Jump to content

Nicolas POISSON

Basic Member
  • Posts

    82
  • Joined

  • Last visited

Profile Information

  • Occupation
    Other
  • Location
    Paris

Recent Profile Visitors

The recent visitors block is disabled and is not being shown to other users.

  1. In a basic system, video cameras have a gamma of 0.45 (OETF: Opto-Electronic Transfer Function) to compensate the inherent gamma 2.2 of CRT (EOTF: Electro-Optical...). The whole chain has a gamma of 0.45*2.2 = 1 (OOTF), which means the brightness output by the CRT is simply proportional to the scene brightness. The problem is the rather limited DR, which is often described as a clue of "video-ish look". Even in the analogue era during the 90's, some pro video cameras had highlight compression, which means the OOTF was not linear. Consumer cameras, however, mostly used the gamma 0.45 curve. A Sony TRV900 - a popular consumer DV camera - could deliver like 6 stops of DR. Things started to change in the late 2000s, especially when photo cameras entered the video market (the famous Canon 5DmkII was released in 2008, IRC). Both cameras and displays improved, but backward compatibility was important. The solution was to keep a 0.45 gamma curve for levels up to the mid-tones, and then use highlight roll-off ( =compression). What was formerly a pro feature became very common. At the beginning, this was presented as an additional "cine profile". But nowadays, the limited "pure gamma 0.45" must have disappeared from most cameras. All built-in looks do have a more or less pronounced roll-off. No video is faithful to the scene luminance since the last 15 years.
  2. DR is tied to bit depth only if the coding is linear. But this is not the case at all for most displays. Common colour spaces, such as Rec709, sRGB, or cinema projection, all use a gamma curves to define the relationship between luma (0-100% form black to white) and real world brightness. Codes in the lower end are much closer to one another. Although the origin was related to the way Cathode Ray Tubes work, this also more or less follows the capacity of the human vision to distinguish light levels. This is a very rare case of a technical non-linearity being favourable, even decades later while CRT have disappeared. LCDs still mimic the response of CRT, and this is not only for backward compatibility: even if we could completely forget about CRT, it would still be a good idea to use a non-linear curve. We might use a Log curve instead of gamma curve, but that would not be that much different. Display DR depends on the definition you choose. To my knowledge, there are at least three: - display contrast: this is the ratio of peak white to "black" level. LCD displays (whether TN, IPS or VA) are not pitch black when luma is 0. A consumer IPS panel is typically around 1200:1 to 1500:1. Some other technologies, as OLED, have true blacks and can be considered to have an infinite contrast ratio. - ratio of the peak brightness to the tiniest difference between two adjacent levels in the whole brightness scale. Usually this tiniest difference is between the lowest and second lowest levels. The value depends on the gamma curve. In Rec709 with a theoretical 2.4 pure gamma curve on a display with infinite contrast, the order of magnitude is a 1:1000 ratio (10 stops). - ratio of the peak brightness to the threshold above which adjacent levels are close enough so that the human eye cannot distinguish them. This is the criteria used in the BBC white paper WHP309 about HLG. The order of magnitude is 1:40 ratio (5.27 stops). There are luma codes below, but they translate into differences in brightness that the eye is able to see. There could be an additional definition: real world DR. For example, if the light output by house lamps falls on your screen, it will hide the darkest details. This cannot be determined as a specification since it depends on the viewing conditions. But this is of up-most importance and this explains why, when producing for consumer TVs, web or (worst case) smart-phones, one might prefer to stay away from the darkest levels since there is a high risk that the viewers will miss them. What happens if the camera can record more stops than the display ? This is up to the production. It can be simply clipped. But most of the time, this will be "compressed": differences in high brightness levels will be reduced compared to what they would be if the curve was a pure gamma. This is the common "shoulder" part of so many tone mapping curves.
  3. Answering to myself. It seems that the default calibration of the Ninja V targets a pure Gamma 2.4 law, ignoring the non-zero black level of the display. Here is a comparison of the Ninja V in "Native" mode and the pure gamma 2.4 law: Reading right to left: the measured curve (plain yellow) matches the reference (dashed white) until the display reaches its inherent black level. Then the lower 5-6% luma levels cannot be distinguished. Now here is the same "native" mode compared to the expected BT.1886 curve considering the black level of the device: Using a LUT, It is of course possible to calibrate the device to follow BT.1886: Lower-but-not-lowest levels are raised, so the image appears a bit washed out. For a medium to high key image, it moves away from what a user watching the same image on a high end OLED panel would see. However, the Ninja V has a 1300:1 contrast ratio, which is very common value for an IPS panel. Calibrated to BT.1886, the image is much closer to what a user watching an decent IPS display would expect. Note for those interested: to calibrate the Ninja V properly, you have to activate an "Identity LUT" first, otherwise the calibration will be off. The Ninja V deactivates some internal calibration when using LUTs, and once you calibrate it, then, well, you will be using LUTs.
