Tyler Purcell

Shots look great, as I suspected, perfectly lit. :) Toss me a short clip from the camera via drop box. I'll gladly take a few minutes to record a video of the problem when in DaVinci.

Looks like a lot of fun... For a movie shot digitally that is. LOL :D

Digital sucks. :rolleyes:

:P

Well, the C300 has the same mush. So does every 8 bit 4:2:0 camera I've ever colored.

Bit depth and lack of Cr (chroma/red) data. Basically what it does is limit the dynamic range of the imager AND color reproduction. Is this a big deal for youtube movies? No... but it's a big deal if you go outside and shoot random stuff that needs to look good. If you don't have the ability to fix issues in post, the camera is worthless in my eyes.

If you shoot perfectly, they look "fine". Just shoot something not perfect and try to correct. Yep that's very true. But it's not about delivery, it's more about being able to correct when you make a mistake.

Give me a on/off button, manual iris, manual focus, manual zoom and I'm happy. I hate menu's, I hate switches and knobs. Cameras these days are too complicated, which is why I really like the blackmagic cameras because they program everything into a very simple menu. It really makes using the camera much easier for most situations. Yea if you're outside shooting ENG (electronic news gathering), you'd use a nice big shoulder mount camera with all the functionality on the side and NO menu's. However, when you're shooting in a controlled situation and you can spare the 10 seconds it takes to make a minor change, its great to loose those buttons and knobs.

I didn't see it in the theater, just on BluRay... which is 8 bit 4:2:0, so it's hard to know. I shot a feature that was shown in the theaters on Canon 5DMKII which is the same 8 bit 4:2:0 codec. It looked ok, but it did not look great. The layer of "mush" on top of every image was constant thanks to the compression and the lack of our ability to do any serious correction. If everything is perfect and you don't need to do any post work what so ever, 8 bit 4:2:0 can be made to look ok, nothing like 12 bit 444, but good enough that the average user wouldn't know the difference. However, most people can't shoot perfectly NOR do they have the time. So everyone today needs the ability to make some serious corrections in post. YES recording in S-Log or whatever variation of compressed "wide dynamic range" you wish to use, does make a difference, it's not enough. I edit and color to pay my bills and I've worked with quite a bit of C100/C300 material (including shot my own stuff on C300) and I'm finding myself up against a wall constantly with the ability to make corrections in post.

Just remember the post I made earlier about the bit depth and chroma sampling. The C100 captures in "substandard" 8 bit 4:2:0. Even though the imager is very good looking, (I think Canon imagers look pretty darn good). Imager is only half the story, the other half is the recording format. You can have the best imager in the world, if you can't get what it's seeing into your editing bay, what's the point? If you want good audio, no camera is going to do it internally. So the "audio" benefits aren't relevant. Generally camera operators are too busy running the camera to make small audio changes whilst recording. Plus the C100 is a big camera, it's tall and fat. The form factor doesn't work well for hand holding because it's so heavy. I've shot a bunch of stuff on the C300 and even though the output looked good, I didn't much care for working with it. Canon has to many little buttons, knobs and panels that flip out/covered up, it's a real pain to just do things with it. Mind you, menu-only operation isn't much better, but at least there is only ONE WAY to get into the menu and make changes. Plus, in the C100 isn't much better in quality then the 5DMKII still camera, unless you want to shoot super high ASA. Obviously the 5DMKII doesn't have the audio function, but it's a lot smaller and less expensive used.

Very cool! DPX is 10 bit BTW and it's a very old codec. Pro Res XQ is 12 bit and pretty much lossless. With B&W, the bit depth becomes more important and honestly, if you do 12 bit, I'd go for it. Come monday, you'll get your ear talked off by the "scanner" guys on here. :D

Why not get Pro Res XQ? Its a full 12 bit RGB format and most transfer houses can deliver it. The great thing about XQ is that Avid, Premiere and Final Cut X use it natively. So you don't need to do any conversion, drag, drop and edit. You'd edit at 24fps for theatrical and going back to print. Are you shooting B&W negative or reversal? Sounds like a cool project, I'd like to know more about it.

