Axel

How did I miss this?

It's a kit you have to assemble and program. But it's the cheapest, as far as I can see. There was a list of about ten cheaper MoVi alternatives, and it didn't list the Came. What is more, there are already a lot of youtube demos that show the Came 7000 in action. This one from steadicam-guru Tom Antos: See especially his own programming: Instead of remotely triggering pan and tilt to follow a motif, he rather made the camera follow his own movements in a smoother flow > genius. I don't expect to be able to repeat this myself. I've seen different glidecam reviews by Antos, some of which I owned or borrowed myself, and he is just a steadicam expert. But see how in this video the woman walks behind the kid (starting at 7:08): She clearly hasn't the faintest idea of how to walk steady, and yet it works quite well. Others point out, that all parts of which such a device is constructed cost $ 300 at most. So maybe within one year there will be some stabilizers for $ 500. The catch is that they are then assembled by some obscure chinese rat-shops, you don't have comprehensible manuals and no community that helps develop the junk. This looks to be more than a cheap alternative. $ 950? No-brainer. I am only having nightmares of sitting alone at home, my cats play with the screws, and the thing refuses to work. I'm not Gyro Gearloose.

You mentioned Came. Look on Youtube for Came 7000 or so. It's $ 950!

This is my fifth attempt to find an affordable stabilization solution for the BMPCC, particularly for the heavy setup with Speedbooster and Sigma 18-35. It works perfectly, and the only two things that are missing are 1. Pistol grip to be clicked underneath the cage with a quick release plate and 2. Lanc-controller with button on top of the grip for START/STOP. It consists of an old, self-built shoulder pad with counterweight (aluminum rack, ~$ 5,- , bent in a vice and padded with foam rubber, ~ $ 2,-), a micro goose neck with 3/8" connections, ~ $ 7,- , filled with a steel rod, ~ $1,-) and a ball head (~$ 30,-). The ball head allows for fast height- and angle-adjustments. I found this to be the most crucial aspect of all shoulder supports. You both stress your neck with only half an inch of unergonomic proportions and run the risk of unintentionally rotating your frame. Not with this ball head. If you loosen the ball, you can screw it on or off the cage within 10 seconds. The counterweight could be substituted with a battery pack, and the rod that helds the goose neck stiff could also be a steel tube (well, perhaps, it could turn out to be too weak then) to hide the cable. The shoulder pad doesn't look particularly elegant, I'm sure you can do better.

Let's be brutally honest. Of all the videos in the Screening Room, only, say, ten impressed me at all. My own stuff would also go unnoticed, for it isn't any better. A contest could be a good way to focus our ambitions. Set a motto, a subject, set a date, when all films have to be online on the same day (not earlier, not later). Create a poll for ten days (or so). Allow people from outside EOSHD to vote (don't know how, perhaps via facebook?).

I subscribe to the 'evolution' theory above. There are so many changes in the video market, one just can't google everything. I receive newsletters from Adobe, Avid, BMD, Apple, Panasonic, Sony, CreativeCow, NoFilmSchool, Wolfcrow and a bunch of german sites, but I learn faster and more from EOSHD. Here is how it works: Andrew shovels in articles. People leave their comments, many in the way of dogs who set their marks on new trees. But very quickly you have a broad scope of opinions, of aspects. They aren't reliable, many aren't even on topic, but in the end, one knows more.

Basically with the Pocket you need to get three things (technically) right: Stable, sharp and (roughly) well exposed shots. Everything else can be fixed in post, at home. For vacations, I only use my iPhone. You could shoot with your 60D. You have a lot of options. Don't despair.

May I comment? You can set the display to VIDEO, but in Recording Mode you should choose FILM. Why? Because the limited values are baked into the file with VIDEO, and you have no chance to recover highlights asf. in post. To get rid of the ugly flat looking FILM image, it's better to use a Rec709 LUT for BM in Premiere. Can't tell you exactly how, in FCP X I use LUTutility, I think it may be LUTbuddy or so. Apply the LUT to an adjustment layer over the whole timeline, underneath you have access to the values that are hidden by the LUT. Also, in bright daylight you usually have big contrasts, and using ISO 400 or 200 will lower the DR (despite what the BM manual says, you just have a dramatically decreased range, as instantly seen in a waveform), so better buy an ND fader to control exposure. But you can use ISO 1600 for night shots.

