Axel

Well, image-wise, if there is nothing specific to express, just visual beauty (as many of us wish to capture), this imo is as good as it gets. Usually I don't like glossy photography, but with your eye everything seems to be a few levels deeper than the usual stuff. Requires superlatives. This is awesome.

As I see it, this thread is not about BM vs. Panasonic. For my part, I have no income with what I do, for the odd wedding I shoot I just ask for a piece of equipment as compensation. I had a GH2 for over two years, it was hyped everywhere, not least at EOSHD, and rightly so, because for me it turned out to be 'THE ONE' (title of a BMPCC short EOSHD-member Frank Glencairn shot when the camera had no RAW yet). My thumbnail still shows the GH2 with my self-built pistol grip. So when I finally decided to jump to BM, I was aware that the comparatively low price of the Pocket was deceiving, and I figured I needed a lot of trial-and-error and frustration tolerance. I just didn't expect how much. I can't allow myself to follow the obsolescence cycles of the industry, making every new camera a one night stand. A daunting example of where this can lead you is my video buddy who at least buys two new cameras a year, always praising their strenghts and soon becoming disappointed by their weaknesses. Once you have invested time and money and dedicated yourself, it's hard to read, well, you're no longer up to date. You can feel hurt. Judging from the clips we saw so far, the GH4 is a winner in almost every respect. Can't get my friend to buy one, because he thinks he can't show up with a Lumix on a professional set, took quite some time to convince him to try 5D M3 with MagicLantern RAW (which proved, btw, that one needs to have the same patience to make it work to ones expectations). 4k is about resolution. In this respect, the BMPCC (and also the BMCC) is tricky, to put it mildly a hundredfold. You'd be well advised not to try and show off high resolution but instead to find out the limits. On the other hand I begin mistrusting the obvious virtues of cameras, be they dynamic range and 'gradability' or high resolution. Have you noticed that almost every BM or ML-RAW clip on the net looks like some aged slide from the sixties? Have you further noticed that the best looking skintones (in the 8-bit version we see on Youtube or Vimeo), after serious CC work match (but rarely surpass) EOS skintones? The sole feature of the GH4 (since it's not a noteworthy better crop factor or nicer colors) right now favors blossoms and bees. We are going to see better images with the GH4 once their owners manage to overcome the obvious virtue. Everyone stay loyal to his camera of choice, in good times and bad times, until it's time to part.

You mean you want to keep the motion information of every frame when not in slomo (i.e. when a time ramp starts at 100% speed)? At first, you have to interpret the 60p footage as 25p footage (2,4 x slomo during playback in 25p timeline). Now if you speed up the stuff, there are several methods: frame skipping, frame blending and what Adobe calls pixel motion (method to be chosen in some contextual menu, don't work with Adobe much). With the latter, not every second and then addiotionally every fortieth frame are skipped, resulting in jerky playback, also the said "cut throughs" are not crossfaded, resulting in occasionally visible ghosting (but the method can look good, depending on the evenness of the movements). Instead, the whole motion within 60 frames is analyzed and completely new 24 frames are being interpolated. Though obviously the most advanced method, it works best with non-solid, organic motifs. If you follow someone on a bicycle, for example, the software may have problems to distinguish the stiff frame changing perspective ever so slightly from frame to frame from autonomous motion. It may look similar to rolling shutter. Allegedly there are some differences in quality to the Twixtor plugin, but all in all, the problem persists. You should perform your own tests.

Supposed to be a satire, but don't think you can keep a secret. Maybe Google doesn't take advantage of all those personal informations yet (can you believe that?), but just the fact that it practically owns them is clearly wrong. Without much resistance, we allowed the Matrix to take control over us. Wake up, Neo!