I made a comment about that earlier in the thread, but you probably skipped it. Today's digital audio recorders are inexpensive, Zoom makes 3 great models. (2 channel, 4 channel and 8 channel) They're pretty much all 24bit 192khz and record to SD cards of some kind. They're reasonably priced and the newest 8 channel one has timecode as well, which is pissah'. What I do is run all my audio to the recorder and take the output from the recorder, send it to a wireless transmitter and put a receiver on the camera. This will give you a decent mix in camera as temp whilst your editing. Then when you're done editing, you can export just the clips you use and there is this great program called Pluraleye's, which automatically sync's everything for you. Adobe Premiere has this function built in as well, but it would require syncing everything which is more time consuming. Once synched, all you've gotta do is export an OMF file and quicktime reference and you're all done. Everyone uses wireless lav's today for almost everything. The new flat/pancake mic elements are taped to individuals chests and a little fanny pack is strapped to their waist under their garments, with the wireless transmitter. The receivers will be in the bag with the recorder, plugged into XLR inputs. Then you'll have one boom mic per two people as backup. So you can see how even 8 channels can be tricky and a lot of bigger films will use even more. I'm very much a fan of getting it right on set and having two channels for each person talking (lav and shotgun) is ideal. For smaller shoots with 2 people or less talking, you can get away with a 2 channel recorder if all you wanna run is lav's and if you want a boom as well, buy a 4 channel recorder. I rarely run more then 2 mic's, so I have the 2 channel one, which I regret purchasing, but it was on sale. One thing to remember, sound is an art and it needs to be monitored. So if you buy a lot of equipment, you'll probably regret it because most sound guys come with their own stuff.

Yea, if you think about size/weight and complexity of a 3 CCD imaging block, CMOS is quite efficient in comparison. Yep, they also give a wider field of view, which is what MOST people buy them for. This way on cameras with smaller imagers, you can achieve a wider field of view with like glass. ASA = ISO F stop = T stop It's all the same stuff. Not quite... Most manufacturers has their own mount and some cameras have interchangeable mounts as well. Canon uses EOS Nikon uses F mount Sony uses E mount Panasonic has relied heavily on the Micro 4/3rds mount, so has many other manufacturers. Yes it's true, in the world of cinema cameras, the vast majority of them are PL mount. However, as you can imagine, PL glass has gone up in value as these digital cameras have hit the market and the demand has increased. Buying a camera that uses PL, prevents you from using anything else because the flange distance is too great and unless the camera has an interchangeable mount, it... then you're kinda screwed. This is the great thing about the Blackmagic Pocket Camera, it has a very short flange distance, which allows you to run almost any glass you want from PL to Arri B and Nikon/Canon/Sony, etc. Plus, the super 16mm sized imager, allows you to run super 16mm glass! I bought some very inexpensive Super 16mm primes for my film camera which work flawlessly with the Pocket camera. People don't want old Arri B mount Super 16mm glass, and ebay is full of the stuff.