I can understand that. For anyone, who is used to *shoot video* with a classic camcorder, it's terribly unpractical. The same was true, back then, for shooting with a bulky 35mm adapter or later with a DSLR. But as the bible knows, you love the prodigal son the most. Imho it can best be compared to ancient mechanical 16mm cameras, with which people also had their trouble. To achieve good results, you had to know them by heart, to a point, where knowledge about stock sensitivity (and characteristics) in correlation with aperture and the (mostly fixed) shutter angle turned into intuition. I don't use a light meter, and according to some comments in other places, only few still do, even in professional cinema. You can learn to judge correct exposure with the Pocket by doing a lot of tests and comparing them to what you saw during recording. Here are my rules of thumb: 1. ISO 800 offers the best possible DR for ProRes. To have a fixed ISO also facilitates things, less factors to be calculated on the fly by your CNS. 2. ETTR is also - partly - valid for ProRes. Fill the well, avoid the noise floor. It's 10-bit, so you are not simplifying skin tones by slightly overexposing, like with 8-bit codecs. Use 95% zebra then. 3. For RAW, use 100% zebra. If the image appears too bright on the display, lower the ISO (has no influence in RAW). 4. When I still had a MFT system lens that set exposure automatically when I pressed IRIS, I realized that I always was in the same ballpark manually. No wonder: The automation uses ETTR as well. My suggestion regarding your trip to far east: 1. Make at least ten test shots every day, document them, evaluate them, you will improve your intuitive skills. 2. Buy a used GH2 or G6 (ridiculously easy cameras compared to the Pocket). They can use the same lenses. Compare, as I did, these cameras in the same situations. At first, the results with these Lumixs will be three times better. But not for long. 3. Learn when to use RAW (seldom) and when ProRes (almost always). 4. Learn how to use Speedgrade. Not just in an afternoon. Watch whole video trainings by Lynda or the like, and delve into CC. As I wrote before: The Zacuto Pocket finder (quite affordable) makes you see every single pixel on the muddy display and helps only a little bit as far as focussing is concerned. But you need it anyway when you shoot in daylight. And it adds a comfortable stabilization point. What you also need is some kind of additional support. Mine weighs below 100g and cost 14 € (new!).

Made interviews in an "impossible" place, narrow, dark, with bright sunlight in broad rays through the windows. ProRes with 95% zebra, filmlog and iso 800. Never could have used a GH2 or G6 for it. Expected the worst whilst filming, because the display is so shitty. Everything worked out fine. Missed the focus in about 5% of takes, had no viewfinder yet. Surprisingly, the Zacuto Pocket viewfinder does help for stabilization, but not so much for focussing. I recommend to work with hyperfocal distances and a DoF calculator (i.e. phone app).

First of all, I'm no expert here, I just share my momentary state of comprehension. I was corrected by Sean Cunningham, f.k.a. Burnet Rhoades, that a LUT, and especially anything done with Filmconvert, should be before CC. I didn't buy Filmconvert, but I tried LUT utility with this advice in mind. By some a LUT is used as a look-layer before final output (or even after). Some LUTs are clearly intended to create a shrill look. Not so Filmconvert. It is meant as a starting point, even before primary CC. That's why it has the input profiles for different cameras. And does it "rob detail"? It changes original data (visually, non-destructive), but it doesn't for instance prevent you from recovering highlights the LUT appeares to have swallowed. On FCP X: If you really want to grade in the NLE, you use an adjustment layer for the LUT. This guarantees that all clips will have the same basis. If you grow unsure if the LUT was a good idea, you can hide the layer with "v", or mix with the clip below from 0-100. On the order of color effects in FCP X: That effects should be applied in a certain order is no news. You can, against all reason, apply a look effect as the first link in the chain. But it will not be rendered in that order. Now that FC demands to be the first in line (if it really does, I don't know) confirms Burnet Rhoades' statement regarding the purpose of that app. Further: There are a bunch of third party plugins for FCP X (I mention the two most useful ones, Lock&Load and Neat) that also need to be the first effects. Not possible if you apply both. This may result in occasional green or pink frames. The companies say, it's a bug Apple is working on, but is it really? The answer is: Apply one or none. No effect built in Motion (like LUT Utility) ever slowed down performance or caused render bugs. Stay with these, and look for another host if you desperately need something not yet available in the API collections.

This contradicts the evidence. Adobe made serious improvements since CC. It's an evolutionary process. That has always been the Adobe politics. And it's ridiculous that one can't afford it anymore. Everone has his reasons to stick to his NLE. In the end what counts is if you reach your goals. Comparing software, interesting though it is, is only significant if you really know the NLEs. I know a guy who had to learn Avid at media school, but who never gave up Vegas. My best friend works professionally with Premiere, since 12 years, he also teaches it. The complexity of his projects would never have overcharged FCP X (indeed, from my point of view, his life would be easier with it), but he simply doesn't like it. I respect that.

