Great job! Nikon has a lot of work to do on their N-LOG which is understandable given its so new. Should also note that its worth using Arri Log and Panasonic V-LOG for conversions as they may render the Nikon files with more pleasing skin tones and contrast than Nikon's own. Also on the over-exposure, its great to see such a good example on why you should overexpose these sensors to the right. You can clearly see on the blueish pot (bottom left) how much cleaner the shadow areas are when captured at +2 expo vs "normal" exposure. You always want to try to push these captures to the right as far as possible without clipping any unwanted areas.

@ZuZusGramp

Жыл бұрын

I find that the N-log transform in Davinci Managed Color gives me better skin tones, and a more malleable grading environment than any LUT I’ve tried.

David, I was intrigued by this test- thanks for doing it! This morning, I decided to see if I could underexpose by 4 stops and grade the N-Raw Log on an HDR Resolve timeline. At the starting point, of course, the clip was extremely noisy, as expected, and there was the color shift that you highlighted. However, I'm happy to report that, with some gamma, gain, temp adjustments, and noise reduction- more so in the chroma- the clip looked good, and was quite usable. So, at least in this context, 4 stops under is possible. UPDATE: Just tried it as rec709, also cleans up nicely.

The more I watch this, the more I feel it confirms that N-LOG really wants to be overexposed which makes it much trickier to monitor in-camera. Pulling up this video in two windows side by side, it seems that N-LOG = SDR + 2stops. i.e. N-LOG +2 over is closer to SDR correctly exposed, and N-LOG -2 under, performs closely to SDR -4 under. If I set my zebras in camera to a midpoint value, I have to overexpose N-LOG by two stops (+2) exactly to have zebras match SDR. Also likely explains why the Leeming LUTs want you to push that exposure so far to the right.

Thanks for this. Coud you make the file available for download?

very cool - thank you

Any dynamic range difference if both shot at native iso? 64iso on sdr, 800iso on nlog?

I wonder why the middle grey point shifts so much between the different modes. If they're both at ISO 800, what's changing between the modes to reduce and increase tolerance to different exposures?

Hello David, you can advice me about color setup N-RAW SDR in DavinciResolve?

It's a very valuable video, thank you for your efforts ، I'll be grateful if you upload one second or less from 8k n-raw sdr - high quality , so we can download it and try it in davinci resolve , Because I am very interested in the capabilities of this camera in video raw ، and I do not want to buy it before I try it , Thank you

I have to point out the conclusion you get is false. The raw footage of iso 100 SDR and iso 800 Nlog are exactly the same, the brightness information are no different (you just need to switch the gamma and colour space to REC709 in Camera Raw in davinchi resolve and lower the exposure by 3 stops). When shooting nlog for example at iso 4000, the amplifier of the sensor actually works at iso 500, then via the Nlog curve, the brightness information is mapped to equivalent of ISO4000. However, under raw footage, this affects nothing because raw is the original data of the sensor. The reason why you get the conclusion that nlog is more friendly to shadow is because the camera was actually operating at ISO100. You see, Nlog ISO 800 is exactly 3 stops higher than SDR ISO100 (I don't know why Nikon make the base iso of Nlog 800 rather than 8x64=512). Similarly, the second base iso 4000 of Nlog is 3 stops higher than ISO500. Thus, to get maximum dynamic range, SDR iso 64 should be the best option.

@ZuZusGramp

Жыл бұрын

I think you are correct-or close to correct. Nikon’s LOG and SDR are a very odd overlays for a raw file, and don’t help understand Nraw’s latitude and dynamic range at all. But I don’t think your recipe is quite right. Lowering the exposure in the Raw development stage by three stops is not something most colorists would do unless the image was overexposed by three stops. And it’s only if the image is overexposed by three stops that what you describe will yield the result you describe. And in doing that, you burn three stops of Nraw LOG’s hilight latitude. So while I do believe you are technically correct, that doesn’t lead to a better understanding of how to get the best images out of Nraw-because one has to shoot using either SDR or LOG mode. Given that, it’s important to know how they behave.

Like your video will contact you soon .I have z9