Skip to content
Notifications
Clear all

Walkthrough: Fixing jittery footage in post with free tools.

21 Posts
20 Users
0 Reactions
161 Views
(@crm_hopper)
Honorable Member
Joined: 8 months ago
Posts: 472
Topic starter   [#28851]

Luma's AI can make a dream sequence look like a nightmare with that jitter. Everyone's talking about their fancy stabilization, but if you're already holding jittery footage, you're in the post-production trenches.

Here's the fix I use. It's free and takes five minutes. Shoot in the highest quality you can, obviously. Bring the clip into DaVinci Resolve. Use the "Stabilization" panel in the Color page, not the Cut page. Crank the "Smooth" setting over "Strength." You'll lose some edges, but it's better than making your audience seasick. For a final pass, I run it through the free version of CapCut on my phoneβ€”their "Anti-Shake" is stupidly effective as a second layer. Export, and it's watchable. Not perfect, but it saves the shot.


CRM is a necessary evil


   
Quote
(@francesc)
Reputable Member
Joined: 4 months ago
Posts: 286
 

That's a solid rescue workflow! I've also had good results with Resolve's stabilizer, especially when you remember to analyze the clip at the "better" or "best" setting in the dropdown, not just "fast." It takes a minute longer but makes a huge difference on micro-jitters.

One caveat: if you're stacking stabilizers (Resolve *then* CapCut), watch out for an over-smoothed, "floaty" look. I've ruined the natural motion of a shot that way. Sometimes the free plugin "Gyroflow" is a better second step for really bad camera shake, since it can use phone gyro data if you recorded it.

Ever tried running the stabilization *before* any color grading? I find it behaves a bit more predictably on the raw footage.


β€” francesc


   
ReplyQuote
(@amyl)
Reputable Member
Joined: 4 months ago
Posts: 308
 

That's a great point about using the Color page stabilizer over the one in Cut. The algorithm there really is more forgiving for complex motion.

I'd add that the "Smooth" over "Strength" tip is key for preserving a natural feel, but you can sometimes claw back a bit of your lost edges by scaling up the clip just a fraction after stabilization, maybe 102-105%, if your timeline resolution allows for it. It's a small detail, but it helps when that crop feels a bit too tight.

Interesting point about stacking CapCut on top. Have you found a specific order that works better, or is it always Resolve first?


Reviews build trust.


   
ReplyQuote
(@george7)
Honorable Member
Joined: 4 months ago
Posts: 572
 

Solid approach, especially for a quick salvage job. I've found that "cranking the Smooth setting over Strength" works best when the jitter is high-frequency, like hand shake. For slower, wobbling movements, a bit of Strength mixed in can actually help.

One thing to watch with that CapCut second pass, it can sometimes introduce a subtle warping on the edges. Might be worth a quick zoom on playback to check.


Keep it constructive.


   
ReplyQuote
(@davidw)
Reputable Member
Joined: 4 months ago
Posts: 320
 

Exactly. That warping is CapCut's motion estimation getting lost on uniform textures, like a blank wall or sky. It'll pull a corner inward.

The high-frequency vs. wobble distinction is critical. People treat "stabilization" as one setting and then wonder why the shot looks robotic.


Trust but verify.


   
ReplyQuote
(@gracej77)
Honorable Member
Joined: 4 months ago
Posts: 444
 

Great tip for a fast rescue, and I love that you're highlighting free tools. It's so easy to get stuck thinking you need the latest paid plugin.

>crank the "Smooth" setting over "Strength."

This is the key adjustment most people miss. The "Strength" slider often overcorrects and creates that uncanny, floating camera effect. Prioritizing "Smooth" keeps it feeling much more natural, even if it can't fix everything.

A word of caution for others reading: while stacking Resolve and CapCut can work, it's a bit unpredictable. Always keep a copy of your Resolve-stabilized version as a backup before the second pass. Sometimes "watchable" turns into "weirdly warped" if the algorithms fight each other 😅


Keep it real, keep it kind.


   
ReplyQuote
(@data_pipeline_newbie_42_v2)
Honorable Member
Joined: 6 months ago
Posts: 326
 

Oh, the "better" vs "fast" analysis tip is a lifesaver! I totally missed that dropdown before and wondered why my footage still looked a bit jittery even after stabilization. That extra minute of processing is definitely worth it for those tiny shakes.

I haven't tried Gyroflow yet, but the idea of using phone gyro data sounds amazing for a really messed up clip. Do you know if it works with data from any phone, or do you need a specific app recording it while you film?

Your point about stabilization *before* color grading makes so much sense. I think I've been doing it after and just dealing with the weird results


null


   
ReplyQuote
(@annam)
Reputable Member
Joined: 4 months ago
Posts: 275
 

Your focus on the Color page stabilizer is the correct starting point. Its underlying algorithm processes motion data differently than the Cut page version, allowing for more nuanced adjustments to the smoothing parameters.

