Aapo Lettinen

I had a 70(me) / 30(rental house) contract with a local rental house for some lenses, I had my own insurance and they did all the work so a 60/40 split could be OK if they have their own insurance and will pay the expenses if the equipment is damaged. renting stuff out is good idea if you have two of them or are using them only on certain time periods. for example lenses and some pro cameras are quite difficult to replace/repair quickly if they are damaged so your own shoots are also screwed if someone accidentally messes up your gear

basic answer it that the digital cameras still have too much noise to be able to capture very low signal values at usable quality. Digital camera's usable dynamic range falls between the clipping point and the base noise ('noise floor') . The clipping point is absolute so if you want more dynamic range you have to get more stops from the lowest side of the range, you basically dig those extra stops from the darkest tones which can just barely be extracted from the base noise. Film's dynamic range is quite a lot based on two emulsion technique, you have one sensitive layer and one insensitive layer per primary color. It can be seen as a kind of HDR technique because it combines two different exposures together inside the film emulsion. I think the newer Varicam has a little bit of similar technique where two different sensitivity photo sites are combined to one pixel and thus you can also switch the camera to a higher "working ISO" by switching the signal processing to use more extensively the larger and more sensitive photo site. Film negative's dynamic range is opposite compared to video: it has quite absolute black point but it has no absolute white point, the film just gots denser and denser until your scanner can't see through it and the scanner outputs only its own signal noise.

the basic problem with video cameras is that the saturation rises very high just before clipping point and when you get clipping, it is very rough transition from very saturated colour to plain white. That's what is usually said to be "THE UGLY VIDEOISH CLIPPING" and can be at least somewhat corrected in colour science if the manufacturer decides to do so, like Arri did with the Alexa. on current Sonys you can have varying amount of that clipping, much more with the lower end models like FS7 because it has slightly different colour science than for example F5 even if the sensor is claimed to be about the same. you can somewhat try to correct the saturation issue in colour correction by lowering saturation curves over the middle gray , it is nowhere near perfect solution but at least helps a little #auto correcting is making lots of typos, sorry :lol:

Oblivion used Epics quite a lot for b-cam/plates work. For example the rear projection plates were shot with 3 epics. Choosing cameras for a shoot is like choosing lenses, you don't shoot anything including handheld and gimbal shots with the Ang 24-290 just because it is "a great and versatile lens" :blink: I really like the Dragon sensor image, a big step up from the MX family. -------------- With video cameras, you don't just choose the right sensor and format for you, you have to take the whole package which includes form factor, weight, how in can stand difficult shooting conditions like flying sand, salt spray, hot/cold environment/humidity; power options, ease of use, frame rates, compression (codecs, ratios), signal processing/color science, possible glitches like rolling shutter & moire, workflow compatibility, light sensitivity... (actually NOT the ASA/ISO the manufacturer claims the sensor to be rated on. With sensor data you have noise floor and clipping point, and thas it. Everything in between of them is colour science/lut/color correction issues and based quite much on personal opinions and testing the most acceptable balance between shadow and highlight latitude. You can't rate an absolute ISO of a video camera. All you can do is decide where you want the middle gray to be and adjust that with gamma/color science or correct in post. It's exactly like the Exposure Index rating or a film stock, the film manufacturer decides based on testing to recommend the film stock to be shot at certain EI to get results which they liked most. Usually their EI rating is roughly one stop under the absolute ISO of the film stock and rated so that you get about from 3 to 4.5 stops of shadow detail before absolute black when the film is normally processed. Video camera's light sensivity is mostly based on how low/high the noise floor is. That's why you can have five video cameras which are all rated 800 ISO and all of them have different light sensitivity because some of them have lower noise floor than others and have thus more stops of latitude below middle gray set by signal processing) ----------- You can't choose a camera based on only what the output image looks like and forget all the rest :blink:

people shoot with REDs and Sonys because they weight less and are smaller than Alexas. They are much more owner operator cameras because of the more complicated menus/user interface and less sturdy mechanics. You can't, for example, use an full sized Alexa with octacopter or gimbal (unless it's a really big one) and the camera mass and lenght may complicate operating in some situations. The Alexa Mini will be nice for these uses when it is common enough in the market. people may have missed the main point of the Raven camera: it is VERY lightweight for a cinema camera so it would be very nice in gimbal and copter use even though it surely has some glitches (I suspect cooling problems might be present but can't know before the actual camera is out. and nowadays when raw recording is not that special anymore it would be nice to have lower compression ratios available)

