RomRaider

The car is a USDM 2005 2.5RS. Only mods to the engine are a hybrid intake (MAF has been correctly rescaled) and headers. When you crank the car from a hot start (as in, drive until hot, run into the store, come back out, refire car), the car will run very poorly for a few seconds after restart before recovering to run perfectly fine. Today, I took a log of a bunch of parameters and I'm not entirely sure what do do with the data. I created one graph to demonstrate the problem. I'm hoping you all can give me some insight on how to resolve the problem.

In the graph, from 0 to 2kms, the car is in the On position without the engine running. At 2kms, I cranked the engine for about 1 second. From initial start until about 8kms, the car ran very rough, idled very low, and acted like it wanted to stall. At about 8kms, the car suddenly perked up and began to idle smoothly.

The LC-1 doesn't have valid output for several seconds after power is cut/reapplied. The heater circuit has to finish initializing before it spits out data. During this time, it pegs the output at max.

I've wired my LC-1 not to go off when the engine is cranking. I turn on the power, let it warm up and begin outputting valid data, then start the engine. This results in continuous valid data throughout the process.

The unit 'kilo-mili-seconds' sounds fun First I've read kilometres. Were hot starts always like this?

Hot starts have always been hard. When I first got it, I had to learn to let it just sit without touching it for 5 seconds after a stall before I tried to refire it or it wouldn't start. That's actually still true but separate from the issue here. The only other weird thing is that after I have ~200mi after an ECU reset, it stops pulling this crap and maintains a stable AFR through cranking and the first few seconds of running.

There are many things to consider..
I wonder if this issue is header-related. Or injector latencies that're way out
It would be good to log A/F learning, A/F Correction and A/F Sensor#1.

AFC will 0 because the CL/OL status is 7 throughout this whole event. AFL will be whatever it was when the car was shut off (usually -1.4 to -2.6). AF1 should match the LC-1 pretty closely under those conditions.

so, it won't come in play at >8000ms where AFRs become right... I've logged hot and cold start some time ago, but deleted those logs and can't remember how it looks like.
How about battery voltage after startup?

Correct, the car stayed CL/OL status 7 for the entire duration of the log. Where the AFR recovers to stoich after 8kms, that's just because the MAF scaling is bang-on in.

Battery voltage is an interesting idea. That's one I'll definitely check both with a multimeter and by logging.