maxmercy

The Low Frequency Content Thread (films, games, music, etc)

3,976 posts in this topic

The Equalizer's extension, 2 stars?  How does this rank lower than TMNT or GotG?  :huh:

 

b360afa733a7f9dbdb1687164a6ad1bd.png

yeah something isnt right

Share this post


Link to post
Share on other sites

Nothing is amiss.  It's just that the huge hit at 54Hz creates an inescapable burden on the green trace, and the average content isn't enough to garner a higher rating.  It's rating is appropriate, colorful scenecaps notwithstanding.  The Equalizer had some nice, loud moments, but it's not a monster by any stretch.

Share this post


Link to post
Share on other sites

I must be misunderstanding what extension means then.  The strongest hit at 54Hz discounts the content at 5Hz?  Never said it was a monster, I'm just confused why this would get the same extension rating as say Godzilla which has a blank screen at 5Hz on the waterfall. 

Share this post


Link to post
Share on other sites

Even if there's content everywhere on the graph, extension is measured at the -10dB point on either peak or average traces - whichever results in a better rating.  Been this way since the beginning.  :)

 

I often think -10dB is too generous.  Sometimes rather inconsequential effects result in a much better rating than the movie deserves - it's rare, but it does happen. The Equalizer has two very minor (short duration, small bandwidth) effect components  of note below 15Hz that are masked by concurrent higher frequency effect components at higher levels. 

 

In your scenecap of the tanker explosion, the 10.5Hz component is a whopping 14dB below the 54Hz component of the effects (albeit 15s later, though concurrently that 10.5Hz has a bunch from 35-60Hz to contend with).  Even the 17-19Hz components of that scene are ~11dB below the 54Hz stuff.  I'm not saying it's completely irrelevant, but it's close.  The second scenecap has essentially nothing below 15Hz.

 

The average trace really tells the story - the mix has little content below 19Hz, on average.  If you like the movie and the bass in the mix, it's all good. But, it's not like The Equalizer's Tanker Explosion scene is in the same league as Olympus Has Fallen's Washington Monument scene, the Hulk vs. Abomination scene, the How To Train Your Dragon Dragon Crash scene, or the War of the World's Pods Emerge scene.

Share this post


Link to post
Share on other sites

Every measurement system will have films the 'fall through the cracks'.  I think we objectively capture how much LF a film has most of the time with the system we have.  That's why the Execution category exists to help shore some of that up, but some films will simply be a great watch, but not score as highly as some would like, Star-Wise.  I know I have my own '5-Star' list....

 

JSS

Share this post


Link to post
Share on other sites

2f765106b8715b33ac129ecdce20e626.png

 

I gave this movie top honors for 2014. Not a chance I would do that if it's extension is 23 Hz. :rolleyes:

 

Yes, you can clearly see the big spike at 54 Hz (actually, it's a dual spike at around 52.X & 53.X Hz), but when the floor ripples, there's plenty of <20 Hz content.

 

The low end effects in this movie rival any movie in my collection for originality, variety and extension. It just isn't a constant drone of low end like some enjoy more, but there is no category for minutes of LFE per minutes of length of movie. Plus, the scene in ch 13, worth the price of admission in itself, is nearly a minute of mayhem.

 

For comparison's sake, the FOTP scene that makes it a 5 star movie lasts 1/2 as long. It has some content below everyone's radar and a spike at 32 Hz, but not much else in the scene or in the rest of the movie.

 

I understand that metrics and anomalies co-exist. I'm just sayin' don't let the ratings fool you on this one. It has the goods and it's a tad better movie to watch than the FOTPs of the cinema world. ;)

1 person likes this

Share this post


Link to post
Share on other sites

Exactly. I wonder often if the way some categories are rated needs some rethinking.

 

A quick glance at the PvA for Equalizer shows it is very well extending in bass but gets a '2' in extension. Others like Lone Survivor get a '5' yet show that it rolls off ~30hz but with a couple of handful of high intensity effects skew the whole rating.

 

Anymore I have to ignore the number rating for these movies and have to go off the PvA (as I always have).