It's a bit different to watch a movie trailer than a simulated camera test. In a trailer there is much more different lighting conditions and shooting situations and you can better see the camera's performance in production environment. I usually watch the trailers from itunes using calibrated display and best playback quality. Comparing a single image or two might not be enough but in a trailer you can seen 30-50 shots and then you will usually see very clearly what is going on. I've been working a year and a half as a dit/assistant editor in a nature documentary which is shooting on lots of different cameras, the main camera being F5 in raw mode so I might be overly sensitive to Sony image, and also Bmpc, bmcc, c300, Epic mx, gh4, and so on... Try to watch this movie's trailer and spot the most obvious Red One shot (hint:two boys on schoolyard.look at the skintones) http://www.imdb.com/title/tt1340107/?ref_=nm_flmg_dr_5

actually one usually CAN tell quite reliably which camera is which, especially if DP's and colourists use the same kind of style for certain camera models but also when they are trying to hide the differences. It does not necessarily mean that the other camera's image is better or worse than the other but the differences are impossible to hide completely. I, for example, watch a lot of movie trailers and can usually tell which camera a movie is mostly shot on, especially when there is lots of skin tones which try to look natural. For example, Tomorrowland trailer looked like Alexa more than a RED but there was something on the colours and skin tones which was more screaming than usual Alexa image (like halfway between Fuji and Kodak in terms of colour response) so I was sure it was something different from those two and thus most likely a Sony camera. It is quite interesting that like in the film era, Japanese manufacturers are more interested in bright lively colours and western manufacturers are more interested in contrast/latitude and forget the colours completely (except Arri which is aiming for subtle natural colour response like Kodak film and is mostly interested in drama uses (skin tone response) ) I sometimes confuse low end Alexa stuff with the Red One MX, you can usually spot REDs by their lifeless skin tones and highlight handling but sometimes low budget Alexa stuff looks quite a lot like that with horrible colour correction and blown out highlights :P "low cost Sony camera" =at least FS7 and, if somebody still shoots on them, the F3

mostly the white balancing difficulties and not very good basic LUTs. most Sony cameras seem to have this ugly video-ish highlight/saturation handling which you have to try to tune out if wanting pretty pictures. I have seen lots of wonderful material shot with the F5 in both xavc and raw mode so you can definitely get great results out of them if you have the time for all the testing and adjustments. out of the box they are like "in your face RED, look how much saturation we can bake to the image before it breaks up" :lol:

2.8K sensor on Alexa was decided to be the sweet spot at the time, Arri specifically did not want to add more resolution at the cost of light sensitivity and dynamic range. I think RED's images, at least from the MX and Dragon sensors, look reasonably good. It's the backups, post workflow and reliability where these cameras fall short compared to higher end brands like Arri. RED ideology is more towards indie productions and special occasions like vfx work. For normal productions, for example drama/comedy films, it is usually more practical to take more reliable and sturdy camera which can stand the elements and has easier post workflow. Even Sony has done this a way better even though their Slog has some glitches and the cameras have sometimes reliability problems

of course does not work for arriglow markings but if the markings are directly on the ground glass it works fine :)

you can roughly test the gate/groundglass alignment in dark room by shining flashlight etc bright light to the viewfinder and projecting the image through the taking lens to a paper. Draw the markings you see, and then put some opaque material over the film gate without moving the camera and project similar image to the same paper (you can draw it with different color pencil so it clearly differs from the previous one) . This way you can instantly see alignment problems and which markings actually represent the gate corners. No need to wait for a film test to be developed :)