  4. I guess many modern still lenses would fall in that category. It is common now to have them super sharp, with great resistance to flare and high micro-contrast. Are you searching lenses for still photography ?
  5. Can you have access to a "Parade" measurement, for example on an external monitor ? In which case you could check the red channel. By the way, because of the Bayer CFA, deep red lighting will also strongly decrease the perceived sharpness. The same would happen with deep blue lighting and in a lesser extent deep green lighting. You may be interested by this thread:
  6. I think there may be a question of wording here. In physics, the concept of "diffuse reflection" means a beam of light made of parallel rays will be scattered in every direction. However my understanding is that in the film-making world, "diffuse light" refers to a source that is both large and with a rather wide beam. That may not be scientifically correct, but I think this is the way it is commonly understood. Although it is possible to create a big source emitting strictly perpendicularly to its surface, such a device will rarely be encountered in practice. LED panels emit with a certain angle. An object close to these panels will have soft shadows, because it still receives light from a great surface. Even panels with very directive LEDs will create rather soft shadows at proximity. The angle of each individual LED is wide enough to blend with a significant amount of neighbours. Soft-boxes with honeycombs will act the same. You can find exceptions, such as this DIY parabolic reflector that is both huge AND creates hard shadows.
  7. A diffuse light is not the same as, nor is contrary to, a directional light. Soft shadows are the result of a light source being big compared to the distance to the object. This is just basic geometry. It does not matter if it is a medium size source close to the object or a bigger source a bit farther away. A 60cm square source at 1m will give the same shadows as a 1.2m square source at 2m, or a 1.8m square source at 3m. Diffusion very close to the source will not make shadows softer, it will spread the light. Adding diffusion on a narrow beam source will light a greater area. Some areas that were not in the beam will now receive light. The trade-off is that areas that were previously in the intense beam now receive less light. In theory, there is no benefit adding diffusion close to a point source that emits at 360°: the beam is already as wide as it can be. There is also no strong benefit adding diffusion close to a source that is already huge and with a wide angle (like a LED flat panel that emits at 120°, a kinoflow, or a COB LED with a softbox). Adding diffusion right on a softbox because you find the shadows still too hard does not help. You have to bring that softbox closer, or use a bigger softbox, or add diffusion at some distance - provided the softbox enlighten the whole surface of this additional diffusion. Well, in practice... one often use additional diffusion close to flat panels with wide beam because they are composed of multiple small sources that create cross-hatched shadows. And of course one can use diffusion close to flat panels that have a narrow beam just to make it wider. Also, adding diffusion close to a narrow beam source might spread the light to other reflecting surfaces like walls or ceiling, all of these acting like a fill light, reducing the contrast of shadows. And of course, if the diffusion is light, the beam at the output can be a mix of diffuse and directional light.
  8. Hi, I am joining a short movie project with no budget and relying on hobbyists. A rather bright scene has already been shot with a Sony FX6 in Cine EI mode (Slog3). The team used mainly natural light. It was shot at 800 ISO, then they switched to 12800 + some ND at the end of the day when light started to miss. I am to work on a bar scene, which will be inter-cut with the first already shot scene, although they are totally different visually. The bar will have a moody reddish atmosphere, and I expect most of the objects to fall in the lower IREs. The camera will be a Sony A7sIII in Slog3. We bring additional lighting on location but not that much. Given that the A7 is supposed to have the same sensor as the FX6 but has analogue gain, what would be the best strategy concerning ISO? - stick to the native ISOs as in the first scene (640 or 12800 on the A7) ? - or set the ISO to expose the bar scene properly? My understanding is that the consistency in noise and dynamic range is already killed by the use of two different native ISOs for the first scene. I would choose one ISO for the bar scene that gives me enough exposure, not taking into account what was used before. Is this correct ?
  9. I do not know the BMPCC 6K, but on many cameras the intensity of the focus peaking can be adjusted. When set to "strong" or "high" or anything like this, the peaking is easy to spot, but also very permissive. It displays as "in focus" objects that aren't.
