2023 big year for Fuji?

norjens wrote:

io_bg wrote:

According to Fuji rumors, the next camera to be announced will be X-S20. I find this slightly strange, as I'd expect Fuji to release their higher-end cameras first (therefore, the X-Pro 4). Perhaps they're still figuring it out in terms of screen design and the possible inclusion of IBIS.

We could also expect a X-T30 successor in the later half of the year, and a X-E5 in 2024. The X100 is difficult to predict. Given the series 'cult status' Fuji will probably not be in a rush to iterate it this year.

Unfounded speculation on this: Might it be that they want to put the new stacked sensor in the X-Pro4, but supply of it is insufficient to launch another camera with it yet?

It’s certainly an option…

norjens wrote:

io_bg wrote:

According to Fuji rumors, the next camera to be announced will be X-S20. I find this slightly strange, as I'd expect Fuji to release their higher-end cameras first (therefore, the X-Pro 4). Perhaps they're still figuring it out in terms of screen design and the possible inclusion of IBIS.

We could also expect a X-T30 successor in the later half of the year, and a X-E5 in 2024. The X100 is difficult to predict. Given the series 'cult status' Fuji will probably not be in a rush to iterate it this year.

Unfounded speculation on this: Might it be that they want to put the new stacked sensor in the X-Pro4, but supply of it is insufficient to launch another camera with it yet?

When I first heard this, I thought it a bit far fetched.  On the other hand, it might very well be there is not sufficient room in the Pro4 for IBIS and a mechanical shutter.  With a 40 MP sensor - which is a 90 MP equivalent FF sensor, IBIS would probably necessary.  That has a bit of credibility given Leica has point blank said this for the Leica M11.

So eliminate the mechanical and make room for IBIS and and release the Pro4 with a stacked sensor and with only an electronic shutter.

I doubt it will happen and I'd rather see a Pro4 with a 40 MP sensor, but it is a thought.

The issue here is that the current 40 MP sensor would need a mechanical shutter, as it has far too much rolling shutter to be use only with ES. If IBIS is a necessity, then it's a catch-22 situation.

yayatosorus wrote:

The issue here is that the current 40 MP sensor would need a mechanical shutter, as it has far too much rolling shutter to be use only with ES. If IBIS is a necessity, then it's a catch-22 situation.

I have no doubt that Fuji will take its time and get the X-Pro4 right.  It may cost more than ever too.

yayatosorus wrote:

The issue here is that the current 40 MP sensor would need a mechanical shutter, as it has far too much rolling shutter to be use only with ES. If IBIS is a necessity, then it's a catch-22 situation.

Sorry - that's wrong. My wife hasn't used anything by the electronic shutter with her XH2. Rolling shutter is a bigger problem in video, i.e., panning in video.

I would be interested in an XPro4 with a 40 MP sensor with only ES and IBIS. After all the XPro is a photographers camera - not a video camera.

Truman Prevatt wrote:

yayatosorus wrote:

The issue here is that the current 40 MP sensor would need a mechanical shutter, as it has far too much rolling shutter to be use only with ES. If IBIS is a necessity, then it's a catch-22 situation.

Sorry - that's wrong. My wife hasn't used anything by the electronic shutter with her XH2. Rolling shutter is a bigger problem in video, i.e., panning in video.

I would be interested in an XPro4 with a 40 MP sensor with only ES and IBIS. After all the XPro is a photographers camera - not a video camera.

Thanks for that real-world perspective, much appreciated. I've also used the ES on my X-H1 numerous times and have never had a problem with it.

I am however uncertain the scan rate is good enough in all scenarios for Fuji to launch the X-Pro without a mechanical shutter. Panning is a thing in photography as well, and I am not sure they would like the possible bad press this would induce.

Still from the Cameralabs video review of the X-H2

michaeladawson wrote:

Strongly disagree. The camera that doesn’t easily allow me to choose my own ISO is the camera I don’t buy. I usually agree with you Truman, but not here. I’ll use auto-ISO at times when it fits how I’m working. Most of the time, no.

An ISO dial is a thing of the past. The only reason there was one on a camera is because there were different films available with different sensitivities.  The ISO was tied to the film sensitivity in that to get a haven film density, the ISO programmed the light meter so that the appropriate exposure value would be used to render a neutral gray card to that density.

A photo sensor has a fixed sensitivity. One can not change the sensitivity of the photodetectors.  Today if you shoot an XT5, your sensor sensitivity is fixed. That is X amount of light energy will result in Y photons captured by the photodetector. No amount of spinning an ISO dial or wheel will make it anything but Y.

ISO is really meaningless in modern digital cameras. It is little more than an amplification level for the amplifiers between the output of the photodetectors and the ADC.  If one is wanting jpegs with a “standard brightness” out of the camera, then one needs to pay attention to ISO. In fact that is how ISO is defined in the standard - jpeg brightness. For raw ISO is more or less meaningless.

I'd have to respectfully disagree here. Comparing film ISO to digital ISO is fine. However talking about sensor signal to noise ratio and then Gain as if they don't interact is a bit of a problem in looking at how sensors function and how they readout to generate the RAW data.