I don't know, but your clip proves little. One time there is a green cast, one time a blue cast, one time sepia. I would have preferred a whole sequence in one style. Two weeks ago, I had my first (unpaid) job with the Pocket, a series of interviews under different light conditions. It had to be edited within two days, and I was glad to be able to apply the Rec709 LUT on an adjustment layer prior to grading. What I found out is, that the LUT doesn't cut off highlights and crushes blacks or whatever, it just gives you all at once a 'normal looking' video as a starting point. Below that, it was still possible to recover seemingly clipped areas (I used 95% zebra for ProRes and more or less ETTR to avoid noise).

No color curves in FCP X. Also no color swatches. They sacrified a lot of features people were used to for the sake of - what? Simplicity? Indeed: The new HSL-colorboards FCP X introduced look unprecise (which they are, but not because of the way the ranges are represented) only for one who comes from, say, Premiere or FCP legacy. However, what they say is, the logic behind changing the appearance of your image is shoving shadows, mids and highlights over these three boards. That's the extend of it. Newbies have no problem to understand that. I knew Color for years, it has everything you can dream of, and the levers, curves and pucks in the swatches are highest precision. Will this give you better results at the end of the day? I tested this over and over again, and I often found my off-the-cuff intuitive grades from FCP X more pleasing. It's crucial for every good CC to follow the order of operations: Primary (scopes needed for this step!), secondary, look, consistency. You try to do everything in one step, you fail. I'd guess if a professional colorist would be asked to grade a cinema trailer or a high-end TV ad in FCP X, he would laugh at you. If you can't afford a professional colorist, if you are basically a one-man-band, if you are not a minor contributor in a big industrial pipeline, you will have better chances with FCP X. That's valid for many other aspects, like sound and effects too. For audio, there are three stages: 1. automatic enhancement (terrible, useless like automatic color balance), next level: 2. equilizing, leveling and repairing with adjustable presets (what you hear is what you get, looks amateurish, but is actually of high quality), and then: 3. filters and effects from Logic. I'm no audio expert, but let's assume they resemble those available in Premiere. Now isn't it so, that anyone who knows about the intricacies of those wouldn't bother to treat audio in the NLE in the first place? Grading in Premiere (not Speedgrade or Color Finesse) is completely unintuitive and way too complicated to perform. A lot of options, but no workflow-integration possible. The Adobe way. If you asked a professional colorist to grade within Premiere, he'd probably say, no problem, I charge $1000 a day, brace yourself for the bill. That said, if someone is really determined to start grading in earnest, but on a narrow budget, there is no alternative but DaVinci.

This indeed shows how better resolution may influence the way you see things. There are a lot of wide shots I would have subconsciously avoided with my Pocket (as well as with my GH2). To make the clip even better, there should have been more close ups of simple structures in between. The camera can resolve the chaos, but after a minute or so this gets somewhat boring.

FCP X, but know Premiere as well. Reason: Best NLE for existing OSX, affordable. Of the Adobe suite (I have access through a friend), I sometimes miss After Effects, particularly for it's puppet tool, but also for the very inspriring Videocopilot tuts. The very good warp stabilizer can be substituted by the Coremelt plugin Log&Load. Color correction - wise Adobe has no advantage. On the contrary, in FCP X you can jump from clip to clip in the timeline, like in a grading suite, without loading an effect window. The Adobe CC is a very advanced package, and they are better than ever, but frankly, they didn't have the balls to get rid of the track concept (the clip-based magnetic timeline at once became the reason for many conservative editors to change to Adobe or Avid, when FCP X was launched, and both immediately saw their chance. They made special offers for the turncoats). Once you successfully weaned yourself from all those useless video tracks, you only find it annoying to go back to them. And though Adobe has a hundred times more tools on the shelves, I only rarely miss anything in FCP X.

I am trying to avoid noisy shots. If I have one, I load it to FCP 7, where I have a stable Neat. When cows fly. Even if FCP X doesn't crash, it gets very slow with just about every plugin you install. I have accepted that. FCP X is a very mighty footage organizer - outsiders just can't fathom how advanced it actually is in that respect. It also is the fastest editing tool - render benchmarks and GPU acceleration set aside, just from the way the edits are organized. It also offers a very intuitive and easy, yet limited, color correction. Everything else, to tell the truth, is in it's infancy.

I agree with ergopossum. Use the HDMI and a specialized software like Dragonframe (or other). Then of course a camera with higher resolution offers you some possibilities like "digital zooms". What you should not do, imho, is use frame interpolation functions or added motion blur. I loved the old Wallace&Gromit films, but I hated the cinema releases. They looked too close to 3D GGI, and that was very disappointing.

Yes, FCP X does not like third party plugins, to put it mildly. One of the most useful, the stabilizer Log & Load, is an exception. With Neat, it depends on the Neat and FCP X version. Google for the experiences of others before you ruin a project.

