Jump to content

Avengers: Age of Ultron Discussion and Poll - CLOSED


maxmercy

Avengers: Age of Ultron   

35 members have voted

  1. 1. Execution?

  2. 2. Recommendation?



Recommended Posts

Avengers: Age of Ultron


 


Level - 2 Stars (105.93dB composite)


Extension - 3 Stars (18Hz)


Dynamics - 5 Stars (27.68dB)


Execution - 3 Stars


Overall - 3.25 Stars


 


Recommendation - Buy


 


Comments - Still carries enough dynamics to just require a volume boost to get proper levels back.  BEQ may help this one, as there is no clipping I could  find on cursory examination.  Good sequel, hard to top the first film in terms of watch-ability.  Horribly filtered ULF, just like the first, but to a lesser extent, but with no clipping.


 


post-20-0-03985300-1443838573.jpg


 


JSS


Link to comment
Share on other sites

  • Replies 67
  • Created
  • Last Reply

Yeah, everyone on AVS is bitching about this. I just finished it and I think it sounded pretty good. Besides having to turn the volume up it sounded pretty good. It's not the heavy hitter that it could have been, but as you pointed out, neither was the first one. I liked it and actually enjoyed it more the second time around. I'm still a little bummed at the audio, but it sounded much better then when I saw it in the theater. I'd say it's worth a buy. There are a few demo worthy scenes, IMVHO.

Link to comment
Share on other sites

Thanks for the graph!

 

How much quiter is this movie than average?

 

If I usually listen at -15 how much more would I have to turn this up?

 

Edit - just looked at the levels for mad max - 113.9 - so I would have to turn up this move +8db more than mad max to get the same volume?

Link to comment
Share on other sites

Edited the first post.  No clipping on this track.  Outside of being quieter and filtered, at least it didn't get clipped, which would make it un-salvageable.

 

JSS

 

Is this an indication that the loudness war on movies is on a decline..

 

Time will show, but is this show up in more new releases, it is a good thing, and will lead to better sound quality and hopefully make Bass-EQ obsolete.

Link to comment
Share on other sites

How is the bass execution? I keep hearing complaints that bass does not hit in spots where you would expect it which a volume bump wont help.

Most places it's spot on, however, there are places where it seems there should be bass, and there is none or not enough.

Link to comment
Share on other sites

I experienced the same thing in the theater.  Some scenes had titanic visuals without enough LF support given the action onscreen.  No different than Avengers 1, when the underground SHIELD HQ implodes or when the giant armored flying slug-things w/ grilles slam into the ground/buildings and.....meh.

 

JSS

Link to comment
Share on other sites

 

 

This explains everything. 

Monkey_doing_sound.gif

 

I would say director's intent would be more probable considering it was similar in theatres (and it (and the first Avengers movie) was mixed by Christopher Boyes and Lora Hirschberg and edited by Frank Eulner, who all have great experience and mixes under their belt).

 

Gave it three stars since it was good enough after raising about 10db.

Link to comment
Share on other sites

Yes, I watched it at reference and it was not loud at all, especially just after watching MMFR.  I will boost the MV today and then compare.  I mean you do feel the bass all the time at regular levels but only one scene hit very hard and moved your pants.  With the MV up 8 dB I bet this will have lots of bass.

Link to comment
Share on other sites

Not as bad as I thought.  I've seen/heard worse than this for sure.  These are loopbacks.

 

 

If those are calibrated so that 0dBFS == 100% then this shows that peaks actually hit 0dB occasionally, and if the average volume level is perceived as 8dB lower than other typical releases of today, this indicates an improvement in dynamics of 8dB.

 

Hope this is the new trend, as it would will lead to some quite awesome sound tracks in future releases.

Link to comment
Share on other sites

The graphs are not calibrated to 0dBFS.  I always adjust speclab so that the highest peaks go to +2 to +5dB on the scale, that way you can see the most resolution.  The overall level of the movie is indeed less than most, just look at the peak hold at the top of the 1st page.  I don't see the reason for that.  Dynamics aside, the lower in level you go from 0dB, the more distortion increases and less dynamic range you have so I don't know why they wouldn't mix so that the peaks are hitting just under 0dBFS.  Don't really see why a director would intend for that or why a sound engineer would prefer that.  It's almost like it is a mistake. 

Link to comment
Share on other sites

Sorry, I don't see anything anywhere on this thread that shows how close the mix gets 0 dBFS.  I'm curious as to whether or not it does get to 0 dBFS, but I won't hold anything against it if it doesn't.  As for why a mix may not exceed 0 dBFS, the answer is that it may not need to in so far as director intent is concerned.  Don't forget that there is a standard calibrated reference level for playback.  Some tracks of intentionally quieter movies that don't hit 0 dBFS may sound way too loud if they are peak normalized.  Peak normalization is totally unnecessary, and the mentality that sound quality may be degraded by not using peak-normalization may be a big factor in perpetuating the loudness war.

