alisterchapman

Rather than Pelican cases I prefer to use hard shell suitcases (normally Samsonite) as there is nothing that shouts out - expensive stuff here - more than a Pelicase. Pelicases are great when the gear is always in your control, for loading in and out of a van and general use, but they are a target for thieves and over many years of flying with kit there have been several times where equipment in pelicases or flight cases has disappeared never to be seen again, but (touch wood) I have never lost a suitcase, even though more than a few times they have arrived late or temporarily gone missing. Also as Doug has observed many newer tripods (including my Miller solo's) are short enough to fit diagonally in a suitcase. I prefer to keep a low profile when travelling with kit, especially if overseas or in parts of the world I don't know so well and suitcases help me to do this. Re 150Wh batteries, you are supposed to obtain the airlines permission for any lithium batteries between 100Wh and 150Wh. They are permitted, but only with permission. I've been stopped with a couple of 150's many times, but if you can show you have gained the airlines permission or show that the airline blanket permits anyone to travel with 100-150Wh batts then it's not an issue. But as this varies from airline to airline or even route to route you cannot assume that you will always be allowed through security or onboard with 100-150Wh batts.

Here's a table of the Base ISO values for the Cinema Line cameras. The FR7 is the same as the FX6.

If you have an FX6 or FX9 you can use the built in intervalometer to shoot great time-lapse. This can be combined with the slow shutter to get long exposures if you want motion blur. You can also use the FX6 or FX9's programmable AE settings to add limits to the amount of auto gain that will be applied, or the range of auto iris that will be used and where the crossover between each auto function occurs. In addition you can use the variable ND on auto to further extend the range of light that you can get an acceptable exposure over. What you don't get is raw still frames, but S-Log3 is highly gradable.

I first came across the new FR7 at IBC in Amsterdam in September and at first I was a little confused by it. Who would actually want one and what for? But now I am convinced that this will be a big hit - what's changed my mind? For those that haven't seen it yet the FR7 is in essence a Sony FX6 camera that has been adapted and modified to fit in a remotely controlled pan and tilt head. You get all of the FX6's features and performance including onboard recording, S-Log3 and CineEI, raw output, built in variable ND etc etc but in a form factor that allows you to operate the camera remotely via either a control panel with a joystick or via any device with a web browser. Remote PTZ (Pan, Tilt, Zoom) cameras get used a lot in shows like Big Brother, Love Island, First Date etc but until now, generally the image quality hasn't been as good as most mainstream broadcast cameras. When you cut between the PTZ camera and other main cameras the difference is often quite obvious. So, straight away there is the obvious use of the FR7 for these types of shows, so that the image quality will be as good as any other cameras being used. Additionally PTZ cameras get used a lot in wildlife productions. A PTZ camera can be placed close to an animals nest or placed closer to a feeding area without disturbing the animals natural behaviour as a camera operator might. Again, often the quality of these cameras was different to the main cameras and more often than not the small sensors in most PTZ cameras doesn't do well in low light. The ability to use the FR7's high sensitivity mode and shoot S-Log3 at 12,800 ISO with a fast lens will really open up a lot of new possibilities for wildlife film makers. The FR7 is controlled over an ethernet connection, so with a simple 4G router and a connection to the internet the camera could be controlled from the comfort of a warm studio anywhere in the world. But what about other applications? Would a freelance camera operator like me benefit from one? Well I think the answer may be yes. For 2 camera shoots such as interviews the FR7 can be used as a second camera and you have the ability to control it from the main camera position. This would be so much easier than having to walk over to the second camera to make a simple adjustment or reframe it. Instead of being a simple locked off shot that never ever changes your B camera can be moved and adjusted more frequently to add more variety to your shots. And what about getting shots from places or angles that aren't normally possible? If you shoot conferences or events, being able to place the camera on the front of the stage or in front of the podium opens up a lot of new possibilities. The FR7 won't obstruct the audiences view in the same way that a camera on a tripod with an operator will. I'm going to be shooting some live performances with one very soon and it will allow me to get the camera into locations around the stage where you just can't normally get a shot. You can hang it from the ceiling or a lighting truss for overhead shots. You can even mount it on a jib. You can also use it for timelapse or other shots where you need to repeat the same move over and over. By setting up preset positions for the start and end points you can perform extremely slow moves all the way to very fast moves from point to point and each time the move is the same. Like the FX6 the FR7 has a built in intervalometer (interval record), so shooting timelapse is easy. The big deal about the FR7 is that while PTZ cameras are not new, they have never offered the image performance possible from a large sensor camera. Because the images from the FR7 closely match the rest of the Sony cinema line it opens up more possible uses. And the cost isn't prohibitive, it's not that much more than a normal FX6.

