A question about Zebras

[Pete_W]

I am not a regular user of zebras on my LUMIX cameras when doing train photography as I find them distracting in the viewfinder when trying to frame (and quickly reframe) moving trains.

While taking some general landscape/street photos a few days ago in bright conditions I had some instances of blown highlights that were annoying to deal with in post so I thought I should just add zebras to My Menu for quick/easy access in future. And given that the sky is almost always going to show zebra stripes I wanted to set the level to 100% or 105%.

So that's what I did on my S5II. I then went to change my S9 settings to be the same but found that I could only set the zebras at maximum 95%!

I checked the S9 manual (the latest version is from firmware v1.3 even though firmware v1.4 was released last week) and it shows that zebras can be set up to 105%, same as the S5II:



I did some Googling but could not find mention of this anywhere. On this site I found this post from @CharlesH just over a year ago where the photo style can impact the zebra setting limit so I checked my photo style. It is set to 'Standard'.

So I am guessing this must be a bug?

 

[CharlesH]

I don't have my S9 anymore so I can't check this. That said, I never use 105%, but 100% or 95% to leave a little margin, and then I will leave some zebras just showing when I take the picture. I would go with the 95%.

 

[Pete_W]

Thanks Charles.

 

[pdk42]

My strategy is this:

- Set photo style to "Flat"
- Turn zebras on a 105%
- Set camera in A mode
- Assign a dial to exp comp
- Spin the dial until any important highlights show no zebras
- Fix shadows later in PP (assuming shooting in raw)

i never let the sky go to overexposure. If that leaves the shadows too dark, I use an EV bracket.

 

[Jan.Wedekind]

My strategy is this:

- Set photo style to "Flat"

Does that matter when you shoot RAW?

 

[CharlesH]

My strategy is this:

- Set photo style to "Flat"
- Turn zebras on a 105%
- Set camera in A mode
- Assign a dial to exp comp
- Spin the dial until any important highlights show no zebras
- Fix shadows later in PP (assuming shooting in raw)

This is a good strategy, but Pete's original question is about the S9. The question is why zebras are limited to 95% on the S9? One possibility could be because the S9 uses 12 bits, and not 14 bits like the other Lumix S cameras. When I was using the S9 for video, compared to the S5II, it seemed like one bit was taken off the high end and one bit was taken off the low end. I don't know if this is true, but again that's what it seemed like when processing S9 video. If Panasonic does something like that for photos it may be there is less dynamic range at the top. And if they use the same firmware for zebras as the S5II, then they might limit the zebra range that can be set at the top. This is speculation, but based on my video experience this might be what is going on. No matter, I always had enough dynamic range with the S9 for the kind of photos I took.

 

[Pete_W]

My strategy is this:

- Set photo style to "Flat"
- Turn zebras on a 105%
- Set camera in A mode
- Assign a dial to exp comp
- Spin the dial until any important highlights show no zebras
- Fix shadows later in PP (assuming shooting in raw)

i never let the sky go to overexposure. If that leaves the shadows too dark, I use an EV bracket.

Thanks Paul. That is essentially what I have done on my S5II. I've found that even 95% causes too much of the scene to be blanketed in zebras in the bright sun here in Australia, so 100% or 105% is preferable.

This is a good strategy, but Pete's original question is about the S9. The question is why zebras are limited to 95% on the S9? One possibility could be because the S9 uses 12 bits, and not 14 bits like the other Lumix S cameras. When I was using the S9 for video, compared to the S5II, it seemed like one bit was taken off the high end and one bit was taken off the low end. I don't know if this is true, but again that's what it seemed like when processing S9 video. If Panasonic does something like that for photos it may be there is less dynamic range at the top. And if they use the same firmware for zebras as the S5II, then they might limit the zebra range that can be set at the top. This is speculation, but based on my video experience this might be what is going on. No matter, I always had enough dynamic range with the S9 for the kind of photos I took.

Yes, the S9 is the problem. I usually take the S5II out for train photos but it bothers me that the S9 is different and the manual doesn't show this lower limit. Your point about 12-bits vs 14-bits is a good one so maybe that's the explanation. I am surprised though that I couldn't find mention of this anywhere with a Google search.

I will see if I can get a question through to Panasonic somehow, maybe on a future LUMIX Live session if I can get myself up early enough!

 

[Markuswelder]

Does that matter when you shoot RAW?

Yes it does. The highlight information (that you see in the EVF) is based on the jpeg representation of the shot. Not on the raw file. So you use a picture style that is as flat, and un-tweaked as possible, hopefully without the camera doing too much in the way of pushing shadows, and pulling back highlights. That way you get a consistent way of knowing, as close as possible, the actual dynamic range in the scene. Not what the camera is creating in its own internal conversion for the EVF, or saved jpeg.
I hope that makes sense
Edit -It's actually easy to see. Look through the EVF with the camera locked in manual mode, and flick through different picture styles. Watch the shadows and highlights change, without changing any of the exposure parameters

 

[Babylonia]

Does that matter when you shoot RAW?

Zebras, as well a histogram is just a "helping tool" to find the borders for what exposure settings can be used without clipping areas.
(Zebras as for the highlight areas alone).

Both interpretations of these "helping tools" are based on lower quality "JPG" data.
"Tweaking" the highlight zebra output as for "JPG" output depends also to profile setting.
As profile settings mainly are based to lower bit depth JPG data output (and e.g. video output without post processing).

In spite you have more latitude within RAW picture data.
It doesn't mean you do have the same amount in latitude as for correction within RAW processing,
in comparison to the output "helping tools" data, as shown by these zebra's.
A JPG based "standard" profile, do have different representation, than a "neutral" or "flat" profile.

The only proper way to find out if the representation of these zebra's (as well histogram) is reliable to your work-arround,
to find the borders, is just by testing and try different values.

Keep in mind to, that the real latitude within RAW picture data, can be different which ISO is used.
So another variable to take into account.