Bill_de Loc: US

Duh! Obviously not true.

Your whole premiss that you couldn't have made the wrong settings because you didn't know what you were doing is ludicrous.

I know you are not following the thread anymore, so I don't expect a response.

---

Harry0 Loc: Gardena, Cal

Yeah, kinda sorta . There are too many variables inherent in the system to call it an absolute standard. I can put my kid's (manual) Canon lens on my Nikon, and compare. Same lens, f8, 1/500, ISO 1600. Same shot- hers is brighter. Tho hers is noisier at ISO 6400 than mine.


Exactly my point. I can get a UK stove, set it to Gas4, and it'll be within 5% of most others, I can set my oven to 350 degrees, and it'll be within 5% of 350, and within 5% of my kid's oven.
Can I trust that my Nikon will be within 5% of the signal/noise ratio of my kid's Canon at a given ISO? Will there be a variation between the 2 at low to high ISO settings? Is my D7100 close enough to yours to give you the same shot at the same settings? Or do we accept that, for now, it's just kinda sorta close enough? An analog value expressed as a digital setting?
Or like my old radio- the voltage, frequency and volume are dead on within 1%. The gain ... varies.


He may admit to possibly have been mistaken a time or two ,,,


It may have. IF you preset the shutter and aperture, the camera had no choice but to set the ISO in a range of adequate exposure. which it sounds like it did.

Grahame Loc: Fiji

So I can only assume by this that you consider that "NEGATIVE comments" (that have stung you) are those for which I have made within my posts as follow;

a) That the reason for the unnecessarily high ISO (10,000 at 1/4000s for a sunny scene) in the image that was posted was not due to some clever engineer knowing better but due to the USER not using the camera correctly.

b) That the image posted at ISO10,000 was neither a good result or one that used 'sensible' settings.

So to recap on the theme of your thread of which the title is "Discussion on the Need to impose Upper Limits to AUTO ISO" ................................

The photographer should always aim for an ISO that is appropriate to the exposure required taking account of the aperture and speed used. How he chooses to do this can be manually or automatic.

But what is MORE important, and so well demonstrated in the D5300 ISO10,000 image you posted is that the photographer should also be 'aware' of what his camera is doing PLUS be aware of how his intervention with respect to having a 'User set ISO' can negatively affect the operation of 'Auto ISO'.

Harry0 Loc: Gardena, Cal

I'be known fools good enough to gobsmack the geniuses.
There's always one ... and there's never one of anything.
'Tis why "USER" is a four letter word.

Larryshuman

Everybody makes mistake with their cameras. Just yesterday I got shots that varied in brightness wildly. Upon checking I realized I had turned on bracketing. Turned it off everything is well again. Knowing your camera speaks volumes about your experience. Knowing what's wrong and being able to correct it. In my experience with the D5300 line of cameras they are not able to perform the adjustments that a D7200 or D7500 can. Also the folks who have the camera expect it to take the perfect picture without a whole lot of input on their part.

Mongo Loc: Western New York

Tony is right, and wrong. It depends upon the camera system designer, and the sensor designer.

User ID

Time for a timely reminder

Mongo Loc: Western New York

It is not just the algorithms (I've written them), it is more about the desired performance of the camera system, the sensor to be used and the architecture of the sensor. It might be about how the sensor handles higher ISO settings, for example, whether it aggregates pixels to reduce noise, and increase area the area of the pixel.

Granted, today most cameras are CMOS, but just a few years ago, CCDs were much more common and they had all kinds of trades which could be made for higher sensitivities.

Also, unless you know for a fact how your camera manufacturer establishes the lowest ISO, I would not assume that it will give you the lowest noise. It might, or it may not. It is a function of how it was implemented.

It is clear from the frustrations that people have, that AUTO-ISO needs another generation of development. I for one would like to define weighting or costs for the trades of ISO/APERTURE/SHUTTER, which might also include noise considerations, and so on.

Just one final word...allot of what we see about ISO ranges is a manipulation of the camera design for the benefit of marketing.

Dikdik Loc: Winnipeg, Canada

I hear that camera manufacturers are adding noise to higher ISOs to give photographers the impression that ISO means something.

Dik

Mongo Loc: Western New York

Ha ha, good rumor!

Camera manufacturers have also added algorithms for film effects, including grain, scratches, etc. About the only film effect I have not heard implemented is exposure reciprocity. And someone probably has done that, I just didn't find out about it.

Larryshuman

All cameras have a native ISO. When ISO is changed upward it is gain that is increased and with increase of gain come increased noise. Read Steve Perry's new ebook on metering and in the last 20 or so pages he goes into this very subject.

Mongo Loc: Western New York

And my point is that for a designer, faced with marketing expectations, it is easy to have the ISO knob be a knob that does not in reality, for each click, change sensitivity (quantum efficiency perhaps) for noise.

Depending upon the architecture of the sensor, there are several ways to increase gain. However, noise is always present. So if there is less signal, the signal to noise level is likely to be poorer.

speters Loc: Grangeville/Idaho

I never do that, if I take delivery of a new DSLR, I go and shoot it in manual and then, when I'm going to get a feel for it, I try one of the auto settings!

Mongo Loc: Western New York

I would agree. I just picked up a dSLR, about four weeks ago. I mostly shot in manual mode for two weeks. Then aperture preferred. But I am still playing around with it. Unfortunately, I wish that some of the settings are more documented as to what they do. In a few cases I have done experiments to characterize the performance with changes.

The first three days I had with it were cold, rainy and with wet snow, so I stayed inside on the couch, and played with all the features and settings. Three full days. I think I checked out about 80% of it's capabilities, but perhaps not.

There are still things I want to do with it, and am pretty sure I can, but I haven't quite figured out how to accomplish them yet.

Mongo Loc: Western New York

A more useful metric commonly used in the design of sensor systems would be the spatial resolution and the quantum efficiency of the sensor. The spatial resolution helps drive the well size impacting gamut. The quantum efficiency gives the ratio of electrons per photons. Loosely.