The short answer. Not on a Mac, or want one application for everything: darktable, and there is nothing to weigh. Want the conversion itself to be the whole interface, or have scans that fight a model (clipped channels, expired stock, redscale): this one, because there is no model to fight. Both are free and GPL, so price decides nothing here.
darktable's negadoctor is the serious free option for inverting colour negatives, it runs on Linux, Windows and macOS, and it is the best public reference on the physics of C-41 inversion that exists, and this application's own notes say so. No code is taken from it: the physics of the process is public, and that module was read, not copied.
What this one is. This one is Open Negative Initiative, a free macOS application for converting and editing scanned film: colour negatives, positives and black & white, from a camera scan or a scanner. It runs on its own, with no host application to buy. No film to describe and no stock database. That it runs on macOS and nowhere else disqualifies it here more often than anything written below. macOS 13 or later. Apple Silicon and Intel.
Converting your own film? An alpha costs you little here: your originals are never modified, every earlier version stays downloadable, and no update installs itself.
negadoctor asks you to describe the film
Its model is the printing process: you give it the density of the film base, a paper black, a paper grey, a paper exposure. Those are the right parameters (they are what an enlarger does) and when you know your film they get you there quickly and repeatably.
The cost is that they are parameters of a model. If your scan does not fit it (a partly clipped channel, an unusual stock, a scanner TIFF with a curve already baked in) you are adjusting the description of something instead of the picture in front of you.
Open Negative Initiative asks you to look at the data
There is no film model here at all, deliberately. The inversion is arithmetically dumb: convert to density, and that is the inversion. Everything after it is you placing landmarks on a measurement: a black point where the data starts, which may be below zero, a white point where it ends, and a midtone. Every control on screen is about the negative; none of them is a general photographic control that happens to also work here.
The cost of that is that nothing is inferred for you. There is no film-base detection and no stock database. The six automatic balances read quantiles and model nothing, which is why there are six of them and not one right answer.
darktable negadoctor vs Open Negative Initiative
| Open Negative Initiative | darktable negadoctor | |
|---|---|---|
| Cost | Free, GPL-3.0 | Free, GPL |
| Platform | macOS 13+ | macOS, Windows, Linux |
| Approach | Landmarks on a measurement | A model of the printing process |
| Scope | Film conversion only | Full raw editor |
| Maturity | Beta, one person | Mature, large project |
| Development | Private repository, a complete source archive per version, no pull requests | Public repository, public history, public review, pull requests |
| If its authors stop | Fork the last archive under the same licence | A project with governance carries on |
That Development row is the structural difference between two GPL projects, and leaving it out of a comparison written by one of them would be convenient rather than honest: darktable has a decade of public history and a project that outlives any one contributor, this one has a source archive attached to each release and nothing before it: the model, and what it costs you.
Run your own roll through both
Nothing above decides it as well as the frame that has been giving you trouble, converted in each and judged side by side. Both are free, so the test costs only time. macOS 13 or later. Apple Silicon and Intel. If the answer turns out to be darktable, you have lost a download and nothing else, and the six-step guide walks the steps here, in order, if you would rather see what the trial involves first.
Free, no account, no trial limit.