4k exists because: "If you make it, they will buy it" Bigger is always better anyway right? How about Super 16? Low ASA stocks are only roughly 2.5k worth of information, yet a good S16 product looks better then MOST digital cinema. What's that all about? Each pixel has 1 red, 1 green and 1 blue photosite behind it in RGB RAW mode, according to Blackmagic design's chief technology specialist. - It has a 16 bit image processing system and 12 bit uncompressed RAW recording for full-dynamic range capture. - Actual 12 stops of latitude recorded to card, unlike Panasonic who measures the imagers capabilities. - Over-all soft/smooth image with Alexa-like, super soft highlight clipping. - Film-like color science (copied from the Alexa) - Has industry standard controls (Precise variable kelvin control, shutter angle and ASA) Cinema is 1.85:1 (2048x1080) and internet, television, BluRay is 1.75:1 (1920x1080) To present 1.85:1 on the internet, television or BluRay, you are putting black bars at the top and bottom. Thus your image is slightly smaller then your display device. To present 1.75:1 on the internet, television or BluRay, the image is presented 1:1, filling the screen of your display device. What's funny is that digital cinema projectors, use the same 1.75:1 aspect ratio imagers our home electronics use. So they actually crop slightly for 2k, 4k 1.85:1 presentations. So theaters have no problem projecting 1920x1080, they just don't want to because they'd have to change the screen matte. This is why everyone upscales 1920x1080 material to fit into the 2k 1.85:1 frame for theatrical release. It has nothing to do with technology, it has to do with theaters being lazy. Mind you, nobody is going to notice the 128 missing pixels when you're 20 feet or more away. Again, I've had theatrical runs of two movies I shot in 1080p and they looked pretty good on the big screen, better then I expected. Panasonic doesn't have cinema color science, it has "video" color science. This is the biggest problem with it, the images always steer towards being cold. The imager also has harsh highlight clipping issues, when it clips, it can actually tear the image. The camera has a 10 bit processing system and bakes in REC709 both internally and externally. So even when you make the image flat using V-log, you still don't have the dynamic range necessary to expand to 12 stops of latitude. Yes, it's better with an external recorder, but not great. It's true, the difference between 10 bit and 12 bit isn't that big when you don't have 12 bit capabilities. When you DO have 12 bit capabilities, the difference is night and day, especially in the coloring suite. One other thing to note DCP is a 12 bit format. Blackmagic have done a fantastic job with the compatibility of Cinema DNG, which is what their cameras use. Premiere and Final Cut X can use the Cinema DNG natively. For Avid (what I use) we simply use DaVinci, select the folder we wish to transcode and it spits out Avid MXF or Pro Res. I've converted full 64gb cards in less then 10 minutes on my bay. So it's SUPER fast to convert because there is no scaling going on, it's a simple 1:1 conversion with a LUT applied. For my personal shoots which are mostly for internet use, I work in 10 bit Pro Res 4:2:2 HQ. It's a 220Mbps codec which is 99% lossless and has a wider dynamic range then standard Rec709 codec's like Long GOP MPEG. So I shoot with a LOG output or "film" mode, edit using the Pro Res LOG footage with no grading. When done, I will grade the whole show and export a final from DaVinci. I don't do any scaling and my final output is 1 generation off my camera original. I also don't use "looks" of any kind on my stuff. I color everything by hand to match what the camera looks like natively. I have used film "looks" to fix shoots that used substandard cameras, but for the pocket camera, it's unnecessary. A good multi-layer grade will look fine. In terms of Audio, I do my mixing with Avid. It uses the same rtas plugin suite as Pro Tools and it allows me to tweak my final mix as I edit. It's one of the reasons I continue to use Avid, the audio portion is WICKED powerful and all the effects are real time. I record 90% of my audio using the 1/8" input of the pocket camera. I have a Sennheiser wireless lav kit which screws onto the top of the camera. Yes, the pocket camera has very noisy preamp, but unless you're doing an absolutely silent scene, you will never notice them. The rare occasions I need more then 1 mic and/or critical noiseless recording, I have a Zoom 4 channel recorder that screws into the top of the camera. I will then run the 1/8" stereo out from the Zoom into the pocket camera and have one mic on left channel and one on right channel. If for some reason the noise floor is too high, I will grab the audio from the Zoom to compensate. If you have more then two people mic'ed and audio becomes more critical, it's time for a sound guy. However, I always give the sound guy my wireless transmitter and they will send me a mixdown that will plug directly into the pocket camera. This way I've got good audio on my editing suite without having to sync anything. I generally only sync what I need, maybe one or two lines of dialog that I need separate. If I use someone else to mix my audio, I will use pluraleyes on the finished shots and hand the mixer an OMF file of all the audio used, synched to the clips. This way, when they import into Pro Tools, everything will be lined up with 60 second handles. It's a great trick to save time in prepping for post AND doesn't really take much time in the back end to prep. Why sync 300 clips, most of which you won't use, when you can sync just the clips you need when finished with the project. The battery issue is one of those things you have to learn about. Once you figure out a good workflow, it's not a problem anymore. The batteries are $12 dollars and if you start recording and let the camera run, they will last the length of an entire card at Pro Res HQ resolution (42 min). If you're constantly starting, stopping and turning the camera on and off, the battery will stop working. On bigger shows where you need the camera to be on for long periods of time during the day, run external power. There are many inexpensive solutions for shoulder rigs which work great. In terms of the cards, yea the camera isn't a toy, it records at a high bitrate. No other camera on the market allows you to record RAW or Pro Res HQ onto standard SD cards. It's by far the cheapest route to go, considering the bigger cameras all run CFAST or SSD's, both of which are tremendously expensive. The standard 95MBps, 64gb SD cards are always on sale at Amazon for $48.00. I have a Mac Pro Tower with double Xeon 5400's and Harpertown chipset, which is around 8 years old. It's an 8 core 2.8ghz machine with 12gb of ram, NVidia 680 4gb video card, AJA Kona card and Red Rocket. I have a tiny it of internal raided storage (around 4gb) and I use external drives for most of my clients projects. I can playback Red Code @ 4k no problem. I can playback Pro Res XQ 4k no problem. I can playback cinema DNG 4k and from my pocket camera, no problem either. Heck, I'm doing a very complex Red rendering job right now using firewire drives which are very slow and I'm still up at around 30fps converting 4k to 2k, with 4 nodes and two mattes per shot. Clearly it's time for a mac? LOL :) I understand how it works. I just don't believe you can extrapolate enough red data to fill in the gaps. There will always be substantially less Cr information then Y and Cb channels.