Share this post


Link to post
Share on other sites

Are you guys forgetting that the red trace, or the average, is showing precisely how much content is there?  This one rolls off steadily, albeit not as steeply as some, from 39Hz on down.  On average, there's just not much content in The Equalizer below 20Hz.  That's not interpretation, that's an analytical measure of the retail BD's content.  What's interpretation is how much we like it, which is what the execution score's for, as maxmercy said.  :)

 

Check this out.  These were all measured with exactly the same settings.  Look at the red trace, and compare it for yourself:

 

Comparison.gif

Share this post


Link to post
Share on other sites

Fury (5.1 DTS-HD MA)

 

Level        - 4 Stars (111dB composite)
Extension - 1 Star (31Hz)
Dynamics - 5 Stars (28.14dB)

Execution - 4 Stars (by poll)

 

Overall     - 3.5 Stars

Recommendation - Buy (by poll)

 

Notes:  If it weren't for that one blip at 38Hz, this would be a 1Hz movie, because of a blip at 1.5Hz.  Otherwise, it's rolled off pretty consistently from 34Hz.  Good weight to some ofthe 50cal and front tank guns, but not what you'd expect for so many explosions and such heavy equipment.

 

I liked the movie, though.

 

PvA:

 

post-17-0-98526600-1422403005.png

Share this post


Link to post
Share on other sites

Personally, I think there should be some smoothing to the average graph to stop narrow peaks scewing the results.

Interesting idea, but I think that is kind of what the average graph does no?

Share this post


Link to post
Share on other sites

The average is taken over the entire length of the movie, so soundtracks with more content will show a higher average even if the content is identical.

Share this post


Link to post
Share on other sites

The average of the content of the entire length is still subject to peaks.  With The Equalizer, the amount of content @ 29hz is causing a peak in the average graph.  Since the extension score is derived from an -10dB point, these peaks should be somewhat smoothed.

 

If I was to judge the extension of The Equalizer based on the unsmoothed average response (with mental smoothing), I would place the extension @ 17hz.  The frequency span of the peaks in the 50-70hz range, and the overall level of the content in that frequncy range still keeps the extension rating fairly low, so it's certainly not an ULF monster.  But an extension to 17hz is probably more subjectively accurate.

 

Fury is another example.  It's quite clear to see that the average level is -45dB, and the -10dB point is more like 19hz.  The peak @ 39hz is exactly that, a peak above the average.  Deriving the -10dB from this peak is not subjectively accurate.

 

There's a reason why all things audio are measured as +/- some value.  Not - some value from an peak.

 

Anyway, I'm just being an grumpy old bastard.  Thanks for the work you guys do.

Share this post


Link to post
Share on other sites

The average of the content of the entire length is still subject to peaks. With The Equalizer, the amount of content @ 29hz is causing a peak in the average graph. Since the extension score is derived from an -10dB point, these peaks should be somewhat smoothed.

 

If I was to judge the extension of The Equalizer based on the unsmoothed average response (with mental smoothing), I would place the extension @ 17hz. The frequency span of the peaks in the 50-70hz range, and the overall level of the content in that frequncy range still keeps the extension rating fairly low, so it's certainly not an ULF monster. But an extension to 17hz is probably more subjectively accurate.

 

Fury is another example. It's quite clear to see that the average level is -45dB, and the -10dB point is more like 19hz. The peak @ 39hz is exactly that, a peak above the average. Deriving the -10dB from this peak is not subjectively accurate.

 

There's a reason why all things audio are measured as +/- some value. Not - some value from an peak.

 

Anyway, I'm just being an grumpy old bastard. Thanks for the work you guys do.

Oh - I like the idea of the -10 from average!

Share this post


Link to post
Share on other sites

Man, just watched wwz again last night and remembered that grenade scene. Just absolutely awesome. I actually paid way more attention to the whole mix second time around and there are some killer surround effects in that one too. There was one point where someone coughed in the rear channels and I got up because I thought some one was sitting behind me.

Share this post


Link to post
Share on other sites

Fury looks pretty good. Ignoring some extra heft in the 30-40hz range it actually is well extended to 20hz at which point it rolls off.

 

1 star? Ugh.

 

 

 

 

So...movies graphed here for bass....

 

Flatness to the single digits? Good!!!

 

Movies with a steady or pronounced rolloff from 30hz and then have a huge peak at 1-2hz? Not very useful or exciting in anyway.