I can't see how to upload an avatar. I can change my avatar to one of the preset images but I don't see any way to upload one.

Often this is a question that is led less by how you shoot and more by how (or who) will be doing the post production. If you do all your own post production then it really is a matter of personal preference. Do you wish to give yourself the best possible "negative" to work with in post by shooting with S-Log3 or do you want a fast turnaround with minimal extra post production effort, in which case you may chose to shoot with S-Cinetone or a baked in LUT. As not every project you may do will be the same, perhaps sometimes you will shoot S-Log3 and others S-Cinetone. Don't be afraid to switch depending on each projects individual needs. There is no right or wrong with either. Where the decision becomes more complex is when you are handing off your footage to someone else to edit/grade. Do they know how to handle S-Log3 correctly or are they going to make a mess of it? For the vast majority of TV news type applications they will want something that can go direct to air with as little work as possible, so they are unlikely to want S-Log3. But for other types of productions the water is much muddier so a conversation needs to be had with whoever will do the post production to make sure that what you are giving them is what they want. Generally for S-Log3 I will not bake in a LUT in camera. I don't see any reason to bake in a LUT as adding a LUT in post production is such a simple and quick process, depending on the software being used it sometimes happens automatically or a LUT can be applied to an entire timeline or project. If you have been careful with your exposure and have set the white balance correctly then no further grading should be necessary to get to the same look that you had at the time of shooting.

I often hear the term Over Exposure used to describe shooting with S-Log3 (or S-Log2) exposed more brightly than perhaps what we would call normal. But is this really over exposure? First of all - shooting with log and exposing more brightly isn't something that always must be done. It may often be beneficial as it is a way of reducing noise and increasing shadow range (at the expense of some of your highlight range), but it shouldn't be seen as something that is essential in every case. Sony's earlier S-Log cameras, the FS700, F5, F55, FS7 and FS5 had sensors that were much noisier than the sensors in the newer FX30, FX3, FX6, FX9 and Venice. These newer, lower noise sensors are capable of producing perfectly acceptable results without being exposed more brightly than the suggested base levels. But, noise is a personal thing and how much noise one person deems acceptable may be different to what another might be happy with and some projects may need to have less noise than another. So exposing more brightly to reduce noise is still a useful tool for some projects, but it doesn't need to be the default way to shoot. So, getting back to my original point: Exposing more brightly to get a better final image is not over exposure. It is in fact the "correct" exposure - it just happens to be brighter. Generally over exposure is something that most would consider undesirable, a mistake perhaps. When you look at conventional footage that has been over exposed it generally doesn't look as good as it should. With Log, if you have exposed it brighter deliberately to gain a better looking final image this shouldn't really be considered to be over exposed (the final image won't look over exposed), it is simply a deliberately brighter exposure, that's all. If something is exposed darkly to look great this isn't under exposure, it's simply a dark exposure. But, if you have exposed too bright and the image looks bad as a result then in this case, it is over exposure. Why is this even important? Using the term "over exposure" to describe a sometimes beneficial shooting method is quite intimidating and a little confusing for those that don't understand the process. "If you over expose your footage it will look great" is a contradiction in terms, because in many peoples minds over exposure is a bad thing, so how can doing a bad thing make footage look great? But "if you expose more brightly your footage will look great" is a more easily believed and perhaps easier to understand concept.

Over exposing is rarely a good thing with something where grading or colour management isn't an integral part of the delivery process. S-Cinetone is designed to look great directly out of the camera, but it is not vanilla Rec-709, so it works best with slightly different (lower) brightness levels to 709. The FX3/FX6/FX9 etc with S-Cinetone all behave exactly the same.

