-
Posts
67 -
Joined
-
Last visited
alisterchapman's Achievements
Top Contributor (7/9)
-
FX3 Feature Request: 3:2 Open Gate recording
alisterchapman replied to govision's topic in Cine Gear
I suspect there are 2 issues: Can the sensors actually be read at 3:2 or 6:5 video frame rates without overheating or other issues and do the processing paths have sufficient bandwidth. Then the other possibly bigger issue is how do you record it? There is no codec in these cameras that can do anything other than 16:9 or 17:9 and the codec is a hardware chip that likely has very limited upgrade options. Hopefully this is something that will be included in the next generation. -
Sony rate the ND filters in most of their cameras using a fractional value such as 1/4, 1/16, 1/64 etc. These values represent the amount of light that can pass through the filter, so a 1/4 ND lets 1/4 of the light through. 1/4 is the equivalent to 2 stops ( 1 stop = half), 2 stops = 1/4, 3 stops = 1/8, 4 stops = 1/16, 5 stops = 1/32, 6 stops = 1/64, 7 stops = 1/128. These fractional values are actually quite easy to work with in conjunction with the cameras ISO rating. If you want to quickly figure out what ISO value to put into a light meter to discover the aperture/shutter needed when using the camera with the built in ND filters, simply take the cameras ISO rating and multiply it by the ND value. So, 800 ISO with 1/4 ND becomes 800 x 1/4 = 200 (or you can do the maths as 800 ÷ 4). Put 200 in the light meter and it will tell what aperture to use for your chosen shutter speed when shooting at 800 ISO with 1/4 ND. If you want to figure out how much ND to use to get an equivalent overall ISO rating (camera ISO and ND combined and added together) you take the ISO of the camera and divide by the ISO you want and this gives you a value “x” which is the fraction in 1/x. So, if you want 3200 ISO then take the base of 12800 and divide by 3200 which gives 4. If you set the camera to 12,800 ISO and the ND to 1/4 the sensitivity of the camera becomes in effect 3200 ISO.
- 1 reply
-
- 2
-
What’s the difference and which should I use? On the FX3 and FX30 the default is Quick Format, but you can also do a Full Format, on the FX6 you can select either from the menu. Full Format erases everything on the card and returns the card to a completely empty state. All footage is deleted from the card and it cannot be recovered later should you perform a Full Format by mistake. Because Full Format returns the card to a completely empty state removing any junk or other clutter it also ensures that the cards performance is maximised. You should do a Full Format periodically to clean up you media and restore any lost card read/write performance. Quick Format erases the file database on the card but it does not actually remove your video files. When you then start a new recording on the card the new recording will fill any empty space left on the card if there is any. If there is no empty space then the new file will overwrite the existing files on the card. In some cases if you have accidentally done a quick format you may be able to use data recovery software to rescue any files that have not already been overwritten. But file recovery is not guaranteed. As quick format does not clear all data from the card, over time the performance of the card may be degraded, so a Full Format should be performed periodically to ensure the best card performance.
-
Most of us are probably aware that Lithium batteries should be treated with great care to keep them safe. But one thing I wasn't fully aware of is the damage that can be done and the resulting safety risks associated with trying to charge a very cold lithium battery. When you charge a very cold lithium battery some of the metallic lithium in the battery gets plated onto the anode of the battery. You can't normally tell or see that this is happening. This will very slightly reduce the capacity of the battery but more importantly it greatly increases the risk of an explosion or fire. No matter how well the battery is made, if sufficient lithium ends up on the anode an impact shock or any high temperature usage can cause the lithium to ignite causing a battery fire. The batteries management and protection circuits cannot protect against this type of failure, although some manufactures do include circuits that will prevent a cold battery from accepting a charge, many do not. You must never try to recharge cold lithium batteries, they should always be allowed to warm up to room temperature before charging. Repeated charging at cold temperatures is dangerous and must be avoided.
