willie_901
Veteran
Neither dynamic range or signal-to-noise ratio would be affected. For M8/9 cameras at ISO 640 and below, thrse characteristics depend primarily on the exposure (light reading the sensor when the shutter is open) and the analog technical characteristics of the sensor's photo sites (quantum efficiency, full-well capacity, read noise) and the ADC read noise. Above ISO 640 the data stream read noise increases for reasons I don't understand.
The difference, if there is any, would be in the quality of changes in tone and color throughout the image.
Many companies use non-linear, lossy compression for raw data. The idea is if you only compress highlight regions, the discarded information mostly eliminates shot noise. In essence the non-linear lossy compression filters the shot noise from the brightest regions in the image. This can not improve the analog signal to noise ratio, but it can make bright, featureless areas such as the sky or water appear smoother. If you are going to discard information by implementing lossy compression, you might as well choose the least useful information. Besides, the mathematical model for shot noise is unambiguous. So one could even add the shot noise back into the image (of course no one does this).
It turns out SONY may need to improve their lossy raw compression. There are endless posts on other photography forums demonstrating and complaining about IQ degradation in SONY's raw, lossy-compression algorithms.
Everything I've read reports Leica's lossy raw compression methods do not have a significant impact on most images. I know from personal experience the same is true for Nikon. No doubt other brands also use lossy raw compression that has little or no effect on rendered IQ.
The difference, if there is any, would be in the quality of changes in tone and color throughout the image.
Many companies use non-linear, lossy compression for raw data. The idea is if you only compress highlight regions, the discarded information mostly eliminates shot noise. In essence the non-linear lossy compression filters the shot noise from the brightest regions in the image. This can not improve the analog signal to noise ratio, but it can make bright, featureless areas such as the sky or water appear smoother. If you are going to discard information by implementing lossy compression, you might as well choose the least useful information. Besides, the mathematical model for shot noise is unambiguous. So one could even add the shot noise back into the image (of course no one does this).
It turns out SONY may need to improve their lossy raw compression. There are endless posts on other photography forums demonstrating and complaining about IQ degradation in SONY's raw, lossy-compression algorithms.
Everything I've read reports Leica's lossy raw compression methods do not have a significant impact on most images. I know from personal experience the same is true for Nikon. No doubt other brands also use lossy raw compression that has little or no effect on rendered IQ.