Mount a 50mm f/1.4 behind a 0.71× speed booster and the maths says f/0.99. Point a light meter at the result and you will not get a full stop back. Check the camera’s own LCD and it may still read f/1.4, as though nothing changed at all. Forum threads on DPReview and Blackmagic’s own forum argue about this constantly: some people swear the booster “only boosts T-stop, not F-stop,” others insist the camera is lying about the aperture it shows. Neither is quite right, and the confusion comes from mixing up four different numbers that all get called “aperture” when they answer four different questions. This guide walks through the t-stop vs f-stop speed booster question layer by layer: the geometric calculation that is genuinely real, the glass transmission loss that makes T-stop a little slower, the firmware rule that decides what your screen shows, and the sensor-level effect that can quietly clip even more of the gain at very fast apertures. It builds on the speed booster effective aperture maths and gives you a worked calculator so you can see exactly where the stops go for your own lens and ratio, with every figure checked against the site’s own rounding rules so nothing here drifts from what the calculator itself shows.
Last updated:
Table of Contents
- The Four Numbers Everyone Calls “Aperture”
- Geometric Effective F-Stop: What the Maths Actually Says
- Why T-Stop Is Always a Little Slower
- T-Stop vs F-Stop Calculator
- Why Your Camera Screen Shows the Wrong Number
- The “Missing Stop”: Microlens and Pixel Vignetting
- Depth of Field Still Uses the Full-Frame Equivalent, Not the T-Stop
- Quick Answers About T-Stop vs F-Stop
- Frequently Asked Questions
The Four Numbers Everyone Calls “Aperture”
Every argument about speed booster aperture comes from treating these as one number when they are four. Simply put, a speed booster changes your lens’s physical f-stop for real, but what gets measured, displayed and felt downstream of that change depends on which of these four you are asking about.
| Number | What it measures | Changed by the booster? |
|---|---|---|
| Geometric effective F-stop | Focal length ÷ entrance pupil diameter, the textbook definition | Yes, exactly by the ratio |
| Estimated T-stop | Light actually transmitted through all the glass, including the booster elements | Yes, but slightly less than the geometric drop |
| Full-frame equivalent (DoF) F-stop | Depth of field and total-light noise level, for comparing across sensor sizes | Partly, it factors in the sensor crop too |
| Camera LCD / EXIF f-number | What the body chooses to display and record | Only for f/1.8 or slower lenses, on most adapters |
The rest of this page works through each row with real numbers. If you want the short version of whether the extra stop is real at all, the do speed boosters really add light guide covers that question on its own. This page goes one level deeper: it separates the geometric calculation from what actually reaches the sensor and what the camera reports.
Geometric Effective F-Stop: What the Maths Actually Says
F-number is a geometric ratio: focal length divided by entrance pupil diameter. A speed booster shortens the focal length by its ratio while the physical iris opening stays the same size, so the f-number drops by exactly that ratio too. A 0.71× booster behind a 100mm f/2.0 lens leaves the 50mm iris opening untouched but shortens the system to 71mm, which gives a true geometric f-number of 71 ÷ 50 = f/1.42, as Metabones’ own FAQ confirms when it explains the F-stop increase as real, not a T-stop-only effect.
This is why the maths is not a trick or a marketing rounding. On an optical bench, measuring the diameter of a defocused point of light with a microscope, a 0.71× booster delivers exactly the predicted geometric reduction. Here is the formula and three worked examples using the site’s standard rounding.
| Master lens | Ratio | Geometric effective F-stop | Light change |
|---|---|---|---|
| 50mm f/1.4 | 0.71× | f/0.99 | +0.99 stops |
| 50mm f/1.8 | 0.71× | f/1.28 | +0.99 stops |
| 50mm f/1.8 | 0.64× | f/1.15 | +1.29 stops |
Geometric effective F-stop is what the homepage calculator shows as “effective aperture.” It governs exposure, meaning shutter speed and ISO at a given brightness. It is real, and it is the easiest of the four numbers to verify yourself.
T-Stop vs F-Stop Speed Booster: Why Is T-Stop Always a Little Slower?
No optical glass transmits 100 percent of the light that hits it. Adding a multi-element booster group introduces a small amount of absorption and surface reflection on top of whatever the master lens already loses, so the real-world T-stop sits a little behind the geometric F-stop. This is the gap people are actually reporting when they say a booster “feels” less bright than the formula predicts.
