range. Because you are putting less light on the sensor if you use 2000 ISO as your base setting the shadows and mids are now not as good as they would be  in 709 or with the other standard gammas.

This is part of the reason why I recommend that you shoot with log between 1 and 2 stops brighter than the base levels given by Sony. If you shoot 1 stop brighter that is the equivalent to shooting at 1000 ISO and this is closer to the 800 ISO that Sony rate the camera at in standard gamma.  Shooting that bit brighter gives you a much better mid range that grades much better.

 

Why is exposing log brightly beneficial?

I have been asked whether you should still expose log a bit brighter than the recommended base levels on the Sony PXW-FS5 now that Sony have released new firmware that gives it a slightly lower base ISO. In this article I take a look at why it might be a good idea to expose log (with any camera) a bit brighter than perhaps the manufacturer recommends.

There are a couple of reasons to expose log nice and bright, not just noise. Exposing log brighter makes no difference to the dynamic range. That’s determined by the sensor and the gain point at which the sensor is working. You want the camera to be at it’s native sensitivity or 0dB gain to get that maximum dynamic range.

Exposing brighter or darker doesn’t change the dynamic range but it does move the mid point of the exposure range up and down.  Exposing brighter increases the under exposure range but decreases the over exposure range. Exposing darker decreases the under exposure range but increases the over exposure range.

Something that’s important when thinking about dynamic range and big dynamic ranges in particular is that dynamic range isn’t just about the highlights it’s also about the shadows, it isn’t just over exposure, it’s under exposure too, it’s RANGE.

So why is a little bit of extra light often beneficial? You might call it “over exposure” but that’s not a term I like to use as it implies “too much exposure”. I prefer to use “brighter exposure”.

It’s actually quite simple, it’s about putting a bit more light on to the sensor. Most sensors perform better when you put a little extra light on them. One thing you can be absolutely sure of – if you don’t put enough light on the sensor you won’t get the best pictures.

Slide01 Why is exposing log brightly beneficial?

Put more light on to the sensor and the shadows come up out of the sensors noise floor. So you will see further into the shadows. I’ve had people comment that “why would I ever want to use the shadows, they are always noisy and grainy”? But that’s the whole point – expose a bit brighter and the shadows will be much less noisy, they will come up out of the noise. Expose 1 stop brighter and you halve the shadow noise (for the same shadows at the previous exposure).  Shadows are are only ever noise ridden if you have under exposed them.

This is particularly relevant in controlled lighting. Say you light a scene for 9 stops. So you have 9 stops of dynamic range but a 14 stop sensor. Open up the aperture, put more light on the sensor, you get a better signal to noise ratio, less noisy shadows but no compromise of any type to the highlights because if the scene is 9 stops and you have 14 to play with, you can bring the exposure up by a couple of stops comfortably within the 14 stop capture range.

S-log-levels Why is exposing log brightly beneficial?
Chart showing S-Log2 and S-Log3 plotted against f-stops and code values. Note how little data there is for each of the darker stops, the best data is above middle grey. Note that current sensor only go to +6 stops ove middle grey so S-Log2 and S-Log record to different peak levels.

Look at the above diagram of Sony’s S-Log2 and S-Log3 curves. The vertical 0 line in the middle is middle grey. Note how above middle grey the log curves are more or less straight lines. That’s because above the nominal middle grey exposure level each stop is recorded with the same amount of data, this you get a straight line when you plot the curve against exposure stops. So that means that it makes very little difference where you expose the brighter parts of the image. Expose skin tones at stop + 1 or stop +3 and they will have a very similar amount of code values (I’m not considering the way dynamic range expands in the scene you shoot as you increase the light in the scene in this discussion). So it makes little difference whether you expose those skin tones at stop +1 or +3, after grading they will look the same.

Looking at the S-Log curve plots again note what happens below the “0” middle grey line. The curves roll off into the shadows. Each stop you go down has less data than the one before, roughly half as much. This mimics the way the light in a real scene behaves, but it also means there is less data for each stop. This is one of the key reasons why you never, ever want to be under exposed as if you are underexposed you mid range ends up in this roll off and will lack data making it not only noisy but also hard to grade as it will lack contrast and tonal information.

Open up by 1 additional stop and each of those darker stops is raised higher up the recording curve by one stop and every stop that was previously below middle grey doubles the amount of tonal values compared to before, so that’s 8 stops that will have 2x more data than before. This gives you a nice fat (lots of data) mid range that grades much better, not just because it has less noise but because you have a lot more data where you really need it – in the mid range.