I think the mount flexibility matters much if you use both PL and Canon EF or Nikon cameras. If you have a e-mount or mft mount camera you should buy PL mount lenses and rod mounted mount adapter for your camera. e-mount and mft are so small mounts and have so much play/bend so much that they are real pain to use with follow focus if supported only by the camera mount. So I recommend buying either PL lenses or EF lenses depending on your use and use a very good quality adapter to adapt the lens to the smaller e-mount or mft. that way you can also change the lenses more quickly (don't have to attach support directly to the CP2 primes because of the strong adapter which is supported by the rods) and don't have to shim the lenses all the time because of the mount changes, just shim them once and then use correctly shimmed adapter with your camera which is perfectly adjusted for that particular camera body (camera mount's ffd:s can vary a little if the camera is using still lens mount (EF, e-mount, mft, etc) so it is not practical to shim all your lenses to fit that particular camera and have off marks with all the other camera bodies, if you just shim your adapter you can always use lenses which are shimmed according to mount standards not cameras

they are made for much bigger market and have more optical compromises, like varying max stops across the range. they breathe less than UPs and the mechanics are more modern. one thing to consider are Cine Xenars, I've heard lots of good about them. I don't know about their Xenon series which is near the price range of Compact Primes. I think the CP2's are very good owner operator lenses especially because of the interchangeable mount. They also should SOMEWHAT match to other zeiss lenses so you can rent the ones you don't need very often and only buy a basic set which you'll use all the time

the REDs are generally speaking FIX-IT-IN-POST cameras, they are specifically made for people who want to use lots of time in post tweaking raw adjustments before they do anything else with the material. If compared to still photography, it's not like comparing raw still and jpeg, it's more like comparing raw still and 16bit or 32bit full resolution tiff recorded in-camera. you can do practically all the same things with the tiff if the in-camera image processing is working well. (and most cameras don't record uncompressed raw so you still get artefacts with it, for example jpeg artefacts. because of the raw-->rgb processing, the artefacts don't look like the same after processing than if shooting normal video so they don't add up in the post chain and are therefore usually manageable although can look weird and lower apparent resolution etc) I think the Mini will be good camera if it'll be reliable enough (not blackmagic's speciality) and the new sensor is usable. Most cameras, especially the lower end ones like Blackmagics, have usually had serious problems with the usability and user interface so I hope their design is practical in the field (not like FS7 for example which has bad viewfinder ergonomics, too complicated menus and not rigid enough lens mount)

I hope they have got rid of the FAT filesystem, it is one of the biggest disadvantages of REDs although not such a pain in the *** than cinemadng or ML raw. I don't find any info but maybe they are at last managed to switch to UDF, I don't know :) compressed rgb formats are just mandatory in today's production environment, of course you can still use raw if you have time and resources to backup and post process it. I think about 90% of people will shoot prores444 or XQ with the Ursa Mini anyway and the raw option is mostly for special use and indie films where you are either shooting very limited amount of material or where you have unlimited time in post to tweak the material (like shooting with Epic and tweaking every clip 2 hours in RedCine before even making previews :lol: )

could be easier at airport security check at least :lol: (Yes I have this Weapon and couple of batteries for it, and also couple of Tactical Cage parts and some Battle Tested gear in those bags. What do you mean I can't bring this Bomb Viewfinder to the plane :lol: )

As long as you're not using windows10 you'll be fine I think :D Anyway, if the scene is about some kind of terrorism I think it is quite unlikely for them to get military grade explosives, more likely scenario is ammonium nitrate, stolen dynamite, acetone peroxide or even mekp. But maybe you don have to be very realistic on this one and it might be even bad to make the scene too realistic

you could watch couple of Mythbusters clips, they blow up lots of stuff with C4 so you can see what a real blasting cap etc looks like so you can imitate them more easily. https://youtu.be/AwyniA5ryhY?t=14 and so on. I think almost all action films have a C4 scene or two. the coiled wires and a timer with a display and a flashing light are cliches though so you may want to think if you want to use them in your film or not :) close up of a blasting cap (detonator): https://youtu.be/01pjt_K-94M?t=2166 I think if you have an item which is closely looking like a blasting cap and just some white clay like substance it will sell the scene easily

Yeah, almost all of them are big budget films from huge studios... The list should have been named " The Greatest American Blockbusters of All Time" or something like that :D There is lots of smaller excellent American films and also lots of better films from other countries, for example Japan, France, Denmark, Soviet Union / Russia, GB, etc. One thing is that without big studio the film is not likely to get wide distribution in the US and is thus very unlikely to end up on lists like these. People usually vote films that everybody has seen, it is also much easier to justify the decisions when most of the audience has seen the films they are talking about. It is usually the same thing in Academy Awards for example, a widely distributed blockbuster is very likely to get most of the awards