Capitalism (of global dimensions) best prospers in democratic societies, something that Putin and the chinese 'communists' will soon learn. Their days are numbered, as are those of the islamists. The idea of freedom (does it really mean plurality and more chances for the individual, I mean the individual individual, who doesn't strive just to assmimilate himself?) is a tsunami that tears away every resistance on the long run. It's funny how we can believe that we live in free countries, as long as it is unthinkable to restrict the power of companies like Google, Apple asf. Not the american people rule the world, not the president, not the military. We are all highly indebted. The banks own us. We start to become obsolete.

Right. It's western capitalism, founded on a nation-independant european mindset that their 'culture' is the most advanced and that all others need to assimilate anyway. Don't you think the freedom has become rather relative insofar as access to information is (I am sure), er, guided by big companies (recently I accidentally downloaded 'Genieo' which disguised itself as a midi-editor software, my start-up homepage was then filled with all goods I ever remember to have seen reviews of or visited their homepages, it was a shock, took some time to get rid of it)? Isn't it like the conspiracy theorizer's nightmare, 'MindHead' (Bowfinger) acting in open daylight? Also, as I see it, you are free to say your irrelevant opinion on something, but you shall not act accordingly, challenging 'our' way of life.

Impressive. What makes the comparison questionable is that you didn't provide information about how the images were treated. Imo it would be more appropriate to say, well, friends, I tried my best to get out the most out of both shots with grading (I did not try, for instance, to match the BMPCC, which needs grading, to the RX100, which you apparently did), and you can have a quick impression how this might look through this vimeo upload. But for your own consideration, you can download this folder, containing short clips in the original codecs ...

You wrote "m43 sensor as well", but the BMPCC has a S16 sensor, in a way the worst of both worlds, neither fish nor fowl. Imo it's not too hard to rig the BMPCC to get rid of the shakes, but anyway, a m34 sensor (like that of the GH4) is a good compromise. The obvious solution if you want the best quality for a budget would be the 5D M3 with MLraw, but that is even less a camera 'for a quick job' then. There is no allrounder that is easy to handle, produces images that survive serious grading and shoots great stills as well.

You didn't write what camera you shot with. Some cameras have psf modes (progressive segmented frames), for instance to get 1080 25p (european standard framerate, which subjectively looks the same as 24p - or 23,98p -, but which doesn't cause flicker with 50Hz power frequency) in order to make the video compliant with the bluray standard, which includes 1080 50i, but not 25p. Also DVDs for PAL can contain 25p "as" 50i. The Canon XHA1 shot psf, the HBR mode of the GH2 was psf (in Europe at least), these are the two I know of. Be it as it may, the psf is actually nothing else but a flag. It's progressive video, but it tells the NLE that it isn't. Afaik, you shouldn't manually tell Premiere that you shot "p". If you did, it is like AKH said: Premiere will deinterlace it and introduce those artifacts. You can however, unless you want to burn a BD, export as progressive.

Reasonable enough. I didn't say you should transcode to an intermediate. I was interested if export speed improved with ProRes with heavy post manipulations, many layers, keyframes asf. Because, if the only advantage of ProRes (mpeg2) over any mpeg4 master file was to soothe some festival receptionists or snoot-nosed Apple fanboy clients, that would not be very important. You could as well say, I give you an mp4, transcode it yourself! If ProRes was not an alternative to something you usually would render uncompressed, than it really makes no sense.

Okay, please tell me if it allows you to create a desktop project with ProRes. And if so, if you feel a performance increase. ProRes actually is not about quality*, it's about ease-of-use, a classic intermediate. I recommend you don't transcode anything in advance, you just try to have ProRes as render & master codec but import native codecs. *at least not directly, only if you have to decrease playback quality for better realtime and judge your project (with effects, compositing and CC) based on that view.

