March 12, 2026 | Samuel Crowe
At some point in almost every camera conversation, resolution comes up. Someone mentions a number (8K, 12K, 17K) and the room treats it like a score. Higher is better. More is more. The assumption is so common that it barely gets questioned.
It's wrong, and the industry knows it's wrong, and it keeps happening anyway. This article is about why.
Resolution is a real specification that describes something real about a sensor. But it describes only one thing, and it trades against several others. Once you understand what those trade-offs actually are, the number on the spec sheet starts to look a lot less important than the decisions behind it.
Why the industry keeps pushing higher resolution
The answer is pretty simple: resolution is the easiest number to put on a box, and it's a huge-sounding win for clients.
Pitching an 8K deliverable to a studio of execs and number-punchers sounds like an unbelievable win when all they really understand is that "4K is better than 1080p so 8K must be great!" Pitching a 1080p deliverable with 16+ stops of dynamic range won't get you a standing ovation and thunderous applause.
Dynamic range requires explanation. Noise performance depends on lighting conditions. Color science takes years to develop an eye for. Resolution is just a number, so manufacturers compete on the number they can most easily communicate, and the market rewards them for it.
There's also a real pull from the display side. The transition from HD to 4K was visible and meaningful. The push from 4K to 8K is a lot harder to see in practice, but it gives manufacturers a story to tell and retailers a reason to sell new TVs. Streaming platforms publishing acquisition guidelines with resolution minimums has added to this. Those specs get misread as "higher is better" when they're actually just setting a floor.
On the production side, VFX-heavy workflows created a perception that more resolution always helps post-production. That claim deserves a lot more scrutiny than it usually gets, and we'll come back to it. For now, it's worth noting that it's been repeated enough times that a lot of people take it as true without really examining it.
The result is a cycle: manufacturers build higher-resolution sensors, market them on that number, and producers chase the spec because they've been told it matters. And hidden inside that cycle is a real cost that rarely gets discussed.
If resolution is everything, why does everyone want to shoot on an ARRI?
The ARRI ALEXA 35 platform is arguably the most sought-after cinema cameras on the market right now. Working DPs at the top of the industry put them at the top of their wish lists. It has been used on major studio films, prestige television, and high-end commercials since its release.
It shoots 4K.
ARRI's first true 4K camera for consumer-purchase came out in 2018 with the release of the ALEXA LF, (the only earlier ARRI with a resolution greater than 4K was in 2014 with the ALEXA 65 at 6K resolution, which is rental-only) while RED, Sony, and others have been offering 6K, 8K, and beyond for years prior. ARRI has been sitting at a fraction of those numbers. By pure spec-sheet logic, they should be losing the market because they're behind the technology. They're not. The ALEXA 35 and its predecessors dominate the high end of the industry while being significantly outgunned on resolution by cameras that cost less.
That's not a coincidence or a brand loyalty story. It's a design philosophy. ARRI has been consistent about this for a long time: they optimize for what each individual photosite does, not how many photosites they can fit on the sensor. Resolution comes second to the quality of the light capture itself.
The fact that the camera nearly every serious cinematographer wants to shoot on tops out at 4K while its competitors are marketing 12K is the clearest possible indication that resolution and image quality are two different things. The question is why.
What resolution actually is, and what it isn't
Resolution describes how many individual photosites exist on the sensor. More photosites means a finer sampling grid. That's all it means on its own.
It says nothing about how much light each photosite collects. It doesn't describe how the sensor handles noise, what dynamic range it can achieve, or whether the lens feeding it can actually resolve that level of detail. Resolution is a count. Image quality is a system.
The concept that ties this together is pixel pitch: the physical size of each individual photosite. On a fixed sensor area, more resolution means smaller photosites. You can't change one without changing the other. A 60-megapixel sensor on the same physical die as a 24-megapixel sensor has smaller photosites. That's the trade, and everything that follows from it is just physics.
Key Takeaway
On any fixed sensor size, more resolution means smaller photosites. The two are in direct competition. Understanding that trade-off is where the resolution conversation actually starts.
How higher resolution works against you
Light performance and noise
Photosites collect photons. A larger photosite has a larger surface area and collects more photons during any given exposure. This isn't an engineering problem you can solve with better software or a faster processor, it's just geometry. A bigger bucket catches more rain.
