tenia

Regarding Disney's recent HT mixes, I can only second what others are saying about them : they are disappointing, and no, it's not just a question of loudness or master volume.

What's interesting is to compare how it pored through video releases : the Forbes article has the starting point wrong : AOU was already problematic on BD. It was under-cranked, but also flatter than expected and with empty bass (I suppose the track doesn't go particularly low in frequencies).

Its unfortunate the article is imprecise regarding this (though I guess that can only be expected from a non-technical publication like Forbes), but having now listened to many of the recent Buena Vista BDs, they do feel disappointing outside from the master volume being too low. They certainly aren't all the same, and some sounded better than others, but AOU, Black Panther, Last Jedi, Thor Ragnarok, GoG 2, but also things like Coco.

Infinity War sounds a bit better and mostly is slighlty under-cranked, but even once the MV is increased, it still feels less impactful that one would expect.

I can't say if it's a creative issue more than a technical limitation. But what I do know is that something definitely changed in the way these tracks are mixed, and despite whatever people might say (especially regarding how other studios have clipped or filtered tracks, as if it explains why the BV tracks are picked up but not the other ones, or as if it would somehow makes it better to know that), it's telling to see in the first post of this very thread the scores obtained by earlier MCU discs (Winter Soldier, Incredible Hulk, Thor 2) and then look at the scores for more recent movies.

I don't know what Disney changed, but clearly, if they didn't change anything, we wouldn't have this discussion. If they changed it in a way that didn't feel this obviously debatable, we wouldn't have it either.

So if Disney wants to simply go back to whatever they were doing 6 years ago, I'm all for it, and I DO hope they listen to the complaints to just go back and do that.

I like your enthusiasm.  If you would like to re-vamp the ratings system based on the data we have gathered, I'm all for it.  I know the rating system is flawed, but at this time, even if I or you did change it, I would not have the time to go back and re-rate all the films, update all the links, etc.

 

If you would like to take on this task, please do so.  My time is much more limited now than when I started this project; I can barely watch one or two films in the HT at greater than -10dBRef monthly, if I am lucky.

 

JSS

I like what this forum is doing, there isn't many forums focused on such a specific part of a BD (sound) and then such a specific part of this part (bass). I get often through data at work so I'm not too bothered working this kind of things around. I'm certainly not a statistician expert but if I can help to slightly fine-tune data systems, why not ? It's more of an exercice to me to keep practicing working data and figures and if this helps somehow...

This being written, I probably won't have the time to re-rate individually the movies individually. So far, I'm more looking into the spreadsheet as a whole, cluster by cluster.

The first thing I see is that the Dynamics data have basically no correlation with the final averaged score (movies with dynamics at 30dB can end up with final scores of 1.5 up to 4.5), so that would be one thing to look at, because Extension and Level are much more correlated to the final score. There is way too few Dynamics score below 2 (only 3 out of 400). It's too easy to score 3+ but even a 5 on Dynamics can yield an overall score of 2 or 3 (53 movies with 5 on Dynamics score 3 or less in overall). I would take the following scoring instead of the current one :

1 star < 25 dB

2 stars < 27.5 dB

3 stars < 30 dB

4 stars < 32.5 dB

5 stars > 32.5 dB

With this new scoring, it's down to 7 movies scoring 5 on Dynamics but 3 or less in overall.

An other more important thing would be the extension. There are MANY movies below 10 Hz (180), so it's also helping good extensions to score very high very quickly. It doesn't seem realistic to cut down the <10 Hz into 2 scores (because there's no way a 7 Hz extension shouldn't score a 5), but there also might be something to do there. Half points or decimal scores if the extension is between 1 and 10 Hz ?

Finally, I'm not sure the Execution should be taken into account in the average score. It's the only thing very well correlated to the overall score, but it's also a subjective element. It might be better off outside the overall score calculation, as some kind of cross-checking value ("OK, the figures are good BUT this one sounded better than that one to me so I would push this one rather than that one"). It's correlated in a rather clean way anyway, so I wouldn't worry about sending too often contradictory messages.

Level seems fine. There seems to be a bit too much movies scoring 3 or less not correlated to the final score compared to 4+ movies, but why not ? There's a slight shift maybe to do there.

Is there a more recent spreadsheet available ? It's an easy way to navigate through the data available, so if there's one, I'm all for it, it'll be a more complete work for me. I don't want to propose changes which turn out to be inadequate.

As a disclaimer : I'm not trying to pick a fight or anything. I was just genuinely curious about Kvalsvoll's original statement and wanted to discuss it (and now does the same with maxmercy's one). I've been reading this thread for quite some months now, but I admittedly never discussed on it before, so I apologise if this reads out in the wrong way somehow. This is not my purpose at all.

