Skip to guide
Twinbly
Reviewed Read Markdown

Sharpen a scene: when it helps and why it may be unavailable

Sharpen is an optional, paid upgrade for eligible finished Twinbly captures. It refines a large scene in overlapping sections and combines the results into a scene that can stream as you explore. It is also called Tiled refine in some workflows.

It is not an image-sharpening filter, and it is not a promise to recover detail the camera never recorded. Twinbly checks whether the processed capture is suitable before offering paid refinement.

Does Sharpen need video or unused images?

No. The current Twinbly web action is not video-only, and it does not automatically pull unused photos or extra frames from your original video. It works from the images and reconstruction retained for the finished capture.

These counts are different:

  • Uploaded: the files you brought to Twinbly.
  • Processed: the photos or selected video frames used for that run.
  • Registered: the usable views successfully placed in the reconstruction. These may be fewer than the processed images.

For example, uploading 1,000 photographs and processing 250 does not give Sharpen 1,000 usable views. The other photographs do not automatically join the upgrade. Likewise, a long original video does not mean its finished scene has enough usable views.

For ordinary photo and flat-video captures, the current check requires at least 400 registered images, followed by other checks. This is an eligibility floor, not a recommended capture target or a guarantee. A 360 capture is assessed using capture positions rather than treating every derived view as an independent position.

When is it useful?

A large, well-covered scene may benefit when different parts have their own useful groups of views: for example, a long outdoor route or an extended site. Twinbly checks whether it can divide the scene into useful overlapping sections.

A small room or a tight orbit around one object may already use most of its images to describe the same area. Splitting that capture may offer little benefit, even when the image count is high. Other checks also consider the reconstruction, the size of the finished model and whether required processing files remain available.

Sharpen eligibility is not a score for how attractive or sharp your scene looks. A refusal does not mean the scene is perfect; it means this particular upgrade cannot be offered for that capture.

How to check and run it

  1. Find the finished capture in your Library and open Sharpen when that action is available.
  2. Request or view the plan. The eligibility check is free; allow it to finish.
  3. Read the result. If eligible, review the proposed refinement and displayed price before confirming paid work. If the price changes, review the new quote.
  4. Follow its progress. Processing completion and the optimized streamed result becoming ready are separate stages.
  5. Compare the new version with your original. The original remains available alongside the refined version.

Do not start merely to discover whether a scene looks better. Review eligibility and cost first. Paid work already completed may still be billed if you cancel.

What the common refusals mean

Not enough images or positions

There are too few usable views in this processed capture for the refinement path. For 360 footage, positions matter; several views from the same position are not equivalent to moving through the space.

Check the original capture and the processing choices. If you have useful additional coverage, you may be able to prepare a larger, better-covered new run, with its own estimate. Sharpen itself does not add the unused files. Duplicating images or adding many near-identical frames is not a useful fix.

Too compact, or too small to refine

The scene may not divide into useful sections, or the finished model may be too small for this upgrade. Keep the current result if it serves your purpose. If an important area is weak, a closer, connected capture may be more useful than trying to force the whole scene through Sharpen.

The reconstruction needs attention

Poor overlap, blurred frames or disconnected coverage can prevent a useful refinement plan. Review capture for coverage and the capture troubleshooting guide. Refinement cannot reliably repair missing surfaces or source motion blur.

Required files are missing, or this capture is unsupported

Some older captures lack retained processing or camera information. An imported finished splat does not necessarily have the source data required for Sharpen. A previously refined result or an individual MegaSplat tile is not a normal starting capture for another Sharpen run.

If this is unexpected, keep the scene/job reference and the exact message and contact Twinbly support. Do not repeatedly pay to reprocess the same capture without understanding the missing requirement.

Does GPS or an SRT file make a scene eligible?

No. Supported GPS data helps with location; Sharpen separately checks the processed capture. A video with a supported DJI SRT sidecar can retain location information while following a video workflow, but neither the sidecar nor the length of the video guarantees refinement.

See MegaSplats and video location data for the distinction between ordinary video uploads and the photos-plus-video One solve workflow.

Is loading more detail the same as Sharpen?

No. A streamed scene may load different detail as you approach or move around it. That is playback behavior. Sharpen is an explicitly confirmed processing upgrade that creates a new result.

Start with good coverage, inspect the finished scene, then use the eligibility check to decide whether this upgrade is appropriate.