The result
How much of the time the person talking was on screen, over the 70 measured minutes of episode 29 (78 minutes, two hosts, five cameras):
| Cut | Analysis time | Speaker on screen | In their close-up |
|---|---|---|---|
| SOLD CUT, from the raw folder | 4 s | 93.5% | 77% |
| SmartSwitch, microphones linked by hand, audio only | 1 min | 87.2% | 37% |
| SmartSwitch, microphones linked by hand, full analysis | 46 min | 80.6% | 69% |
| SmartSwitch, the cameras as they are | 46 min | 72.3% | 61% |
The camera stayed on the person who was not talking 6.5% of the time with SOLD CUT, and 12.8%, 19.4% and 27.7% with the three SmartSwitch runs, in the order above. A new speaker appeared more than two seconds late never with SOLD CUT, and in 9%, 11% and 21% of the changes of speaker with SmartSwitch.
The same minute of both cuts, made on the same Mac from the same files: SOLD CUT on the left, SmartSwitch on the right.
How we measured
- One episode of our own podcast: two hosts, five cameras, 78 minutes, of which 70 are measured (the time an illustration fills the screen is left out).
- DaVinci Resolve Studio 21.1, Multicam SmartSwitch at its default settings, on the same Mac. The audio-only row uses its “Use Audio Only Fast Analysis” option.
- “Linked by hand” means each close-up camera was first given its person’s microphone as its sound, which SmartSwitch does not do by itself.
- Who is talking is read from each person’s microphone with one loudness rule, the same for every cut. These are the microphones SOLD CUT listens to, so the measure suits it; we say so plainly.
- When both talk at once, mostly short interjections, SOLD CUT stays on the person who has the floor and showed both people 25% of the time. SmartSwitch’s audio-only run showed both 62% of the time, because it stays on the wide shot for about half of the episode.
It is one episode: yours will differ.
Why the difference
SmartSwitch works on a multicam clip you have already synced and built, and decides from the cameras’ own sound and the lip movement it sees. SOLD CUT reads each person’s own microphone, so it knows who is talking without guessing, and then applies editing rules tuned on hundreds of our own episodes: hold on real turns, absorb short interjections, go wide at the right moments, show both people when they talk over each other.
It also does the parts SmartSwitch leaves to you: it starts from the folder of raw files, syncs everything by sound or timecode, builds the project, keeps every angle on its own track under the cut, and builds split screens when two people talk. After the cut, the same app adds subtitles, YouTube titles, react layouts, jump cuts and clips. Read how a podcast edit works in SOLD CUT.
When SmartSwitch is enough
If your show has one or two cameras, you already build multicam clips in Resolve, and you are happy to fix the cut by hand, SmartSwitch comes with Resolve Studio and costs nothing more. If you edit a podcast every week, with several cameras and a microphone per person, SOLD CUT gives you a cut you only need to polish, in seconds, with every angle one click away.
DaVinci Resolve and Multicam SmartSwitch are trademarks of Blackmagic Design. SOLD CUT is not affiliated with Blackmagic Design.
Questions
Is SmartSwitch free?
It comes with DaVinci Resolve Studio, so if you own Studio you already have it. SOLD CUT also needs Resolve Studio and costs €150 once.
Can I use both?
Yes. SOLD CUT never changes your timeline: it builds its cut on a copy, so a SmartSwitch cut and a SOLD CUT cut can sit side by side in the same project for you to compare.
Will my episode give the same numbers?
No one episode speaks for all of them. Ours had two hosts, five cameras and mostly one person talking at a time. Try SOLD CUT free for 3 days on your own episode and compare.
Does SOLD CUT need a multicam clip first?
No. It can start from the raw folder: it syncs the cameras and microphones, builds the project and the timeline, then cuts. It can also start from a timeline you have already synced.