You know, Andrew is right when he states that image quality (in ML raw!) is potentially much higher than that of BMPCC. In the meantime, I accompany my video buddies' efforts to build an ergonomic rig for his M3, specially for raw. I thought (and often wrote here) that it was naive fallacy to believe that the BMPCC was a pocket camera. I did try a lot of add-ons to overcome it's limitations. Now I have the most basic 'rig', consisting of a viewfinder and a breast pad, and I am no longer aware of a serious limitation, serious meaning inhibiting to get shots right. The so-called 'issues' with the BMPCC cease to be important when you have a closer look: 1-You can't delete clips in-camera Oh, I see. Have been doing this all the time with my former cameras. Always checked in between. Very good habit. 2-You can't erase a memory card. You have to go back home or find a mac store to do it when you're out. Let me think. I am on a journey, I have with me a camera that can format the card. Now as we all know, you sometimes think, hey, what I recorded yesterday at the Taj Mahal was actually not worth keeping. Shall I copy it to my laptop? No, delete the whole crap! EDIT, BTW: The clips on the card can be deleted (or copied) individually after viewing them on a laptop (admittedly, for raw it has to be a capable laptop). This often is a problem for common card structures (the BMPCC puts all videos in the root directory), especially with long ("spanned") clips or with certain NLEs (i.e. FCP X). 3-Camera stops recording without warning when the card is full, god knows when, Not entirely true. This depends on the card obviously. I repeatedly got a warning "CARD FULL" in the display (SanDisk Extreme Pro 64 GB), and the Pocket allowed me to finish the recording. Afterwards I found there was still over 1 GB space left. I admit this could be better, like "remaining minutes" or so. Since this is, according to the BM team, 'no problem', I'm looking forward to a firmware update within few years. 4-You need to carry a little bag full of batteries to shoot a normal day Yeah, indeed a little bag, with tiny and really cheap batteries. This again depends on the batteries you bought. Those with 1080mAh last way more than one 35 min (ProRes) or 18 min (raw) card, with usual interruptions (without shorter). You change the card, you change the battery, and everything remains small and lightweight. Now I foresee the argument, the cards are shat full too quickly. Well yes, who told you to leave AVCHD behind? EDIT A propos 'normal day': The BMPCC ist not the right camera if you wish to shoot very long takes without time in between to let the camera cool. Wedding and concert videographers beware! At room temperature and left continously "on" (somebody checked with 8 1,2V AA batteries and an HDMI monitor), after a little under an hour the body's temperature exceeds 108°F. Then the dead pixels (every BMPCC sensor has them) appear in the video (If you put the camera off or not then really doesn't matter, firstly because the camera uses battery power to transport the heat from the sensor to the metal body and secondly because the hotter it get's, the faster the battery is empty, and the camera will go off anyway). This is a good argument against an external battery solution. It also makes 'long play' solutions like external field recorders questionable. Can't tell if that's a problem for the 5D with MLraw, must be considered. If, which I didn't follow, the 5D can record noteworthy lengtths of raw video in the first place. 2-No waveforms or historgrams Are we talking about raw here? This is indeed a minor issue for ProRes, but not for raw. Afaik there is also an ETTR assistance in ML, and that's the way to expose for raw (unless you denoise in post, which will give your final image a video look, no matter with which tool you denoise). Works flawlessly with 100% zebra with the Pocket. 3-No audio. Period. This camera can not record audio. I see. Whereas the 5D can. Internally. I tried several ways to record directly to the video file, and it's all a matter of the right pre-amp settings. Whenever you record dialog, it's better anyway to record sound externally. Period. 4-No live WB adjustment. Are we talking about raw here? 5-No variable framerate above 30p True. 6-Very limited ISO range, poor sensitivity Compared to what? Again: Are we talking about raw here? And are we talking about a cinema camera or an ENG camcorder suitable for existing light? You don't ETTR enough with the 5D M3, and you will have ugly noise where you never expected it ... 7-Too big a format for the SD media (ProRes HQ only, no LT) or Cinema DNG (2:1) The latter it has, and the files are much smaller than MLraw. On the standard PR and LT I agree, but this is a skinflint's complaint. Arri currently charges a few thousand bucks for it's Amira to step up from the standard PR to HQ. BM seem to have abandoned the support, otherwise I think the implemenation would have been 'no problem'. Again, you want to compare this to 5D with MLraw. I doubt very much, if austerity was a virtue for raw at all, that the comparison would be favorable for the Canon. 8-No external controls for adjusting settings, only through menus Which settings are we talking about? The (practically) non-existent WB? The useless audio levels? The switch between ProRes and raw? It does have aperture control, focus assistant. It has no fancy picture styles, for a reason. Add metadata to a clip? Aww, come on! 9-Camera adjusts Iris for you everytime you turn it on For a reason. This only happens with MFT system lenses, and the automatic iris E'sTTR in raw. You could never be more exact manually. Why? Because you only have the friggin' zebra. Are we talking about raw here? I will add THE killer argument against the BMPCC, something that has not yet been solved looong ago by the users: Moire. Particularly in raw, and particularly in comparison to the M3 with MLraw. From a Wolfscrow newsletter I found this video, demonstrating a rather sophisticated rig: I won't buy all that stuff (sold my cage recently), but I am willing to bet that there never will be a more compact and more ergonomic rig for the 5DM3.