When you use LUT's you do not loose any of the original image information, it's all still there. But if you grade your material by adding corrections down stream (after) the LUT then the LUT will have baked in its look and then you are adjusting that look, you are not adjusting the log. So, you need to do your adjustments in a node or layer before the LUT, that way you are adjusting the Log before it gets to the LUT and then you will have access to the full tonal range that was captured. One thing well designed LUT's also do is take care of the colourspace transform that is necessary to bring the S-Log content into the Rec-709 colourspace that most are working in. ACES and other colour managed workflows do generally simplify the grading process as they take care of the conversion from the shooting colour space (S-Log3/SGamut) to the desired final colour space - Rec-709, HDR10 or whatever it is you need to view and deliver in. Simple direct grading of S-Log3 material in a non colour managed Rec-709 project does not include any colourspace transforms, so sometimes you will get funky colours or odd behaviours that make getting exactly what you want tricky. Colour management eliminates these issues and generally the footage is handled in a way that makes changing the white balance or exposure much simpler. BUT - ACES in particular is designed to add a film like look. This can work for you very nicely if you want a film style roll off and lower level skin tones etc. But it can also work against you if you are working with content that you want to "pop" and be bright, the ACES roll off that is applied on the output signal will try to stop your content from getting bright. So, there is no one fits all workflow. ACES is great for film style projects. LUT's are great for fast turn-around with pre determined looks. A colour managed workflow such as Resolves own may be better for brighter and more vibrant projects. In the near future when we will need to deliver in multiple standards so that you will have an HDR version for the multitude of HDR devices from phones to TV's that are already out there, as well as Rec-709 for legacy devices. Or use processes such as Dolby to provide a single multi-standard compatible deliverable colour managed workflows will become both necessary and normal.

I do use auto a lot, and generally it works well, but I hate those quiet moments in the audio where the auto gain creeps up and the background noise increases, then someone speaks and the background noise abruptly changes as the auto compensates for the louder voice. So for interviews or other situations where the audio won't vary that much I make use of the limiter. The limiter is a variable slope soft limiter, so they will almost never hard clip doing their thing, which is to prevent audio peaks from clipping or distorting. Instead once the audio level goes above the limiter set point it is very gradually compressed, much like a video highlight roll off. The further into the compressor range it goes the more it is compressed. The greater the limiter range, the softer the effect. So, as an example, I will set the limiter level to -9dB, this means that the soft limiter will start to kick in at the -9dB level on the audio meters, which is approx 3/4 of the way up the scale. Then set the manual audio level so that the loudest normal audio peaks just reach 3/4 of the way up the scale. If the audio then suddenly becomes much louder (for example the person you are interviewing raises their voice) the peaks will start to go into the limiter and will be gently compressed preventing them from distorting, the person will sound louder as the average sound energy will increase (as it should) but there won't be any pumping of the audio level as there can be with auto gain. Obviously there will eventually be a point where the compression becomes audible if the sound level significantly increases but with a -9dB limiter the sound has to become 8 times louder before you'll start to notice anything. -12dB and you need a 16x increase before the limiter is even detectable, let alone objectionable.

I think there is a bit of a common miss-conception about grading flexibility with different gammas. S-Log doesn't have any greater grading flexibility than any other gamma. However so many of the commonly used tools such as LUT's or other plugins found in most edit applications evolved as simply ways to grade log footage within the constraints of a 709 workflow so many have gained the impression that Log has more flexibility. In reality if you take log, 709 or any other gamma into a decent post production workflow such as a proper colour managed workflow like ACES and you eliminate the gamma by transforming it into a linear space then there is no difference in the grading flexibility, all will grade equally well as in essence all contain the very same image information. Gamma curves are simply a mathematical calculation applied to the original material to make it fit in a limited recording range. They do not make an image more or less gradable and being a simple mathematical calculation they can be reversed to remove the gamma. What S-Log does allow for is a greater dynamic range and a larger DR might be important for many shots, but this isn't always the case. This greater range is achieved by squashing the original material using log math and this allows more stops to be stored in the same size data bucket. But, that comes at a price. S-Log3 and 14+ stops recorded into code values 95-886 means each stop on average has 56 code values (you actually have more for brighter stops but less for stops below middle grey). If you use vanilla Rec-709, with no knee (not S-Cinetone or any of the other modified versions of 709) then you will have 6 stops in code values 64 to 1023 so on average each stop will have 159 code values, almost 3 times as many as log. I'm quite sure that most reading this will already be aware of the advantages that more code values per stop brings, after all it's why we prefer 10 bit over 8 bit recordings. The difference in code values per stop between S-Log3 and vanilla Rec-709 is not that different to the difference between an 8 bit and a 10 bit recording. However in a considerably under exposed situation where neither the log or the 709 can be fully exposed it isn't going to make a huge difference what gamma you use because the limiting factor is the actual sensitivity of the sensor and the noise it produces. To go from a certain scene brightness to a like for like final screen brightness the cumulative gain used (sensor gain + gamma gain + post gain) will be the same whatever gamma you use, so the ratio of signal to noise will be broadly the same whatever route you take and it's the noise that will likely limit how well you can grade or manipulate the material rather than the gamma choice. But if your gamma choice allows you to fill your recording range rather than only using a small part of it then you will have more code values per stop with a narrow range gamma than a wide range one and that will provide more textures and that will grade better in just the same way as a 10 bit recording with 970 code values will grade better than an 8 bit recording with 235 values. So for some studio applications where you have control over your lighting there can be advantages for using 709. But all of this also requires care with your workflow, simply dropping the clips into an edit applications timeline isn't going to be ideal as you will be constrained by the working gamma. You really want to use a proper grading tool with good colour management if you really want to get the best results.

