Leaderboard

What you're talking about is for EM-CCD sensors, not CMOS. "Pixel clock rate is an important specification for CCD and EM-CCD cameras, but not for CMOS cameras because of chip architecture. Unlike CCD and EM-CCD cameras where photoelectrons from each pixel are converted into voltage one-at-a-time via a single amplifier, in CMOS cameras each pixel has its own amplifier, so conversion of photoelectrons happens in parallel in all the pixels." From Hamamatsu company website. You know more than they do?

Cined's interview with Fuji that I posted above pretty much confirms that the Eterna camera is "basically the same" as the GFX 100 II in terms of capability. So I don't think it's even safe to say that the RS improvement is "likely." It's possible, but I wouldn't bank on it. It's believed by just about everybody that the GFX 100 II continues to use an IMX 461, given the complete lack of meaningful change in photo capabilities and the fact that Sony doesn't have any newer sensor available. It's not clear whether it's been tweaked by Sony or whether Fuji is using some other trickery (such as using an 11-bit readout instead of 12-bit) to achieve the increased frame rates. Anyway, the point isn't what it's safe to assume about this camera, but that is a nice attempt to reframe the argument so that you'll be right. The point is that the readout modes chart in the sensor data sheet doesn't give the entire picture about the capabilities of the sensor and nothing in that datasheet or elsewhere specifically excludes the possibility of overclocking. You know, the argument that you were actually trying to make and failed at.

What a pity for me and for every other current user of the GFX 100 II that, according to the table on page 105, the sensor being used by it has absolutely no capability to do 4k at 60 frames per second, 8k at 24-30 frames per second, or 5.8k at anything resembling 24-30fps. Thanks so much for enlightening all of us as to how useful the sensor data sheet is for understanding the actual capacity of the sensor being used in the GFX 100 II. You also get a gold star.

Here you go bro. This is an example of a sensor spec sheet. By the way, the really quick way to tell if the rolling shutter is “fast” is to look at the fastest frame rate the camera can do. It’s not complicated. But thinking you can just “overclock” without it being DESIGNED to actually do that or have those modes designed from the ground up. This sensor is designed around being a stills sensor. Not a fast RS motion sensor. Page 105 shows the maximum frame rates depending on the mode being used. https://www.dropbox.com/scl/fi/85986570odkx36kc7tw9e/GFX100-Data-Sheet-2.pdf?rlkey=6ek4lgy5lqcx138jz3afliuks&e=1&dl=0

The pocketable part is the trickiest bit... Best of luck getting an S5 or R5 or anything of that nature in your pocket. Unless you are Matt Granger. S9 springs to mind and maybe with that new 18-40? Even then, still need a decent sized pocket! I've been guilty of it myself and that is chasing a 'grail' solution that does not exist and you probably are better off with that S5 + GX85 combo?

Mmm, maybe the Fujifilm XT-4 or 5 or X100 XI, or the Panasonic S9?

Good job redefining the argument so that you can be right. Literally nobody said "edit a few lines of code" or "let the sensor consume more power." You win today's gold star for bad faith argument loss tactics.

FWIW, the S5 and R5 are nearly identical in size and while the R5 is 10-bit 4:2:0 in H.265 mode, it also shoots raw internally - and unless you're doing special effects work, you won't miss it - that or just record in 8K and when you downsize it to 4K, you'll have all the luma/chroma info that you need anyway. Regardless, sounds like you chose the S5. It's a great camera and I'm sure you'll enjoy it! Happy shooting!

I'm not sure if I have the patience to watch the entire video - did he say whether he's using the anti-newton glass carrier for his Coolscan? That makes an absolutely enormous difference in scanning resolution. I use an 8000 and when using the glass carrier at 4,000 dpi, I can see grain pretty clearly defined when zooming in (though not as much with some films like Velvia 50, but that's not an especially common stock for filmmakers). Anyway, the standard carriers for the Coolscan do a pretty piss-poor job of keeping film flat and the scanner has a really tiny DOF - so non-flat film = unsharp scans (but still usually usable for a print or whatever, an 8x10 print at 300dpi is only about 8 megapixels after all). Either way, it's just plain untrue that most S35 film has much more resolution than 4k. At the point where the grain structure is clearly visible at 100% (including the bigger and smaller grains), you aren't going to get a lot more detail. The main reason that most people think 4K is good enough now is that unless you're sitting really close to a huge screen, we're well past the threshold of declining returns on increasing resolution. The average human eye can't perceive any difference whatsoever between 4K and 8K on a 50" screen viewed from a distance of around 2-3 meters - and the perceived difference, even between 2K and 4K content at that distance is not huge. Upgrade yourself to a 75" TV and at that distance, the sharp-eyed can perceive a difference between 4K and 8K... but most people won't notice the difference or care at all. I'm not sure how things are in every market, but in my local market, I can say that many cinemas are still projecting most content at 2K. Even on a huge screen, people don't seem to mind the difference when watching from 15 meters away. Also, also - at 4K, lots of little flaws (make-up seams, etc) become more visible and VFX can take a while to render. Step up to 8K and those problems are compounded yet again.