tellure

Here's the 100% crop comparison of that head portion of the image Pavel posted above (A7R2 S35 on the left, NX1 on the right). It's definitely not as big a difference as the playing cards above, but I'd say the A7R2 in S35 looks sharper and more detailed.. e.g. the hairs on the top of his head and the detail in the eyes. Of course the number scale on top is quite different. Definitely agree that lens plays a huge role.. in their NX1 review DPReview says they used the 16-50 OIS since it provided the best quality: "We found that within the lens range that keeps us at the proper shooting distance, the Samsung 16-50 OIS provides the best overall performance (corner to corner). That's high praise for a zoom lens, but most systems' primes offer better performance. Hence, our comparison as it is is not the final word in sharpness for NX1 files - it's entirely plausible that a better lens for the NX system will be available in the future." I could certainly understand some people wanting to see test results with the exact same lens via an adapter. Personally this test is still pretty valid for me since I like using the best native lenses on the system so I can get the full range of image stabilization and autofocus capabilities.

Weird, I had the opposite reaction when I saw the NX1 4K framegrabs compared to the A7R2.. the NX1 looked softer and more compressed to me (especially around the grey-on-grey "DPREVIEW" text). I downloaded the full-size JPG's DPReview provided and did some A/B.. what am I missing here? Left is A7R2 Super35, right is NX1.. I see more softness and a lot more compression artifacts on the NX1. These are saved at JPG quality 10 in photoshop and hosted on my own site to avoid re-compression by the forum software BTW. Here are the playing cards in the upper right..

I'm referring to the % zoom ratios you get in Photoshop or Premiere. When I look at a 4k framegrab in Photoshop full-screen on my 1900x1200 monitor, Photoshop shows the zoom ratio at 50%.

I've been pretty happy with the resolution of my A7RII coming from the HD world, but I don't have a 4K monitor so I've basically been viewing all the footage scaled down 50%. Today I decided to pull out some screengrabs and I have to say it does look soft at 100% to me. The screengrab this was taken from was shot with the Sony / Zeiss 24mm 1.8 APS-C lens in Super35 mode (effective focal length 35mm). Don't recall the exact settings from this shot but I think it was 1/60 at 25p. 4K 100Mbps of course. Some of the traffic was moving of so I was looking mainly at static elements like the "Speed Limit 45" sign in the crop, which looks quite soft to me. And it's fairly close to the center of the frame too so I'm guessing it's not corner softness. The "End Road Work" sign next to it is even worse. Maybe XAVCS 100Mbps is still too high compression / low bitrate to keep up with the fidelity needed for 4K? Sharpening these in Photoshop as a test does improve things, but it does feel like it should be sharper out of the box. Incidentally, when I first attached these images they looked worse on the site than the jpg's did on my computer.. I suspect the forum software applies some pretty aggressive compression to keep attachments small. So I put them on a server and linked to the jpg's directly, which preserved the same image quality I'm seeing.

Like Andrew said this problem exists on the A7RII as well. It was one of the first issues I ran into when I went down to the beach with my new A7RII to do some test shots and found I couldn't even see what I was shooting after I hit record. It's terrible Sony doesn't provide an option for this. Like the OP I shoot very shot clips and I would far rather risk overheating than not be able to frame a shot. And there are definitely tons of times where you cannot use the EVF. I bought a gimbal so I can get some nice stabilized shots for example.. no way to EVF with that. Here's a video I made showing off the screen darkening if anyone wants to see it in action:

I agree Watchtower of Turkey doesn't have a story in the classic sense, but I found it just as powerful in its own right. Is a Rothko inferior to a Michelangelo because the latter is figurative and chooses to tell a story? They have different aims and each succeeds in its own way. Looking at the medium of moving images as more than cinema / storytelling, to me this short explores some aesthetic qualities of the medium extremely well. Thanks for posting the question OP, and thanks to all the other members for the excellent explanations and links; this is a great thread!

Thanks for your first impressions article Andrew, super informative as always. Quick question here from an A7R2 owner: does the A7S2 have the same issue the R2 has where the screen, when set to the brightest setting ("Sunny Weather"), will darken when switched to movie mode while in 4K (or when starting a 4K recording while in stills mode)? This one is proving to be a killer for me when shooting in bright sunlight. Looking forward to your Cine4 PP settings, I'd love to be able to get Canon colors out of these Sony sensors. It's only after coming here that I realized what I was missing on my Sony's compared to my old 5D colors.

