Reader Comment: Sony A7R V vs Sony A1 Pixel Shift, Focus Bracketing for Focus Stacking
Related: botched design, digital sensor, focus stacking, focusing, imcompetence crisis, pixel shift, reader comment, Sony, Sony A1, Sony mirrorless, Sony Pixel Shift, true-color sensor or image
re: focus stacking
re: botched design
re: pixel shift
Reader Nick M writes:
RGGB Bayer matrix in typical digital sensorI’ve been enjoying your Sony A7R V pixel shift content.
Have you tried the new Sony software with motion correction on your Sony A1? I’d be interested to know your thoughts regarding whether the improvement you are seeing is due to the processing software or the A7RV hardware (or firmware I guess?).
On a related note, I was also interested to see your comments about not being able to combine focus stacking with pixel shift on the A7RV. After being on the lookout for an automated focus stacking solution for the A1 for a long time, I have recently tried the iPhone app ’Shutter’ (https://shutter.dev/) with some success. Unlike other solutions, it is purely app based - no dongles or other appendages.
Of course, this does mean connecting the app to the camera via the WiFi which I have sometimes found to be flaky in the past, but that seems to be working pretty well so far. The app allows you to set a near focus point, a far point and the number of shots. From my limited testing it seems to work pretty well. Only bummer is that it doesn’t work with my Sigma 14-24, but does work with all my Sony lenses. And if you set the camera to 4 shot pixel shift and then control focus stacking from the Shutter app, it combines the two really quite efficiently.
DIGLLOYD: my impression is that the Sony A7R V has the best pixel alignment ever delivered in a Sony camera. Meaning that when the sensor is shifted for pixel shift, how accurate/precise is the alignment... I’m guessing something on the order of 0.2 pixels.
With its predecessor, the Sony A7R IV, I always felt that pixel shift was sloppy somehow, as if the pixels were only coarsely aligned eg ±0.5 pixels, leading to far less than the expected improvements; Sony marketing seems to confirm this; see Sony Pixel Shift: Sony A7R V vs Sony A7R IV a Major Advance in Capture Detail.
I should test the Sony A1 on the scene below, after our rains last night the creek should be roaring. I don’t recall Sony making claims about IBIS precision as with the A7R V. Still, maybe the A1 is similarly precise and it’s worth checking on. However, I don’t have much need for the A1 at this point, so am selling it to get the A7R V. Takers?
UPDATE: similarly outstanding results are possible with the Sony A1, as confirmed with a side-by-side comparison.
As for phone-assisted focus stacking, I’ve had so much wireless flakiness with every solution I’ve tried that I gave up on it for field use. I can’t even get my Sony wireless remote to work properly with the Sony A7R V so I am loathe to depend on a wireless solution that is likely to fail me. Yeah I 'get' the part about trying it, and probably should, but after seeing flaky solution after flaky solution fail, I do not feel motivated. A big problem is being out hiking, turning on the camera, and the thing not connected to Bluetooth and taking 30-60 seconds to connect it. Non-starter. Maybe WiFi is better, but would it also drain battery more? Dunno.
Sony implementation for a 'C+" instead of an 'A'
Do camera companies collude, with each vendor agreeing to at least three botched design aspects?
The whole Sony implementation is a science-fair project that could be so much better. As implemented, it’s a make-work project for the photographer, lacking all elegance. And yet, it could all be fixed in firmware.
- Uncompressed raw for pixel shift (huge files) when should be lossless-compressed.
- Sony does not allow pixel shift + focus bracketing.
- Broken focus bracketing to INF
- No in-camera motion-corrected DNG derived from the pixel shift frames. Panasonic had this 3 years ago and in a much more complex 8-shot form!
- Sony Viewer software with a horrible interface for efficient workflow that outputs needlessly huge ARQ file that must be further converted to compressed DNG to avoid the massive space wastage. And Viewer chews up 2 CPU cores when idle, a sign of incompetent programming.
Not a crowd pleaser. And with outright bugs to boot.
Sony A7R V + Sony FE 12-24mm f/2.8 GM @ 12mm
ENV: Alpine Creek, altitude 650 ft / 198 m, 58°F / 14°C
RAW: Camera ST, LACA corrected, vignetting corrected, +10 Whites, +10 Clarity
[low-res image for bot]
