Lots of misunderstandings and irrelevancies there...and I won't address most of them.
If crop your images to a few hundred pixels, then that might even matter. For example, it might make a difference in this picture of one of my cats.
View attachment 228732
But...that's not really much of a picture now, is it?
As I mentioned previously, at the light levels where those ISOs are used shot noise dominates and read noise is essentially irrelevant. All of the 'improvements' that you think are meaningful (bigger pixels, BSI, per-pixel S/N, etc.) affect read noise, albeit in a functionally insignificant way. Shot noise is dependent on the amount of light being collected by the sensor, meaning all of your arguments about this issue are irrelevant.
If you want less noise at high ISO, you need a bigger sensor to collect more light. It's really that simple, despite your unwillingness or inability to accept that.
While you're hoping for that meaningful improvement in high ISO noise performance in the R7II, you should also hope for unicorn rides and for Canon to set the price for the R7II at $14. Those are all at just about the same level of likelihood, the only difference being the first one is not going to happen because of physics while the last two are just silly fantasies.
I understand the point about shot noise dominating at higher ISOs and that total light collection is a primary driver of overall noise—that’s well established.
Where I think we’re talking past each other is in how that translates to real-world use. In practice, many of us aren’t evaluating downsampled images—we’re working with native files and often cropping, particularly in action and wildlife scenarios. In those cases, how noise presents at the pixel level still matters in the final image.
It’s also not accurate to say that factors like pixel architecture or sensor design are irrelevant simply because shot noise is present. While they don’t change the fundamental photon statistics, they can influence how efficiently signal is captured and how noise manifests in the actual file. Those differences may not be dramatic, but they’re not nonexistent either.
So yes, a larger sensor collecting more total light is one way to improve noise performance—but it’s not the only path. Sensor and processing design choices can influence how efficiently that light is used and how noise ultimately presents in the image.
I’m not disputing the physics there. What I’m saying is that, within those constraints, design tradeoffs still matter—and I believe Canon may have prioritized resolution and broad feature support over high-ISO performance for action use cases.
For example, pixel density itself is a design choice. As I mentioned earlier, a lower pixel count on the same APS-C sensor would increase pixel area, improving per-pixel signal-to-noise and typically resulting in cleaner output at higher ISOs—particularly in scenarios where images are viewed at native resolution or cropped. That would come at the expense of resolution, but it illustrates that noise performance is not solely dictated by sensor size; it’s also a function of how that sensor is configured.
More broadly, the statement that these factors are “irrelevant” because shot noise dominates is an oversimplification. While shot noise sets a fundamental limit, sensor design still affects how efficiently signal is captured and how noise presents in the final image. Those distinctions matter in practice, even if they don’t change the underlying photon statistics.
Part of the challenge here is that cameras like this are being asked to serve multiple use cases—both video and high-speed action photography—which can constrain how aggressively certain optimizations are pursued.
On the “reach” point, I don’t fully agree with the framing either. Reach only has value if the image holds together. If noise degrades detail, then the effective advantage of that reach is reduced. In practice, most photographers are looking for an image that holds up in terms of detail and clarity, not just one that appears closer on paper.
This isn’t just theoretical—in applied optical and sensor system design, these tradeoffs are well understood and routinely balanced depending on the intended use case.
I’m simply saying that I’m not convinced the right balance is being achieved here for action-oriented shooting, and I’m also trying to help clarify the tradeoffs between noise and resolution for those following the discussion.
At the end of the day, clarity—actual image quality—is what most photographers are aiming for.
phillipreeve.net
).