Sorry, my mistake. Transcoding 60p to 24p looks fine (just tried it). I think I had been doing some 30p --> 24p transcoding before which generated a bunch of artifacts. Thanks, and sorry for the confusion.

One thing I wish the A7R2 had was a 48p 1080 recording option. Let me elaborate.. Normally I want to shoot at 25p in 4K for a cinematic look. But occasionally if I want to shoot something in slo-mo I'll go down to 1080p so I can capture a clip at 50p, which allows me to slow it down to half speed and have the framerate match the rest of my footage. However tonight I ran into a problem with this approach. I live in North America (LA) and thus our fluorescent lights run at 60Hz. When shooting inside my friend's house I saw a super noticeable strobing effect on all the footage I shot in his kitchen where the fluorescent lights are. The solution of course is for me to shoot in NTSC at 24p, which would be fine for 4K, but whenever I want to shoot a slo-mo clip I'm now shooting at 60p, which does not divide into 24 evenly, creating some noticeable transcoding artifacts when you try to output the footage at 24p. Since the camera can obviously shoot at 50p or 60p in 1080, why not give us the option of shooting at 48p so we have even multiplier of 24? Apologies if I missing some obvious solution to this problem.. help me out here

I don't have an NX-1 but here's my day-1 review of the Pilotfly H1+ with the A7R2 (scroll to the bottom of the thread for an updated video): http://www.eoshd.com/comments/topic/9232-pilotfly-h1/?do=findComment&comment=107768

It's definitely a great camera, I'm loving my A7R2 a lot but there's one issue which is particularly troublesome for shooting video that I wish was getting more coverage from these big-name reviewers: the screen (and EVF) darkens substantially when shooting in 4K (but not in 1080p). Presumably this is done to reduce the heat, but it makes shooting in bright sunlight extremely difficult. I documented the issue here: https://www.youtube.com/watch?v=fAHGsw_ca9A&feature=youtu.be Note that test was done in Manual but the issue also occurs when in Movie Mode (the screen darkens immediately when you switch to Movie Mode). Also note that this is most noticeable when the screen brightness is set to the brightest setting (which they call "Sunny Weather"), which is of course what you'd want to use when shooting in bright conditions..

I updated the settings to the ones recommended by Pilotfly for the A7R2 + 55mm 1.8 FE combo and also tried some smoother walker technique. Seemed to smooth out the bobbing a fair bit. Running still has a few jitters though.. think I might need to up the pitch power. Here's the updated walking + running test (second part with tracking a moving subject is the same):

That shot is pretty easy to achieve with any of the 3-axis gimbals. I took one like that on my first day testing my new Pilotfly H1+ (first day review here): https://youtu.be/Q0k9QmOkYdg?t=37 5-axis sounds great. I was definitely seeing noticeable up/down bob in my first day tests of the Pilotfly, although I hope to smooth it out with some better walking technique and tweaks to the settings. Since I'm looking for small size and weight for traveling personally my ideal handheld stabilizer is a 4-axis with the size and weight of a Pilotfly.

Lens used was the Zeiss 55mm 1.8

I recently spent some time researching the most compact and light gimbal stabilizers for my A7R2. The top choices seem to be (these are all pistol-grip stabilizers since everything that's not is a fair bit larger by necessity): CAME-TV Single Pilotfly H1+ Nebula 4000 I bought a Pilotfly H1+, just got it a couple days ago.. posted my first-day results here on the forum. Here are some video comparisons with footage for more info (mostly Pilotfly since that's what I decided to buy): Dave Dugdale's Pilotfly vs. CAME-TV Single review CheesyCam Pilotfly vs. Nebula vs. CAME-TV Single comparison Viktor Studios Pilotfly review (in spanish with English subs) GH4 user w/ Pilotfly - unboxing / setup / test footage

