Today’s Question: Am I correct that the histogram that you see on the camera refers to the JPEG and not the RAW capture? I’ve been told that you actually have more headroom than it shows you on the histogram if you’re shooting RAW.
Tim’s Quick Answer: The histogram you see on your camera’s LCD display is based on an in-camera conversion of the raw capture to what is effectively a JPEG image. That does mean that there is a degree of latitude when it comes to processing the raw capture, and that even if there is some clipping shown on the histogram, it may be possible to “recover” the clipped data in post-processing.
More Detail: As noted in an Ask Tim Grey eNewsletter last week (https://asktimgrey.com/2019/06/17/is-camera-preview-a-jpeg/), the preview image you see on your camera’s LCD display is essentially a JPEG image created in-camera based on the raw capture. As such, the preview does not necessarily reflect the full potential of the raw capture.
Similarly, the histogram for a raw capture is based on the in-camera JPEG rendering from the capture. That does mean that the histogram isn’t completely accurate relative to the potential for the final image.
For example, if the histogram shows a small amount of clipping for the darkest shadows or brightest highlights for a photo, that doesn’t necessarily mean that the detail is permanently lost. As long as the clipping is not too severe, you may be able to process the raw capture in a way that will preserve the detail that seemed to be lost based on the histogram viewed on the camera’s LCD display.
That said, I don’t recommend relying too heavily on the ability to recover detail beyond what the histogram may indicate. In other words, I recommend avoiding clipping of shadow or highlight detail based on the histogram display, even though you may have some ability to recover that detail when processing the raw capture. If you aren’t able to avoid clipping highlights and/or shadows based on the histogram display, I recommend bracketing to create a high dynamic range (HDR) image later, rather than depending on the ability to recover detail based on how the raw capture is processed after the capture.