Detailed Guide Coming Soon
We're working on a comprehensive educational guide for the ISO Noise Calculator in your language. The content below is shown in English.
Apa itu ISO Noise Calculator?
▾
The ISO Noise Calculator estimates the signal-to-noise ratio (SNR) and effective image quality at different ISO settings for a given camera sensor. ISO sensitivity (originally defined for film by the International Organization for Standardization in ISO 5800:2001 and updated for digital sensors in ISO 12232:2019) measures the sensor's amplification of light signals. Every doubling of ISO (e.g., ISO 200 to ISO 400) represents one stop of additional sensitivity — but also one stop of reduced signal-to-noise ratio, because the amplification increases the sensor's own electronic noise alongside the image signal. Modern digital sensors have a base ISO (typically ISO 100–400) at which they achieve maximum dynamic range and minimum noise. Above base ISO, the sensor applies analog gain (preferred) or digital gain (worse), amplifying the signal but also amplifying noise. The SNR at a given ISO depends fundamentally on the number of photons captured — a sensor with more exposure (more photons) always has better SNR regardless of ISO. ISO is a compensation tool for inadequate exposure, not a substitute for it. The 'ISO invariance' concept, relevant to many modern Sony and Nikon sensors, means that lifting shadow exposure in post from a low-ISO image can produce similar or better results than shooting at a higher ISO — because the noise added by high ISO analog gain can be equivalent to or worse than the noise introduced by digital brightening of a low-ISO exposure. Understanding ISO noise helps photographers choose optimal settings for low-light photography, compare camera sensor performance across models, and make informed decisions about noise reduction in post-production.
PrimeCalcPro provides professional-grade tools trusted by businesses and academics.
Rumus
▾
SNR (dB) = 20 × log10(Signal / Noise)
Noise ≈ √(Shot Noise² + Read Noise² + Dark Current²)
Shot Noise = √N_photons (fundamental quantum noise)
Read Noise = inherent amplifier noise (electrons RMS, specified by manufacturer)
ISO equivalent DR loss = 1 stop per ISO doubling above base ISO
Effective ISO Noise Score = Read Noise (e-) × (ISO / Base ISO)^0.5Keterangan variabel
▾
| Simbol | Nama | Satuan | Deskripsi |
|---|---|---|---|
| SNR | Signal-to-Noise Ratio | dB | Ratio of signal power to noise power in decibels. Higher SNR = cleaner image. |
| N_ph | Photon Count | photons/pixel | Number of photons captured per pixel during exposure. Increases with exposure time, aperture, and scene brightness. |
| RN | Read Noise | electrons (e-) | Inherent electronic noise of the sensor readout circuit, independent of exposure. |
| FWC | Full Well Capacity | electrons (e-) | Maximum number of electrons a pixel can hold before clipping (highlights blow out). |
Cara ISO Noise Calculator
▾
- 1Step 1: Identify your camera's sensor specifications: base ISO, read noise (e-), full well capacity (e-), and pixel pitch.
- 2Step 2: Estimate photon capture: proportional to aperture area, shutter speed, scene luminance, and pixel pitch squared.
- 3Step 3: Calculate shot noise: √(photon count). This is irreducible quantum noise.
- 4Step 4: Add read noise in quadrature: total noise = √(shot noise² + read noise²).
- 5Step 5: SNR = signal / total noise. Convert to dB: SNR_dB = 20 × log10(SNR).
- 6Step 6: Each stop of ISO increase above base approximately halves the SNR (reduces it by ~6 dB). Modern cameras maintain usable quality to 2–3 stops above rated high ISO.
Contoh Terpecahkan
▾
6400/100 = 64 = 2^6 (6 stops). Each stop reduces effective SNR by approximately 3 dB. At ISO 6400, the A7 III's measured SNR drops to approximately 36 dB — good for 8×10 prints but challenging for large commercial use.
A full-frame sensor with 8μm pixels collects proportionally more light per pixel than APS-C with 5.5μm pixels. More photons = less noise relative to signal. Full-frame SNR advantage at high ISO: approximately 0.8–1.3 stops depending on specific sensors.
