What do I need to do with the light metering to get these highlights under control?
It's not really a metering issue, but mostly a scanning issue. The way these images were scanned has chopped off the higher part of the curve, leaving you with highlights that were blown out. They are not blown out on the actual negatives, I can assure you. It's nigh impossible to really blow out highlights on color negative!
Your metering of the scenes is certainly on the generous side, but overall this is preferable over the opposite, which would have left you with large, blank shadow areas. Contrary to the dense highlights, those shadows would be unrecoverable. With these particular negatives, getting the desired result is just a matter of re-scanning them so that the entire density range is scanned (nothing is being truncated) and then adjusting the curves until you're happy.
To illustrate, this is the Waffle House image with a curve adjustment as shown:
I suspect that the tonal values are more like you would have imagined them (of course I'm merely guessing at your intentions, so feel free to point it out if I got it all wrong!)
What I've done is two things:
1: suppress the green curve; the overall very green appearance is an artefact of Vision3 film run through processes (both developing and scanning) that are geared towards C41 color negative. This can mostly be corrected for as I've done here.
2: Suppress the high values very drastically to reduce overall contrast, especially in those high values.
This didn't yield a perfect outcome, and as indicated above, this is because of choices made in scanning the negative. These choices may have been made automatically by the scanning software. In this case it's a matter of choosing a different route to scan these; a route that offers more manual control.
The effects of the unfortunate scanning choices are visible in how the highlights render. Note in particular:
A: The coarse/'pixelated' rendition of the transitions in the upper highlights; e.g. in the "Waffle House" signage.
B: Color crossovers esp. in the green/magenta channel.
This is due to severe compression of the upper part of the curve that results in an inability to render tonal transitions smoothly. The problem is mostly apparent in the green channel because that's the "iffy one" when it comes to Vision3/ECN-2 film - and it's in my experience exacerbated if this film is run through C41 developer (but it's possible to mostly fix the issue in digital post processing - but much less so in optical enlargement).
A further problem is the limited bit depth I worked with here; if you have a 16-bit TIFF or PNG original scan to work with, you may get somewhat better results. But the actual scan will still be a limitation.
So all considered, this is mostly a problem with the digital part of the process. This is not to say that
@MattKing is wrong in suggesting a different approach to metering - what he says about striking a compromise certainly is true. However, in this particular case and the metering choices that were made for these exposures, a better end result is still very well possible thanks to the latitude of the film used. Had you exposed these a stop or so less, the 'problem' of the highlights would also have been slightly reduced, although if the negatives were scanned in a similar way, the result would still have been compromised. So taking more control over the scanning step is critical in getting closer to where you want to get.