Those are awesome factors. Who wants to deal with 4k media? Nobody... and since 1080p is the "industry standard" for deliverables, it kinda makes sense to shoot with it. Small imager means you only use the center of the glass, which you may not know, is the sweet spot. That's where the image is the most clear, the further out you get from the center, the more distortion you get. Commercial/professional glass focal lengths are a fixed measurement that have zero impact on the imager size. Most everyone uses S35mm glass on S16mm size imagers and film cameras for that matter. It's up to the cinematographer to know what a 50mm lens looks like on S16 vs S35. Yes, it's hard to get shallow depth of field, but it's not impossible. RAW is another world. Once you debayer in camera and bake in the limited color space of Rec 709, you can compress the color space all you want into that Rec 709 image, it will never look anywhere near actual RAW. Not only is the raw 12 bit 4:4:4 with lossless compression, but it does the debayer in DaVinci. So you have much more information to work with color wise. All of the aliasing nonsense you get with in camera debayer 4:2:2, is all gone. The colors no longer bleed, they are crisp and the 12 bit color space delivers an ACTUAL 12 stops of dynamic range, not a made-up number. The GH4 shoots 8 bit 4:2:0, that's the recording format. That's what you will be stuck coloring with. Sure, technically you can scale it down to 2k, but you still have the same limitations in color space. You don't magically gain color space, you would simply be reducing the aliasing lines between the channels. Plus, I don't think you can get anything out of that red channel, it's missing so much data, you'll be lucky to get 4:4:2 out of it. Ohh and you're wrong about a native imager not getting 4:4:4. You don't need very many extra/spare pixels to achieve that. Red, Arri and Blackmagic have been doing it from inception. The blackmagic URSA mini's aren't anywhere near the price tag of the GH4. To do an URSA mini package properly, you're looking at between $6 - $8k depending on which body you buy. The GH4 and A7SII kinda stand alone in the world of pocket-sized 4k cameras. I wouldn't consider either one "cinema" by any stretch of the imagination. The closest thing for the price to an actual cinema camera is the Pocket camera. Also, the whole idea of shooting 4k, scaling down to 2k and stuff, what a pain. Far better to link original camera media in your editing software, edit the show, import to DaVinci for color and export from DaVinci a final colored piece in the native camera resolution and format. Again, trying to solve camera problems with upgrades, what a joke. It just shows what a toy the GH4 is, you can't even record with it properly unless you add outboard gear. Now you need to have a shoulder rig to mount the stupid Ninja to, cables running from the camera to the recorder and you're STILL only capturing 10 bit 4:2:2, which is again, the lowest acceptable quality for broadcast. All of that money spent because somehow you NEED 4k... for what? Give me 1080p 12 bit RAW any day of the week over 4k 8 bit 4:2:0 Long GOP MPEG. You should really rent a decent cinema camera and learn what it's like to use one. The experience is polar opposite to the GH4, which I've used... I've colored... and I've bitched and moaned about it's worthless results. Great toy to play with when bored, but not a serious camera. Far better to buy a more standard 1080p camcorder and save the money for something else.

