[Rover]

Okay folks, I'm a little weak with theory but... If I find that I'm not having enough DOF all the time with my Z9 (and mostly the 16-35/4 lens) but can't stop way down because ISO will be into five digits when shooting in iffy light indoors, will switching to APS-C mode (semi-)permanently help? The noise wouldn't get any worse because it's the same sensor, just losing the edges, right?

Since I'm mostly shooting in medium resolution, I could just combine high resolution with APS-C crop to get back to the same # of pixels, give or take.

Oh, and I'd have to get a new ultra wide because 16 will not be that wide anymore, but I've been eyeing the Laowa 10mm anyway... And it'd turn into a monster if I switch back to FF mode.

[stoppingdown]

My take: noise won't change, but I think that more pixels (with the same noise level) can produce an inferior noise level on the output (assuming the output image size don't change), so you might lose something. I'm trying to explain with an example and round numbers. If you start with 8000x5300 pixels and produce an output of 4000x2650 pixels, each output pixel is the average of 2x2 captured pixels (this is rough and somewhat inaccurate, but I think you get the point). Averaging reduces the noise on the output. If you start with 4000x2650 pixels the output doesn't get any averaging. If you start with 5300x3500 (APS-C crop) you have averaging, but with a smaller window than the original.

But the best way to deal with the issue is to shoot FF and APS-C with the same framing and processing and compare the outputs.

[Rover]

Medium size FF is 6192*4128
Large size APS-C is 5392*3592
No deal breaker I think especially since I downsize everything for publication - even in the newspaper - anyway. I'll need to try and see how DOF changes when doing 16mm on APS-C instead of 24mm FF. Should not be a great deal of difference of course, but any little bit helps. And I'm not one of those users who agonize over numbers and dismiss anything that isn't a 800mm f/0.95 level of shallow DOF.

[mike]

 I'll need to try and see how DOF changes when doing 16mm on APS-C instead of 24mm FF. Should not be a great deal of difference of course, but any little bit helps. And I'm not one of those users who agonize over numbers and dismiss anything that isn't a 800mm f/0.95 level of shallow DOF.

I'm a little confused on what you're trying to do. If you put a 16mm lens your camera and put it in aps-c mode and compare it to a 24mm lens on the same camera in FF mode for the same aperture, you'll get ~1.5x greater dof and the same fov. Likewise, if you use an aperture of 2.8 on the FF, you will need to use 1.8 (~2.8/1.5) to have the same dof. But I assume you knew that already. 

I would expect the aps-c and the ff noise to be approximately the same at 100% on the screen - ignoring minor difference due to exposure times for say a typical shot. That, however, translates to more noise, ~1.5x, on an say anA3 print just by the virtue there are less pixels to cover the same area for the aps-c. The question is does it matter for your application?

[Rover]

Why would there be more noise if I'm not changing the ISO, and the sensor is the same? ?

[mike]

Why would there be more noise if I'm not changing the ISO, and the sensor is the same? ?

At 100% on the computer screen, they would be comparable. You are looking at the same pixels. Only minor differences if the exposure times were slightly different (If the shots were setup for the same FOV and DOF, the aperture would be bigger on the aps-c - more light coming in - and the exposure shorter). For practical purposes we can say they're the same.

The difference comes when these images are printed. There are less pixels/cm^2 on the actual print on the aps-c. Ergo, the pixels are "blown up" more. There are less pixels/cm^2 to cover a given area for aps-c. It's analogous to viewing the image at 50% vs 100% on the flat screen. At 50% you might not see any noise. Whereas at 100% you can. The aps-c print is closer to being that 100% than the FF for any given print size. Again, I'm assuming the same FOV to compare apples to apples.

It's similar to how a high ISO images looks just great on Instagram or a web page where the image is only viewed within 1200x1200 pixels or less area on the flat screen, but if you printed the image to an A3 size it would look horribly noisy.

YMMV. For your purpose you may never notice the difference.