There have been many reports of application-specific or custom designed high dynamic range (HDR) CMOS image sensors. To achieve their extended dynamic range, these sensors utilize techniques that can significantly degrade their signal-to-noise ratios (SNR). We utilize a simplified sensor model to compare two HDR techniques with a conventional APS sensor regarding their SNR and dynamic range (DR). We perform a new analysis of a mixed APS and time-to-saturation sensor that shows that it can detect similar high illuminations levels that the multiple capture sensor without degrading the SNR at lower levels. Furthermore, the time-to-saturation sensor can be adjusted on-the-fly to detect specific illumination levels with optimized image quality.