ShortVi Beta
Return to Topic List

Why does my 4K footage look soft and lack that 'cinematic pop'? (Sony A7S III + DaVinci)

Video Critique Created: 02.08.2026 03:10 6 Views
Hey everyone! I've been shooting with my Sony A7S III in 4K (XAVC S-I, 10-bit 4:2:2, S-Log3) and grading in DaVinci Resolve Studio. My footage is technically sharp, but it always looks a bit flat and soft, especially compared to some of the cinematic footage I see on YouTube and Vimeo. I'm using a 35mm f/1.8 prime, usually stopped down to f/2.8, and I've got my sharpness set to 0 in-camera. I grade using a CST from S-Log3 to Rec.709, but the result still feels... digital and lifeless. I've attached a still from a recent shoot (a simple indoor interview) – you can see the skin tones are okay, but the overall image just lacks that 'pop' and depth.

I'm not sure if it's my lighting (I'm using two Aputure 120D II's with softboxes), my grading workflow, or something about my sharpening settings in Resolve. I've tried adding a bit of contrast and saturation, but it either looks too crunchy or still flat. I've also tried using a LUT, but that just makes it look worse. Does anyone have a solid workflow for getting that clean, sharp, yet organic cinematic look from S-Log3? I'd love to hear your go-to steps for sharpening, contrast, and color – maybe I'm missing a key step like a proper sharpening pass or a specific node structure. Any advice would be massively appreciated!

Replies (5)

Hey! I've been down this exact rabbit hole with my A7S III. The 'soft' look is often not a sharpness issue – it's a contrast and micro-contrast issue. S-Log3 is designed to be flat, but when you do a simple CST, you're not adding any local contrast or texture. Here's my go-to node structure:

1. **First node**: CST from S-Log3/S-Gamut3.Cine to Rec.709 (or use a proper IDT if you're in a color-managed workflow).
2. **Second node**: Add a slight contrast curve – an S-curve, but keep it gentle. Don't push it too hard or you'll clip highlights.
3. **Third node**: Add a 'mid-tone detail' pass – go to the 'Sharpening' tab in Resolve, but use a very low radius (0.3-0.5) and moderate amount (20-30). This adds that 'pop' without halos.
4. **Fourth node**: Add a subtle 'glow' – use a soft light blend mode on a blurred copy of the image (or use the 'Glow' effect in Resolve 18) to add a gentle bloom on highlights. This mimics the organic roll-off of film.

Also, check your in-camera settings – you said sharpness 0, but make sure you also have 'Detail' set to -2 and 'Noise Reduction' to -2 for maximum detail retention. And for lighting, try to add a bit of negative fill on the shadow side to increase contrast – that will add depth more than any grade.




I'm using a 35mm f/1.8 prime, usually stopped down to f/2.8, and I've got my sharpness set to 0 in-camera.



That's good, but also consider using a diffusion filter like a Tiffen Black Pro Mist 1/8 – it softens highlights and adds that cinematic glow, but it can also make the image look softer. If you want to keep it sharp, skip the filter and do the glow in post.
Focus on retention, the rest will follow.
I agree with zoomwizard38 on the contrast curve – that's the first thing I'd check. But also, are you using the full range of your codec? A common mistake is exposing S-Log3 too dark. For S-Log3, you want to expose to the right (ETTR) – typically around +1.7 to +2.0 stops over your normal meter. If your footage is underexposed, you'll have to push it in post, which amplifies noise and makes the image look softer and more digital.

I always use a waveform and aim for around 55-65 IRE for skin tones, and I make sure the highlights are just below clipping (about 90-95 IRE). That gives you plenty of information to work with. In Resolve, after your CST, add a 'Color Space Transform' to 'DaVinci Intermediate' if you're not using a color-managed workflow – that can help with color accuracy.

Also, for sharpening, I prefer to use the 'Sharpen' effect (not the 'Sharpening' tab) with a low radius and high threshold – that only sharpens edges and leaves skin smooth. Set radius to 0.7, amount to 0.5, and threshold to 0.1. That's my secret for that 'crisp but not crunchy' look.
Always testing the YouTube algorithm.
Since you mentioned the Aputure 120D II's, I'm wondering about your light modifiers and distance. Even with softboxes, if the light is too close to the subject, the falloff can be harsh, but if it's too far, it becomes flat. I'd suggest using a 4x4 diffusion frame or a large bounce to create a bigger, softer source – that adds a natural contrast gradient on the face, which gives a more three-dimensional look.

Also, don't forget about the background. A common mistake is lighting the subject and background evenly. I like to underexpose the background by 1-2 stops – that instantly adds depth and makes the subject 'pop'. In Resolve, you can use a power window to darken the background slightly if you didn't do it in camera.

For the grade, try adding a subtle 'film grain' effect (like the 'Grain' tab in Resolve) – even a light grain (about 4-6 on the 'Amount' slider) can mask any digital artifacts and give it that organic texture. And make sure you're working in a properly calibrated monitor – if your monitor is too bright, you'll grade too dark.
TikTok growth is all about the first 3 seconds hook.
I'll add a sound perspective – but not audio, actually. The 'cinematic pop' is also about the motion cadence. If you're shooting at 24fps, make sure your shutter angle is 180° (so 1/50s for 24fps). If you're shooting at 25fps, use 1/50s as well. A wrong shutter speed can make motion look too smooth or too stuttery, which can affect the perceived sharpness. Also, in Resolve, try adding a slight 'motion blur' effect – but that's more for VFX.

But for the grade, one thing that's often overlooked is the 'Hue vs. Luma' curve. In S-Log3, skin tones can get a bit desaturated and shifted. I use the 'Color Warper' or a Hue vs. Sat curve to boost the saturation of skin tones slightly without affecting the rest of the image. That makes the subject look more alive and gives that 'pop'.

Also, consider using the 'Lift, Gamma, Gain' controls instead of a simple contrast curve – I find that lifting the lift slightly (around -5 to -10) and pulling the gain down a touch (around +5) creates a more filmic contrast than an S-curve. It's a subtle difference, but it adds up.
Creator burnout is real, take care guys.
Thanks everyone – these tips are gold! I tried the node structure from zoomwizard38, and the mid-tone detail pass was the game-changer. My footage now has that sharp but organic look. I also exposed to the right on my next shoot, and the noise is virtually gone – that was a huge part of the problem. I also added a subtle grain and darkened the background a bit, and now the interview footage looks like it came from a film set.

I think the biggest lesson was that S-Log3 needs a deliberate sharpening and contrast workflow, not just a simple CST. And the lighting advice about negative fill and background separation really helped add depth. I'm finally getting that 'cinematic pop' I was after. Thanks again to everyone for sharing your expertise – this community is amazing!
Shorts and Reels creator | DaVinci Resolve fan!

You must register or log in to reply to this topic.

Related Topics & Similar Discussions

Why Does My 4K Drone Footage Look Soft in DaVinci Resolve? Category: Video Critique
My cinematic travel short is flopping – is it the pacing or the algorithm? Category: Video Critique
Need Honest Critique on My First Cinematic Travel Short – Why Does It Feel Flat? Category: Video Critique