At ISO 3200, the Canon R5's read noise contribution becomes negligible compared to sky glow photon shot noise. Going above ISO 6400 adds amplifier noise without improving effective SNR for dark sky imaging — 1600–3200 is the 'read noise floor' crossover.
ISO-invariant sensors like the Sony A7R IV have such low read noise at base ISO that digital brightening adds minimal noise compared to analog gain at high ISO. For critical shadow recovery, shoot at base ISO and lift in post rather than using very high ISO.
Aplikasi nyata
▾
Photographers selecting cameras for specific shooting conditions (wedding, sports, astrophotography).
Cinematographers evaluating camera sensor performance for low-light narrative or documentary work.
Camera reviewers benchmarking sensor noise performance across competing models.
Astrophotographers planning exposure sequences for optimal noise in stacked images.
Kasus khusus
▾
Expanded ISO settings
Most cameras offer 'expanded' ISO settings (often labeled H1, H2 or numerical values beyond the rated maximum) that apply digital gain beyond the analog circuit's range. These settings dramatically increase noise and reduce dynamic range, and should generally be avoided except in extreme necessity. The rated maximum ISO is the highest value where performance is considered acceptable by the manufacturer.
Stacking for noise reduction
Multiple exposures stacked in post-processing reduce random noise by √N (where N is the number of frames). 4 identical exposures averaged together halve the noise (2 stops improvement). This technique — image stacking — is widely used in astrophotography, microscopy, and scientific imaging to achieve SNR levels impossible in a single exposure.
When input values approach zero or become negative, the Iso Noise Calculator
When input values approach zero or become negative, the Iso Noise Calculator calculation may produce undefined or misleading results. Always validate that inputs fall within the model's valid range before interpreting outputs. Extreme values should be flagged for manual review.
High ISO Performance Comparison by Camera Generation
▾
| Camera / Sensor | Base ISO | Usable Max ISO | Read Noise (e-) | SNR at ISO 3200 |
|---|---|---|---|---|
| Sony A7 III (Gen 3 BSI) | 100 | 25600 | 2.4 | ~38 dB |
| Sony A7R V (61 MP BSI) | 100 | 32000 | 3.1 | ~35 dB |
| Nikon Z6 III (partial stacked) | 100 | 64000 | 2.1 | ~39 dB |
| Canon EOS R5 (FF) | 100 | 51200 | 4.2 | ~34 dB |
| Fujifilm X-T5 (APS-C 40 MP) | 125 | 51200 | 3.8 | ~33 dB |
| iPhone 15 Pro (1/1.28") | Varies | ≤12800 equiv | ~8 | ~30 dB (processed) |
Pertanyaan yang sering diajukan
▾
What is ISO invariance and why does it matter?
ISO invariance describes sensors where the read noise at base ISO is so low that amplifying the image digitally in post-processing produces similar or better results than using analog gain (high ISO) in-camera. For these sensors, shooting at ISO 100 in a dark scene and brightening in Lightroom can yield equivalent or better noise performance than shooting at ISO 3200. Not all cameras are ISO-invariant — Canon DSLRs historically have higher read noise at base ISO, making actual in-camera ISO adjustment more important. Check your specific camera's SNR curves at sites like Photons to Photos.
What is the 'base ISO' and why is it significant?
Base ISO is the sensor's native sensitivity — the ISO at which it achieves maximum dynamic range and signal-to-noise ratio. For most cameras, base ISO is 100 or 200. Some cameras have a dual native ISO with two base points (e.g., Sony FX3: base ISO 800 and ISO 12800 for video), where a second amplification circuit engages at the higher value. Shooting at base ISO gives the cleanest files; any ISO increase sacrifices some dynamic range in exchange for exposure sensitivity.
How does pixel size affect noise performance?
Larger pixels collect more photons per unit of exposure, resulting in more shot noise photons (absolute) but much better signal-to-noise ratio because signal increases faster than noise. A pixel with twice the area captures twice the photons, giving √2 more shot noise but 2× the signal — resulting in √2 better SNR (approximately 3 dB). This is why full-frame cameras with large pixels generally outperform smaller sensors with tiny pixels in low-light situations, even if the smaller sensor has more megapixels.
