6.5 is the figure used at PhotonsToPhotos to determine Low Light ISO.

Here's a link to the table with relevant values: www.photonstophotos.net/Charts/PDR.htm#resizableT

It's important to realize that such sensor testing reports dynamic range in linear space.
But the final image, and human vision, operate in a more logarithmic fashion.
So any comparisons would be "apples to oranges".

The pixel is an analog device with a dynamic range.
We determine (measure) that dynamic range (and other values) by repeatedly sampling the pixel to get, for example, and average (Signal) and a standard deviation (Noise).
These values, the Signal and the Noise are not integral, they are decimal numbers. Both Signal and Noise are measurable below 1 Digital Number (DN aka ADU aka LSB)
They are not measurable all the way down to zero; that's where quantization error fits in; quantization error says you can't go lower than ~0.288 DN
But, in practice, because of inaccuracies in the bias (BlackLevel) we only see reliable results down to about 0.6 to 0.7 DN (the Panasonic being an extreme case at 0.54DN)

Regards,

Sorry, this is exactly where you are wrong, at least regarding the application in digital cameras.
In my read noise measurements of over 200 cameras I have definitely encountered cameras with read noise of less than 1 LSB.
Most of these cases are 12-bit ADC that really should be 14-bit.
Here's one example, the Panasonic Lumix DMC-GX80
See www.photonstophotos.net/Charts/RN_ADU.htm#Panasonic%20Lumix%20DMC-GX80_12
Note the curve starts to "misbehave below 1 (0 on logarithmic y-axis) due to the influence of quantization error.
But it ultimately gets down to about 0.54 DN; that's actually 13 stops

Substantially correct. I have a few tweaks below...

Analog to Digital Converter (ADC) bit depth is simply bit depth (eg. 14-bits). Regarding dynamic range, see below.

When you raise the ISO setting (one stop for example) you lose an equal amount of highlight dynamic range because the ADC gets clipped.
Surprisingly, you actually get a small increase in shadow dynamic range (noise goes up a little less than signal so Signal to Noise Ratio (SNR) actually improves in the shadows).
The net effect is that you loose less than one stop of dynamic range per stop of increased ISO setting.
WWhen the shadow improvement reaches diminishing returns you achieve "ISO Invariance".

Canon cameras often use either digital scaling or a second analog gain to accomplish intermediate ISO settings which can give dynamic range curves a "jagged" look.
At PhotonsToPhotos I detect digital scaling and other manipulations which are indicated by various symbols on the charts.
For example, an open triangle (pointing up) indicates digital scaling whereas an open triangle pointing down indicates noise reduction.

Quantization error is not added to the noise. As you indicated it is a limiting factor on the smallest measurement that can be made.
Theoretically quantization error is limited to 1/sqrt(12) = 0.29 bits but in practice things fall apart below about 0.6 to 0.7 bits.
So, strange as it seems, you can get about 1/2 stop more dynamic range out of an ADC than it's bit depth. In other words 14.5 stops for a 14-bit ADC.

Regards,

For normalized measures, the one we normally care about, what matters is sensor size rather than photosite size.
FWIW, PhotonsToPhotos also has a Low Light ISO measure.
It's listed in the sort-able table below the interactive Photographic Dynamic Range (PDR) (www.PhotonsToPhotos.net/Charts/PDR.htm) chart.
Comparisons are best made with the EV (stops) value.

No. Photographic exposure is determined by the Exposure Value which is the aperture and shutter.
The lightness or darkness of the resulting photograph starts out with the photographic exposure but also has had tone curve and other adjustments applied.
Think about it, you can't overexpose a picture by using the making it too bright in Photoshop.
The exposure is baked in when you choose you aperture and shutter speed.

They also often care if a lens is parfocal, does not change focus when focal length is changed (zoomed). Most of our zoom lenses are not parfocal.
I have one such lens, it can be very helpful.

You could be right but I'd like to see a citation on that.

I don't know if this distinction is made at Ugly Hedgehog but shutter speed and aperture are your exposure and the ISO setting affects the lightness of the image, not exposure.

I don't have my birds in flight example handy but this old post is similar:
www.photonstophotos.net/GeneralTopics/Exposure/ISO_Auto_on_Parade.htm

Oh, here's a bird in flight:
For the 16 shots ISO varied from 220 to 640.

I don't shoot Canon but the only time I don't use ISO Auto is when I'm using flash because ISO Auto isn't smart enough to know about the additional light.
(There's no harm, remember, nothing happens unless you can't gather enough light by aperture and shutter speed.)

With the 7D Mark II they sky is the limit although you won't gain much above about ISO 1280.
See www.photonstophotos.net/Charts/PDR_Shadow.htm#Canon%20EOS%207D%20Mark%20II

The focal length of a lens is determined at infinity focus.
As you focus closer the only way to maintain that infinity focal length is to focus by extending the entire lens away from the sensor plane.
(One common current example is the 50mm f/1.8D AF-Nikkor (D not G) )
But that would mean that lenses could extend very far out on a long helicoid.
So quite some time ago lens designers started making the lenses extend less but decrease in focal length as you focus closer.
For most lenses that also meant the air would move in/out of the lens; hence the term "focus breathing".
But it's a misnomer because even internal focus lenses that don't "breath" almost certainly decrease in focal length as you focus closer.
Because the focal length decreases as you focus closer you get a larger Field Of View (FOV) so yes, it might appear that you are zooming.

Well, I think you misunderstand my position.
I never said don't take a picture at higher ISO settings.
I say use the lowest ISO setting that your photographic situation allows.
That means don't take a picture at ISO 800 that you could take at ISO 100; but do so if you can't gather enough light for ISO 100.
You may get more Image Quality (IQ) than you need.
On the other hand, you might be happy to have the "excess" IQ if afterwards you decide to crop more aggressively than usual, for example.

Often when a range is cited it's the range of ISO setting for which only analog gain is applied by the camera electronics.
For the ILCE-6500 from DxOMark numbers I judge that as ISO 100 to ISO 6400. (I haven't got A6500 measurements of my own yet.)
Beyond ISO 6400 appears to be digital scaling.

"Native" ISO is a confusing subject. It's generally the highest ISO setting that has the lowest analog gain applied inside the camera.
Often that's the lowest ISO setting excluding "extended" ones, but there are exceptions.

For your Canon EOS 5D Mark II ISO 160 is a good choice but actually ISO 100 could also be considered native.
Canons screwy treatment of intermediate ISOs has always confused this issue.

For the Sony ILCE-6500 it's ISO 100; period.

There is never a range of native ISO settings, but more and more often you will see two rather than one.
Dual conversion gain technology makes the pixel behave in two different ways each with it's own native sensitivity.
For these cameras I would say there are two native ISO settings. (This can be a nuisance to those who use ISO Auto)
An example of such a camera is the Sony ILCA-99M2; one native ISO at 100 and the second at ISO 500.