Sekhar

I really like Warp Stabilizer. It has produced decent to fantastic results for me when run with detailed analysis and enhanced RS reduction (under advanced). The problem is that it takes forever, so anything close to it in camera (like DIS or IBIS) is going to be way more preferable. Also, WS results look strange when there is a lot of motion blur (the object looks still, but has blur on different sides), so IMO it only works well with low shutter angles or when the movement is limited.

Looks like I hit a nerve, mission accomplished I guess. Don't take it that badly man, my point wasn't about Canon per se (I have a 6D that I love for stills, and all but one of my lenses are Canon).

Actually, you made my point. You can't just pick fuzzy stuff like "color science" and "it just works" (easiest things to defend because they are vague) and leave out the hard facts that don't fit. Are you saying Sony, Nikon, etc. don't "just work?" Did you forget Canon's recent messy sensor issue with t6i already? Nobody would say Canon is junk, all I'm asking is: let's give the full story and stay balanced.

The sad part is nobody seems to be truly objective anymore, you can spin things any way you want. Canon (especially the hallowed, beyond-any-reproach-of-any-kind 1Dc) can have any number of flaws, but you can always take refuge with two words: "color science" (whatever that is). And be sure to ignore what you can do in post, it don't matter. Kind of like the $8K Leica M Monochrom buyer ignoring all objective comparisons and calling it worth the price because it's just "magical" (whatever that is). And it's not just with cameras. Last night we had a Republican debates with 10 candidates, want to know who the winner was? Any of them, depends on who you ask. Heck, we don't know anything for sure anymore, from if there is a multiverse to if OJ was guilty to if a7RII or NX1 can render color as well as 1Dc. Reminds me of what Jack Nicholsen tells Diane Keaton in Something's Gotta Give: "I've never lied to you. I've always told you some version of the truth."

No, mostly cloudy for daylight (gives a nice warm look) and depends on indoors, I've found fluorescent white to be good for many common indoor lighting situations. When all else fails (and I have time), custom.

Custom profile, with shaprness -10 and contrast -5. Problem with profiles is they burn in the look. For WB, I don't like auto because it changes from shot to shot, making shot matching unnecessarily tricky.

Uh oh, you just declared war on the Canon crowd. Say what we may about Canon and video, people thought they had at least the stills corner secure.

Fair enough, I thought you were looking for feedback on how the Match Color worked. Once you're done with all your tests, please do post your findings/conclusions on the pros/cons of different settings.

It looks good overall, but I still see issues with color (they don't look natural) and contrast (not much shadow detail).

Clearly, the guy had real low shutter angle, see the passing car at 0:12, there hardly any blur. These days, I'm regularly shooting at high shutter speeds with my NX1 myself and not bothering to stick to the hallowed 180 degrees. Two reasons: You can get some amazing 8MP still frame grabs. I went to Costa Rica recently and some of most exciting stills of wild life I got were frames from video because you have so many to choose from with just the right timing and framing.I don't like using ND filters anymore. I have a pretty good variable ND filter (Tiffen 77mm), but it reduces the quality (especially adding flare) and is not the most convenient to use. Putting it on/off was not really an option in the Costa Rica rain forests (BTW, both my NX1 and 16-50 S lens took the rain well).Given the low angle footage, I'm experimenting with getting the blur. So far, I've found the pixel motion blur in AE to be quite good (you need to bump up the frame count though), but it is dog slow, so you'll need to add that to your workflow time. The ReelSmart motion blur is really fast (especially when using graphics card), but has artifacts (at least in my testing). If anyone else has suggestions, please share.

I use 0-255 and MBL 0 (defaults) with DR gamma, sharpness -10 (lowest), and contrast -5 (anything lower or higher can cause artifacts). Keeping at 0-255 and MBL 0 makes sure you use the entire range of the (tiny) 8 bit range. As I see, the only reason to use 16-235 is when you plan to hand off the footage directly without post processing. Check out my thread NX1 Dynamic Range Settings that has more on this.