 

I will need to watch this movie to judge its level for myself, but I suspect this is another case of us needing to recalibrate our ears for what "reference level playback" *should* sound like versus what most Blu-ray's sounds like at "0" on an 85 dBC / 20 dBFS pink-noise calibrated system.  I believe it's very likely that the studios have been mixing most Blu-ray content way too hot and have been degrading the sound quality in the process.  From what I've seen, many Blu-Ray titles are likely mixed as much as 12 dB louder.  Mixes with less loudness should be welcomed!

Link to comment
Share on other sites

Dammit SME, I never said anything about exceeding 0dBFS or peak normalization and I sure as hell didn't imply that loudness wars are good, you argumentative mental case. 

 

This is an example of what I'm talking about.  Look at these 2 stereo files.  The one on top is rendered so that it's HIGHEST peak hits just under 0dBFS.  The one on the bottom is mastered so that it's highest peak is 9dB under 0dBFS.  This takes 9dB of dynamic range away and the audio is closer to the noise floor.  This has nothing at all to do with putting a limiter on the track to make things louder and decrease dynamics. 

d91e2b17f42be33c0ac49df315d34783.png

 

I've done tests with exporting audio files 10dB under 0dBFS, level matching them and then blind A/B testing them with a full strength rendering and there is a difference.  Exporting digital audio to take advantage of all headroom available is no different than properly dialing in gain structure. 

Link to comment
Share on other sites

The importance of scene caps, calibrated levels notwithstanding, is shown in the single scene that causes the huge 30 Hz peak in the peak hold graph:

 

4a32Rpz.jpg

 

No matter what the offset is set to the colors will reflect the peak at 28-35 Hz relative to the rest of the scene. With the peak hold graph overlaid, it's apparent that this single effect forms the final shape of the peak hold graph of the entire movie.

 

As Shred mentioned above, we set the offset option to cover the spectrum of the color scale to show maximum resolution in the SL graphs. The calibrated levels are covered in the LEVEL metric of the movie rating. With a soundtrack like this one, with lower levels, if the spectrograph is calibrated to 0dB, the result is the same as using a completely different color scale which defeats the purpose of a higher rez (more color bands in the spectrum) color scale.

 

I also wanted to mention that this is a memorable occasion for Max, Nube and myself. The original Avengers movie was steeply filtered at 30 Hz. I mentioned in the AVS MWB thread at the time that I would refrain from even renting the flick because it was my only viable response to such a negligent audio production in a 1/4 billion dollar product.

 

We were severely pissed on in that thread for not agreeing with the herd. And, being that we were the top contributors to the actual posted data in the forum thread, we decided that wasting time counter posting against flat out derision from the peanut gallery for posting the facts was a bit much. It was then we decided to move our stuff here. Thanks to Josh & Kyle for their support and great space. I hope it's worked out to their advantage as well.

 

So, the original Avengers soundtrack started it all. Great to see the huge improvement in Avengers Deux, moving the filter down an octave. Even though Gwyneth and Natalie were only mentioned in the script and missing from the flick, and I def missed Tom Hiddleston as Loki who is also MIA in this one, the movie is a classic, 4 stars and a strong buy, IMO.

 

Maybe the next one will arrive in full bandwidth glory.

Link to comment
Share on other sites

This is an example of what I'm talking about.  Look at these 2 stereo files.  The one on top is rendered so that it's HIGHEST peak hits just under 0dBFS.  The one on the bottom is mastered so that it's highest peak is 9dB under 0dBFS.  This takes 9dB of dynamic range away and the audio is closer to the noise floor.  This has nothing at all to do with putting a limiter on the track to make things louder and decrease dynamics.

I understand what you are doing here and why, and my comments were directed toward your other statement about the notion that having a soundtrack peak below 0 dBFS is somehow a mistake.  I do think you should label each spec accordingly because otherwise it's misleading.  It also makes it impossible to compare two different specs with one another.  Ideally, your label might say (e.g.) "the level was increased by 9 dB to bring peaks to dBFS so as to maximize use of the color map", even though comparisons would still be difficult.

 

I've done tests with exporting audio files 10dB under 0dBFS, level matching them and then blind A/B testing them with a full strength rendering and there is a difference.  Exporting digital audio to take advantage of all headroom available is no different than properly dialing in gain structure. 

 

What kind of difference did you notice and with what content?  How many bits per sample were used in the original and in the export?  When exporting at 10 dB under dBFS, was any dithering used?  If so, what kind?

 

There are a lot of variables here, but if I had to guess, the difference you noticed had to do with raising the noise floor of your upstream analog electronics, assuming the gain control you used for level matching operates downstream from some analog signal source.  If so, this test of your reveals nothing about the dynamic range of the digital content and everything about the dynamic range limits of your analog signal path.  Unless this is especially bad though, I doubt it has much impact on the listening experience for all but the quietest content.

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.


×
×
  • Create New...