In stock trading softwares, the volumes underneath the candlestick chart of stock prices are typically marked by one color, either green (buy volume) or red (sell volume), depending on if the closing price on that day is higher or lower than the opening price. But in fact the stock price can go up and down within a day. So why not accumulate the buy and sell volumes per minute separately and display a two-color volume bar each day instead of painting the entire bar either green or red? Does anybody see the down side of a two-color volume bar system?
Trivially, buy == sell volume, as discussed in other answers. Yet this is a fair question, as there is benefit to look at lift volume vs hit volume. The former balances over longer horizons, and latter may have value in intraday moves. I think that is what ZH suggests.
It is likely that, when prices move rapidly, there is a significant weight on one side of the markets (i.e lift >> hit and v.v.).
Then this metric would only be useful in quieter, intraday periods.
Your question has two further legs: (1) i see no value in transaction velocity that you describe, but would see volume-tick ratio as some kind of appetite metric (2) note that the bar colour is not representative of volume, but of O-C direction.
By definition buy=sell volumes. You need a counter-party to make "volume" happen, therefore the cumulative buy=cumulative sell volume.
Volume is predicated on your time frame, if you make 1 minute charts the volume will be determined on the minute open and close price for coloration.
there is no "buy volume" and "sell volume".
"Buy=sell" on actual transactions, but market sentiment is related to sell order volume (possibly unfilled) and buy order volume (possibly unfilled) at a particular time. This is not easily available.
The down-side is that the two-color outcome depends on the sub timeframe you choose.
For example, 1-minute and 15-minutes volume aggregation will have different results, because a 1-minute bullish bar with a abnormally high volume within a bearish 15-minute bar would be counted to down-volume in a 15-minute aggregation but to up-volume in a 1-minute aggregation.
So one would have to adjust two timeframes. I guess the effort and confusion is not worth the little gain of information in most standard charting tools.
This might be overcome with using the lowest timeframe possible, hence using every single price tick.