Well, I like the settings. E.g., see the fashion show I covered using these settings at https://youtu.be/PsHTzuANSHo. What you see is the way it came out of the camera.

My suggestion: DR gamma, sharpness -10, contrast -5, and others at default (that has black level at 0). See my thread NX1 Dynamic Range Settings, which has details.

There are two reasons for reducing sharpness to -10 in the camera. The default sharpness of 0 actually has in-camera sharpening that results in artifacts that cannot be removed in post. If you zoom in closely, you'll see the issues at the edges, try getting rid of them. Next, the in-camera sharpening introduces extra (false) information that stresses the codec, further reducing quality. Take a video file, sharpen and re-save, you'll see that the sharpened version is larger. Many filmmakers (me included) prefer the natural look of hi res un-sharpened images (which are more film like and less like video), so it's also an aesthetic choice. But for the two reasons above, even if you want well defined edges, you're better off sharpening in post. If you're interested, check out my thread NX1 Dynamic Range Settings, where I discuss my finding with NX1 experiments and conclusions (that sharpness -10 and contrast -5 with DR gamma give the best results). Thanks for viewing my fashion show video and for your comment. Yes, I did have sharpness at -10 and contrast at -5, with DR gamma (applies only to NX1 though, it's not in NX500). Try out reducing the sharpness, you might end up liking it!

The first part (normally exposed, till 1:50 or so) looks great, I liked it the best. Natural looking, with nice detail. Seemed a tad contrasty though, for example at 1:00.

Mark, placing 4K onto a 1080p to get a crop doesn't soften the image, the result will have the same 1080p resolution as that of 4K scaled down to 1080p. So in this case, you're going tight without losing detail. That's the point. And it's not just about zoom, you can fake two-camera shots, slider/jib moves, slow zoom, etc. without worries. Of course, all this is assuming you're delivering in 1080p. See my post Using 4K to simulate two-camera shoot (and other things) for more details, including a video I did of a fashion show that uses these techniques.

Whoa, why the hostility? This is a friendly community where we share ideas to improve each other's work, we aren't competing. If you have input that helps the rest of us, you're welcome to share. Having said that, you're mistaken in many ways. First, mercer here also got the NX500, and we're comparing notes on the effect of sharpness/contrast settings at default and at -10/-5, you likely didn't know that. Next, my goal is to cross cut NX500 and NX1 footage, and I'm checking to see if setting sharpness/contrast the same would help on that. And you can always get it sharper either in camera or outside (see below for a sharpened version of the same video), the point is to evaluate NX500's capability as a teleconverter, this is not a production video. Finally, it's great that you are happy with your video, but none of your shots go as tight as I went, which is the point of this thread. If you do have anything that goes as close, please post so we can compare apples to apples.

No, this time I had the sharpness at -10 and contrast at -5 like on my NX1. I'll use those for everything going forward, I still need to check how the NX500 video compares to NX1's with those settings, I want to be able to crosscut.

Wow, that is some amazing stabilization in the camera or you have rock steady hands! Footage looks pleasing, I like the colors more than those in the test footage I've seen this far, but they still seem muted and on the muddy side in some spots, I'll check on big monitor tomorrow. Also, there is a lot of banding in the sky and macroblocking in low light shots (towards the end for example), not sure if it's YouTube doing its thing.

You know NX500 as a great little video camera that shares many of NX1's awesome qualities. What may not be obvious is that it can work as a great teleconverter. Using its APS-C sensor and 4K crop and placing its 4K footage onto a 1080p timeline, you can turn a 200 mm zoom into a 1,000+ mm ultra zoom (200 x 1.54 x 1.58 x 2.13 = 1,037 mm). I got an NX500 yesterday and thought I'd post some footage of birds on our street that I shot today with my Canon 70-200 L lens. Try it out if you have the NX500.