Different LUTs may have different brightness ranges compared to other LUTs or commonly used gammas such as Rec-709. As an example where normal Rec-709 might be exposed so that a white card is around 90%, the s709 LUT is correctly exposed when the same white card is at 83%. So, it is important to know what the expected brightness level for each LUT that you use is so that you can expose it correctly. If you have a new LUT that you haven’t used before. How do you find the correct exposure level for that LUT? If your camera has a waveform display as the FS7/FX6/FX9 do this is easy. It's a little trickier with the FX3 and FX30, but still possible. Set the camera up in the CineEI mode. Start by turning on the cameras waveform then turning OFF the LUT. If you don't have a waveform monitor set the zebra level to 61% (On the FS3 and FX30 use the C1 custom zebra setting to set the zebra level to 61% and the range to +/-2% - the narrower range makes the zebra indication more precise). Now with the LUT OFF you are measuring the S-Log3 and you should expose the S-Log3 so that a 90% reflectivity white card is at 61% and middle grey card at 41%. If you don'r have a proper grey card a piece of white paper can be used but in this case it should be exposed a touch brighter at around 63-64%. Now turn ON the LUT for the SDI and HDMI (the waveform measures the signal on the SDI/HDMI what the waveform is measuring is indicated directly above the waveform display), check that the EI is set to 800 to match the base 800 ISO (if you camera has a different base ISO set the EI to match this different base ISO). Now look at the waveform and wherever white and middle grey are, that is the correct level for the selected LUT. Skin tones will typically be half way between white and middle grey. So, if with the LUT ON white ends up at 80% and middle grey at 45% then skin tones will be approx 62%. On the FX6 and FX9 If you are unsure of where 61% or 41% are then go to the Zebra settings and set Zebra 1 to 41% and Zebra 2 to 61%, this will add orange lines to the waveform at 41% and 61%. You can also then experiment with the levels of the zebra values and the orange lines , moving them up and down up to confirm the levels you are seeing on the waveform when you measure the LUT. On the FX3 and FX30 you will need to move the zebra level up and down to discover the point where the zebras appear over the white card with the LUT on. This can be done using the C1 custom zebra setting and you will be able to move the zebra level up and down until it appears over the white card while viewing the live image on the LCD.

