So a 5.1 track can have tremendous bass but a 7.1.4 must not because it needs more headroom? Really?
Not at all. The amount of headroom (peak SPL per channel) in the track is chosen by the mixer based on the monitoring level he or she selects before doing the mix. Even a mono track can have "tremendous bass" if it's monitored and played back at a high enough level (and the playback system is up to the task).
Note that it doesn't make sense to talk about a 7.1.4 track, even though it appears some titles may advertise "7.1.4" simply because they were monitored on a "7.1.4" playback system. The soundtrack for Atmos home mixes is really only 7.1. The objects are encoded via matrixing with metadata that allows the decoder to lift the objects out of the mix so that they can be rendered on the playback system, which could be 7.1.4 or whatever.
For the same amount of per-channel headroom a 7.1 track actually has more potential for bass than a 5.1 track does because there are more channels available that can contain bass. However, a theatrical Atmos track has even more potential compared to a standard 7.1 because each of the objects can contain bass that gets bass-managed to the appropriate subs (front or surround) in the playback venue. The problem comes when it's time to down-mix, that is, recreate the same soundtrack using fewer channels. Assuming the original track had one or more channels hitting digital full-scale (not at all unusual), either the down-mix will require more headroom (and a higher monitoring/playback level) or else some combination of compression, limiting, or clipping must be used to make the waveform fit. This is already a problem when down-mixing from 7.1 to 5.1 and may explain why often the 5.1 track has a bass filter where the 7.1 track does not. For a theatrical Atmos track which has 7.2 (IIRC) beds plus over hundred simultaneous objects, the problem is much greater.
One might ask at this point what this has to do with loudness. Very roughly speaking, loudness correlates with sound power. If a sound is played in more than one channel at once, it is roughly true that the average SPL and sound power (and hence loudness) increase by 3 dB. By mixing the sound into each channel at a level of -3 dB, the same loudness is achieved as by playing at 0 dB through one channel, but the peak SPL still goes up by 3 dB. If those channels get down-mixed to one, then the peak SPL required of that channel is still 3 dB higher. So in other words, the more channels you have to begin with, the more peak SPL required in each channel for the down-mix, even for a soundtrack having the same loudness.
A majority of the Atmos movies cataloged here just so happen to be the big, bombastic type mixes and most of them have had lots of compression just because they wanted them to be/sound loud.That's it. Atmos has NOTHING to do with that. There are plenty of good sounding ones that are not compressed to shit.
I realize this is a very confusing subject, which is why my argument has been misunderstood so far. To the extent that what Infrasonic here says is true (and I don't disagree), the original theatrical Atmos mix is already compressed quite a bit and made to sound loud. And these kind of mixes suffer *even more* when down-mixed to 7.1 (for Atmos @ home) with 3 or 6 dB less headroom per channel than the theatrical reference. If the original mix was not so loud to begin with, then the down-mix might have only a few peaks that exceed full-scale that need to be compressed or limited. The sound quality impact may be quite monitor. However, for a mix that was highly compressed to begin with, a lot more content is going to be pushing above full-scale and the compression and limiting will have to be much more aggressive to keep it contained. Hence, the down-mix will sound worse than either a "quiet" Atmos mix or a "loud" 7.1 theatrical mix.
Note that we should expect this problem to be there whether or not the "home release" was done in Atmos, as long as the original theatrical track was done in Atmos. Simply because the original mix was prepared for systems with much higher *total headroom* (taking all channels and objects into account), the sound quality in the down-mixes will tend to suffer more for these unless the headroom is increased by choosing a higher monitoring level (and requiring home viewers to turn up the master volume more to achieve parity with other content).
Whether or not the studios want their home mixes to sound like clipped garbage is not clear. We are told that the director "signs off" on the home mix, but that doesn't mean that the severe clipping that turns up is necessarily desired. This issue is complicated by the likelihood that mixers are monitoring home mixes with substandard equipment that may not be up to the task of accurately reproducing the mix even at at the lower monitoring levels typically used. It may not be possible for either the director or mixers to critically evaluate what they've done. OTOH, I see some evidence that "home mixes" are starting to fall out of favor in some cases, partly evidenced by the number of recent home releases (e.g. "Avengers 2") that sound "a lot quieter". (Another piece of evidence is a comment posted by FilmMixer on AVS, sadly deleted a few hours later, that suggested "about 60%" of his recent contracts have included near-field mixes). Most likely "Avengers 2" wasn't intended to be 8 dB less loud but instead we got the theatrical mix instead of a home mix. Or those mixers realized what I'm saying above about headroom with Atmos and opted to not abide by "the (not publicly available) standard" for home mixes that says they should be done with monitoring level 6 dB (or whatever) lower.
All I'm saying is that there is a very real technical reasons why an Atmos home-mix will likely sound worse than a home mix done from 5.1/7.1 original content unless the monitoring level used for the mix is increased. Those mixes that are done to existing standards will sound like trash, but mixes done by studios who actually listen and note that their mix-downs sound like ass will opt to push the monitoring level up, the result of which is cleaner more dynamic mixes that require the home viewer to turn the MV up a bit more than usual. That's it.