-
- 1
-
Just be careful doing this not to exceed 2 amps as you can damage the cable. When you run 12v instead of 24v the maximum power than can be passed own the cable safely is halved. At 12V you have a maximum power draw of around 24W but at 24v you have 48W. It's better to put any voltage converters at the output from the Rialto rather than at the camera body end.
-
Quite possibly the issue was that 12v is a bit on the low side. The specs are 11v to 17v, but ideally the camera wants 13.8v to around 15v or 24v/28v. At 12v even the slightest voltage sag in any of the connecting cables will cause the voltage at the camera to drop too low. Given the high current draw some voltage sag is inevitable even with good quality cables. I regularly power Venice using a 13.8V 100w power supply without issue, but you must ensure the DC cables are of suitably high quality and actually capable of carrying at least 8 amps.
-
To make 3D LUT's a manageable size 3D LUT's don't have an adjustment for every possible input and output value, if they did they would be massive. So, 3D LUT's divide the image into ranges, typically 33x Red 33x Green 33x Blue. Each of the 33x segments has the same correction value, so within an image there will be 33 steps between each correction. These steps can often show up within the output image as banding or odd sudden colour/brightness shifts. To prevent these steps in post production additional calculations are used to smooth out the steps by interpolating between each of the ranges within the LUT, but this needs a lot of processing power to do well. Especially if you not only interpolate within each colour input channel but also 3 dimensionally across all 3 output colours. In a camera there often isn't sufficient processing power to perform these interpolation processes so banding is seen on the cameras output. The .art system was designed as an alternative to 3D LUTs for the original Venice camera so that you get a transformation function similar to a 3D LUT but without introducing the commonly seen step/banding artefacts that were a result of the limited amount of interpolation available in the original Venice camera. In Venice 2 the LUT processing capabilities have been greatly improved so that normal 3D LUTs now have much better interpolation and banding is rare.
-
Don't forget as well that camera like the FX9/FS7 and many others have a special colour matrix called "FL-Light" that can be used under fluorescent lighting to eliminate the green bias without affecting the tint or hue.
-
Will we ever see full frame slowmo on the FX9....?
alisterchapman replied to Danny_Schmidt's topic in Cine Gear
I think it's pretty safe to say no. The 6K sensor readout in the FX9 clearly can't go above 30fps which is why Sony had to include the 5K scan mode and the 2K full frame scan modes (to read 6K requires double the bandwidth of 4K). I'm quite certain that if it was possible it would have been done early on in the cameras life. It appears to be a sensor issue and the penalty you pay for having a 6K sensor that allows FF + s35 with at least 4K of pixels. If you were to read only 4K of pixels at FF there would be some nasty aliasing and moire artefacts, to have a decent looking image you need to read all 6K of pixels. It's also worth observing that the FX9's 4K 120fps raw mode is limited to s35 scan and is only 10 bit. It's also interesting to look at the FX30 which is another lower cost Sony camera with a 6K sensor and observe that to shoot 4K 120fps with the FX30 you have to crop by 1.5x so that only 4K of pixels are read out. -
Every now and then I'll come across someone struggling to grade a shot. Often the problem is a result of a having a light source that has an incomplete or very narrow spectrum. A lot of cheap LED lights as well as many types of discharge lights such as sodium street lights or neon lights only output light at certain specific wavelength's or only encompass an extremely narrow part of the light spectrum. These kinds of lights will result in images that will be strongly coloured and next to impossible to colour grade because the narrow bandwidth of the light means the footage only contains a single colour or extremely limited range of colours. You will be able to change that single colour to a new colour, but because there is only one colour everything else in the image will change by a similar amount. It will be next to impossible to pull out subtle skin tone hues from a face lit by a sodium street light as every part of the face will be the same hue due to the single wavelenght of the light source. Unfortunately there is no simple fix for this. So, it is something that needs to be considered when shooting under discharge lights. If shooting using S-Log this is one of those times where monitoring and only seeing the S-Log image can be a little dangerous as it may not be obvious that your colour palette is extremely restricted. Monitor via a LUT and it will normally