When you shrink photosites to fit more of them on the same sensor area, each one is collecting fewer photons. Fewer photons means a weaker signal. A weaker signal has noise that becomes proportionally more visible. This is why cameras with larger photosites consistently outperform higher-resolution cameras in low light, even when those higher-resolution cameras are on nominally the same format. It's also why shooting a 60-megapixel stills sensor in a dark restaurant and shooting a 24-megapixel sensor in the same conditions produces noticeably different results. Not because of resolution, but because of what the resolution cost.
This is the quiet truth about ARRI's sensor philosophy. When they choose fewer, larger photosites, they're choosing cleaner signal, better low-light performance, and more usable shadow detail. The 4K number is a side effect of that choice, not a limitation.
Dynamic range
Each photosite has a limit to how many photons it can hold before it clips and registers as pure white. This is called the full well capacity, and it scales with the physical size of the photosite.
Smaller photosites fill up faster in bright conditions. That means they clip sooner in the highlights, which compresses the top end of the sensor's dynamic range. So packing more resolution into the same sensor area doesn't just increase noise in the shadows... it also costs you headroom in the highlights. You're getting squeezed from both ends.
Dynamic range is one of the most important image quality specifications in cinema. It's what lets a camera hold detail in a window behind an actor, or keep texture in a bright sky while preserving shadow detail in a face. Giving up stops of dynamic range for a higher pixel count is a real cost, and it's one that never makes it onto the marketing materials.
Diffraction
When light passes through any opening, it bends at the edges. The smaller the opening, the more pronounced that bending becomes relative to the size of the hole. At very small apertures, this effect known as "diffraction" spreads each arriving point of light into a soft disk before it reaches the sensor. That disk is called the Airy disk.
At wide and moderate apertures, the Airy disk is small enough that the sensor resolves it cleanly. It doesn't affect the image. As you stop down further, the disk grows. At some point it becomes large enough that the sensor can no longer render a point of light as a point, and sharpness starts to fall off from the small-aperture side.
Here's the part that directly connects to resolution: when that Airy disk becomes visible depends entirely on the size of the photosites. Smaller photosites on a high-resolution sensor start resolving the diffraction pattern at a wider aperture than larger photosites on a lower-resolution sensor. A high-resolution camera on the same size sensor becomes diffraction-limited sooner. The apertures you'd normally stop down to for depth of field control are the same apertures where diffraction starts hurting you.
If you've read Every Lens Has a Sweet Spot, this is the same diffraction ceiling discussed there. The connection to resolution is that a higher-resolution sensor pulls that ceiling down to a wider aperture, which tightens the range of stops where your lens is performing at its best.
Key Takeaway
Higher resolution on the same sensor size makes diffraction visible at wider apertures. You're not gaining optical freedom by shooting higher resolution. In some cases, you're losing it.
The lens ceiling
Resolution on the sensor only matters up to the point where the lens can actually resolve detail onto it. A sensor is a sampling grid. If the optical system feeding it can't produce detail finer than a certain threshold, adding more photosites below that threshold doesn't capture anything new. It just records the same softness at a higher file size.
High-end cinema glass at its sharpest working aperture may resolve enough detail to meaningfully feed a 6K sensor. The same image through a soft consumer zoom will look like a high-resolution capture of a soft image, because that's exactly what it is. You cannot resolve detail the lens didn't deliver in the first place.
Pairing a 12K camera with a lens that can't support 12K doesn't produce a 12K image. It produces an expensive, very large file of a soft image. Resolution is the last variable to optimize. If the glass isn't right, if the exposure isn't right, more photosites simply record the problem in higher definition instead of fixing it.
Compression and codec performance
More resolution means more data per frame. More data at a fixed bitrate means more compression per frame. More compression means more artifacts and less retained detail per recorded pixel.
A 4K recording at a given bitrate is a better-encoded 4K image than an 8K recording at the same bitrate, because the codec has more bits available for each pixel. The 4K camera is doing less work to represent each pixel faithfully. The 8K camera is spreading the same number of bits across four times as many pixels.
This matters for image quality, but it also has downstream effects that rarely show up in the resolution conversation: larger files, more storage, more expensive media, and heavier editing timelines. None of these are reasons to avoid high resolution when it's genuinely warranted. But they're all part of the cost, and they should be weighed against an actual benefit instead of being an assumed one.