Note: Skin tones can cover a wide exposure range, but typically the mid point is around 1 to 1.5 stops above middle grey. In a high contrast lighting situation skin tones will start just under middle grey and extend to about 2 stops over. If you accidentally under expose by 1 stop or perhaps don’t have enough light for the correct exposure you will seriously degrade the quality of your skin tones as half of your skin tones will be well below middle grey and in the data roll-off.

Now of course you do have to remember that if your scene does have a very large dynamic range opening up an extra stop might mean that some of the very brightest highlights might end up clipped. But I’d happily give up a couple of specular highlights for a richer more detailed mid range because when it comes to highlights – A: you can’t show them properly anyway because we don’t have 14 stop TV screens and B: because highlights are the least important part of our visual range.

A further consideration when we think about the highlights is that with log there is no highlight roll-off. Most conventional gamma curves incorporate a highlight roll-off to help increase the highlight range. These traditional highlight roll-offs reduce the contrast in the highlights as the levels are squeezed together and as a result the highlights contain very little tonal information. So even after grading they never look good, no matter what you do. But log has no highlight roll-off. So even the very brightest stop, the one right on the edge of clipping contains just as much tonal information as each of the other brighter than middle grey stops. As a result there is an amazingly large amount of detail than can be pulled out of these very bright stops, much more than you would ever be able to pull from most conventional gammas.

Compare log to standard gammas for a moment. Log has a shadow roll-off but no highlight roll-off. Most standard gammas have a strong highlight roll-off. Log is the opposite of standard gammas. With standard gammas, because of the highlight roll-off, we normally avoid over exposure because it doesn’t look good. With Log we need to avoid under exposure because of the shadow roll-off, it is the opposite to shooting with standard gammas.

As a result I strongly recommend you never, ever under expose log. I normally like to shoot log between 1 and 2 stops brighter than the manufacturers base recommendation.

Next week: Why is a Sony camera like the FS7,F5 800 ISO with standard gamma but 2000 ISO in log and how does that impact the image?

 

Beware Exposing To The Right With Log.

That may seem like quite a sensational headline – beware exposing to the right with log – but let me explain.

First of all, I’m not saying you can’t or shouldn’t expose to the right, all I am saying is beware – understand the implications.

First of all what is normally meant by exposing to the right? Well it’s a term that comes from the world of photography where you would use the cameras histogram to measure the exposure levels. Exposing to the right would normally mean setting the shutter speed and aperture so that the levels shown on the histogram are as far to the right as you can get them without going beyond the right side of the histogram. This would ensure a nice bright exposure with lots of light falling on the sensor, something that is normally highly desirable as you get a nice low noise picture once you have adjusted and processed it in your photo editing software.

You can expose to the right with a video camera too. However when shooting with Rec-709 or conventional gammas this can often result in nasty looking highlights thanks to the default knee settings, so it’s not normally a good idea for 709 and standard gammas.

With log or raw as there is no highlight roll off you can expose to the right and it should give you a nice bright exposure… or will it?????

The problem with exposing to the right is that you are exposing for the highlights in the scene. If shooting a low contrast or low dynamic range scene this isn’t going to cause any problems as exposing to the right will mean that everything in the scene is nice and bright.

But if shooting a high dynamic range scene, say an outdoor scene with bright clouds in the sky but large areas of shadow, the exposure will be optimised for the highlights. The mid range and shadows may end up too dark. On a sunny day if shooting a person with their back to the sun the sky could easily be 6 or 7 stops brighter than the skin tones. If you expose for the sky/highlights the skin tones will be 1 or 2 stops darker than the basic exposure level recommended for most  log curves.

(S-log2/3 has 14 stops. At the base exposure you have 6 stops above middle grey and 8 below. Skin tones are normally between 1 and 2 stops above middle grey. So if the sky/highlights are 6 stops above the skin tones, then exposing for the highlights will put the skin tones where middle grey should be, which is 1 stop under exposed and 2 stops below where I would normally like to see skin tones when shooting with log or raw).

The first thing a viewer will notice when they look at a scene with faces or people will be the skin tones. If these have been under exposed they will be grainy and less than ideal. The viewer will notice noise and grain and poor shadows long before they look at the brightest highlights. Shooting log and protecting the highlights or exposing to the right will often compromise the all important mid tones because you are exposing for the highlights, not the midrange. In addition exposing for highlights with a high dynamic range scene can often push the shadows down in level and they will end up noisy and grainy. The biggest issue with exposing to the right is that it’s extremely difficult to estimate how many stops there are between your mid tones and the highlights, so you never know quite where your mid tones are falling.