You are essentially correct in everything said seperately but the system doesn't work in the dicrete way you imply. That the amount of light a photo site captures on a sensor is set for the sensor is correct. However the sensitivity of the photosite is defined by the very amplification circuits you mention. That is the response to the signal from the sensor can be increased or decreased as a function of the gain in the readout circuit. The way you put it it's like the flow is: -

Light --> Sensor --> RAW --> Signal Gain --> Image processing? Its really like so: -

Light --> Sensor (with Gain/Amp applied (as pretend ISO) ) --> Image Processing --> RAW Output 
If signal is too high in bright light it'll max out the sensor readout and information is lost if it's too low it'll bottom out sensor readout of the signal and again information is lost (because the sensor is not a 0-whatever volt signal it floats say as a 1-5v signal (example only)). When increasing gain or reducing it too much, more noise is introduced. Signal to noise Ratio increases in either case (Overly large signal or overly small signal both create more noise) . You can't really de-couple the amplification of the sensor signal from the sensor sensitivity/snr the of the signal from sensor to the image processor. RAW output is a function of both aspects of the system.

In essence ISO on film and ISO on digital sensors have the same effect less/more noise and less/more sensitivity the only significant difference to me is you can change ISO without changing film.

For reference (and not a flex!) I come at this topic as an engineer of many years who spent 14 of those years using CCD and CMOS sensors on Robots to look at things in 3D and get them to act accordingly so my thoughts are not random I've gone quite deep on programming these things and knowing how they work and how they 'see'.

Other information that may be of use or interest is sensors generally perform multiple readouts into multiple matched amplifiers. The signals are in pairs and one is inverted. The 1st is then added to the 2nd signal to try to establish a noise signal (a + -a). If you add a to -a you should get only noise. You can then remove this from the original signal in a kind of feedback loop.

Problem is that inherent noise in the system will only grow as gain goes too high or too low (Like high or low ISO does (with digital cameras) and high film ISO does with Film) so noise will become an increasing problem as you magnify this issue by ramping up gain. Also the less the strength of the original signal the more of that signal is noise (of a more significant kind) so the signals are harder to discern from noise and clean-up hence noise at low light is way more an issue. You can even notice high ISO at higher light levels has less noise as it's still a decent signal (except where it maxes out).

By maxing out a signal I mean in the digital sense. In simple terms if the readout of the digital sensor can only report 0 as black and 255 as white and you up the gain too much so everything is 255 (Gross overexposure from) all your information is lost from an image point of view. Dropping ISO (gain of sensor readout) might recover that so the sensor output can represent light and dark again properly. Obviously Shutter Speed and Aperture also play a central role in that but that goes without saying.

D

gdanmitchell wrote:

Truman Prevatt wrote:

michaeladawson wrote:

Strongly disagree. The camera that doesn’t easily allow me to choose my own ISO is the camera I don’t buy. I usually agree with you Truman, but not here. I’ll use auto-ISO at times when it fits how I’m working. Most of the time, no.

An ISO dial is a thing of the past. The only reason there was one on a camera is because there were different films available with different sensitivities. The ISO was tied to the film sensitivity in that to get a haven film density, the ISO programmed the light meter so that the appropriate exposure value would be used to render a neutral gray card to that density.

A photo sensor has a fixed sensitivity. One can not change the sensitivity of the photodetectors. Today if you shoot an XT5, your sensor sensitivity is fixed. That is X amount of light energy will result in Y photons captured by the photodetector. No amount of spinning an ISO dial or wheel will make it anything but Y.

ISO is really meaningless in modern digital cameras. It is little more than an amplification level for the amplifiers between the output of the photodetectors and the ADC. If one is wanting jpegs with a “standard brightness” out of the camera, then one needs to pay attention to ISO. In fact that is how ISO is defined in the standard - jpeg brightness. For raw ISO is more or less meaningless.

I disagree that it is a "thing of the past." Regardless of the theory about all ISOs being the same (not quite the case), the ISO dial lets the photographer make a determination of whether to prioritize better IQ performance (specifically for noise and DR) at the low ISO (thus lengthening exposure time and/or increasing ISO) or sacrifice that in order to keep shutter speeds short and/or apertures small.

ISO is not "meaningless" in modern digital cameras, though the specifics of what it means may be somewhat different. In fact, you would do better to try to make the case that the meaning has changed than to stick to the "Meaningless" stance.

All amplification associated with ISO takes place after the photons are converted to electrons.  It doesn't matter that such processing of the voltages or currents take place in circuits on the same chip as is the case of CMOS sensors or marched out of the photodetectors off chip for processing as done in CCD sensors.  The voltage contains signal and converted photon shot noise.  Photon shot noise is the limiting factor in imaging performance.  The ISO gain is nothing more than turning up the volume on the radio.  Both the signal and the shot noise are amplified equally.  It might make the EVF brighter - but it does not improve the Photographic DR.

Truman Prevatt wrote

The ISO gain is nothing more than turning up the volume on the radio. Both the signal and the shot noise are amplified equally. It might make the EVF brighter - but it does not improve the Photographic DR.

It is still more like turning up the preamp gain - at least in certain situations- than turning up the master volume to stay in the picture. I find gdanmitchell‘s post quite convincing.