Missed the whole series. Judging by this trailer ... it has some kind of music video aesthetics. The comments below seem to confirm my suspicion that it may have an interesting visual concept, but with some scripts off the shelf. Just imagine if some of the masters from the past had these post production opportunities! But then again, properly restored, their work holds up well. They had to do everything with lighting and set design / wardrobe / make up. Maybe the whole grading affair is more the poor man's chance to enhance otherwise boring images.

Now as I mentioned, the ugly headphones were designed specially for the BMPCC. This is what they write about the D800: As I understood (but I may err), this is the situation when the microphone jack delivers power for the Earworm, but will distort the sound when not dialed down to "0". The characteristic of this device is to capture quiet sounds as quiet (the sculpture across the brook with the birds, with little noise, s/n 82dB @ 1kHz) and loud sounds as loud (the jazz band, without clipping up to 138dB maximum input). Please contact the inventor to learn if the traditional little blue headsets suffice for your camera, I would've prefered them too, but Winne needed place to build in an additional miniature preamp for the Pocket that's powered by the mic jack.

Winne wrote, if your GH4 has a mic in jack that provides 1-10 V power (as most do), the classic Earworm suffices (then the input level needs to be set to "0"). But hurry, he has only few items in stock and plans higher pricing.

I will link the thread with your question to Winne. You could be lucky, and the original Earworm, which is much nicer and cheaper, works with the GH4: Here's the site. He says, orders from outside the EU need to contact him in advance.

That's what Stu Maschwitz calls 'memory colors'. Among them the blue of the sky, the green of the grass, a yellow banana and (no joke, but typical for Maschwitz) coffee beans. Usually, if the skin tones hit the 'skin line' (vectorscope), the others fall into place. Here is a catch: One can easily be tempted to achieve a certain look at this early stage, making the image appear darker (usually also described as 'more punchy') or brighter (more hazy, as if the strength of the light shines through). If the shot was an early morning landscape, backlit by the sun, you may very well have recorded neither blacks nor whites. If you shot in a night club, you can have crushed shadows and clipped lights, and then all that remains to alter the exposure in post are the mids. It also depends - of course - if you are correcting in 8-10-12 or 14-bit color depth. 8-bit forbids a too much stretched ratio between input and output levels (spreading the midtones). But one rule stays valid whatever your footage is: Expose correctly, capture everything. See point 5. This will accentuate the color contrast and make the image look sharper. However, if not done with care and taste, it dominates the overall look and limits one's options. I don't consider it primary CC. The point is to make an informed decision how your finished image should look. This should be clear already during shooting. With all (also mine) 'test shots' people have no clear *concept* of that. But of course, to just correct and bring to full impact a well-shot, powerful scene is preferable to apply some exchangeable, stylish look for the sake of it. Not sure about this one. The goal rarely is an HDR-look, that's right. But the question is, when to decide how much I take away, because of: Yes, I've seen Erin Brokovich a few days ago. Heavily graded film. Many highlights (windows when shot from inside) appeared to have been allowed to clip unashamedly. But recorded clipping mostly consists of unbalanced color information. Somewhat hard to color those areas convincingly, at least with my very limited experience. The downside of grading within an NLE is that the tools are accessible only as effects. So you have to load them individually and apply them either to one clip at a time or to a clip selection. Both not very practical. In a special CC software, you grade the whole sequence, the tools always visible (sometimes you have to one-click a tab, that's the extend of it), you jump from clip to clip, and you have a better overview of how your manipulations affect a scene in it's entirety.

