The Exposure Triangle Is Misleading And Unnecessary.

+1. Don't do in camera what can be done at least as well in post,

my raws go strait into ACR and adjust levels and save a preset based on the raw file then into photoshop for my last post processing which is always sped up using one of my action presets i have created. then do that 500 times over , the boring part 😒 do a test print on the first to make sure i havnt forgotten or missed anything than away i go.

I thought I remembered that. So you don't know the characteristics of the raw files.

To be honest i try to shoot the best image quality out of camera possible with very little post processing ,my lighting and wb are near perfection and has to be because i shoot live with twin 24 inch monitors for the client. I save the 16bit psd file for cases where my very finely filtered grey backgrounds might posturize in print but the 8 bit jpegs have amazing amounts of smooth tonal grades from the a74 not like my past m43 cameras. my canon pro 10s printer is excellent as well and sell around 1000 prints a year from it. Ive just stumbled across this amazing satin photo paper during the last week, skin tones are on another level compared to some other brand paper i use.😁 just testing colour fast at the moment im hoping it holds up well in direct sunlight for at least 6 months.

Live histogram is based on the JPEG output, so if you set up the JPEG processing to be closer to the RAW, then the histogram will reflect the RAW output. For example, with my Olympus E-M1 II, I have switch to the "Muted" picture mode and set -2 contrast and -2 saturation. If you search or experiment you might find the right settings for your particular camera to get an accurate histogram that you can use to ETTR.

If you do not care how the image looks in the EVF, why not use UniWB instead? It is very easy to set up with Olympus.

In what way is it false?
Are Bob's statement about the ET false
Or is the ET in itself false?

Morning Jim and Illiah,reconvening after catching up with some sleep. Before I plough on, I want to thank you for your time, I can imagine how tedious and frustrating it must be answering kindergarden questions, especially as you don't get paid. I do appreciate it.

Recapping the state of my knowledge so far...

(Assuming I'm shooting raw only and using aperture priority mode)

• when I twist the ISO dial by one stop, the camera meter selects a 1 stop higher shutter speed (meter now thinks I'm shooting ISO 200 "film")
• This results in a one stop less exposure compared to standard (resulting in more shot noise and reducing SNR)
• The camera may use analogue amplification before ADC and/or multiplication of the raw data after ADC (this may reduce read noise, but this is minor component of overall noise and will be outweighed by the increase in shot noise from the reduction in exposure. It may also risk early clipping)
• The alternative to raising the ISO, is to dial in -1 exposure compensation. This will also raise the shutter speed by 1 stop but will not apply any analogue amplification before ADC or alter the raw values after ADC (this will result in higher read noise than raising ISO and the same shot noise increase but avoids the risk of clipping. As shot noise outweighs read noise, the extra read noise may not add anything visible to the increased shot noise)
• If the above is correct, it seems to me that raising ISO makes things overall worse by potentially introducing clipping without doing much to reduce noise. Therefore I conclude that the ISO dial is nothing but an alternative, but inferior, exposure comp dial and effectively pointless and I should ignore it and use exposure comp instead and use raw convertor tools to apply whatever curves are needed to achieve the required visual appearance of the image.

No doubt there is far more complexity to it than this...

p.s.

I like the editor's preview function. Something DPR never implemented forcing you to post to review what you had written or compile the text outside the site and paste it in when you had revised.

A triangle is a plane figure with three sides.

A set of three independent variables is not a triangle. It can define a cuboid, such as a brick, with length, width and height.

Don

I think the exposure triangle is an analogy, a metaphor or simply a simple visual graphic image that communicates something. That's more important than geometric accuracy. A better know example of pretty much the same thing is the Fire Triangle

en.wikipedia.org/wiki/Fire_triangle

I think this kind of thing is a legitimate communications device if it is easy to grasp and gets the message across. The problem with the exposure triangle device is that it may not actually work in the context of digital cameras. It's more useful for film.

I would suggest that the ubiquity of the exposure triangle motif suggests that it is something that resonates strongly with a lot of people and makes intuitive sense with them (even though it is incorrect). In that narrow sense, it is clearly a successful device in that it communicates where other devices don't. The fact that the concept doesn't apply to digital cameras is made all the more unfortunate by its success as a communication vehicle.

However, I suspect that this mis-fortunate step has more to do with the complexity and confusion round how digital ISO works than the motif itself. I don't consider myself a beginner photographer by any stretch but I believed the exposure triangle was a correct-enough statement. It wasn't the exposure triangle that led me to this incorrect belief, but other writings, especially the early writings on DPR. Also, camera designs reinforce this error. The ISO dial on my old Nikon D100 looks exactly like the ISO dial on my old Nikon F80. In fact, it is probably the same part from same part bin. Manufacturers have encouraged (maybe by accident) film photographers to believe digital ISO works the same as film ISO. I don't know if it is semantically correct to describe an accidental confluence of events as a conspiracy, but it almost seems as if there has been a deliberate attempt to obfuscate and over-simplify digital ISO. Given the difficulties I've been having getting my head around it, it's not entirely surprising that mass marketers don't want to make the effort to get the truth out there. Maybe a similar dilemma afflicted Sigma marketing when they gave up trying to explain the Foveon sensor and resorted to the lowest common denominator of making convenient sounding stuff up. It's not entirely fair to pin the whole mess on the exposure triangle - that motif is popular because it is an effective communications vehicle but it's incorrectness is more a symptom of a wider malaise than the cause, IMO. Hence the attempts of articles like the DPR one you advised on to correct the record.

Fortunately, the good folks here are slowly correcting my own misconceptions and I'm grateful for that.

Having spent 37 years of my working life as a "computer man", I can confidently say that there are many different kinds of computer men and they don't all think alike. I started my computing career writing scripts for a DOS based office package for our finance team, moved on to writing audit random sampling and totalling programs in Filetab running on the Met Office's weather forecasting IBM mainframes, moved on to being an Oracle data adminstrator, then on to being a web developer and content manager for our corporate website and intranet. For many people in my office, I was "The Computer Guy". But the reality is beyond the specific things I worked on, I don't really understand a lot about how hardware or software works. There's all kinds of computer man.