Leaderboard

  I decided to put the Sony A7S and Atomos Shogun to use shooting some of the snow over Christmas. Shot in Derbyshire (North of England) in 4K with family and at Chatsworth House. Music by singer-songwriter Anne Haight of Berlin (hear more on SoundCloud here) Read the full article

Happy New Year!!! I'm finally happy with this setup. Not perfect but it works so far. I'll try and shoot some tests this weekend.

Thanks for this review. I had pre-ordered the Shogun when it was first announced, but canceled it when the delays started cascading. Based on your (and other) reviews, this version of the product clearly has applications -- but it seems Atomos missed more than one target in terms of the design, and I'd much rather wait for v.2. Unfortunately, I can't see them delivering a substantially improved revision in the near term, as it would completely alienate v.1 owners. And, by then, there will certainly be more cameras recording 4k internally. Finally, I'd like to say that "Ethic Statements" aside, I find your reviews to be honest with absolutely no pandering.

​Here's a comparison to the other anamorphics. Size is an issue, as is the massive weight. I normally shoot with a rig anyway, but this thing is easily noticed.

Maybe the winner, is in the centre of the picture..The GH4 ..easy to use..great 1080p, 4K inside, APS-C option (speed booster), light weight, weather sealed , battery life, al lot of users..new firmware updates will come and also very very nice..it's only 1500,- I have seen so much good footage coming from a GH4.. that we do not have to call it or lets say that it doesn't have to be 'video like'!

​I've tried putting old Pentax lenses on and get the same results. I don't think its lens related. Easiest way to test this is point at a window and overexpose until zebras are filling the frame. Go into movie preview mode. Whip down to the floor. What I'm seeing is that the floor "settles" into a darker exposure over the course of some frames.

1D C A7S Same exposure

Next test. This one is a bit clearer I think. Click to enlarge them. The shot contains a GH4 to represents the blacks / lows and out of the window at the top I've cropped out the tree line / sky for the highs. The shot of the GH4 on the 1D C shot is usable, the A7S isn't, or at least needs cleaning up. There's also moire and aliasing on the lens rings. You get these noisy blacks at ISO 3200 at the bottom of the S-LOG file. What I do is to of course apply a curve to the file to get contrast back, crush the noisy blacks away, get colour looking right and tonality spot on. I have always crushed these parts of the image away by 1-2 stops because that's what you're supposed to do with LOG. Those blacks look awful if you keep them as-is. The blacks are so cinematic and milky clean on the 1D C I can keep them if I want to. If you clean up the noise in the A7S blacks with noise reduction software, the image overall will still look bad. So they need to go. Nobody can say looking at the shot above that the A7S looks better than the 1D C. I am not bringing bias into it, it's science pure and simple. The tree line out of the window represents the highlights. Here the A7S is doing a hard clip to white above the tree branches and the 1D C goes beyond the trees to maintain some of the bright clouds. Well, as bright as the sky gets in January in England anyway!! Subtle difference. To find out how many stops DR each give I need a chart and I am not a chart tester. I am certain the 1D C is giving me more usable dynamic range though.

​I think your examples clearly shows the Sony advantage. In my opinion there are even better areas for comparison like the strap lug and the front control dial on the Nikon. If you can show this advantage on a small web jpeg, the real world advantages on the original 4k files with proper pp could be significant. Maybe Andrew should make the original clips available for download to demonstrate this. On the other hand I suspect that this discussion is not technical anymore.

Happy new year to you too, and to all EOSHD forum users

​If you're trying to compare source files to see differences, and then go on to complain that the 0-255 file has deeper blacks than the 16(32?)-255 file, that doesn't sound like a rational or fair comparison to me. Properly matching your source levels to your output levels isn't grading. And getting your black and white points to match when you are specifically comparing DR and talking about "crushing blacks" would seem to me like the first thing you would want to do. Within the a7s itself, the cine1 gamma goes to superwhites and the cine2 gamma doesn't. If you didn't know/understand this and just looked at them both to "see differences" and declared "the cine1 highlights are clipped, this needs to be fixed!", or "the cine2 highlights are washed out, this needs to be fixed", then that would be incorrect. As far as noise goes, there is clearly more noise reduction going on in the 1dc (regardless of what the camera is telling you) than the a7s which appears to be applying none or little. I personally would prefer to apply some chroma NR on the a7s file in post (just like I do to all my RAW photos) and keep some grain-like luma noise and the extra detail, than have the baked in NR and loss of details on the 1dc. And if I wanted to do luma NR on tha a7s file as well, then I have that choice I can make in post with the trade-off in details. It's starting to sound like a case of post purchase rationalisation for the 1dc at this point. Having said that, I think we can all agree that colours on the a7s are a pain in the ass!