I made some zoom shots with the Sigma. What I found confirmed is that it is not really parfocal (focus changes slightly, at least or most noticeable with open aperture). There definitely was no change in brightness. However, although the zoom is rather smooth for a manual zoom ring, the 2x zoom range imo doesn't justify to use it as a transition (to point to a detail or to reveal it's surroundings).

Since a few years, a little device called 'Ohrwurm' (earworm) had quite a good reputation among nature documentary filmers because it could capture ambient sound so well. It looks like a hearing apparatus. Unfortunately, for the BMPCC the very small design didn't work. But inventor Wolfgang Winne constructed a new mic, by adding highly sensitive capsules to very cheap headphones. He tested it with his own BMPCC: I bought one (no good name for it yet, said Winne) and went to the river yesterday to make some test shots myself: My quick review (as you can see and hear, I'm no expert, this is just about usability, others may comment on audio quality): + it's comfortable to wear + it isolates you perfectly + no extra batteries needed + the recorded sound is very dynamic with no noise + you very rarely capture your own breath or clicking sounds from camera operation or so + with ~200 bucks, it's about as expensive as any other solution that records directly to the video file - the fur looks a little silly ... - ... and doesn't stop wind noises completely - the headphones are noisy. During recording, you don't get the right impression. Winne says it's the Pocket's fault. This is how it looks:

Didn't see that show. Could be deconstructivistic intention. Generally, if someone is on the extreme left, looking to the right, he sees 'into the future'. Accordingly, if he sits on the right side and looks left (into an empty frame), he thinks about 'the past'. In WW2, german newsreel cameramen had the order from the ministry of propaganda, that tanks and troops had always to move from left to right, when they were meant to conquer Russia. And from right to left for France ... There is something arbitrary about the rule that one needs to be able to follow. We follow where the master leads us? Life is chaotic, important moments are fragments, thoughts jump and mix with impulses from outside. An excuse for Nolan? Does he need excuses? Go, recut the sequence so it becomes comprehensible, will it make any more sense then or become 'better'? Decades ago someone wrote a rant about Back To The Future. How the episodes didn't fit if you tried and recut them in chronological order. I was more shocked than when I learned there was no Santa Claus.

It's not my video. Good exposure (please, BM, give us a waveform or at least a histogram!) and roughly appropriate WB in ProRes can let you grade the image to match a raw comparison shot. With much less moire and noise. Among other characteristics the raw video appears much sharper. If you read out the metadata, you see that 25% sharpening is already added by default. Take that out, and it becomes as soft as ProRes (but colored moire remains visible). Add 2,5% sharpening to ProRes (depends on the method, resp. the software, with 'unsharp mask' less is needed), and the same shots become indistinguishable. There is slightly greater DR in raw (given you E'dTTR), but are the problems worth it? The clip above was probably shot in ProRes, that's what I suspect always when the image has that undersaturated look. As far as the reported softness of the 14mm is concerned, that's not true, compared to other MFT system lenses, see here. It just appears to be not sharp enough to resolve moire patterns, which makes it ideal, at least in my opinion.