My point is: except for FF, it's not a slam dunk. Log on 8 bit 4:2:0 isn't exactly great, IMO colors are muddy because of it. This is even assuming NX1 won't get log itself soon. Low light is great for candle light test shots, but does anyone ever do any serious project, especially narrative work, based on low light capabilities rather than light the scene? For doc folks, would be good though. I don't know what you mean by codec, H.265 is an extremely efficient codec that makes great use of the bandwidth, perhaps you're talking about the need to transcode to ProRes? You'll need to do that with a7RII as well. Rolling shutter is just as bad on crop mode, if not worse, as I understand. IBIS and lenses would be a (possibly big) advantage if you use other lenses, but otherwise lens stabilization works quite well with Samsung lenses (though I don't know how it compares to the IBIS in a7RII). And don't forget that NX1 costs less than half of a7RII, which should be mentioned when you list the pluses. Bottom line, we can enumerate all the things on one side to make the advantages seem big, but as I see it, the only headline plus I see is the FF aspect, but as I said the colors on the other hand seem muddy compared to the vibrant colors of NX1, which for me trumps FF. Of course, I'm only seeing the test clips posted, and that's why I said we need to see more before judging.

Not sure why you say that. Sure, it is FF. But NX1 has pluses of its own: e.g., 120 fps in 1080p and true 4k (vs. UHD). At the end of the day, both are 8 bit 4:2:0 internal, so I'm not sure what the headline feature is for A7RII that should get everyone excited over above NX1. Against A7S, it'll be internal 4K; but against NX1, what? Comes down to FF, right? IMO, we need to learn a lot more about its performance and see real world results before passing judgment vs. other cameras. For one thing, I thought the colors are muddy in all the test footage I saw from A7RII so far.

Yeah, I did an extensive test once between 422, LT, and Proxy. There is a HUGE drop-off in quality from LT to Proxy, but not much from 422 to LT. If space is tight, may be LT is enough, but I would go with 422 for all regular jobs. For 420 8-bit video, HQ and above is overkill IMO. BTW, the lens I used is the "S" version of the 16-50, I just realized there is another version. This lens has given me amazing results, I got it with the NX1. I'm going to do a bunch of tests tomorrow, mostly on the big zoom use of NX500, for which I'll use my Canon 70-200 L lens (non IS version), which is also great. Lenses always make a difference, but IMO more so when it comes to 4K video, which is why I kept mentioning lenses.

I convert to ProRes 422, I think it's enough. And yes, you have access to the whole frame, which is why you can cut with 100% and 50% to simulate two cameras; keyframe the zoom to effect slow zoom; keyframe lateral/vertical moves to fake slider and jib shots; etc. For the fashion show video I posted, I only had the 16-50 zoom, and even at 50 I just about managed to cover the stage; but with this crop technique for example I was able to go real tight, and the models (all amateur event attendees and their kids) were very happy. Going to ultra zoom is the main reason I got the NX500.

I went back and saw your leaf shots (4:35). Hard to say about resolution difference since the profiles are different, but I still think colors are off. Anyway, what I think is irrelevant when it comes to your video; if you like it, that's all matters. For my clip, I used the Samsung 16-50 lens. And yes, I shot in 4K 24p, converted to 4K ProRes with FFmpeg, and then imported it onto a 1080p project. One potential source of difference is that there is no scaling on my side: the 1080p project cropped the 4K video, so what you're seeing is a portion of the video frame Whereas you scaled the full frame to 1080p at H.265 conversion time and imported a 1080p clip onto a 1080p project. Just for kicks, try this method and see if you prefer that. Placing 4K clips onto 1080p project is now my preferred method because it lets me do fancy stuff like faking slider moves and two camera shots, like I did with a recent fashion show. I posted it here on EOSHD some time back, but here it is again if you're interested. https://youtu.be/PsHTzuANSHo