be very obvious. Another issue I see from time to time is where one of these low quality narrow spectrum lights is spilling into a scene and causing very odd looking, highly saturated highlights. Our eyes will adapt to these lights, often a pool of light from a narrow bandwidth light won't look all that bright to us. But to an electronic sensor that doesn't adapt the same way as our eyes do it might appear very bright and in some situations may cause one of the sensors colour channels to clip resulting is some very odd looking highlights. This is commonly seen where blue LED up-lighters are used to light up a wall. To our eyes the blue might not appear super bright, but to a video camera that intense but narrow wavelength blue light will cause the blue channel to clip. In a lot of cases there isn't much you can do about it (other than turning down the light), so it's something to watch out for. If you have a Sony broadcast camera many have a matrix setting called "Adaptive Matrix" that can be turned on to specifically to deal with this. But if shooting S-Log you won't be able to take advantage of this, so again, look carefully at you images while shooting, preferably via a LUT if you suspect that there may be narrow bandwidth light in your scene.
-
I love the FX30, it's a great little camera and it produces a lovely image. But one thing a few have noticed is that if you shoot at 100fps or 120fps it can be a little more noisy than it is at other frame rates. When shooting 4K/UHD up to 60fps the FX30 downsamples from 6K of pixels to a 4K recording. This oversampling brings a nice noise reduction, the equivalent to almost 1 stop of exposure (around 4-6dB). When you shoot at 100/120fps the camera reads 4K of pixels, there is no downsampling, so the images will be noticeably more noisy. This is just a limitation of how this camera works, my guess is that it doesn't have enough processing power to convert 6K to 4K at 120fps or perhaps the sensor can't be read at 6K at 120fps. The plus side is that the normal 4K recordings downsampled from 6K really are very good indeed and packed full of detail and texture. Plus, if shooting at 120fps and you choose to use the near equivalent of a 180 degree shutter - 1/250th, you will need 5 times more light to get the same exposure compared to 24fps and 1/48 (180 degrees). So to get the same exposure you need to open up the lens by 5 stops or increase the light level by 5 stops. Any less than this and you will be under exposed and that will make your footage look more noisy. When I shoot at 120fps I will often use 1/125 to gain back 1 stop, then expose nice and bright to help eliminate the extra noise.
-
Syncing 2 cameras: Sony FX6 with Canon C300mk2
alisterchapman replied to Jason12345's topic in Cinematography
I'll try to simplify this as much as possible as there are some different concepts that are often miss-understood. SYNC: Is when 2 devices are connected such that they run or operate at exactly the same rate, at the same time. On a video camera a reference signal is fed to a cameras Genlock port and then the "genlocked" camera runs at exactly the same frame rate as and locked to the reference signal. But as Doug has already commented, the FX6 cannot be Genlocked, so there is no way to regulate the frame rate to precisely match that of another camera. As a result when you have multiple cameras there is no guarantee that even if both cameras are started at exactly the same moment that over a period of more than a few minutes that both cameras will record exactly the same number of frames. TIMECODE: Is a unique, sequential, numerical value given to each video frame by a video camera. Each frame in a sequence must have a timecode value that is 1 frame greater than the previous frame. If you record 1000 frames, the time code count must increase by 1000 in the clip. EXTERNAL TIMECODE: It is not a sync signal. It is the output of the timecode clock of another device (which could be another camera) that is fed to the timecode input of a camera and then the timecode clock in the receiving camera will follow the external time code value. But this external timecode clock may be counting at a very slightly different rate to the number of frames the camera is actually recording. So to ensure every frame always has a unique sequential number what the camera does is the moment you press the record button the camera takes the last TC clock number that was seen on the external input and from that moment on counts the frames actually recorded and adds +1 to each frame, so each frame has a unique number that is 1 more than the frame before, regardless of the external TC number. Where you sometimes (often?) get an issue is with long takes. The sync clock in most cameras will drift in frequency very slightly as the temperature of the camera changes or due to other factors. If during the record period the external TC clock counts to 1005, but the camera only records 1000 