What is the difference between luminance noise and color noise?
Luminance noise appears as random brightness variations (grain-like texture), while color noise appears as random colored speckles — red, green, and blue pixels scattered randomly. Color noise is more visually distracting than equivalent luminance noise. Most RAW converters (Lightroom, Capture One, DxO) have separate luminance and color noise sliders. Color noise reduction is typically more aggressive than luminance (which also reduces apparent detail). Modern cameras and AI denoise tools (Lightroom Denoise, Topaz DeNoise AI) handle both types effectively at higher ISO settings.
Is it better to underexpose and lift in post, or use higher ISO?
Generally: expose to the right (ETTR — Expose to The Right of the histogram without clipping) at any ISO, then adjust in post. An underexposed image at ISO 100 lifted 4 stops in post will have more noise than a correctly exposed ISO 1600 image, because the shadow regions capture fewer photons. The optimal approach is to use the ISO needed to correctly expose the scene (not underexpose at low ISO), then apply noise reduction in post. The exception is ISO-invariant sensors where the low-ISO read noise is exceptionally low.
How much noise reduction can modern AI tools achieve?
AI-based noise reduction tools (Adobe Lightroom's Denoise, Topaz DeNoise AI, DxO PureRAW, ON1 NoNoise AI) apply trained neural networks to identify and preserve real detail while removing sensor noise. Results are dramatically better than traditional luminance/color noise sliders — typically recovering 2–3 usable stops of ISO range. A noisier ISO 12800 image can often be cleaned to match a non-AI processed ISO 3200 image. These tools work best on RAW files rather than JPEGs.
What SNR level corresponds to acceptable photo quality?
Industry benchmarks: SNR ≥ 40 dB is excellent (clean, commercial print quality). SNR 35–40 dB is good (acceptable for most uses, slight grain visible at 100%). SNR 30–35 dB is fair (noise visible, suitable for web and social media). SNR below 30 dB shows obvious grain, often suitable only for artistic use or heavily noise-reduced output. Camera sensors are measured by DxOMark's SNR 18% metric, which tests at a standard 18% gray target.
Kesalahan Umum yang Harus Dihindari
▾
- !Assuming more megapixels always means more noise — it's sensor size and pixel pitch that determine per-pixel noise, not megapixel count alone.
- !Using expanded ISO settings (H1, H2) without understanding they apply digital gain with no quality benefit vs. analog gain at the rated maximum.
- !Not using ETTR (Expose to the Right) technique to maximize photon capture at any ISO.
- !Applying excessive luminance noise reduction that eliminates fine detail alongside noise.
- !Ignoring that color space (sRGB vs. AdobeRGB vs. RAW) affects apparent noise levels in exports.
Tip Pro
Use Adobe Lightroom's AI Denoise feature (introduced 2023) on any RAW file shot at ISO 1600 or above. It produces dramatically cleaner results than traditional noise sliders by analyzing the full resolution RAW data before demosaicing. Set strength to 50–70% to balance noise reduction with natural-looking detail retention.
Tahukah Anda?
The first digital camera with a full-frame sensor, the Contax N Digital (2002), had a maximum ISO of just 800 with considerable noise at ISO 400. Today's Sony A7 IV achieves comparable image quality at ISO 25,600 — representing 32× (5 stops) improvement in high-ISO performance in just 20 years, driven by BSI (Back-Side Illuminated) and stacked sensor architectures.
Regional Guides
▾
🇺🇸 US▾
🇬🇧 UK▾
🇪🇺 EU▾
Referensi
- ›ISO 12232:2019 – Digital cameras: Determination of exposure index, ISO speed, and noise
- ›DxOMark: Sensor and Camera Noise Measurement Methodology
- ›Photons to Photos: Sensor Analysis and ISO Invariance Testing
- ›Emil Martinec: Noise, Dynamic Range, and Bit Depth in Digital SLRs
- ›Adobe: AI Denoise in Lightroom – Technical Overview
Have a question about this calculator? Get a detailed answer.
Dapatkan Tips Matematika Mingguan
Bergabunglah dengan pelanggan 12.000+ yang mendapatkan tip kalkulator setiap minggu.