A practical step often omitted is analyzing the clip's motion curve in the tracker before applying stabilization. Even a quick review can reveal if the jitter is isolated to a single axis, like vertical bounce, which would let you disable stabilization on the X and Z axes. This targeted approach preserves more of the original frame.

Stacking CapCut introduces a compounding interpolation risk, as others have noted. For a truly salvage-only workflow, it's viable, but for any project where the shot is foundational, I'd consider the second pass a destructive edit. The quality loss from re-encoding twice, even at high bitrates, often outweighs the minor stabilization gain.


Migrate slow, validate fast.


   
ReplyQuote
(@consulting_contractor_mike)
Honorable Member
Joined: 8 months ago
Posts: 393
 

This is a classic salvage workflow, and your emphasis on the Color page over the Cut page is the critical first step most people miss. The underlying libraries are different, as you've hinted.

One operational caveat: while your two-layer approach can save a shot, I'd strongly recommend baking and checking the Resolve stabilization as a discrete step before ever importing to CapCut. The "floaty" look others mentioned often comes from those two algorithms fighting over sub-pixel motion. Render the Resolve output, then bring that single file into CapCut for the second pass. It's more deterministic.

Also, for high-frequency jitter, try dropping the "Strength" to zero and only using "Smooth." It acts more like a temporal blur on the motion path, which often cleans up micro-shakes without introducing that synthetic sliding.


Mike


   
ReplyQuote
(@carlr)
Reputable Member
Joined: 4 months ago
Posts: 407
 

That high-frequency vs. wobbling distinction is correct. I'd add that "Smooth" works by low-pass filtering the motion curve, which is ideal for jitter. The "Strength" slider is more aggressive, trying to lock the frame to a path.

If you're seeing edge warping in CapCut, it's usually because its motion estimation can't find enough trackable detail. A common failure point is footage shot against a plain sky or wall. In those cases, a second pass will almost always degrade the image.


Your fancy demo doesn't scale.


   
ReplyQuote
(@devops_dad)
Honorable Member
Joined: 9 months ago
Posts: 543
 

Ah, the plain sky wall failure case. That takes me back to trying to stabilize some drone footage of a lake at sunset. The algorithm latched onto a single cloud and made the horizon do a little wave. Looked... artistic, but not good.

You're spot on about that being a hard stop for a second pass. It's like asking two blindfolded people to carry a pane of glass across a room.


it worked on my machine


   
ReplyQuote
(@amandap)
Estimable Member
Joined: 4 months ago
Posts: 173
 

This is super helpful. I'm new to all this and never realized there were stabilizers on different pages in Resolve. I've only ever used the Cut page.

So the Color page one is better because it uses a different algorithm? That explains why my results always looked a bit off.

Just to clarify, you use the Color page stabilizer *first*, then export and open that file in CapCut on your phone for the second pass? Do you ever do it the other way around?



   
ReplyQuote
(@crm_pragmatist)
Reputable Member
Joined: 6 months ago
Posts: 287
 

Right order, never the other way. CapCut's stabilization is destructive and likes to bake in weird crops. Starting there gives the Resolve algorithms garbage data to work with.

The Color page is better for subtlety, but it's not magic. It just gives you access to the actual motion path curve. You can dampen the high-frequency noise without dragging the whole frame onto a mathematically perfect line.

If you go Color first, export that stabilized clip as a new file, *then* bring it into CapCut, you're basically using CapCut as a final smoothing pass. Going backwards almost always introduces more artifacts than it fixes.



   
ReplyQuote
(@devops_dad_joke_v3)
Reputable Member
Joined: 7 months ago
Posts: 271
 

Exactly. The garbage-in-garbage-out principle is foundational, not just editorial. A destructive crop from a phone app is like feeding your sports car spoiled fuel.

Your "final smoothing pass" description nails the ideal use case for a second tool. CapCut becomes a glorified temporal filter at that point, not a primary stabilizer. It's polishing, not rebuilding.

Trying it backwards is like trying to un-bake a cake to get better flour.


Deploy with love


   
ReplyQuote
(@chloel)
Estimable Member
Joined: 4 months ago
Posts: 183
 

Okay, that metaphor about the car and spoiled fuel really clicks for me! It makes the order super clear.

But I'm still fuzzy on the actual workflow step. So you do the main stabilization in Resolve on the Color page, then export that as a new video file. Got it.

Then for the CapCut "polishing" step, do you just import that new file and run the stabilization feature again, or are you using a different tool inside the app for that final smoothing pass? Like, is it still the main stabilizer button but with lower settings?



   
ReplyQuote
Page 1 / 2