A useful real-world reference point: True Lens Services, which rehouses vintage Mamiya 645 lenses with a built-in 0.71× speed booster for cinema use, documents an original 35mm f/2.8 lens becoming a 25mm f/2.0 lens geometrically, but calibrates the iris ring at T2.3, a difference of roughly 0.4 stops between the calculated F-stop and the measured T-stop across the combined lens and booster elements. That 0.4-stop figure is on the higher end for a cine rehousing with extra glass; a typical consumer speed booster loses somewhat less, commonly cited in the 0.25 to 0.3 stop range.
There is no universal T-stop loss figure, because it depends on the specific glass, coatings and element count in both the lens and the booster. The calculator below lets you enter your own estimate if you have one, such as a figure printed on a cine-converted adapter, and see the resulting T-stop.
T-Stop vs F-Stop Calculator
Enter your lens and booster to see all four numbers side by side: the geometric effective F-stop, the estimated T-stop with an optional glass-loss figure, the full-frame equivalent used for depth of field, and what your camera’s screen is likely to display.
T-stop vs F-stop calculator
Results update as you type. Nothing is sent to a server.
Speed boosters are made for sensors smaller than the lens image circle. On a full-frame or larger sensor the shrunken image circle will not cover the frame, so expect vignetting.
If your adapter’s spec sheet or a measured T-stop figure lists a light loss, enter it in the adapter light loss field to see the estimated T-stop next to the geometric F-stop. Without a known figure, treat anything from 0.25 to 0.4 stops as a reasonable working range for a glass speed booster.
Why Your Camera Screen Shows the Wrong Number
Here’s the part that causes the most arguments: on most adapters, including every Metabones Speed Booster, the camera’s LCD and EXIF data do not always show the boosted aperture at all. Metabones’ own FAQ is direct about this: the boosted aperture is displayed only when an f/1.8 or slower lens is attached. Mount an f/1.4 or faster lens and the screen keeps showing the original, pre-boost aperture, even though the actual aperture in use is faster.
The reason is the EXIF standard itself. It has no defined way to record an aperture faster than f/1.0, and many camera bodies handle sub-f/1.0 values unpredictably. Rather than risk corrupted metadata or erratic exposure behaviour, Metabones’ firmware simply suppresses the boosted value whenever the result would read faster than about f/1.0.
| Lens + ratio | True geometric F-stop | What the camera LCD shows |
|---|---|---|
| 50mm f/1.4 + 0.71× | f/0.99 | f/1.4 (native, unboosted) |
| 85mm f/1.2 + 0.71× | f/0.85 | f/1.2 (native, unboosted) |
| 50mm f/1.8 + 0.71× | f/1.28 | f/1.28 (boosted value shown) |
There is a second layer on top of this: camera bodies round f-numbers to the nearest third-stop, and most do it by rounding up, not down. Metabones’ FAQ lists the standard third-stop sequence as 1.00, 1.12, 1.26, 1.41, 1.59, 1.78, 2.00, each value the previous one multiplied by 1.122462. So a geometric f/1.15 result, such as the 50mm f/1.8 behind a 0.64× booster worked out above, can show up on screen rounded to the nearest step in that sequence rather than the exact decimal. This is normal rounding behaviour, not an error, and it explains why your camera’s displayed aperture rarely matches a calculator’s output to the exact decimal.
If you are metering with a handheld light meter rather than the camera’s own metering, this matters: set your exposure from the calculated effective aperture, not from whatever number happens to be on the camera screen, because for fast lenses that screen number may not have moved at all.
The “Missing Stop”: Microlens and Pixel Vignetting
A separate complaint shows up in DPReview and Blackmagic forum threads: photographers spot-meter a fast lens through a speed booster and measure less than the full predicted stop gain, even after accounting for T-stop glass loss. Part of this is simply measurement noise from spot metering. But there is a real optical mechanism too.
Every pixel on a digital sensor sits behind a microlens with a finite acceptance angle for incoming light. A speed booster compresses the cone of light hitting the sensor into a steeper angle than an unboosted lens would produce. When that cone gets steep enough, typically beyond roughly f/1.2 to f/1.4 equivalent, the outer rays can overfill the microlens’s acceptance angle, so some of the light that physically reaches the sensor doesn’t get converted into signal at that pixel. The practical result looks like a missing fraction of a stop compared to the geometric prediction, on top of the glass transmission loss already covered above.
This effect varies by sensor generation and pixel design, and LensMath doesn’t have measured figures for specific sensors to publish here. The practical takeaway is simpler: if your fastest lens and lowest ratio combination computes to an effective aperture faster than about f/1.2, treat the geometric number as a ceiling rather than a guarantee, and expect the real brightness gain to fall a little short of it.
Does Depth of Field Follow the T-Stop or the F-Stop?