Thanks Maxotics and Julian. These insider tips are always the best. Same with music. Mainstream is often too predictable. A propos POV. Don't miss Enter The Void. It's a film with very many flaws and I wouldn't call it influential (at least not for me, who a least twice a year opens his 'doors of perception' as a recreative vacation). But that's imo what most interesting films have in common: They have something that makes them stand out for the individual, in this sense: Art is not so much about perfection than about bold exclusivity. The meanest and lowest mainstream movies can be more perfect (technically or in the closeness of narration and narrational form) than the masterpieces. Those tend to have some very strong aspects, but also flaws. Leonardo didn't know about materials, his Mona Lisa had the 500.000 crackles the moment the paint had dried (which took months though). Same with his Last Supper: The wall sucked the colors off, it weren't the farts of admirers over 500 years (source: Donald Sassoon, Mona Lisa). The numerous fakers know better than the master. Beethoven chose improper instruments for his symphonies (Bernstein said, none of his students would have gotten away with such an orchestration). Shakespeare had so many logic flaws in his (predominantly adapted) plays that a modern script doctor would despair.

Probably there were one or more 8-bit effects/filters, of which FCS2 still had a few. Also, Canon then provided an EOS-FC-importer plugin on it's support/download site, originally only for 5D and 7D, but that could also be used with 550D asf., because Log&Transfer didn't convert in the right range (this is what 5D2RGB *allows* to override, otherwise it uses the same encoder). A lot of bugs in initial QT7 releases (and continuing) - yes, no wonder it's better now. With AVCHD in FCP X, quality also shouldn't change (original media compared to optimized media), I didn't transcode a wedding with GH2 last summer, and there was quite a lot of CC needed. It's just that at some time I had to switch to 'better performance' (skipped frames and/or dynamically lowered preview resolution) to maintain real time. During CC you rather want to have 'highest quality'. In the end, this is all about performance.

Stu Maschwitz, in his DV Rebel's Guide, made the definition: If you can ride it, it's a dolly. To put a tripod on rails is not the same. And you can easily build rails for your tripod for under $50 that don't look so elegant and will discredit you on a professional job like a slider shot in a studio for a car ad ;-) - but that actually work the same. As for a jib, watch this:

Better results than working with original data? This was proven to be a myth. 5D2RGB does not improve the quality. On the other hand there is another myth, that converting to ProRes (or other Wavelet-type codecs) would cost quality. Whereas it does in theory, it was done millions of times without known issues. As a consequence, I'd say you should choose what fits your needs. Editing AVCHD is no problem on any modestly fast machine, nor is moderate CC. Also exporting a master in mpeg4 (at high bitrates of course) is a common MO for many Adobe users, particularly as the AME was tested to do a good job here. That's why Adobe never developed it's own intermediate. A completely different affair is compositing with long GOP codecs, i.e. in After Effects or Motion (multiple layers with tons of keyframes on 'virtual' frames). This so dramatically slows down performance on any machine, that imo it is harebrained to do. People wonder then why they get errors in the exported files that did not appear in the preview. My suspicion is that they sacrified preview accuracy to get more real time.

Does one of the adapters fit?

In the chronological order I saw them, with explanations: 1. Goofy Adventures Story The first film I saw in a cinema when I was 5 years old. The initiation. 2. Son Of Godzilla One of a series of japanese monster movies that became my every goddam sunday afternoons excitements. 3. King Kong (1933). I was so thrilled by the stop motion, that my father bought me a Super 8 camera, and I staged plasticine monster fights with a lot of blood, destroying my big brother's model railroad houses. 4. Live And Let Die (and all 007) In puberty, I wanted to be like Bond. I wanted to star in a movie and be like a successful man was supposed to be. Dreams were my reality. 5. A Clockwork Orange Our weekly visit to the local cinema, random choice. I didn't know anything about it in advance, I anticipated some science fiction movie with a time travel story. All of a sudden I saw 'cinema' completely different. 6. Once Upon A Time In America The narrative montage, or the montage of the narration, wonderful. 7. Blue Velvet Absolutely thrilling, I think I saw it ten times in a fortnight. 8. Andrej Rubljev (english spelling not checked) I felt a real artist invited me to meditate about the nature of life and of art. I knew it was not mainstream, but I saw it repeatedly since. 9. Groundhog Day Life is what you make it. Every day is the beginning of a new chance to make things better. A comedy, but with a serious message. 10. Vertigo Revisited after seeing Chris Markers Sans Soleil, with other Hitchcocks, the perfect film insofar as form and content became inseparable. Other examples, not quite fitting into this timeline: Running On Empty, Modern Times, Magnolia, Being John Malkovich, Fight Club. Star Wars would be between Bond and Clockwork, for obvious reasons.