I use M-mode; in fact, I use M-mode a lot. I am also fond of my K-3's built-in bracket function and use it frequently. I guess that makes me proficient with both except that I had not used the two together until yesterday when exercising the combination as part of a troubleshooting request. After several hours figuring out how it works, I am perplexed.
Is my camera broken?! The first thing I noticed that has nothing to do with M-mode is that "one-push" bracketing seems to be active regardless of whether the feature is enabled in the custom setting menu. I did not notice this before yesterday and the behavior persists even after updating from firmware v1.10 to v1.30. Is it supposed to do that? Edit: I figured it out...
Down the rabbit hole...
Normally, I would not have considered using the "Exposure Bracketing" drive mode in conjunction with M-mode. After all, it is manual mode and manual mode is all about full control of exposure. Doing a bracketed exposure series manually is very straightforward and easy. Doing the same using the bracketing feature generally works pretty well too, despite a few quirks that bear being attentive to. I will be posting those findings separately in another thread. This thread is about anomaly, as in Star Trek.
Star Trek anomalies fall broadly into several categories: energy, spacial, temporal, and so on. Thankfully, I don't believe I ran into any of those. What I encountered instead was a vexing "aperture anomaly" where the e-dial behaves according to several unconventional notions of what constitutes 1/3 stop. In short, I found two fairly serious
bugs. To be clear on how to reproduce, here is the setup:
Setup- K-3, firmware v1.10 or higher
- FA 35/2 mounted
- EV steps: 1/3 stop (default)
- Bracketing Order: "- 0 +"
- D-Range Settings: all off
- Manual ISO
- Program Line: Normal
- Green Button Action: Program Line
- Drive Mode: Exposure Bracketing, 3 frames in 1/3 EV steps
Except for the bracketing order, there is not much in this setup that unusual.
To see what I saw with the FA 35/2*
- In Av mode, set the aperture to f/2.8 (one stop down from full open)
- Switch to M mode and note that f/3.2 is now the set aperture. Don't be alarmed. F/3.2 is intentional and expected.
- Move the aperture e-dial one click to widen the aperture 1/3 stop to f/2.8. Cool! It worked!
- Move the aperture e-dial one click back to narrow the aperture 1/3 stop to f/3.2 again. Note that it says f/3.5, not f/3.2.
- Move one more click narrower and note that 2/3 stops from f/2.8 is now f/4.5. Move one step again and see that f/5.6 is now 1 stop (3 steps) above f/2.8. The magic continues if one keeps rotating the dial.
- Now lets step back down and magically f/5.6 is two stops (6 steps) above f/2.8 like it is supposed to be.
Are we amazed yet or is it just me with my camera? If not reproducible, it may be time for me and/or my camera to see a specialist. If y'all see the same, then we have a vexing behavior that some might call a "bug".
But wait! There's more!!- Move the set aperture two steps wider from f/2.8 to f/2.2. The ability to do so should raise a few alarms since f/2.2 may not work properly for a three frame bracket in 0.3 EV steps. It all depends on which step it is. If it is the "-" step (narrowest opening in the series) we have a problem.
- Expose the full bracket series of three frames
- Be amazed to see a series of three with the last two frames (0.0 EV and +0.3 EV) with the same settings (+0.6 EV more light than the first frame)
I stopped exploring at this point and decided I had found enough vexation for one day.
The Work-Around(s)- Changing the bracket order to "0 - +" or "0 + -" will correct the problem of improper e-dial EV steps
- Changing the green button action from "Program Line" to "Tv-shift" will also correct the EV step issue and does so in a fairly intuitive way consistent with the notion of manual exposure control. Update: I did some additional inquiry with firmware v1.30 and found that this workaround also mitigates the bad bracket set problem.
I could not find a work-around for the bad bracket set problem
Comments, confirmations, and refutation are all encouraged and appreciated
Edit: After reviewing the first group of comments, it occurred to me to simply state the issues in terms of the implied "contract" of the camera with its user. The broken contract rules are as follows:
- The e-dial does not consistently increment/decrement aperture in 1/3rd stop steps as per system settings
- The system does not properly bracket exposure (EV) when the set aperture is within one bracket step of the lens maximum aperture.
/Edit
Steve
* The actual lens is not pertinent except that the description assumes f/2 maximum aperture
Last edited by stevebrot; 04-11-2017 at 07:43 PM.