frames because it is running fractionally slower than the external clock source, there will be a 5 frame difference between the external TC and the TC recorded with the clip. Once you stop recording the cameras TC clock will re-sync with the external TC clock so the error becomes zero again. So, the first frame of every clip will match the external TC, but later in the clip the external TC value and clip TC value may be very slightly out. Generally this is only rarely an issue with clips under 10 minutes. But when trying to shoot long takes such as performances the drift can become significant. If the cameras are genlocked, because the frame rates of all cameras will be identical, there will not be any timecode drift. So, when shooting with cameras that can't be genlocked, but do accept external timecode it is a good idea to stop recording from time to time to allow the cameras timecode clock to re-sync with the external TC. If using 1 camera with a sound recorder, if you can feed the TC from the camera to the audio recorder because audio recorders don't have frames, they just place the external TC alongside the audio so going from camera to audio recorder there isn't a sync issue. -
Sony have now released new firmware updates for both the FX3 and FX30. The FX3 now goes to firmware version 2.02 and the FX30 to firmware version 1.02 These are mainly stability releases that fix some minor bugs, but if you have an FX3 on the original version 1 firmware then this version adds the CineEI mode and LUTs. It is a major update that is well worth having. Before attempting to update the camera you should insert a fully charged battery The FX3 is updated via a computer application. While there is a Mac application there can be some hoops to jump through to get it to work, so I would urge you to find a windows PC to do the update, it is far simpler and far more likely to be successful. The good news is that once you have updated to version 2.02 future updates can be done by uploading the update file to an SD card and initiating the update from the camera like the FX30. The FX30 is updated by placing the downloaded BODYDATA.DAT file on to an SD card that was previously formatted in the camera. Then place the card in the camera and go down to the SETUP – SETUP OPTION – VERSION page of the menu. Here you should see the cameras current firmware version plus a “SOFTWARE UPDATE” button. Press (select) the software update button. On the next page it will say “Update ?” and show the old firmware version and the new firmware version. Then just below this is a box where it says “Please follow these precautions until the very end”. What isn’t clear at this step is that you need to scroll down inside that box and read the full list of precautions before the camera will allow you to do the update. If you don’t scroll down and just press the “Execute” button you get a large popup telling you to “Follow the precautions to the very end” and pressing “OK” simply takes you will go back to the previous page. So do make sure you scroll down through the full list of precautions before you press execute. Once the update starts the screen will go blank, the only clue that the update is happening will be the slow flash of the media LED on the back of the camera. The update takes about 10 minutes to complete and the camera will reboot when it’s done.
-
- 3
-
It's a difficult one. While there is no doubt that 48 or 60 fps will deliver smoother motion it will also appear more "video" like (not that that's necessarily a bad thing). For decades feature films and IMAX films shot at 24p have been shown on huge screens and judder hasn't typically been an issue and many consider it part of the "movie experience" as some people have a different emotional reaction to the lower frame rates. 48fps in the Hobbit movies did not get a good reception, Avatar seems to be more of a mixed bag with some people loving it and others not. Personally, for narrative I like 24fps, there is something about it that separates it from the real world, I also like reading books and allowing my imagination to fill in the gaps. For "experience" films I think higher frame rates are better. What is interesting is that there is a lot of discussion at the moment about shooting narrative at 30p. It has significantly less judder and stutter than 24p but isn't as smooth as 48 or 60p. There is some reluctance to 30p simply because many confuse this with 30fps interlace, which has motion closer to 60p. But a lot of people are wondering if 30p may be a serious option for digital movie making as it offers a nice middle ground and fits well with computer displays. If it were me and with the technologies available right now, I would be thinking about 30p or 60p for large screen presentation, in part because computer servers etc tend to output at 60Hz and 24fps or 48fps will stutter more than it should on a 60Hz system. But at the same time 24p on a big screen doesn't scare me.