I received a newsletter from Canon, regarding the popular 5D M3 (my awkward translation from german): 4:2:0 ist not responsible for banding ... EDIT: According to a friend, this firmware was already released, but there were doubts if it really was 4:2:2. The video engineers once decided to compress to 4:2:0, because human vision can't tell the difference anyway.

A lot of videographers had the delusion, not corrected (for obvious reasons) by companies like Atomos, that 'clean HDMI out' would make their DSLRs cinema cameras. Although HDMI out is always 'uncompressed', it stays 8-bit 4:2:0 if that's what the camera records, even if the external recorder writes 10-bit as ProRes or DNxHD. Everytime one is confronted with a shortcoming, he feels the urge to overcome it. With the issue of banding in the skies (and similar 'motifs'), workarounds are discussed. We've seen a lot of okay-looking skies from DSLRs. But with MagicLanterns raw and split-screen comparisons, even on vimeo, one could clearly see that a banding-free sky is not the sky. Avoid the sky? Live with some minor banding (saw it in quite some TV docs that were otherwise nicely photographed)? Downscale and dither 4k? Experiment and find a proper picture profile and proper exposure? Grade conservatively? Shoot raw? Genuine ProRes? We are getting more demanding and critical and tend to see weaknesses clearer than strengths. The receipt to be unhappy. But of course our videos are getting technically better, the industry says thanks.

As Andrew wrote, there is also compression to consider. Actually all sensors of our 8-bit cameras take in raw. Afterwords they have to somehow map all the values anew, which is called, if I recall correctly, quantization. It's more or less applying a curve that transforms all values needed to represent the same image with the same dynamic range to 256 (or less) luma steps, same with color. With different picture profiles, you will get more or less banding in a sky (if there is a preset like 'landscape'), but might lose values in skin tones ('portrait') or in the dark areas ('night'). Grading for 8-bit means respecting what's there and what's not there. If it was there, because there were sufficient values or because you recorded in a better codec in the first place, you can go ahead and reduce information further (which is what grading is about, in a way). It is better to try to get most of things right in-camera if you deal with 8-bit. The last two days, me and my video buddy had the chance to borrow a C300 (free) from Canon. It records in 8-bit 4:2:2, and from what we saw it's real 1920, as close as you can get to the real resolution. I can't believe the image could look any better with downscaled 4k. To overcome the 8-bit limitation for grading, it has a 'cinema-lock' feature, meaning C-log. It's not that flat as i.e. Technicolor picture style, it almost looks okay as it is, you may just say it's too neutral. We put the stuff in Resolve. What we found was: It's not raw. You have to expose correctly, there also should be a correct WB. Other than that, it holds up well. We had a lot of skies, but apparently no banding. My conclusion: You can do well with 8-bit for 8-bit, it's just not as 'forgiving'.

I wonder if anybody sees the writing on the wall when he praises the advantages of shrinking a 4k-8-bit image to HD for an 8-bit player/display. What we had to expect if we watched an HD image with 10-bit color depth on a 10-bit HD monitor was not so much *better colors*, but rather *less artifacts*. Therefore, if we watched a 4k-8-bit 4:2:0 image on a 4k-monitor (no matter if 8-bit or 10-bit) we'd see way more banding. Banding, again, is just the highly visible result of having too few values to spread a soft grade evenly over a big space. It's the tip of the iceberg, indicating that there is unsufficient color depth and only a quarter of color resolution. One mustn't ignore relative size. The higher the spatial luma resolution, the more color information you need. Remember: Digital cinema once was 1k (i.e. Star Wars Ep.I), it was just enlarged. You couldn't appreciate individual sand grains on Tatooine, but it still would look better than 4k 10-bit 4:2:0, side by side. In other words: Yes, 4k may look better than 8-bit HD, but 10-bit HD may very well look better than 8-bit 4k. I'd like to add this to the 4k hype to stress the importance of having better color depth, unless you are content with '4k for better HD'.