(Midtones – generally a white piece of paper or a 90% reflectivity white card would be considered to be the top end of the mid tones. Go down about 2.5 stops from white and you hit middle grey  (18% grey card). This range between middle grey and white is where skin tones, plants, most animals etc will be and it probably the most important part of most images).

An important consideration with log and raw is that there is no highlight roll off. Standard gammas (with the default knee found on almost every camera) , cinegammas, hypergammas etc all roll off the highlights. That is to say that as you approach the peak recording level the contrast is reduced as the highlights are squeezed together to try to extend the dynamic range. This reduction in contrast means that it is very difficult, if not impossible, to recover any nice, useable picture information out of anything close to the peak recording level. As a result with conventional gammas we tend to avoid over exposure at all costs as it looks nasty. This highlight roll off is one of the things that gives video the video look.

Log and raw don’t have this same kind of highlight roll off. The image gets brighter and brighter until it clips. With log the stop immediately below clipping contains just as much picture information as any other stop brighter than middle grey. With linear raw the stop just below clipping has more information than any other stop. As a result in post production there is a very large amount of data that can be pulled out of these highlights, even if they are a little clipped! So don’t worry about a few clipped highlights when shooting log. The other thing to remember is there is no TV or monitor that can show these highlights as they really are, so they will never look perfect anyway.

Another thing that happens when exposing to the right is that grading becomes harder than it needs to be. Because the separation between the mid tones and highlights will vary greatly depending on things like whether you are shooting into or away from the sun, when you expose to the right you mid tone brightness will be up and down all over the place. So in post production as well as adding the look that you want to your footage, you are also going to have to spend a lot of time matching the mid range exposure to balance skin tones etc from shot to shot.

Rather than exposing to the right what I recommend is exposing for the mid range. After all this is the important part of the image. To do this you need to use a diffuse reflective shade. The most commonly used shades are a 90% white card and/or an 18% reflectivity grey card – middle grey. Get the mid range right and in most cases the highlights will take care of themselves. Getting the mid range right might mean exposing the mid range  brighter than the recommended levels. But it’s the mid range we need to measure, not the highlights, this is the important part of the image.

90% white is an incredibly important level in the world of film and video. A typical piece of office paper reflects about 92-94% of the light falling on it. Office paper often uses brighteners and special chemicals to make it look bright and white. This white is the brightest diffuse surface you will likely ever see.  Anything brighter than this is normally going to be an actual source of light. The sky perhaps or a direct bounced reflection off a shiny, reflective surface such as the bodywork of a car. So anything brighter than 90% white would normally be considered to be a highlight and to us humans, highlights are visually less important than the mid range. This is why the knee on most video cameras kicks in at around 90%. Anything brighter than 90% is a highlight so the knee only effects highlights and leaves the all important mid range alone.

Middle grey is also very important because it’s a shade of grey that to most people looks to be half way between black and white. Skin tones fall roughly half way between middle grey and white. In addition if you average all the brightness levels within a typical scene the end result is typically very close to middle grey.  Light meters are calibrated to middle grey. The relationship between middle grey and white is fixed. White reflects 90%, middle grey 18%, no matter how bright the actual light source. So whether you are indoors, outside. Whether it’s sunny or overcast, white and middle grey will always be close to 2.5 stops apart. They are extremely useful fixed reference levels.

There are many ways to measure the brightness of a white or grey card. My preferred method is with a waveform display. But you could also use zebras (use a narrow zebra window if you can).  You can also use false colour. Unfortunately it’s very difficult to use a histogram to measure the brightness of a specific target. The histogram is a great measuring tool for photography, but less than ideal for video. If you can’t get a white/grey card out in front of the camera you could consider using a light meter. It’s also worth noting that skin tones sit just a little over half way between middle grey and white, so if you have no other reference you could simply place your skin tones a touch brighter than half way between the values you are targetting for middle grey and white.

Just to be clear: I do still recommend exposing Sony’s S-log2, S-log3 and raw between 1 and 2 stops brighter than the Sony base levels. But the key take-away is that it’s the mid range you need to measure and expose at this level. Exposing to the right using a histogram or waveform and just looking at the peaks and brightest parts of the image does not tell you what is happening in he mid range. Measure the mid range, not the peak brightness.

 

Guide to Cine EI – Still Current.

Just a reminder that my guide to shooting with Cine EI for the PMW-F5 and F55 cameras is still just as valid today as it was when I wrote it back in 2013. There have been a few tweaks to the cameras menu here and there, but the principles and basic operation have not changed.

So if you are new to Cine-EI please take a look at the guide. It takes you through how to shoot with Cine EI, which LUT’s to use and how to expose them.

Cine-EI Mode when recording S-Log2/3 and raw on the F5 and F55.