Ohh! Well! Now that's starting to make more sense. Here is the skinny. When still cameras started moving towards digital, many manufacturers felt it would be nice to include movie recording as well. It took a few years, but now almost every still camera shoots video and some of them shoot pretty darn good video. So if you walk into a camera shop, you can buy sub $600 still and video cameras, which aren't too bad. The still cameras will have interchangeable lenses and for a few hundred more, you can buy a video-only camera that also has interchangeable lenses. There are two basic resolutions; 1920x1080 (which is 1.75:1 aspect ratio version of 2k) and 3840x2160 (which is the 1.75:1 aspect ratio of 4k) Since movie theaters are 1.85:1 aspect ratio's natively and broadcast/VOD/BluRay aspect ratio is 1.75:1, that's why you hear of 2k and 4k resolutions. In reality, they are nearly identical to the "consumer" versions; HD 1080p and UHD 2160p Today we use CMOS imagers in almost all cameras. They are a grid of photosensitive diodes which generate a small electric charge when hit with light. This data is then pulsed to an imaging processor. The key thing to understand about CMOS is that it's a B&W technology, it has no inherent color. In the past with CCD, most cameras had 3 imagers, each one with a color filter in front of them. However, CMOS imagers are more efficient, smaller (don't need 3 of them) and have far better dynamic range. How CMOS get's its color is through a filter which placed on the imager itself that tells certain pixels what color they are to receive. The end result will be a perfect RGB signal, which is referred to as 4:4:4. The lowest acceptable quality for deliverable is 4:2:2, which zooms up on the imager for the blue and red channels. So you will see aliasing on the image when the blue and red channels meet one another since they are HALF the resolution of the green channel. 4:2:2 is the lowest compression acceptable for broadcast deliverables and all commercial cameras shoot in this format OR better. Bit depth is another issue and it's the defining factor when it comes to dynamic range. You've probably heard of dynamic range from your past experiences, but it's defined by the variance between the highs and lows within a single exposure. So if you can imagine a chart, one side is white, the other side is black, the lower the bit depth the more steps you will see in that image. The higher the bit depth, those steps will basically disappear and you'll see a perfect white to black image with no steps. Broadcast deliverable standard is 10 bit, which is the lowest quality acceptable in the industry today. Camera manufacturers have many choices when they make cameras, but all of them want to keep the cost down. They have to still compete with one another, so most of the cameras on the market are "feature rich". They have very small form factors, slow mo, high resolution, wi-fi, bluetooth, big fancy display, high ASA, etc. As a consequence of the feature set, they have to run a light weight processor so the battery doesn't die and they don't have to run a big heat sync. This limits what the cameras are capable of doing in a pretty substantial way. Today, the vast majority of sub $1000 cameras record 8 bit 4:2:0 color space. Yes, there is a ZERO for the RED channel, so you can imagine how much information is missing in that image. This is the same quality as youtube, vimeo and even BluRay. However, for a "camera original" you really want better quality, yet for sub $1000 there isn't much. These cameras also record highly compressed files using a format called MPEG Long GOP. This format takes one still image very 14 frames. It then only records the difference between the 13 remaining frames. It works great for static shots and still life, it doesn't work great for high motion material and anything where the lighting and exposure isn't perfect. The other way to keep cost down, is to strip the camera of all it's features. Get rid of the fancy display, ditch the slow-mo and remove all bells and whistles. Unfortunately, there is only one camera manufacturer building sub $1000 cameras like that, it's Blackmagic Designs. You know, the guys who use to make video cards, well they make cameras today. The Blackmagic Pocket Cinema camera is the same form factor of an iPhone 5 yet it delivers cinema-grade images using a very powerful 12 bit 4:4:4 processing system, which delivers uncompressed RAW capture or several flavors of 10 bit 4:2:2 ProRes. If you don't know about Pro Res, it's become the industry standard for post production and now cameras shoot with it. Pro Res compression uses wavelet technology, which is similar to JPEG. It compresses each frame individually and is 99% lossless. The only downside to the Pocket Cinema camera is that it's a small imager, just standard 1080p. However, it's a super small camera, so you get best of both worlds. There are a dozen really powerful sub $1k cameras, each of them have their pro's and con's. However, if your budget is $3k, you should be buying a $1k camera and spending $2k on lenses and accessories. Most of the modern still glass today is all electronic, so it's very hard to do repeat focus and manual aperture. This is where older still glass or modern cine glass comes into play. There are three companies making very nice modern cine glass; Rokinon, GL Optics and Veydra. Prices for new glass range from $350 - $3000 per lens. Used however, you can find some smoking deals. Also, there is nothing wrong with some classic still glass, a lot of people use Nikon primes and zooms because there is a huge collection of non-electronic glass out there, fro multiple brands. In today's world, glass is everything because the cameras (imagers) are so good and crisp, the quality of the glass shows through much more then with DV or even HDV. So this begs the question... what to buy! Well, it's really up to exactly what you wanna do and what quality you're looking for. This post was designed to educate about what's changed in technology, but there are many choices you have to make. Is quality more important then features? Is critical audio necessary? Can you get away with a camcorder like a Panasonic AF100A? These are subjects that should be thought about before making a decision. I've been a shooter for over two decades and I do most of my shooting with a pair of Blackmagic Pocket cameras. I'm working on my third documentary feature shot almost exclusively with them and I personally love them for the size, portability, recording format and quality. Sure, they don't have slow mo or 4k, but I don't need those features. I'm very happy with my results and I just love being able to manipulate so much in post production, it's great. There are many other cameras like it in the $3k range used, but if your budget is $3k... it's gonna be hard to get a camera like it WITH GLASS for $3k. So the sky's the limit! I suggest google searching and seeing what's out there. Read reviews, and understand the formatting they use before purchasing. I hope this has been informative! Ask more questions if you want.