-
Something that often gets asked is - how should I clean my camera? My process is this: Start with a good quality soft paint brush and gently brush off any dirt or dust from the outside of the camera. DO NOT use the paint brush on any glass ports or the sensor, just the camera body, handles and other accessories. A small artists brush can be used to get into all the little gaps and crevices, but don't poke it into any connectors as you could damage the pins. Most of the time this is all you need to do. If the camera is extremely dusty then I might use a small vacuum cleaner with a brush attachment, but this needs to be done with care as excessive suction could damage the fans inside the camera. The next step (when needed) is to use a soft polishing cloth to wipe down the camera. If it's very dirty then I will use a solution made up with 1 cup of warm water with 1 or 2 drops of dish soap (you really do only need 1 or 2 drops). This is very effective at removing dirt and grease but shouldn't attack the paint or damage the plastic. Don't soak the camera, just dampen the cloth and gently wipe over the camera. If there is dirt in a connector or similar I will use a handheld puffer to try to blow it out. I do not like canned air, it can make things worse as it is quite powerful and can blow dirt and debris deeper into the camera. What about cleaning the optical port - the piece of glass in front of the ND filters and sensor on full size cameras like the FX6/FX9/FS7 etc? This piece of glass is coated with an anti-reflective coating so needs to be treated very gently - don't use strong solvents as they can strip the coating. To clean this I start with a handheld puffer (get one where the nozzle is part of the bulb to ensure the nozzle doesn't fly off onto the glass). I use the puffer to blow off any dirt or dust. Again in most cases this is all that is needed and I always start with this as it should remove anything that could scratch the glass if you do need to progress to the next steps. If the puffer isn't enough then I use a the brush end of a "Lens Pen". This is a very soft brush designed for cleaning delicate optics, they are available from most camera stores. Use the brush to very gently brush off any dirt. If that still isn't enough then you can use the other end of the lens pen, which is normally a flat swab to wipe the glass port. By be very, very gentle. Start at the center and work your way towards the edge in a very slow, light circular motion. You should now have a nice clean optical port. But if someone has put greasy fingers on the port and you are struggling to get it clean then as a last resort I would use 1 drop of dish soap in 1 cup of distilled water and use a microfibre lens cloth dipped in the solution to gently wipe the port. Lens pens are cheap, you should replace it regularly, especially if it gets dirty or has been used on an oily or greasy surface. For cameras that have an exposed sensor, I will try to avoid cleaning the sensor at all costs. A puffer can be used to blow off dust, but you don't want to ever touch the sensor unless you absolutely have to. To clean the sensor buy a good quality sensor cleaning kit, which will normally consist of special swabs which are gently dry wiped across the face of the sensor. Never rub, never scrub, follow the instructions that come with the swabs. Lenses: Again, puffer first to blow off dust and dirt. Then the brush end of a lens pen, for more stubborn dirt the flat end of a lens pen. For a lens cloth it depends on what I am shooting. For most applications a microfibre lens cloth will work well. But if you are shooting in the rain or a very damp location a soft Chamois Leather (the very soft leather used to dry a car after washing) is good for removing rain as most conventional lens cloths just tend to smear it all over the lens. Finally: Keep your lens cloths in sealed bags to keep them clean and free of grit and dirt. Just 1 spec of hard grit on a lens cloth can ruin a lens if it gets wiped across the glass. You should be very careful to keep your cleaning gear clean. And also, a few small specs of dust on the front of a lens or filter rarely cause an issue, don't overdo the cleaning as any time you wipe a lens there is a risk of scratching it and a scratch will show up a lot more than a few specs of dust.