 

 

We really need to rethink the way this stuff is classified. :(

Share this post


Link to post
Share on other sites

To everyone who is criticizing the measurement system I took the time to re-explain in the first post:  The following is meant with zero disrespect, but with a fair amount of frustration:

 

If anyone comes up with a better scheme (I NEVER said mine was best), feel free to start your own thread, and update it with all the data that already exists but with with your new criteria, see how the stars shake out, create a Google Spreadsheet with same, and then measure the pertinent new releases going forward for me.  I will support such an endeavour fully.  I will also then see how you field questions that imply your measurement system is flawed after the undertaking...and I will not be doing the asking, as I will have understood what a journey it is.

 

I left AVS because it had turned into a mob of 20-30Hz extension capable people extolling the tremendous bass virtues of every new film, no matter how clipped, or how rolled off, the louder the better.  I got tired of the endless BS saying 'Avengers was awesome' despite the obvious filter, etc, etc, etc.

 

I came up with a scheme that nearly guarantees if a film is worthy, it will get lots of stars, with very few exceptions.  EVERY scheme that is decently simple to implement will have flaws.  I have tried several schemes, especially early on.

 

Let's say we smooth to get the 'extension' score.  How much smoothing?  1/6th octave 1/3rd octave? 1 octave?  Til it 'looks right'?  Throw out peaks higher than 10dB above the rest of the graph?  5dB above the rest of the graph?  There's going to be a film with a peak 4dB above what is deemed 'the average' that will throw people into hissy fits, especially if it is a good movie that people WANT to score higher.  Throw the peaks out before or after smoothing?  No matter what you decide, I can generate a trace to screw up the result prompting another "the system needs to be changed!" cry.  You of course can make the system so complex that no one will want to spend the time to actually utilize it, or understand it.

 

I have never said the system was perfect, and I know films will (and do) fall through cracks.  But we haven't seen a bass monster in a while, and I sense the natives growing restless, wanting a film to have one or two more stars in extension (or not)....when in fact, nothing recent has been in the class of TIH, B:LA, and WotW.  

 

I for one, will not update the immense amount of data because ONE OR TWO FILMS DIDN'T MEET SOMEONE'S EXPECTATIONS IN THE STAR RATINGS.  GET  OVER   IT.  Or, put your money where your mouth is, and start your own thread.  I promise I will be very supportive, and help when I can, because I am well aware of the road you are about to travel.

 

Like I have said many times before, my personal favorites do not necessarily score out at 5-Stars, and I have my own personal 5-Star list.  And I have no plans to change the ranking system to make my fav films 5-Star films...and I've gotten over it.

 

 

JSS

5 people like this

Share this post


Link to post
Share on other sites

I hear ya, Max... loud & clear.

 

To be clear as crystal, my recent comments regarding EQ were sparked by Nube's subjective comments that the low end was mid-bass with droning 29 Hz content, which is bullshit, plain and simple, and needed to be clarified, IMO.

 

The problem really is that no one caps scenes using the waterfall/waveform features of SL. Those will ALWAYS separate the facts from the subjective comments and certainly eliminate the need for alterations of the peak hold/average graph, IMO.

Share this post


Link to post
Share on other sites

Hey, John. Sorry if I came off inflammatory about the ratings. I don't mean to be. I just wanted to make it clear that there is a discrepancy between a 5star extension and 1star just by a couple of blips.

 

You know that I know how hard you worked on this system and I don't fault you. Is it perfect? No. Though I don't have that kind of expectation either.

 

I guess just ignore my earlier comments. ;)

Share this post


Link to post
Share on other sites

Max, you are doing great! Nothing is perfect but we do te best we can and what you are doing is amazing and to me just like the spec graphs Bosso has contributed. All I do is build silly systems based on your guys data and know how and compare and give opinions. I mean whether it is rated 1 or 5 stars all one has to do is compare the PVA's and see where a movie will be louder or not. When people tell me a movie has better bass and the PVA's don't back up their opinions I just say it is based on their response and not what is on tw disc. That is OK as we all have opinions on what sounds great. EoT does not sound as loud as others but it does play loud and low and I bet many ported systems won't reproduce it correctly unless tuned in the single digits. I can see where the guys don't like peaks as it is like calibrating your your levels with peaks rather than a flat response.