You don't need 2000 ASA, just focus on renting faster glass and optimizing your lighting. I would focus on lighting for 1000 ASA and if you need a bit more, you've got some room for pushing the camera. I'd just be very careful about fixed pattern noise. I'd absolutely go Red Dragon, that's the right camera to use for something like this. It will shoot REAL 4k (not UHD) @ 120fps at 9:1 compression, which is pretty good.

Umm... why? Are you just looking for something ultra cheap? Youtube today is 4k compatible. So people expect quality.

LOL! That's what happens when all your money and time goes into equipment and having fun, shoes aren't very high on the priority list! HA!

I haven't been keeping up with his stuff evidently. After I saw him shoot Amour digitally, I thought that was the end of that. I totally forgot he had been shooting Allen's movies! http://www.imdb.com/title/tt1602620/technical?ref_=tt_dt_spec

Ohh Khondji's back shooting 35 again! WOOT :) Looks like they're going to be striking prints as well.

...

Well if you work with Super 8 I can't comment, it's a whole other can of worms. My earlier comment was about making a professional product that would be printed for theatrical projection, originated on S16 or 35mm. Digital presentation systems have zero flicker. So each frame stays longer on screen then with a film projector. So what you see is not the same. Obviously, everyone telecine's their films to show people because they have no choice. Well, it's a dead market. Most of the people who shoot film today are directors and DP's themselves, who have their own equipment and shoot their own productions. The days of being a DP for hire that shoots on film... that's a super rare occasion. Right, but in a digital cinema the difference is only in "look", which can be faked. If you did an A/B comparison between a photochemically made film print and a digital presentation, the difference is night and day, its MUCH MORE then a baked-in look.

The only similarity digital has to film is the fact it generates a moving image. Digital has much more in common with television then it does with film. Yes, digital cinema cameras try to emulate film cameras by having similar controls, but how they function has zero commonality.