Terminator,
I just reviewed both files you attached and I don't see what you do, but my perspective may not be the same as yours.
Here is the sequence that is ideal, keeping in mind that ideal is just that. The real world offers varying tones to the palette.
You get FTT and then, while price traverses from FTT back to then current RTL, volume will be decreasing just as usual for a non-dominant. Then, the RTL will b/o, and increasing volume will be seen as a retrace (non-dom traverse) transforms into a reversal (dominant in new direction, also known as "change").
My hunch is that the two different b2b (or r2r) that you see is because you're looking at a lower resolution than full forest level. Keep this in mind: while the ideal picture is each bar in the first half of a r2r (for example) is a red price bar with a volume histogram that is sequentially lower in value, this rarely if ever happens. Flaws occur and sometimes things just aren't ideal. Bar to bar or PRV is not where you need to be to see the forest level change in the biggest perspective. You may have some black bars mixed in, the bars probably won't be sequentially lower. You may get a spike somewhere in the middle. None of that changes the overall effect of the b2b.
In order to "see" the overall effect, you need to do one thing: match the number of price bars in the traverse to the number of bars in the histogram you are referencing for determining increasing or decreasing volume.
This is a gross process. No sophisitication or delicateness is involved. I know first hand how easy it is to NOT see what you can't see. Looking back, I can't believe how stubborn my eyes were. Once you get it, you'll say to yourself, "how the hell did I miss this", it's that obvious (once you see it

)
Hope this helps.