SL3: Discrepancy between LV and playback histogram.

[SrMi]

Advertisement (gone after registration)

3 hours ago, davidmknoble said:

The iDF setting does affect the EVF live view.  So does enhanced live view.’

Thanks for the comments.

iDR does affect the histogram by lifting the shadows. However, in all my testing, I have not found that it affects the clipping of highlights, which is the only information I need from the histogram. Anyway, iDR should affect LV and recorded images similarly, and histograms should not differ. I noted that the discrepancy between histograms occurs regardless of the iDR setting.

3 hours ago, davidmknoble said:

On page 121 of SL3 manual (english) it says the histogram represents the brightness displayed and the settings on the camera affect it, so it may not be the histogram of the final image.  It also represents the histogram of the crop shown, which is likely why it is affected by what is shown.

The manual does not differentiate between LV and replay histograms; I assume they are supposed to be identical.

3 hours ago, davidmknoble said:

I have long sought what I thought we had with the original SL which was more accurate representation of the final image in the DNG.  So, I leave the iDF off, enhanced live view off and leave the JPEG settings at zero watching the histogram from there.  It’s close, but a great shot on the preview is likely dark on the DNG.  Luckily that saves highlights and we have good shadow recovery.

These are two separate issues. The first is how to make the JPEG histograms as close as possible to the raw histogram. The second is the difference between the JPEG histogram in LV and the recorded image.

I have Enhanced LV off. 

The only setting that I can imagine should cause a discrepancy between LV and image histograms is exposure simulation. I leave it on for manual mode.

[hdmesa]

1 hour ago, SrMi said:

The manual does not differentiate between LV and replay histograms; I assume they are supposed to be identical.

They can't be identical, though often they are very close. The histogram in LV is a running average of the feed, while the histogram of the recorded image is exact/finite. If you're on a tripod or in very even lighting, the histogram in LV will be much closer to the final image. If shooting handheld in varied lighting, the final histogram almost always differs from the histogram I see in LV at the moment of exposure.

From the SL3 manual below:

Welcome, dear visitor! As registered member you'd see an image here…

Simply register for free here – We are always happy to welcome new members!

Edited December 15, 2024 by hdmesa

[SrMi]

1 hour ago, hdmesa said:

They can't be identical, though often they are very close. The histogram in LV is a running average of the feed, while the histogram of the recorded image is exact/finite.

They should be using the same JPEG engine to display and, therefore, the same histogram and clipping. All other manufacturers have identical histogram/clipping in LV and image review.

I believe the manual explains why one should not judge the exposure from the brightness in the EVF/LCD. That is true for all cameras. 

An LV histogram that sometimes differs from the histogram of the recorded image is misleading and not very helpful.

[hdmesa]

39 minutes ago, SrMi said:

They should be using the same JPEG engine to display and, therefore, the same histogram and clipping. All other manufacturers have identical histogram/clipping in LV and image review.

I believe the manual explains why one should not judge the exposure from the brightness in the EVF/LCD. That is true for all cameras. 

An LV histogram that sometimes differs from the histogram of the recorded image is misleading and not very helpful.

Look at how fast the histogram changes in LV while handholding in high contrast light. If I'm inside with a bright window in the frame, the smallest movement in framing changes the highlight end of the histogram. Maybe I'm just seeing the delay between pressing the shutter and capture.

If you shoot on a tripod, you get this 2/3 stop difference between LV and playback? I can only see that happening if for some reason the histogram in LV is using a smaller area of the frame but the playback uses the entire image.

[SrMi]

13 minutes ago, hdmesa said:

Look at how fast the histogram changes in LV while handholding in high contrast light. If I'm inside with a bright window in the frame, the smallest movement in framing changes the highlight end of the histogram. Maybe I'm just seeing the delay between pressing the shutter and capture.

If you shoot on a tripod, you get this 2/3 stop difference between LV and playback? I can only see that happening if for some reason the histogram in LV is using a smaller area of the frame but the playback uses the entire image.

I get the difference with static framing (no camera movements), and sometimes there is  no difference. I speculate that difference occurs only in very high contrast scenes, e.g., with overcast sky, but I need more experience.