Share this post


Link to post
Share on other sites

I think that most of the time when the system errs on extension, it tends to give a higher rating that is deserved.  "The Equalizer" (TE) seems to be a rare case of a movie that got a worse extension rating than it probably deserves.

 

I have given a lot of thought to what a "better" system would look like, and I have to say that the problem of objective measurement does not admit a single optimal solution.  A lot depends on preferences too.  For example, one possible way to improve the extension rating is to take into account equal loudness contours.  Instead of choosing the -10 dB point, we might instead look for the lowest frequency that is represented in the soundtrack above the threshold of audibility.  Ignore for the moment the issues of psycho-acoustic masking of ULF content by higher frequency content and the fact that distortion introduced by vibrating room surfaces may make ULF audible that wouldn't otherwise be.  Bossobass reported excellent bass for TE when listening (presumably at theatrical reference) with his subs at +6 dB.  Without a doubt, he's hearing more ULF (potentially a lot more) than he would have if he ran his subs calibrated equal to his mains.  Does this mean that he overrates the bass in TE?  I don't think there's a straightforward answer there at all because we all listen differently!  Even in the studio, as we've found out, tracks are rarely monitored with enough low-end capability to hear everything that's present, and we have reason to suspect that much monitoring happens at lower-than-reference playback levels in order to compensate for the reduced playback levels used in common theaters.

 

Now, one could point out that if Bossobass plays "Pacific Rim" (PR) with subs at +6 dB, it's unlikely he will hear nearly as much ULF as he did in "The Equalizer".  PR has a steep filter compared to the modest roll-off seen in TE, so arguably TE should have a much better extension rating.  On the other hand, PR with BEQ gets rave reviews, so given that PR has good sounding ULF content underneath a steep filter, perhaps it should get a good extension rating too?  As a thought experiment, what happens if you cancel the 20 Hz filter and boost the sub by 30 dB on HTTYD2?  It might sound awesome!  Clearly (looking at the PvA) there's a nice 10 Hz hit buried in there, just waiting to be EQed and boosted out.  Or maybe it'll wreck the score and make it sound bad.

 

The point here is that it's impossible to make an objective measurement system that satisfies all of our subjective qualifications.  I for one mostly ignore the star ratings anymore and I go straight for the PvA and the scene spectrograms.  Even then I can't judge until I've viewed the film because my chosen playback level depends on the overall loudness of the film.  I also don't boost my sub level because it screws up the crossover region and the upper bass (100-200 Hz).  Indeed, I have found that upper bass response has a considerable impact on how real world bass sound effects are heard and felt, even many of those with very low fundamentals.  So merely having a flat sub response with low single digit extension does not necessarily yield a better experience compared to a system that is smooth up in the upper bass octave.  Incidentally, a gradually sloping house curve that is applied to the mains and subs together will likely perform much better than a sub-only boost, provided your mains chain has the headroom for the extra upper bass.

 

What we all need to avoid doing is reading the "stars" ratings as value judgments.  Aside from the "execution" category, all the stars represent is the outcome of an objective measurement process.  Any such process will inevitably be flawed from a subjective standpoint, and the particular flaws will vary depending on the subject!  The only change I would consider at this point is to give more weight to the execution category.  The trouble is that opinions vary widely depending on the capabilities and listening choices of each voter's system.  A movie that's "a 29 Hz drone fest" for one listener is "a top pick" for another.  And there's the larger audience of 30 Hz ported subs that always rate things with high levels like TF4 and "Godzilla" as 5 stars.  Should we restrict voting to "elite" users whose systems have been qualified by an export board as "ULF-capable enough"?  That would surely exclude my opinion, even though I frequently agree with the "elites" in my execution assessments, and I can only imagine the pissing wars arising over whether flat from 3-20 Hz is more important than flat from 100-200 Hz.  (For what it's worth, I'm putting my attention on the latter for the time being.)

 

Alas, in the end, I don't think we'll ever be able to fairly rate the bass in films or come even close.  I'm just happy that there's so much great data posted here that I can use to relate to what I hear in the movies I watch as well as to do useful things like estimate how much headroom my system needs.  Big props for those who are volunteering their time to do and post these analyses!

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!


Register a new account

Sign in

Already have an account? Sign in here.


Sign In Now