I still think for us jack-of-all-trades (masters of none) shooting with >$1k cams, it's okay to put accurate skin tone a bit down on the list of priorities. If we're close, that's fine. Should we put skin tone above production design, writing, directing, or editing for example? I'd rather take an opportunity to implement an incredible story in a slightly technical flawed way than to implement a flawed idea in a technically incredible way. That's just my approach at the level I'm at. As mentioned, some folks have the luxury or inclination to focus on the intricacies. You know that ridiculous saying, "It's above my pay grade?" Well...

If directors want to go to HFR I think they have to be extremely disciplined how they move their camera in this visual space. I feel HFR can be exploited, but in order to do so you have to appreciate how humans utilize their own eyesight. Even though our eyesight has a biological "shutter speed" well above 48fps, we also FOCUS our attention on specific objects in our personal field of view. All else, according to our brain, falls out of focus, so to speak. I really think that we're evolutionarily wired to appreciate a motion picture frame with a strong focusable center of interest. Take that away and you're asking for trouble. AKA: This Hobbit nonsense. Ultimately it's just poorly directed. The sensational motion and hyper-kinetictivity is so incredibly abnormal it's a visual liability. Again, any director going this route better be extremely wary of the technology they wield. Jackson just went for it all as he had that capability at his disposal...and it bit him in the ass. Here's something to think about, do the very limitations of 24fps cinema supports our visual experience? It may be counterintuitive to think so, but consider how a film camera has to stay below a certain pan speed to avoid judder. Couldn't that visual constraint actually reinforce our biological expectations? Obviously our sight shifts much quicker than a slow 24fps pan, right? But, we also "defocus" our perception as we rapidly shift our field of view. Perhaps when you DON'T do that in cinema it's an uncomfortable assault on our senses. Longer takes, slower camera movement, no rapid editing, and a bit of shallow DOF. This combined with HFR might make a more tolerable experience. The whole biological and psychological human visuality needs to be more considered. It's going to take a serious re-think of the cinematography approach, I believe. That said, I think maybe a good compromise would be to shoot 48fps with a 0º shutter. You'd get the benefit of object motion blur combined with a high frame rate. Also, you'd create a source file that would be easily down converted into the traditional medium for those that prefer it. I'd encourage any of you PAL shooters to give that set-up a go with 50fps and a 50 shutter. Then convert to 25fps and see how it plays.

Happy new Year!

Happy new Year to all the FM users

It's HFR 3D that's the problem not 4K. It destroys the mystery of the image and makes it seem real. Still depends on the subject matter though! Reality is good for showing reality not so much for showing a wizard with makeup on.

The Metabones Smart EF adapters for Micro Four Thirds support Canon IS, it works 100% on the GH3 and GH4. I have the 35mm F2.0 IS and stabilisation is great on the GH4 with it.

​The poor Nikon didn't make it to the 4K party

I love that you have the nikon there, as a lab animal

Fuzzy, sorry if my posts are coming off pedantic. You're making a case that skin tones aren't that important. I'm providing many facts (which you can verify for yourself) to make the case that skin tone color is indeed the most import element in your thread "the skin tone holy grail". Back when the 5D3 was released the big issue was it was too soft. I showed that with post sharpening it was pretty good (lots of folks disagreed; many felt post-sharpening was invalid and the camera should be sharp straight from the camera). I spent a lot of time looking for a camera that provided better native resolution and went with the FS700 when the SpeedBooster came out. After many projects with feedback from many people, it was clear that people preferred the look of the Canon over the Sony. There were indeed comments on the 5D3 being too soft for some shots, but overall people love the look of the 5D3 over the FS700+SB. The FS700 still provides value with up to 240fps 1080p (ish) slomo. As I read about lots of different cameras, the recurring pattern for what people love the most is skin tones. We are emotional beings and we evolved color vision, at least in part, to read emotion through skin tone color. Researching the best cameras, I found the same pattern: they produce skin tones which evoke a positive emotional response: "I like it." "I love it!". In narrative and commercials too, color sets the mood, helps tell the story, and conveys emotion at a deep level. Hi Jonesy- Blackmagic cameras have indeed gotten good reports for skin tones, though I haven't used them. A big part of the filmic look is highlight behavior of the camera/sensor and even more important: lighting for narrative (emotion). I have started paying very careful attention to scenes that look really 'filmic'. Even that ARRI Alexa and Red Dragon can look 'videoish' with bright lighting and resulting harsh highlights. Combine such lighting with an oversharp image and you've got the video look. Notice how often narrative scenes have very little lighting, with the most lighting on the face and eyes. The challenge I've had with Sony cameras, especially the FS700, is that after setting proper WB and exposure, even tweaking WB in post (including 3-way), parts of the face can look good/correct, but other parts are too yellow/orange/green/blue and don't look right. So now you have to do secondary color correction for those areas. Time consuming. The final results tend not to look as good as Canon or Panasonic. The A7S is better, but still requires more work vs. the 5D3 or GH4. For video, in my experience Panasonic cameras work pretty well for skin tones, to the point that I didn't worry about them. They just worked. For stills, in my experience Canon looks much better than Panasonic. It wasn't until using Sony cameras and getting feedback from others regarding skin tones that I started looking at skin tones in more detail. I was curious why so many top productions used ARRI cameras when it was only 2.8K/3.4K max resolution. The result of much reading was that ARRI provides the best color and skin tones. How do we know it's the best? We look at what people use when cost is not an issue, which cameras were used in the top grossing/award-winning films, etc. This lead to learning more about color and skin tones. So while many cameras can look great, cameras which look great with the least effort are the most useful and cost effective. It's also why Canon drives Andrew crazy releasing cameras with very limited features which outsell all the other brands with more features. Canon hasn't had to compete on features so much as they compete mostly on (skin tone) color.