Just got my Pilotfly H1+ yesterday, here's a sample of some basic walking / running / tracking moving subject footage. Shot on A7R2, 4K @ 25p 1/60, creative style neutral, picture profile off, DRO lv4, no post processing. Definitely seeing some very noticeable bob when walking, and some jitter at the end of my running test, but I'm hoping I can iron this out with some tweaks to the settings. The A7R2 + 55 1.8 is about 75% of the rated weight limit of 1.2kg so I might need to increase the power on the motors.. the above was shot using the default profile 1 that came with the stabilizer. There's a lot to play with in the SimpleBGC software so I'm hoping I can eliminate the bob after some fine-tuning. Overall Pros: - Does a good job stabilizing; pans are also nice and smooth, something I've never been able to do well hand-held - Lots of tweaker controls; pretty sure I can improve my first day's results with some adjustments - Bluetooth support for the SimpleBGC mobile app, so you can tweak every setting and run calibrations while in the field - Inverted mode is very convenient, also feels more stable to hold the rig when up-side down as gravity is helping - Nice and compact, doesn't look too crazy when in public settings - Fairly lightweight for such a powerful device (780g) Overall Cons: - Pretty confusing to set up initially, especially when you get into the software. For example there are 5 different profiles included but it wasn't until I asked on their Facebook group that I fully understood what they did. There's a lot I'm still figuring out about this device. - Battery life - unlike the CAME-TV which has encoders on every motor that reduce battery consumption, the Pilotfly is only rated at 1-2 hrs of battery (depending on size of camera/lens). I bought an external battery charger with DC output so I can charge the Pilotfly (and my camera / phone / etc.) in the field. So far that seems to be working but of course it adds another pound or so to your bag. - Re-balancing for different lenses can be slow since you need to use a mini allen key and re-set a bunch of screws; thankfully my current lenses are somewhat close in weight right now so I can let the motors just compensate for the slightly un-balanced camera (at the cost of even shorter battery life presumably). The Victor studios video linked below shows how if you really need to though you can crank up the motor strength really high and use a very unbalanced lens in a pinch. - Needs some convenience features, like a way to lock the arms down when you're walking around. Also have to be very careful when rotating the arms with the device turned off around as the camera and arms will flop around and it's easy to bang your lens against the screws. I was trying to switch out of inverted mode and I somehow got into a spot where one of the screws was within half a centimeter of hitting the front element of my Zeiss 55mm Overall it's still a pretty great device and I'm pretty stoked at the kinds of shots I can do with it. I had a cheap old manual counterweight gimbal before (Hague MMC) and it wasn't bad but for almost the same weight this is pretty awesome. Videos I found useful when researching / setting up the Pilotfly: Dave Dugdale's Pilotfly vs. CAME-TV Single review CheesyCam Pilotfly vs. Nebula vs. CAME-TV Single comparison Viktor Studios Pilotfly review (in spanish with English subs) Pilotfly's own tutorial on how to switch to inverted mode, swap pitch and roll GH4 user unboxing / setup / test footage

Here's a short vacation video I shot on my a6000. Lenses used were the Zeiss 24mm 1.8, Sony 50mm 1.8, and Sony 10-18mm F4. Coming from the NEX-6 I appreciated the a6000's autofocus speed and accuracy quite a bit. I also used the face detection focus tracking a couple times and it worked pretty well. Still wishing we could record in XAVC / slog2 though like the A7s.

Here's a short (2 min) vacation video I shot on the A6000 in Jasper, Alberta. Lenses used were the Sony 24mm 1.8, Sony 10-18mm f/4, and Sony 50mm 1.8 OSS. I found the quality pretty decent for most 'normal' scenes with a central subject against a DOF-blurred background, but the codec showed a fair bit of compression artifacts when there was a lot of high frequency detail on screen that was moving rapidly, like forest scenes. Looking forward to either upgrading to the A7s or a future APS-C e-mount camera with XAVC @ 50Mbps.

I have an A6000 and although I find the AVCHD codec quality is very good in most normal scenes when you have a central subject, in other shots when there is a ton of high-frequency information AND a lot of motion in the scene sometimes the compression can't keep up. Here are some examples: Shot #1 - central subject, light camera and subject motion, not a ton of high-frequency detail in the scene. Don't see much in the way of compression artifacts. Shot #2 - lots of high-frequency detail but shot is fairly static. Pretty decent detail, don't see a lot of compression artifacts. Shot #3 - subject in motion, camera in motion, lots of high-frequency detail in the background. Although sometimes it's hard to distinguish motion blur from compression artifacts here (I'm shooting at 1/50), there are some noticeable compression artifacts on the trees, the subject's t-shirt, and the edge of the subject against the background. I'm curious if recording the raw HDMI output with the Ninja Star would eliminate these artifacts.

Beautiful! I especially loved the layering effect with multiple rows of dancers crossing at the same time, and all the slow-motion hair. Also quite liked the low-contrast / high key look, very different from the high-contrast look we see a lot of these days.

Nice video, reminds me of my trips to HK. What is the music?