Why ratio function in DMMs can't show the two voltages?

[NaxFM]

The title says it all.
I know that the ratio function can be very handy when designing or verifying a voltage divider, using a DMMs as transfer standard and other stuff, but if it measures both voltages, why just don't offer the possibility to display them on screen?
The only explanation I could give myself is that with the ratio function the new voltage reference for the ADC is the reference voltage input, so the DMM doesn't really know what voltage it is but can only compare it to other voltages. This can also explain why the range of the reference is always fixed to the usual 10V
However this goes against what the application note of the 3458A regarding ratio measurements says
That note says that the two voltages are first measured and then the ratio is displayed on screen, so in this case the DMM knows exactly the two voltages.

[BeBuLamar]

I am sorry but I haven't seen a DMM that has the ratio function.

[TimFox]

Some bench DVMs, such as the -hp- 3456, have a voltage ratio mode, where the display is the ratio between the voltages at two pair of binding posts.  In the 3456, there is a limit to the voltage that can appear between the two "Low" terminals. 
If I remember correctly, the 3456 does not measure each voltage against its internal reference voltage, but rather measures the ratio by applying one voltage tothe reference of the ADC.  Therefore, there is no accurate voltage to be displayed in this mode.  I assume the 3458 is similar in that mode.

[NaxFM]

I am sorry but I haven't seen a DMM that has the ratio function.

Almost all bench multimeters have this function, even the old 34401A.

[NaxFM]

Some bench DVMs, such as the -hp- 3456, have a voltage ratio mode, where the display is the ratio between the voltages at two pair of binding posts.  In the 3456, there is a limit to the voltage that can appear between the two "Low" terminals. 
If I remember correctly, the 3456 does not measure each voltage against its internal reference voltage, but rather measures the ratio by applying one voltage tothe reference of the ADC.  Therefore, there is no accurate voltage to be displayed in this mode.  I assume the 3458 is similar in that mode.

Thanks, so it is just as i thought.
The application note for the 3458A explains it very badly, i'd even say it's just wrong.

From the application note:
"A “Ratio” measurement performs this same comparison but instead uses a second
measurement of an external reference voltage that is generally known to a much greater
precision"
...
"The DMM will automatically measure the two applied voltages using internal switching
and then automatically calculate the resulting voltage ratio value from the two
independent measurements. "

It says multiple times that the two voltages are measured and only later compared

[Kleinstein]

Some old meters (looks like the HP3456) actually used one voltage for the ADC reference and thus only do 1 converserion (and usually a zero for the auto zero mode). 

Many new meters (e.g. the HP3458, HP34401) make 2 seprate voltage measurements and calculate the ratio. So they internally have the 2 voltage reading and not showing them is just a limitation in the software. It is a balance in what functions to offer and keeping the UI simple. One point there is how to show the 2 results on a single line display.

In many cases they actually do 4 conversions: the 1st input and zero for auto zero mode for the main input and than a conversion each for both inputs for the 2nd voltage (usually the ohms sense terminals). So in theory they could also give 3 voltage reading relative to the common terminal.

[wizard69]

In modern equipment this is likely a pure software issue.    Your question makes me wonder if this is a general question or you have a specific issue to address?    In the end you can always hook up a PC and write a scrip to grab the values, display them and compute a ratio.   That would require two meters in many cases though on some bench meters you could also switch between front and rear terminals.