meant to say "over the noise floor" :lol: anyway, I think 14 stops or more is good enough in 95% of cases and very rarely you would need something like 16+ stops. if you are, for example, shooting high contrast scene with reflective surfaces it may be useful, or with for example drone shots where you can't control the exposure 100% during flight. We had some drone shots in Spain for example which had backlit water and very high contrast, the water reflection was a bit too much for Epic Dragon which ran 3 stops of HDRx :blink:

the usable dynamic range depends a lot of the noise floor and how noisy stops are still considered usable. I think most of the current cameras have at least 1 or 2 stops less usable dynamic range than the manufacturer claims because of this reason. (you record the lowest stops and they can be seen below the noise floor but they are so noisy that they are just not usable for commercial purposes and can be very well forgotten when determining the actual dynamic range :P ) It's just like with film, you know that it can record, say, 4 stops below middle gray but you know that you can't use that last stop for anything important

Hi you all! I just wanted to inform you that according to my tests the Double Data 's manual verification mode DOES NOT WORK AT ALL in the 1.2.1 (current version) of the software. I did not find this info easily from the web so I thought I should share it in case some users are not aware of this problem. This is just for the "Data Manage" ----> "Verify..." -function and does not affect normal transfers. It can cause you trouble however because the software does not read checksums correctly in this mode because of the bug and may show that the files are OK even if you alter them or remove all of them and just leave the first text file in place which contains the checksums from the original camera media. Because of the bug, the software actually copies checksum data from this file and adds a new text file containing these checksums which it is thinking is the checksum data of the verified files. So, if you look manually through the checksum text files they seem to be OK although they have nothing to do with the new verification round. So, it basically falsifies the checksum data in this manual verification mode and green lights all the files no matter if they are altered or not or even are in the same folder anymore. I contacted the manufacturer and they don't have any bug fix release yet but they are aware of the problem and are working on it, so it is best to not use the manual verification (Data Manage -->Verify... -function) at all with this software until they get a fixed version out. As I said, the bug does not affect any normal transfers from camera card to storage, and is only in the additional manual verification which can be used afterwards if needed. You can also trust the checksum files which the program creates during normal verified transfers.

You can get continuous rec with the high bitrate h264, if i remember correctly it splits the files so fat filesystem is not a problem. You still need to test it first, too high bitrate can cause stop of recording

yeah, the range is quite similar to the range of a still raw image captured with 5D. the noise may look a bit different because of the line skipping and lower resolution / scaling so you can't gain the image as much in post as you could gain a still image. I think the compressed mode is best option for general use and raw is usable only for special purposes, for example if you need to use more noise reduction in post or if shooting gradients etc where the compression shows clearly. With ML you can also use higher than normal bit rates for h264 recording so it could solve a lot of problems you are trying to avoid with raw. I think that the ML RAW is too time/work consuming and unreliable for general shooting but for special purposes it is OK. I, for example, use it with 5D2 for underwater tests and when I have to shoot in places where the camera is at risk so I don't have to put a more expensive camera, for example FS7, to danger and can still get usable shots for doc purposes.

1.yes it is possible. you may have drop frames and take length limitations depending on your memory cards, ML version, rec settings and even the specific camera you use (some cameras react to hack firmwares better than others even if identical models) 2. you have to update the firmware of the camera which always has small possibility of "bricking" the camera if the upgrade fails. this is rare so you should be fine 3.you don't get any more range than with h264 using aggressive gamma settings but you get rid of the compression artifacts. other artifacts like moire and noise are still there, although noise reduction works better for less compressed format. it's something between 9 and 10 stops which you can get from raw at maximum if you use the noisy lowest stops 4.it depends from a lot of things and even varies if for example temperature changes or other weird things, see 1. if lowering resolution you can usually shoot at least 50 seconds continuous but it may vary from take to take so only way is to test with your actual setup 5.usually you transcode the ML's own raw format to either CinemaDNG or compressed Quicktime, for example Prores. I have made some low scale stuff by transcoding with Rawmagic to cinemadng, imported to After effects, made raw corrections, exported to 10bit prores 444 and used that for editing 6.you should be aware of possible drop frames and varying max take lengths. I suggest not to shoot dialogue or any kind of stuff which requires lots of takes, you'd be more happy with a black magic for that if you need raw and the workflow is much quicker and easier 7. you should test the noise floor you can live with, you can't expose the camera carelessly even if it's shooting raw. aliasing is also there and you should test if you can live with that, raw creating a clearer image reveals all the other imperfections more clearly