When higher resolution actually helps
Downsampling
Shooting above your delivery resolution and then downsampling to your output can improve apparent sharpness. When you reduce a 6K image to 4K delivery, the scaling process averages multiple photosites together, which can smooth out sub-pixel noise and produce a cleaner final result than a natively captured 4K image on the same sensor die.
This is a real and legitimate technique. The important caveat is that it exists largely to compensate for the weaknesses that come with small photosites. Larger photosites on a natively lower-resolution sensor may already be producing a cleaner image without the downsample step, because the signal was stronger at capture. ARRI's 4K image often looks better than a competitor's 8K image downsampled to 4K for exactly this reason.
Downsampling is also only as good as the debayering and scaling pipeline allows. Poor debayering on a 6K capture can produce a worse 4K output than a well-tuned native 4K camera. The technique works, but it's not a guaranteed upgrade just by virtue of having more pixels to start with.
Reframing in post
Shooting above your delivery resolution gives you real pixel headroom to reframe, stabilize, or correct a slightly loose shot in post without any resolution loss. This is practically useful, particularly in documentary work, run-and-gun production, or single-camera narrative where having a little room to adjust in the edit has direct value.
The honest limit here is that reframing only works as far as the optical quality of the original frame allows. If the corners of your captured image are soft because of vignetting or focus falloff, punching into that area reveals the softness you were outside of before. The headroom for reframing is determined by optical quality, not just resolution.
Large format delivery
Genuinely high-resolution capture has a real home in large format projection, IMAX origination, or any context where the final display surface is large enough that a viewer at the intended distance can distinguish detail that lower resolution wouldn't provide. If you're shooting material specifically for a screen that's forty feet wide, the math for how much resolution you need looks different.
However, this is also exactly why IMAX has a list of approved cameras (see Filmed for IMAX camera list), and why there's no 12K Super35 sensors on the list. A large factor in the image quality that IMAX puts their name on, is exactly that... image quality. A 12K Super35 does not have the same image quality that a 6K 65mm sensor has because of the photosite size and pixel pitch. IMAX is not interested in displaying super crisp digital noise.
This is a narrower use case than the general resolution conversation acknowledges. Most theatrical, streaming, and broadcast delivery is not operating in a range where the difference between 4K and 8K acquisition is visible to anyone watching the finished product.
My wife and I watch most things on a 70-inch TV, sitting about 8 feet back, with a 1080p Apple TV. The lack of resolution has never once pulled either of us out of a film. What pulls you out of a film is bad light, bad audio, a soft performance, or a story that isn't working. Nobody watches a gripping scene and thinks "this would have been so much better in 4K."
Key Takeaway
Higher resolution has real value in specific contexts: downsampling workflows, reframing headroom, and large format delivery. None of these are general arguments for always shooting at higher resolution.
Common misconceptions
"We need it for VFX"
This is probably the most repeated justification for high resolution in professional production, and it doesn't hold up to scrutiny in most cases.
The claim goes that VFX pipelines require high-resolution plates for compositing, tracking, and screen replacements. The reality is that the overwhelming majority of VFX work is done internally at 2K. Most major VFX pipelines (including the ones on large studio films with enormous VFX budgets) ingest and work at 2K or 2.5K because the rendering, compositing, and iteration cost scales with resolution. There's no practical benefit to compositing a shot at 8K that you can't achieve at 2K with a well-run pipeline.
Where the argument actually has some validity: chroma keying to ensure that every fine detail is properly-evaluated, or IMAX-specific VFX work, can benefit from resolution headroom. But this is a narrow subset of VFX work, not the general case. A VFX supervisor asking for higher resolution is usually asking for a specific production reason. "VFX wants it" as a blanket justification for always shooting 8K is not how these pipelines actually function.
"More resolution means sharper images"
Sharpness is a perceptual quality. It's influenced by how good your 1AC is, optically-based on-set decisions, optical quality, contrast, motion, and resolution. Roughly in that order of impact.
A well-exposed, well-lit image with good tonal contrast and a sharp lens will look sharper than a poorly exposed image on a higher-resolution sensor. This isn't a subjective preference but instead how our eyes reads detail. Perceived sharpness is driven at least as much by edge contrast as by fine pixel-level detail. You can demonstrate this by slightly increasing local contrast on a soft image and watching it appear sharper without changing the resolution at all.