Yes, FuzzyNormal, I wouldn't say ever had a "problem" with skin tones until I shot some RAW with a Canon 50D. That is, most video had a "video" look which I felt I was stuck with. But RAW changed everything for me. Then I went to an EOS-M, then a BMPCC and feel that both camera deliver great looks. I never get tired of RAW (just the workflow). Anyway, I've always wanted the nice crisp image, small file sizes, and video ergonomics of Panasonic cameras. What happened is that I wanted the high dynamic range, flat look, of my BMPCC with the GH4. So I used Cine-D and dialed down the contrast and changed the hue TOO MUCH. What I didn't realize is that the low-contrast look I was getting was AT THE EXPENSE of color, which meant the skin tones went to crap. So I can understand how this problem seems silly to you Fuzzy because you don't shoot RAW, far as I know, and you haven't pushed the GH4 too far as I did (and should have known better). The good thing for me, is that it has awaken me to much of the stuff JCS has been writing about for a long time here. I agree with him, people are biologically sensitive to good skin tone. If you're doing a very busy video, with many cuts, and high contrast, etc., you don't notice. But if you're going for a natural look it ain't easy. If you take a family member, and sit them down for an "interview" where the person has to look at their face for minutes on end I think you'll go through the same thing as other people--the image may not do justice to the person in front of you.

After shooting and editing extensively with the 5D3 (H.264 + RAW), Sony FS700, GH4, and A7S, based not only on my personal opinion, but on feedback from actors/models/clients, cameras which produce better skin tones are preferred. I found this out by shooting the same scenes with multiple cameras and reviewing the results with others. Doing research online I found that skin tones were the single most important element for cameras used to make money. Resolution, frame rates, viewfinders, ergonomics, etc., are all very important too, but skin tones are number one. Skin tones affect emotion, and emotion is used to tell a story or sell a product. What cameras provide the best skin tones? ARRI and Canon. What cameras are used the most professionally? ARRI and Canon. What cameras are used the most in Oscar winning films (last few years)? ARRI. What DSLR was used most in feature films? Canon (5D, 7D, 1D): http://shotonwhat.com/?s=5D. Why did the C100/C300 far outsell/outrent the FS700, even though the FS700 has way more features (and can even look full frame with a SpeedBooster)? C100/C300 produce better skin tones with less effort. In the end, it's possible to get similar, sometimes even better (rare/unusual lighting conditions) skin tones from the GH4 and A7S vs. the 5D3, however on average, the 5D3 requires a lot less time and work. How do we know when skin tones are better, when it's so subjective? Shoot the same scene with multiple cameras then show the results to multiple people for feedback. Some DPs do a ton of testing to figure this out before shooting a feature: http://www.hurlbutvisuals.com/blog/2013/10/arri-alexa-vs-canon-c500/

​I like people who ask good questions [Goes off to check...] Yes it's a good fit actually. Only the tiniest little bit of corner shading at 1.3x in 4K. The 1.85:1 crop gets rid of most of it. Easily fixable in post for the shots where it shows up (i.e. on a plain wall or sky). There's no hard vignetting. Any shading is pretty much gone by 24mm. Looks great!

not sure it will look all that great without a lens though :)