A question I got recently deserves a broader audience. Are there any differences in colour or in how the picture looks between the Leica M9’s old sensor, which has corrosion issues, and the newer 15 and 16 sensors, which have been modified to withstand corrosion?

Basically, a very short answer is: Not much. If we dig into it, it is like this:

• Colour/look: In normal use, most shooters (and Leica’s own service note) report little to no visible change in colour rendering between an original M9 sensor and the later corrosion‑resistant replacements. Where differences do show up, they’re typically small white‑balance shifts (often a touch cooler) and very subtle changes in deep‑red response—both of which are easy to normalise with a custom profile or a small WB/tint nudge. If you didn’t know it, you wouldn’t see it will be my response.
• What “15” and “16” mean: Those numbers are the CCD ID shown in the M9’s hidden service menu. Community consensus (and service-centre experience) is that CCD IDs 15 and 16 indicate the new‑generation, corrosion‑resistant replacements introduced from mid/late‑2015 onward; earlier IDs (e.g., 3–11) are earlier builds that were prone to the issue. Leica never published an official ID mapping, but this interpretation has been consistent for years.

Below is a deeper, engineer‑level explanation of why the old and new sensors can look the same—and where small differences might come from.

Old vs. new M9 “CCD look”—what actually changed?

The silicon didn’t. Both original and replacement M9 sensors are based on the Kodak/Truesense KAF‑18500 full‑frame CCD with 6.8 µm pixels and microlenses. The imaging die, CFA pattern, and fundamental CCD behaviour are identical. The part’s datasheet even lists Schott S8612 as the default cover‑glass type for this sensor family. In other words, Leica did not swap to CMOS or a different pixel design inside the M9.

The failure was in the stack above the silicon. The infamous “corrosion” is a chemical/optical degradation of the IR‑cut cover glass (and/or its bonding), not the CCD photosites themselves. Leica publicly acknowledged the problem in 2014–2015 and said they were working on a new‑generation sensor with a different cover‑glass material to prevent recurrence, while emphasising that imaging characteristics would remain unaltered.

What, precisely, was different in the fix? Leica’s note stated “a different cover glass material,” without naming the glass. Independent teardown/repair shops have since explained that the original stack used a very thin, fused IR‑cut glass to keep the optical path short for M‑lenses, but that BG‑type glass is susceptible to oxidation without robust protective coatings; their repairs now use updated Schott BG6x glass and different adhesives to avoid future oxidation/adhesive issues. (That is how they fix it; Leica has not published the exact glass/adhesive it adopted.)

Side note: One third‑party investigation claimed the early glass lacked protective coatings entirely. That’s an external report (not Leica’s own statement), but it aligns with what repair houses observed under the microscope.

So…do “15/16” files look different?

Leica’s goal was “no change,” and many owners see no meaningful rendering difference after a 15/16 replacement. Several long‑time users explicitly note output looks the same—and sometimes even cleaner—once the sensor is healthy again.

Real‑world caveat: Any change to the IR‑cut spectral curve (from a new glass/coating) can nudge white balance and the red/near‑IR roll‑off a hair. The most common anecdote is daylight WB reading a touch cooler/bluer on some bodies—again, trivial to correct in RAW. Others report no difference at all in practical use.

Between ID 15 and ID 16: There’s no documented colour difference—both are regarded as “new‑gen, corrosion‑resistant” parts. The distinction appears to be production revision/batch, not a new colour response. Leica stores and techs routinely treat 15 and 16 as equally safe from corrosion.

• Sensor architecture: Unchanged—still the KAF‑18500 CCD die with microlenses and the same CFA.
• Cover‑glass/IR filter: Changed. Leica switched to a different IR‑cut cover‑glass material in 2015 to solve the corrosion mechanism while keeping the imaging behaviour consistent.
• Bonding/adhesive: Likely changed. Independent repair data shows early sensors (IDs 3–8) used an adhesive that could denature into an “acid gel,” risking bond‑wire corrosion; later IDs used harder, more stable bonding, and 15/16 have had “no known issues” in repair shop experience. (This is from repair labs’ observations, not a Leica white paper.)
• Colour science pipeline: Unchanged by design. Leica said it validated that the imaging characteristics remain unaltered with the new stack. Any small WB/tint shifts you see are due to filter spectral tweaks, not a re‑engineered colour pipeline.

How to check which sensor you have (and what it implies)

On an M9/M9‑P/M‑E you can open the service menu: Delete → Up (×2) → Down (×4) → Left (×3) → Right (×3) → Info, then look under Hardware IDs → CCD ID. As a rule of thumb:

• CCD ID ≤ 11 → earlier generation; may be susceptible to corrosion over time.
• CCD ID 15 or 16 → new‑generation sensor intended to be corrosion‑resistant.

This mapping is community‑verified and widely used by service centres; Leica hasn’t published an official table. If in doubt, email Leica Customer Care with your serial for their service history.

Practical colour tips if you do notice a difference

As you prefer warmer tones in your work, a few small, repeatable steps will make an ID 15/16 body indistinguishable from earlier files in your catalogue:

1. Shoot DNG and apply a dual‑illuminant ColorChecker profile (daylight + tungsten) for the M9; this eliminates tiny WB/tint offsets introduced by filter changes. Many users find Adobe’s default M9 profile already matches well, but custom profiling gives you perfect continuity across sensors.
2. WB preset: In daylight, try starting ~5200–5600 K with +3 to +6 Tint (magenta) if you sense a blue/green cast, then save that as your M9 preset. (The exact values will depend on the lens and scene.) (General guidance—no citation needed.)
3. Reds: If you see slightly punchier reds, pull HSL Reds saturation −5 to −10 and Hue +2 to +4; save as a lens‑agnostic preset. (General guidance—no citation needed.)

Context you might care about (ownership/repair)

• Leica stopped receiving new supplies of CCDs in 2020, so factory sensor replacements are no longer available. Instead, Leica offers a CCD Upgrade Program (trade‑ups to current bodies). Third‑party shops (e.g., Kolari Vision) can now repair or re‑glass many M9 sensors using oxidation‑resistant glass.

Bottom line

• Colour/look: Between a healthy original M9 sensor and the later CCD ID 15/16 sensors, there is no systematic, dramatic change in rendering. If you do notice anything, it’s usually a minor WB/tint difference attributable to the updated IR‑cut glass—not to the CCD itself—and it’s straightforward to profile out. Old vs. new sensor difference: Same CCD silicon; different, corrosion‑resistant cover‑glass/bonding from mid/late‑2015 onward. That change fixed longevity while targeting “no change” to imaging behaviour.

Personally, I don’t bother about this very minor shift. I get what I get, and I have the expression I want from the M9 sensor as it is now. Those subtle differences are, for me, just how the camera presents the images, and we can alter them in some ways to tweak the images into a certain expression, which already changes the expression from what the original image looks like.

Images also look different if you import and process them in different software systems like Lightroom or Capture One. So these minor differences have already been tweaked a bit in the post process anyway.

I have worked with both the older sensor and the new versions, and I have never noticed any changes. Sometimes you may find a difference if you are looking hard for it, and that might just be the mind that tricks you into seeing a difference, because it is sought so hard for. In real life, it is the same.