Running Multiple Tune Files: 91, 93, E85 and Valet Without Losing Track

Almost nobody who tunes a car ends up with exactly one calibration. It starts as one, then a road trip needs a lower-octane version, then someone finds cheap ethanol, then a friend borrows the car and you wish there were a gentler file. Within a year there are five files in a folder and a genuine uncertainty about which one is actually loaded.

As of September 2026 the fix is not fewer files — it is treating them as one family with one parent rather than as five separate tunes. That distinction is the whole subject of this article, because it decides whether a fix applied in June is present in the file you flash in December.

Why octane files exist at all

Pump fuel is not a consistent product. It varies by region, by season, by brand and by how long it has been in a station's tank. A calibration has to decide how much spark advance to command, and the amount that produces the most torque is usually a little more than the fuel can safely support.

Higher octane resists detonation better, which lets the calibration advance timing further under load before combustion turns unstable. That is essentially the whole story of an octane-specific tune: the same car, the same air model, the same transmission strategy, and a different high-load spark table.

The engine has a defence if you get this wrong. Knock retard pulls timing when the sensors hear detonation, so a 93-octane file on 91-octane fuel does not immediately destroy anything — it just spends the whole drive fighting. That protection is real but it is a safety net rather than a strategy, and it is why how much knock retard is normal is worth reading before you assume your files are interchangeable.

What should differ, and what absolutely should not

This is the part that separates a well-managed family of files from a folder of unrelated tunes.

Area of the calibration	Differs between octane files?	Note
High-load spark tables	Yes	The core reason the files exist
Borderline knock / octane-scaled timing	Yes	Where the platform offers a second spark axis
Power-enrichment fuelling	Sometimes, slightly	Higher-octane files often run marginally leaner at full load
Boost targets, on a boosted car	Often	Frequently the largest real difference
Air model — VE or MAF calibration	No	This describes the engine, not the fuel
Idle and cruise behaviour	No	Same engine, same requirements
Transmission shift pressure and torque limits	No	Unless the power level genuinely changed
Injector characterisation	No	Unless the fuel itself changed, as with ethanol
Speedometer, tyre size, gear ratio	No	Vehicle configuration, not fuel

The rule that falls out of this table is the important one: there should be one master calibration and a small set of deliberate deltas from it. When you fix a drivability problem, you fix it in the master and regenerate the variants. When you fix it in whichever file happened to be loaded that day, you have created a divergence that will resurface months later as "the 91 tune drives worse and I don't know why".

Ethanol is the exception that proves the rule. A high-ethanol file legitimately differs in far more places — fuelling everywhere rather than only at load, cold-start enrichment, injector sizing headroom — because the fuel itself is chemically different rather than merely more knock-resistant. That is covered properly in E85 and flex-fuel conversion, and it is the main reason an ethanol file is better treated as a sibling calibration than as a variant.

The four files worth actually having

Most owners are well served by a small, deliberate set.

The stock read. Never flashed for driving, never edited, never deleted. It is your route back to a factory vehicle for a dealer visit, a warranty conversation or a sale. If you have not archived one, that is the single most urgent item on your list — the reasoning is in tune file management and the recovery options when it is missing are considerably worse, as lost your stock tune file explains.

The primary calibration. The fuel your car normally runs, tuned properly, verified on datalogs. This is the master that everything else derives from.

A conservative travel file. Timing pulled back at load for a trip through regions where you cannot count on fuel quality, or for a heavy tow day where heat is the limiting factor rather than octane. This is the same discipline described in tuning a truck for towing, applied as a separate file rather than a compromise in the main one.

An ethanol file, if the hardware supports it. Only if your fuel system genuinely has the headroom — an ethanol blend needs roughly a third more fuel volume for the same mixture, which puts real demand on injectors and pump, a limit discussed in injector duty cycle and fuel pump headroom.

A valet or reduced-power file is a legitimate fifth, though on many platforms a torque limit is a cleaner way to achieve the same thing than an entirely separate calibration.

Naming, and why it is not a trivial detail

The single most common failure in this whole subject is not a bad tune. It is flashing the wrong file.

A filename should tell you everything you need without opening anything:

2026-09-11-silverado-93oct-rev4-trans-shift-fix.hpt

Date first so a sorted folder is chronological. Vehicle, because you will eventually tune a second car. Fuel, because that is the variable that decides whether this file is safe today. Revision number, because "final" and "final2" are how everybody loses track. A short purpose note, because in eighteen months you will not remember what rev4 changed.

Alongside the folder, keep a plain text log with one line per flash: date, file written, odometer, why. It takes ten seconds and it is the only thing that reliably answers "what is actually in the car right now" when something goes wrong. Digital storage is cheap; reconstructing a history is not.

Switching safely

Each flash is a few minutes of writing to a controller that must not lose power partway through. The failure mode is rare and unpleasant, and the precautions are simple.