Every camera in Sony's Cinema Line has the ability to shoot at 2 different base sensitivity levels. On Sony Venice, the FX9 and the FX30 the sensors used have two different operating modes that deliver 2 distinctly different sensitivities to light. In each mode the sensors retain almost identical levels of noise and dynamic range. This allows a film maker using these cameras to be able to switch between either of the dual base ISO's according to the light levels they are working under and achieve near identical results with either (there is generaly just a touch more noise at the higher level). The gap between the upper and lower base ISO tends to be between 2 and 3 stops depending on the camera you are using. For S-Log3- Venice 1 is 500/2500 ISO Venice 2 with the 8K sensor is 800/3200 ISO FX9 is 800/4000 ISO FX30 is 800/2500 ISO Sony's FX3 and FX6 cameras have a slightly different approach where the camera offers 2 distinctly different base sensitivity ranges, Sony don't refer to this as Dual Base ISO but instead call it dual sensitivity. I'm not exactly sure of how this is achieved, but it is likely a combination of adjustments to the way the sensor operates, it's gain and the processing applied to the sensor output. What it delivers is two different sensitivity ranges with a 4 stop gap between the upper and lower range. With S-Log3 on the FX3 and FX6 you get 800 ISO and 12,800 ISO. Unlike dual base ISO there is a noticable difference in the noise between these two sensitivity levels, but the noise difference is nowhere near as large as it would be if you started at 800 ISO and added a whopping 24dB of gain to get up to 12,800 ISO. There is also a small reduction to the dynamic range at the upper base ISO. So, is one better than the other? Well, it depends on what you need. If you need a camera to shoot in extremely low light then the dual sensitivity mode of the FX3 and FX6 means you can get some remakable performance in very low light levels. But if you are shooting a drama, performance or similar sometimes the 12,800 ISO from an FX3 or FX6 will be a bit too noisy or simply more sensitivity than you really need, so you might end up shooting at 3200 or 6400 EI which will result in much less noise but will reduce your highlight range. For this kind of application Venice or the FX9 with it's dual base ISO system might be better. The new FX30 has a dual base ISO sensor and this needs to be considered when comparing low light performance simply by looking at the specifications. On paper it may appear that the FX3 with its upper high base sensitivity of 12,800 will always perform significantly better than the FX30 where the upper base ISO is 2500. But the FX30 at 2500 produces a very low noise image while the FX3 at 12,800 is more noisy. The reality is that while the FX3 and FX6 will remain the better choice for extreme low light the FX30 really isn't that far behind and performs very well in low light.

“Slow shutter” or "SLS" is when your cameras electronic shutter is open for longer than 1 frame of video. Normally a video camera operates at 25 or 30 frames per second. So, the camera sensor normally captures light for 1/25th or 1/30th of a second or some smaller fraction of the length of each frame such as 1/50th or 1/60th. The sensor is then reset, captures for the next 1/25th or 1/30th and then writes the next frame and so on, creating a video sequence. With a slow shutter the sensor is allowed to capture light for more than the length of one frame, so the shutter opening spans several frames. As an example with a 16 frame slow shutter the sensor is allowed to capture light for 16 frames before creating an image. At 30fps, where one frame lasts 1/30th of a second a 16 frame slow shutter will be open for 16 x 1/30th = 0.53 seconds. The sensor is being allowed to capture light for 0.53 seconds before getting reset. If you do not use interval record mode or time-lapse and record conventionally at 30fps, each group of the 16 video frames the camera records while slow shutter is open will contain the same image. So with a conventional video recording the image only refreshes once every 16 frames. If however you were to use the camera interval record function and set the interval to 1 frame, each recorded frame will have a shutter opening of just over half a second. This might be a nice way to add some motion blur to your timelapse to make it flow better. When the shutter is set to off, the cameras shutter is open for the full duration of each frame. So shutter off at 25fps means a 1/25th shutter or 360 degree shutter. If you use S&Q motion and shoot at 1fps and set the shutter to "off" the shutter will be open for the entire duration of each frame, in this case 1 second.

Hi, I'm Alister a cinematographer and film maker based in the UK. I don't and have never been a Sony employee, but I do have a very good relationship with Sony that goes back many years. I've been working in film and TV all my life. I studied photography at school and then went on to study electronics to degree level. A lucky break in the late 1980's allowed me to combine photography, electronics and my passion for rally cars when I landed a job designing some of the first broadcast quality camera systems used to film inside race and rally cars. For several years I travelled the world covering the world rally championship and many other motor racing series. Not just doing in car cameras but also track side filming. After this I became freelance and did many different things including TV news, corporate video, more motorsports as well as filming, editing and producing programmes about windsurfing and snowboarding. Eventually I got fed up with all the travelling, so got a job working as an editor in one of the UK's top post production houses. It was here that I learnt how to colour grade while still doing the occaisional shoot. Then another lucky break came when I got a comission from National Geograhic to produce a programme about storm chasing. The programme was called "Twister Tours" and was a hit that led to several spin offs and a lasting desire to shoot more of natures natural extremes. Around this time I also set up my first website, tvcameraman.info. A place to share hints and tips for camera users. In recent years I have filmed all kinds of things including drama, short films, 3D films, TV adverts, hurricanes and volcanoes. At the same time I still run my own website xdcam-user.com and now write about and hold workshops on many aspects of film and TV production. I never intended to get into the teaching or training, but it seems that the combination of my many years of experience and my technical background allows me to take often tricky subjects and explain them in a way that is easy to understand.