  10. According to this Depth-of-field (DoF) calculator... https://www.dofmaster.com/dofjs.html ...If you are using a full-frame camera, the total DoF in your case should be 1.29m, ranging from 2m50 to 3m80. It should be enough to keep a sharp image on the whole 86cm distance of your slider. However, you have to set the focus when your slider is more or less at the middle of its travel, to use both sides of the DoF. All these calculations are theoretical and rely on the "circle of confusion". They use a default value, which might not be adequate, but its impossible for the common user to decide what better value to use. Most still lenses with auto-focus have other drawbacks: - manual focusing is indeed "fly by wire" type: the focus ring does not move anything, it just tells the camera body what was the rotation, and the body tells a motor inside the lens to move accordingly. This is not very precise - moreover, the relationship between rotation and focus change is usually not linear and depends on the speed of rotation. For a same angle, the fastest you turn the ring, the greater the focus change. This is of great help for stills, as it makes an intuitive "coarse / fine" focus search. But you cannot have a very repeatable action when pulling focus in video. Some camera bodies allow to change the relationship to "linear", but the very common Panasonic GH5 and Sony A7III don't.
  11. I would not pay too much attention to the exact direction of the light. As long as it comes from the left side, I think nobody will notice that you changed something between the two shots. Sometimes even within the frame, a practical lamp is put to justify some lighting, but if you analyse you can see that it is inconsistent. Two examples from "Birdman". In this shot, it's hard to explain how the light coming from the street could reach the interior of the hair and the ear of the girl (should be darken by her own hair). The bright "St. Ja..." sign also helps selling the lighting while it is actually behind the actors. Here again the neon is behind the actors, and should light their opposite side. One could argue there might be other ceiling lights, but the wall on the top left simply becomes darker as if the neon within the frame was the only light source. Moreover, in the real world, who hangs this type of lighting so close to a wall anyway ? In both cases, nothing looks fake at first sight.
  12. You can light from both sides of the stage, together with honeycombs on the soft-boxes. The more on the sides, the less it spills on the background. But going too far you end up with the speaker's face dark in the centre, which is not pleasing (or very theatrical). So you need to keep the sources slightly on the front. Also, if the speaker is not at the centre of the stage, he will quickly receive much more light from the nearest side. Greater distance helps. Having the sources high above and pointing down also helps, because the shadow will be lower and less noticeable. And of course, the farther the speaker from the background, the better.
  13. I did not investigate every single camera released recently, but I guess none of them has a strict Rec709 profile nowadays, as it would limit the DR to slightly more than 5 stops (if I understand correctly). I guess all current so-called rec709 profiles include some kind of "S-shape" tone mapping, or at least some highlight roll-off, because nobody would accept only 5 stops of DR, except for a specific project wanting to mimic vintage found footage style. Thus, the diagram in the first post exaggerates the benefit of log curves over SDR: In a 'IRE vs. stop" scale it would appear less impressive, and the "SDR" used as a reference is a norm nobody follows any longer. I went through the Shogun manual, and I am a bit confused with that chapter. It seems that this diagram has nothing to do with the way the shogun handles log signals or is able to turn into an HDR display. It is simply a representation of various vendors log curves, nothing more.
  14. Imagine a mate white board that reflects a certain amount of the light it receives, sending it to your camera. You define this quantity to be 100%. You set the exposure of your SDR camera with pure Rec709 profile so that this white board reaches 100% IRE. Now changing the camera profile, it may be able to capture objects sending 4 times more light (400%), or 15 times more light (1500%) without clipping. In the real world, you may simulate this keeping the same board and just increasing the light source. This is a bit like estimating the variation of the cost of life through the years: you take the price of various products in 1990 and compute the total amount to be $1200. You define this as your 100% reference. Ten years later, the same products cost $1800. This represents 150% compared to 1990. The absolute values of $1200 and $1800 do not matter, only the ratio is of interest.
  15. This is the bandwidth of the SDI connection, expressed in gigabit per second. This limits the characteristics of the video signal you wish to transfer. Many factors are at play: resolution frame-rate bit depth chroma sub-sampling Typically, 3G-SDI will be used for FHD and 2K signals, while 12G will be used for UHD and 4K.
×
×
  • Create New...