1. Put a charger on the battery. Not because the flash takes much current, but because a marginal battery during a long write is the classic way to create an unresponsive controller.
2. Engine off, key on, accessories off. Follow the tool's prompts exactly rather than from memory.
3. Do not switch fuel and tune in one motion. Flash first, then fill, or fill then flash before driving hard. A tank of 93 in a car running a 91 file is harmless. The reverse is what you are trying to avoid.
4. Write down what you flashed. Immediately, in the log, before you drive away.
5. Verify on the first drive. A short datalog after any file change is how you learn that the right file is loaded and behaving, rather than finding out at full throttle two weeks later.

That verification step deserves more weight than it usually gets. The point of a datalog after a switch is not to re-tune anything — it is to confirm the car is doing what the file says it should, which is exactly the reading described in reading your first wide-open-throttle pull.

When a family of files is the wrong answer

If you are flashing more than about once a fortnight, the files are managing you rather than the other way round.

The alternative is a calibration that adapts. A flex-fuel setup with an ethanol content sensor lets one file blend between low and high ethanol strategies continuously, including at partial mixes that no discrete file can handle. It costs hardware and tuning time up front and removes the switching problem permanently.

On a boosted car there is a second option: a single file with a switchable boost or timing level, where supported by the platform. That trades some flexibility for never touching the controller again.

Both of those are more work initially and less work forever. Which of the three approaches suits you is genuinely a question about how you use the car, not about which is technically superior.

Keeping the family in sync

The failure this whole article exists to prevent is drift — three files that started as siblings and are now three different tunes because each received fixes the others did not.

The discipline is simple to describe and easy to skip. Make changes in the master. Regenerate the variants from it. Compare the variants against the master before flashing to confirm the only differences are the ones you intended. Re-verify with a short datalog on each variant, not only the one you drive daily.

That comparison step is where a second pair of eyes earns its money. TuneVault reads HP Tuners VCM Editor tables and audits them, which means it can tell you what a variant actually differs from its parent in and flag changes that have no business being there — the transmission fix that only made it into one file, the air model that quietly diverged, the timing that crept past what the fuel supports. A guided audit is $39, or from $19 a month if you are maintaining a family of files continuously; the detail is on the pricing page.

The goal is not to have fewer calibrations. It is to be able to say, with confidence, exactly what is in the car and exactly how it differs from everything else you own.

Frequently asked questions

How many tune files does a typical modified car end up with?

Usually three to five. A stock read that never changes, a primary calibration for the fuel the car normally runs, a lower-octane version for travel or bad fuel, and often a high-ethanol file if the car is set up for it. Some owners add a reduced-power valet or winter file. Beyond about five the set becomes hard to keep genuinely in sync and a flex-fuel strategy is usually the better answer.

What actually changes between a 91 and a 93 octane tune?

Mostly spark timing, and usually only in the high-load region. Higher octane resists detonation better, so the calibration can advance timing further under load before combustion becomes unstable. Fuelling under power enrichment is often slightly different too. Idle, cruise, transmission behaviour and the air model should be identical between the two files — if they are not, you have two unrelated tunes rather than one tune with two octane variants.

Can I switch tune files whenever I want?

Physically yes on most platforms, within whatever licensing your interface enforces. The constraint is practical rather than technical: each write takes a few minutes with the engine off and a healthy battery, and it increments the controller's flash counter. Switching once per tank is routine. Switching at a fuel pump every few days is a sign the car wants a flex-fuel setup instead.

What happens if I run the wrong tune for the fuel in the tank?

Running a lower-octane file on higher-octane fuel is harmless and simply leaves power on the table. The dangerous direction is the other one — a high-octane calibration on low-octane fuel asks for timing the fuel cannot support, and the knock control system spends the whole time retarding to protect the engine. That works, until a hot day or a heavy load outruns it.

Do I need separate files for E85 or can I use flex fuel?

Both work and they suit different cars. Separate files are simpler and need no extra hardware, but they require you to flash whenever you change fuel and they cannot handle a partial mix. A true flex-fuel setup with an ethanol content sensor lets the calibration blend between strategies continuously, which is far better if you actually switch fuels often.

How should I name my tune files?

Put the information you need at a glance in the filename itself: date, vehicle, fuel, revision and a one-word purpose. Something like 2026-09-11-silverado-93oct-rev4-trans-fix reads correctly in a sorted folder years later. Trusting your memory or a tool's default naming is how people end up with four files and no idea which is current.

Does flashing between files wear anything out?

Not meaningfully in normal use. The controller's memory is rated for far more write cycles than any enthusiast will use, and the practical risks are elsewhere — an interrupted write on a weak battery, or losing track of which file is loaded. Use a charger for every flash and keep a written log, and frequent switching is not a durability concern.