More Komodo Footage

The video is best watched on an OLED TV. It was shot at 6K 2.4:1, 50 FPS in R3D LQ. RED raw files are a joy to grade, the color and texture of the image are like nothing we’ve ever experienced. The Canon 35mm macro is a crispy little bugger with pleasant enough rendition of out of focus areas, but if you’re recording sound, you’ll have to be careful not to pick up the loud noise the lens produces when focusing. It’s also got pretty horrendous barrel distortion, so if architecture is your bag, you might want to look elsewhere. hehe Tomorrow, we put the Meike 75mm T2.1 s35 EF mount cinema lens through its paces.

20 thoughts on “More Komodo Footage

Add yours

  1. The DR seems great.
    The more I read about RAW, the more think you made the good choice because with Red you have a 16 bits raw codec recorded internally and you can directly use the files in Resolve with all the options available.
    AppleProRes RAW is a “fake” raw which is only 12 bits and unusable in Resolve.
    I should find your articles where you may have compared 422 10 bits vs AppleProRes Raw 12 bits to see if this one really brings something for grading or not.

    1. The contrast might possibly have been improved had I used a lens hood, but Canon didn’t see fit to provide one with their 35mm macro. I’ve seen it selling for as much as USD $65 here in Vietnam, while a third party one goes for around eight dollars. It’s the tiniest lens hood I’ve ever seen, so most likely worthless anyhow! The control ring on the lens adds welcome functionality and since the Komodo doesn’t have IBIS, the OIS is indispensible for handheld work. When it comes to AF, the lens prioritizes smoothness over speed, which is fine if you want to do something like programming focus pulls with the Control app, not so great if you’re photographing action. Concerning the rendition of the lens, when most photographers talk about bokeh, they’re referring to the appearance of out-of-focus light sources in the frame, but bokeh actually refers to the quality of blur throughout the entire image. So, when you see onion rings or hard edges around those circular out-of-focus areas, those same artifacts are in fact influencing everything that isn’t in focus. And although the Canon isn’t a bokeh monster, the blur is surprisingly pleasing, at least at closer distances – though it does have some edging that makes it appear busy at farther distances. The lens suffers from huge vignetting, but using it on a s35 sensor lessens the issue somewhat. I’m seeing a massive improvement in resolution, which I’m guessing isn’t the lens, but the downsampled 6K. I was going to shoot some tests with the Meike 75mm but I’m going do my subscribers a favor and wait till I can find someone to model for me. hehe

  2. Thanks for this feedback.
    Yes, it’s very cool when you make videos with models with a context.
    I remember a video in some kind of arty flat/coffee-flat with natural light I think which was very nice to watch.

    Jon, what do you think we should do for a scene for eg in the street with some people.
    Is it better to ETTR (on the brightest point you want to keep information on) or expose specifically for the skin tones of the people and maybe let some part of the image being too high or too dark?

    1. I’d expose for the brightest point with important information. In one of the shots in my most recent upload, the window at the end of the hallway is clipping, and even if the audience doesn’t notice it, it still bothers me. If you’re shooting rec709, the overexposed area rolls off gently whereas in HDR, it becomes a dreadful white hole that can never be recovered.

      1. So in general, you expose especially for the skin tones only if there’s really a main character?

      2. I usually expose for the talent and if I’m out on the street and the background is too bright or too contrasty, I either change positions, find a different location or shoot at another time of day. 😅When I shot with the a7s III and Ninja V, I just made sure that there was no yellow, orange or red false color on skin that would be difficult or impossible to recover in post. Now that I’ve got the Komodo, the traffic lights alert me if I’m clipping and the false color guide lets me see where the clipping is occurring. If you’re shooting passersby, you don’t have the luxury of changing position or shooting at a different time of day, so I suppose I’d just underexpose the people a bit and pull it up in post if it’s necessary to preserve highlight detail.

  3. Ok thanks.
    Yes I shoot 10 bits but then grade in Rec709.
    If it’s the video of this post, to my eyes, it seems well balanced in general.

    1. Hi Val. I added illustrations to my post on exposing with the Komodo. Important highlight and shadow information would be the texture in the fabric of the mask. It’s also critical to give enough exposure to preserve tonal differences in the shadows (for example, to prevent the mask from blending into the television set). Unavoidable clipping would be in the iPhone. You can also see how much more noise there is at ISO 800 compared to ISO 250.

      https://daejeonchronicles.com/2022/07/25/red-exposure-strategy-in-a-nutshell/

      1. Yes, that’s what I thought about, the window in the background.
        Well, I could not watch it in an HDR Tvs these latest days but on a average smartphone it’s ok.

    1. I would definitely avoid fetishizing ETTR. I found that, with the a7s III, around +1.66 stops was plenty, and even one stop over was perfectly fine. The risk of blowing out highlights, particularly in HDR – and more especially when working in rapidly changing light – is very real. I’m learning the same lesson with the Komodo. Shooting at a slightly higher ISO can provide some extra insurance against clipping and when there is clipping, the transition to clipping will be less abrupt (probably of greater importance to SDR, as any clipping at all in HDR is ugly), which is why I should settle on an ND solution ASAP.

    1. Good question! All my videos are recorded in the R3D file format. From RED.com: “The RAW data is recorded independently of any ISO, WB or other RGB color space settings. Instead, color parameters are saved as reference metadata; that is, color is not burned into the recorded RAW data.”

      1. Thanks for your response and all of the trial and error you put into finding out all of this information. What I meant was the output color space option under the image/lut section. Choices are dci-p3, rec 709, rec 2020, and redwidegamutrgb. Thanks again.

      2. Oh! I’m monitoring on an iPhone 12 using rec709/SDR. Sometimes I use the RED FilmBias LUT. I picked up an SDI module for the Ninja V so I can monitor in HDR but haven’t tried it yet – having the Control app on the phone is just so convenient!

  4. Hi, any chance you could also upload a clip from this shoot in R3D RAW? Much more colour info to play with then the previous clip you provided! Thanks

Leave a Reply to Val Cancel reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Blog at WordPress.com.

Up ↑