The same size as their predecessors, the small two-way M5i standmount now costs £425 per pair (up from £350 for the M5 in 2005), while the floorstanding two-and-a-half-way M16i costs £850 per pair (£800 M16 in 2007).
Both models use what appear to be identical bass, mid and treble drive units and both bass, mid drivers are loaded by similar size enclosures - rear-ported in the case of the small M5i, but a sealed sub-enclosure in the case of the larger floorstanding M16i (which has an additional port-loaded, bass-only driver occupying the lower section of the enclosure).
The tweeters all have 25mm metal dome diaphragms, while the cone drivers have 135mm frames and 95mm diameter moulded plastic cones. Those operating up through the midband also have 'bullet' phase plugs fixed onto their central magnet polepieces, whereas the bass-only drivers have a regular dust cover attached to the cone.
If the outlines are similar, the changes are numerous. The cabinetwork is particularly impressive considering the prices, with curvaceous vertical edges. Improved quality book-matched real wood veneers come with smooth finish for the cherry and a fabric red cherry optional grille, though fixed perforated metal covers protect the metal dome tweeters.
The tweeters themselves are significantly different from before, with metal faceplates in place of plastic and double ferrite magnets in place of neodymium, both of which add to the heatsinking, improving thermal stability and power handling.
The M16i is now supplied with a plinth already fitted to the enclosure. Although the latter ensures secure spike fixing, it has exactly the same small footprint as the enclosure, and therefore does nothing to improve the very questionable stability - the likelihood of it passing the EC 'knockover' test seems remote.
The bi- and tri-wire terminal pairs are now mounted through a flat alloy panel, optionally connected by brass links. Superior polypropylene capacitors in the crossover networks replace the bipolar electrolytic types that were previously used.
Our measurement regime clearly shows that the changes introduced in these i-suffixed versions are much more than skin deep. Examining the M5i first, the impedance is now altogether easier to drive, staying above six ohms throughout (rather than falling to four ohms minimum) and the overall averaged in-room frequency response is less uneven and more extended at the top end.
The 56Hz port tuning ensures plenty of midbass output wherever the speakers are placed, but there are difficulties further up the band. In free space, there's a lack of output at 60-250Hz.
Close-to-wall siting was preferred on balance, as this usefully helps lift the M5i's bass output, but leaves it rather too lean through the upper bass and broad midband (100-800Hz) and more exposed in the upper mid (800Hz-1.5kHz).
Furthermore, the treble (4.5-10kHz) looks significantly stronger than average. Through the bass and midrange, the M16i measures exactly like its predecessor.
However, the change in tweeter and its associated network has wrought a dramatic improvement in the response above 2kHz, completely avoiding the deep suckout centred on 3.5kHz found in the previous version and also usefully extending the top end.
Although the overall in-room averaged frequency response is impressively smooth and flat above 120Hz (albeit a slightly prominent 700Hz-1kHz), the bass alignment again proved problematic under our room conditions.
The 45Hz port tuning interacts with a room to create a substantial excess around 50Hz, even when the speakers are positioned well clear of walls, leaving the 60-120Hz octave too lean and rendering close-to-wall siting inappropriate. Furthermore, the load dips briefly below four ohms in the low bass region.
The various balance anomalies, especially the rather strong treble, proved problematic in the listening tests. A bright top end is sometimes acceptable, particularly with tweeters of the highest quality, provided the bass and lower midband add a corresponding warmth at the other end of the spectrum.
Unhappily that is not the case with either of these Epos models. Both tend to be a little lean through the mid and upper bass, but the situation with the small M5i is much more serious than with its bigger brother.
Here the sound imposes a cool, thinned and bright character to all kinds of music, due as much to a lack of punch and authority through the lower registers as the strong top end.
Diction is open, clear and articulate, but there's an obvious tendency to over-project the voice band in general and to over-emphasize the sibilants and fricatives (forced consonants) on speech.
Although detail is always explicit, the tendency towards edginess and 'splash' on percussion instruments like cymbals inhibits the enthusiastic application of volume for fear of inducing fatigue.
The M16i is significantly more palatable overall and while it does share some of the characteristics (as well as the ingredients) of its smaller brother, these are somewhat less extreme with an end result which is more enjoyable as a consequence.
Sibilants remain quite strong, but are certainly more acceptable, and speech retains that attractive openness, but some 'cupped hands' coloration is also audible in the voice band.
The overall character remains a little on the thin side overall, which can become a trifle wearing over the long haul, while on some tracks a degree of bass 'thump' can become intrusive, especially if the speakers are too close to a wall, though this does go some way towards 'balancing out' the fairly strong top end.
A mixed bag of results
Overall the M16i gives a pretty good account of itself, aided by fine mid-to-treble openness and neutrality, but finding ideal locations for the speakers proved quite difficult under our listening conditions.
While the cabinetwork on both of these Epos models is quite exceptional, the same cannot be said for their sound quality.
The M16i certainly passes muster, especially if the bass end interacts well with the intended listening room, but our samples of the M5i proved disappointingly thin and bright.