In blue-boxed measure, at 1st downbeat the chord seems to be IVMaj7. Beamed notes seem to suggest ii triad.
I have read that IV and ii might come together from time to time and collectively they were called predominant complex. I considered it is (IV7-8, 5-6) but I have never seen such figured bass, normally it is 8-7 and 6-5.
But again the ii chord is at upbeat and no bass is played.
Do people consider such chord as a single chord(IV7-8, 5-6) or two separte chords(IV7 ii)?
Short version: Just because some notes line up vertically does not mean we need to explain them as a chord.
When we analyze, we have to look at the context, and not just examine each "stack" in isolation. In very simple homophonic texture, it's easy enough to look "stack by stack" and call them chords. But once composers start throwing in nonchordal tones and doing interesting voice leading, we have to take off the blinders and look side to side a bit.
In this example,
... it's enough to say we have "V, I." We don't have to say that we have a V, then a V9, then a V, then a Vmaj7, etc. It's just a simple cadence with some melismatic scalar passing motion, and that's enough for a harmonic analysis. (If we want to dive into melodic or textural analysis, we can progress to that.) Harmonic analysis means asking "What do the chords do," and it works best when we start with our widest lens and make the biggest, most crude summary, and then zoom in and refine it as needed.
"What do the chords do" at the cadence of your example? Counting from the 4th measure, they do I, V, IV, V, I. No need for anything more exotic. Now we just dig deeper to explain anything that's added to that framework.
The ties are a very strong clue that these notes don't need to be explained as belonging (chordally) to measure 5. Instead, they're "left over" from the previous chord. It just so happens that the altos' Bb is also a chord member in Eb, but the sopranos' D is explained as just a suspension. Suspensions resolve downwards, and indeed the sopranos get to the C, but they indulge in an escape tone of Eb first, and the altos come along for the ride.
^7 or its final ^2 is the important tone resolving to ^1. But it's often "six of one, half dozen of the other" or I go with which ever choice avoids some voice leading "problem." Anyway, yes, this is just C: V I.
Commented
Jun 6 at 16:22
It depends what level of harmonic analysis you want to do.
There is the deepest, structural level of just analyzing a whole piece - I can't see the beginning, but this one probably included - as Bb: I V I.
You can go to smaller and smaller levels of analysis.
You can analyze just the cadence of each phrase (being sure to include key labels.)
The next levels can be a bit of a mix, but you could analyze at the one bar level, probably the first beat of each bar, or possibly divisions of the measure down to the beat level. This depends a lot on the music. You would be trying to find the harmonic rhythm of the piece and analyzing the chord changes accordingly.
You're question is whether to analyze at the eighth note level. You can, but I would not. Not unless it was required by some exercise, either self-directed or from a teacher.
Personally, I ran into this problem the first time I tried to analyze Bach's first two-part invention. You could analyze that down to the sixteenth note level! I was really confused about how to do it compared to a simple SATB chorale. The solution is to analyze at the level you think is appropriate for your purpose. I'm most interested in how phrases are used to construct form. So I tend to analyze phrase cadences and then fill in important beat one changes. Also, in contrapuntal music, I tend to do Roman numeral analysis just at cadences, everything in between is contrapuntal texture far better understood via counterpoint rather than root progressions.
Also, in recent years I've been influenced by thorough bass, so I've moved away from thinking that analysis has to be Roman numeral root progression analysis (RNA.) There is an adage, I think I read it from Schoenberg, that to understand harmony you should "follow the bass".
So, finally, the moment in the music you highlighted is the final cadence, the bass is ^4 ^5 ^1, the first beat harmony is clearly IV, anything of ii is weak beat decorative motion. I would analyze it IV V I. Like Andy Bonner said in comments, I wouldn't even care about labeling the 7 on IV. It's just a retardation of the E5 and so is melodic/contrapuntal and no real concern for RNA root progression analysis.
FWIW, it think the interesting thing is the soprano ending on ^3 instead of ^1 so the final cadence technically is not a perfect authentic cadence. Sometime pieces that end in this way on ^3 are described as "pastoral".
7-8 and 5-6, called retardations (upward-resolving suspensions), and are "legit" if less common than the usual 8-7 or 6-5 resolutions. I would interpret this passage as ii[6-5] - V, especially since it's a classical setting where 7 is a dissonance that must resolve.
The harmony in bar 6 is B&flat major, and the majority of the previous bar 5 is F the V of said B♭. So it's a perfect cadence there.
But also in question is the start of bar 5, so IVmaj7 >V fits that bar.