As far as focusing on one thing, well that's pretty much what this forum is for. I doubt anybody would deny that we focus on bass maybe too much, but that's the fun stuff so why not?

Which I totally understand.   I mean : this is the "Low frequency content" thread after all ! But it nevertheless shouldn't prevent people to avoid hyperboles when possible. I'm not defending cheap mix and disappointing AQ, just challenging Kvalsvoll's statement.

The reason for the semi-Gaussian distribution is that only 3 people, who receive zero compensation, have measured the films listed.  The rating system was developed and is flawed in that most films will rate 3+.  There would be far more <3 Star rated films if people would take the time to measure every release.  I tend to only measure the releases I think will have significant LF effects, hence the number of 4+ Star films listed. If anyone wants to, they can measure more dialogue-heavy films to flesh out the list, or come up with a new ratings system.

I'm now even more baffled by this answer than Kvalsvoll's original statement.

It's all a question of the sample you're looking to analyse. Do you care about bass content in catalog movies or the latest Pedro Almodovar movie ? 

The movies listed here seems to often be blockbusters or action movies (ie movies viewers expect to have powerful soundtracks). If you're looking to discriminate adequate powerful AQ from perfect powerful AQ, it seems adequate to test these rather than, say, Cafe Society. You score soundtracks with powerful bass where you can find them. You have a relevant sample and are trying to discriminate within it. Adding non-relevant movies will only artificially flesh out the list (like Only Lovers Left Alive).

If the powerfully sonic movies measured mostly get 4+ scores while they shouldn't, then, either movies with significant LF effects are actually objectively good (and they then shouldn't draw negative generalization), or the rating system is indeed so flawed that it fails to properly discriminate movies even within the scope of what it wants to achieve (ie guide people towards Edge of Tomorrow but not Star Trek Beyond, in opposition to Children of Men but not Cafe Society).

I don't know what Star Trek Beyond final score will be, but I suppose it shouldn't be 4+. As of now, it should be 3, which actually is pretty meh for such a movie. One would have hoped for it to be much better, so I guess that's why it's being measured here. Is Star Trek Beyond supposed to be worse than 3 ? If it does, then, one should either ponderate the weaker point more heavily, or set up new thereshold for the highest ratings. From my readings here, it seems that the extension is often a focus point, so I'd suppose ponderating it harder in the average might be a solution.

I don't know what Star Trek Beyond final score will be, but I suppose it shouldn't be 4+. As of now, it should be 3, which is pretty meh.

EDIT : I've downloaded the Google spreadsheet. I'll have a look tomorrow to shift the Gaussian distribution towards 2.5 instead of the current 3.5. The aim of this list, I suppose, is to select the best of the bests. Even within 3+ or 4+ movies, some movies are better than others, so this looks like a distribution methodology issue more than anything.

When I sit down in the media room I end up listening to music instead, other entertainment can be anything streamed from the net, just as good to watch something on youtube when this is the current level for sound quality in movies.

 

A combination of ignorance and incompetence has destroyed what could have been a great business opportunity for the whole supply chain - from content providers to equipment manufacturers - and a great experience for movie enthusiasts.

Just my 2 cents because it feels like such an hyperbole.

I understand people being disappointed by what looks to be underwhelming bass, but should we say stuff on Youtube is equivalent to that ?

Bass isn't the only component of a soundtrack, I'm quite certain people here know that better than me. I'm unsure if it's still the case, but Youtube was not long ago maxing out at 192 kbps AAC. I don't think this will ever come close to most of the BD lossless soundtracks (and actually many unfiltered lossy soundtracks).

There are reviewers praising filtered lossless tracks over unfiltered lossy ones (catalog movies) or people throwing the baby out with the washwater because they only focus on 1 component (here : bass). I don't think it's helping the whole cause.

Even when looking at the first post here, notes simply seems to follow a classical Gauss curve. More than half of the BDs listed have a 4+ score. 16% are scoring 3 or below, 38% 3.5 or below.

Yes, it's quite disappointing to see blockbusters that should get reference AQ to be rolled off at 30 Hz, but it certainly hasn't destroyed "destroyed what could have been a great business opportunity for the whole supply chain and a great experience for movie enthusiasts."

Because, again, at least half of the BDs listed here should content people, but also because even within movie enthusiasts, such a deep focus on bass only concerns a small part of the enthusiasts (and they would also need to be equipped adequately to detect this type of things).

Don't get me wrong, I enjoy a lot the insights people are are displaying and it helps me a lot understanding what I'm hearing when watching the BDs listed here. But it's only part of the soundtrack, which itself is only part of the experience.