Resolution is a real contributor to sharpness, but it's the last one on the list, and it's the one that matters least when the others aren't handled first.
"It future-proofs the footage"
The argument: shoot in 8K now so the footage holds up when 8K delivery becomes standard.
This requires believing several things simultaneously: that your footage will still be in active distribution when that standard arrives, that your lenses actually resolved to 8K in the first place, that your codec and pipeline handled the data well enough to preserve that resolution, and that the image holds up at 8K given all the photosite trade-offs described above. A noisy, compressed 8K image doesn't future-proof anything.
Future-proofing is a real concept in archival and preservation work. As a production argument for shooting maximum resolution on a commercial, a music video, or a narrative feature with a normal distribution window, it's mostly a way of making a spec-sheet decision sound like a strategy, and let's be honest... clients eat it up because they don't understand what's really going on under the hood.
Summary
Resolution is one variable in a system that has several, and it competes directly with others on any fixed sensor size. Smaller photosites mean weaker light collection, more noise, reduced dynamic range, and earlier diffraction limits. You don't get those things for free just because the resolution number went up.
The manufacturers chasing resolution are responding to market pressure. The ones holding back are responding to physics. ARRI's position in the industry isn't an accident; it's what happens when you decide that the quality of what each photosite captures matters more than the count of how many you have.
Resolution matters. But it matters least when everything else isn't already right, and it is not a substitute for good glass, good exposure, or good light. The spec sheet tells you how many photosites are on the sensor. It doesn't tell you anything about what those photosites are actually capable of.
Written by
Sam Crowe
Director of Photography · Colorist · Camera Operator
I'm a cinematographer based in Nashville with over a decade of experience shooting across the Southeast. I care about images that serve the story — not the other way around. Outside of production, I spend a lot of time thinking about the technical side of the craft and building tools that help other cinematographers work smarter on set.
Frequently Asked Questions
No. Perceived sharpness is driven primarily by contrast and optical quality. Resolution is one contributor, but it's the last one that matters and the first one that gets optimized. A well-shot image on a 4K sensor with good glass will look sharper than a poorly exposed image on a 12K sensor with soft glass. More pixels don't fix problems that happened upstream of the sensor.
Because more resolution on the same sensor area means smaller individual photosites. Smaller photosites have less surface area to collect photons, which produces a weaker signal. A weaker signal has proportionally more noise. This is physics, not an engineering failure. A sensor with fewer, larger photosites collects more light per photosite and produces a cleaner result in the same conditions.
Pixel pitch is the physical size of each individual photosite on the sensor. On a fixed sensor area, higher resolution means smaller pixel pitch. Pixel pitch directly determines light-gathering capacity, noise performance, dynamic range headroom, and diffraction sensitivity. It's arguably a more useful specification than resolution for evaluating real-world image quality, but it rarely appears on marketing materials.
In most cases, no. The majority of VFX pipelines work internally at 2K or 2.5K, regardless of acquisition resolution. Compositing, rendering, and iteration all scale in cost with resolution, and there's no general benefit to doing that work at 8K. There are specific cases (extreme crops within a comp, IMAX-format VFX work) where higher resolution plates are genuinely useful. But "VFX wants more resolution" as a blanket production argument doesn't reflect how most VFX pipelines actually operate.
Because ARRI has made a deliberate design choice to prioritize photosite quality over photosite count. Fewer, larger photosites mean better light collection, cleaner noise performance, and more dynamic range. The 4K output of an ALEXA 35 is produced by photosites with significantly more light-gathering capacity than the photosites on a competing sensor with four times the pixel density. The result consistently outperforms higher-resolution cameras on the metrics that actually matter to cinematographers.
Sometimes, with caveats. Downsampling from a higher capture resolution to your delivery resolution can produce a cleaner image by averaging out sub-pixel noise during the scale. But this benefit exists primarily because of the weaknesses that come with smaller photosites -- and a sensor with larger photosites may already produce a cleaner image natively, without the downsample step. Downsampling is a technique for managing the trade-offs of high resolution, not a general argument for always capturing at maximum resolution.