One more mix-up worth closing off here: T-stop and F-stop both describe what happens at the sensor and govern exposure. Neither one is what you reach for to predict depth of field or background blur across formats. That job belongs to the full-frame equivalent aperture, which multiplies the effective aperture by the camera’s own crop factor.
Quick example: a 50mm f/1.8 behind a 0.71× booster on 1.5× APS-C gives a geometric effective aperture of f/1.28, which is what your exposure meter cares about. But depth of field and total-light noise follow the full-frame equivalent, f/1.92, worked out by multiplying the effective aperture by the 1.5× crop. The geometric F-stop, the T-stop and the full-frame equivalent are three different numbers answering three different questions, and that is the whole source of the forum arguments this page opened with.
Quick Answers About T-Stop vs F-Stop
Simply put, what is the difference between T-stop and F-stop on a speed booster?
F-stop is a geometric calculation, focal length divided by entrance pupil diameter, and a speed booster changes it exactly by its ratio. T-stop is the light actually transmitted through all the glass, lens and booster combined, and it is always a little slower than the F-stop because no glass is perfectly transparent.
| Aspect | Details |
|---|---|
| Symptom | Measured exposure gain is less than the calculated stop gain from the ratio |
| Root Cause | Glass transmission loss in the booster elements, plus possible microlens/pixel vignetting at very fast effective apertures |
| Fix | Use the estimated T-stop, not the geometric F-stop, for critical exposure work; meter directly where possible |
| Performance Gain or Result | Typically 0.25 to 0.4 stops less than the geometric prediction, more at very fast apertures |
| Applies To | All glass speed boosters; more noticeable on cine-grade or multi-element adapters |
When Does This Apply?
Mainly for precise exposure work: cinematography with a light meter, matching exposure across multiple lenses, or any case where a third of a stop actually matters. For casual photography, the camera’s own metering compensates automatically and the geometric vs T-stop gap rarely matters in practice.
Frequently Asked Questions
Does a speed booster only increase T-stop, leaving F-stop unchanged?
No. This is a common misconception. The F-stop increase is real and geometric, since focal length shortens while the iris opening stays the same size. T-stop also increases, but by slightly less than F-stop because of glass transmission loss.
Why does my camera still show the lens's original aperture after adding a speed booster?
Most adapters, including Metabones, only display the boosted aperture for f/1.8 or slower lenses. Faster lenses would compute to below f/1.0, which many camera bodies and the EXIF standard itself don't handle reliably, so the firmware shows the original value instead.
How much light does a glass speed booster actually lose to transmission?
There's no single figure, since it depends on element count and coatings, but 0.25 to 0.4 stops is a reasonable working range for consumer and cine-grade boosters based on documented examples.
Can a T-stop ever be faster than the F-stop?
No. It's optically impossible, since the F-stop is the geometric best case and T-stop measures real transmitted light, which is always equal to or less than that geometric figure.
Why does my fast lens seem to gain less than a full stop through a speed booster?
Part of it is normal T-stop glass loss. At very fast effective apertures, typically faster than about f/1.2, microlens and pixel vignetting at the sensor can also clip some of the compressed light cone, reducing the measured gain further.
Should I use the geometric F-stop or the T-stop for exposure?
Use the T-stop when you have a measured or documented figure and precision matters, such as cinematography. For everyday photography, the camera's own metering accounts for the difference automatically.
Does the full-frame equivalent aperture use the T-stop or the F-stop?
Neither directly; it multiplies the geometric effective F-stop by the camera's crop factor. The full-frame equivalent is for comparing depth of field and total light across formats, not for setting exposure.
Why do camera bodies round the displayed aperture differently from a calculator?
Cameras round the f-number to the nearest third-stop step, commonly upward, following a standard sequence. A calculator shows the exact decimal, so small mismatches between the two are normal rounding, not an error.
T-stop vs f-stop confusion comes down to treating one word, aperture, as if it only had one meaning. It doesn’t. The geometric effective F-stop is a real, bench-verifiable number that a speed booster changes exactly by its ratio. The T-stop is what actually reaches the sensor after glass losses, always a touch slower. The full-frame equivalent aperture is a different calculation again, built for comparing depth of field across sensor sizes rather than exposure. And the number on your camera’s screen may not match any of them, for reasons that have nothing to do with the booster’s optics and everything to do with how the EXIF standard and camera firmware handle fast apertures. That’s the whole t-stop vs f-stop speed booster puzzle in one place. Run your own lens and ratio through the speed booster calculator above to see all four numbers for your setup, and bookmark this page next time a forum thread tries to tell you a speed booster “doesn’t really” do what the maths says in 2026.