S-Cinetone is a rather clever gamma curve and the way you expose it alters the way your final images look. Exposing brighter doesn't just make the picture brighter, it also reduces the contrast, making the image look flatter. Exposing a little darker has the opposite effect increasing contrast. This happens because the S-cinetone curve has a very gentle highlight roll off that starts a bit lower than most other gamma curves and a toe at the lower end that increases contrast as you go down into the shadows. It is also worth considering that S-Cinetone was designed to produce a more film like look than normal Rec-709. If you look at most movies you will find that faces and skin tones are typically much darker than is normal for daytime TV. Where a face my be exposed around 70% for TV, in movies it is not at all uncommon for faces to be closer to 50%. In my opinion S-Cinetone tends to work best when faces are exposed around 60%. If you expose faces at 70% with S-Cinetone they will be in the highlight roll off area. This causes the skin tones to flatten, have reduced contrast and generally look less natural. Bring faces and skin tones down to 60% and the contrast will increase and the face will look much better. This lower brightness isn't under underexposure, S-Cinetones mid range is supposed to be a bit darker. So, it won't be noisy and you'll get the added benefit of a larger highlight range. When using zebras with S-Cinetone I set mine to 60% as this gives me a look that I like. I would never go as high as 70%, 65% would be my upper zebra limit. At the lower end I may consider going down as far as 57% for a drama or narrative piece. There is no "one-fits-all" exposure for S-Cinetone so you might need to experiment a little to see what works for you. Perhaps for TV news you might go for 65% as this won't look dark against regular 709 cameras shooting at 70%. If you are shooting a drama using S-Cinetone and feeling brave you could go as low as perhaps 55-57%.

If you have never used CineEI before it can be difficult to get your head around it. So for my first post here I thought I would share some things that may make it easier to understand this mode. It certainly took me a little while to figure it out the very first time I came across it. If you are someone that has shot with conventional video cameras all your life, then using CineEI is going to be very alien and you do need to think very differently. You need to discard most of what you have learnt about video cameras and instead think "film camera". For me the lightbulb moment came when someone told me to pretend the camera is a film camera that only has one film stock. CineEI stands for Cine Exposure Index. EI or exposure index is used by just about every digital cinema camera. It is used because a video sensor will only perform at its very best when operated at or very close to its native sensitivity and when the full output range of the sensor is precicesly matched to the recording range. Adding gain to the sensors output will reduce the dynamic range that can be recorded. So, when shooting S-Log3 or raw, to really get the best from the camera you must record at the cameras native (or "base") sensitivity. When shooting using Exposure Index you will come across two very similar sounding things, but they are very different from each other. One is ISO and the other is EI. Both have a numerical value and sometimes that value will be the same for both. But each has a different meaning and they should not be muddled, each has a specific meaning and I find it helps me to understand the CineEI mode if I think of each as the different things they are. ISO = The sensitivity of the sensor - often called the "base" ISO. EI = The sensitivity of the LUT that you are using. When the two match, when the LUT is correctly exposed then the sensor will be exposed at a normal level. When the EI is higher than the ISO the LUT becomes brighter than normal. If, looking at the LUT image, when you see a brighter image the normal response is to close the aperture until the LUT looks correct again. This will then result in less light falling on the sensor and a darker recording. A darker recording will be more noisy, have fewer shadow textures but also have a larger highlight range and this might help with a scene with a lot of very bright highlights. An EI that is lower than the ISO will make the LUT darker. When you see that darker LUT image in the viewfinder you will open the aperture to make the image look correct. This puts more light on to the sensor and this will increase the amount of shadow information but decrease the highlight range. More light going on to the sensor also means the final image will have less noise. If you were to use a light meter you would set the light meter to the EI to achieve the same result. A lower EI will result in a brighter and generally less noisy exposure with increased shadow range but reduced highlight range and a higher EI will result in a darker and generally more noisy image with increased highlight range but reduced shadow range. So... something to think about - when shooting under low light don't raise the EI. It won't make the camera more sensitive and the brighter image might make you think that your exposure is fine, when in fact it might be excessivly noisy. Instead keep the EI at the same value as the base ISO. This way you will be better able to deteremine how under exposed you might be. If you can, use an EI a bit lower than the ISO as this will encourage you to find more light or put more light on the sensor which will help overcome any noise.