Jump to content

Analyzing Waveforms of Heavy Hitting Movies


Shredhead

Recommended Posts

I picked a volume level on a bluray disc player with analog outs and recorded the waveform of the Voltage on the subwoofer output while playing various movie scenes.  This voltage includes the re-directed bass from all 7 channels.  The vertical scale on waveform grid is 1 Volt per division for all of the following movie scenes.  How many Volts this particular player outputs for the content isn't important, it just shows the difference in output for different movie mixes.  The readout next to the waveform below tells what the peak positive voltage level is and then the peak negative voltage.  When you add the 2 together, you get the peak to peak value (Vpp).  I tried to pick the hottest scenes of the following movies. 

 

There are some lower level movies included here to give a frame of reference but mostly there are sub crushing monster soundtracks.  Seeing the difference in the voltage levels of the low level films vs. the heavy weights, it's no wonder why when you pop in certain soundtracks like HTTYD you have to be very careful where your sub trim and MVL is set to so that you don't overdrive your amps and drivers. 

 

Let's start with a lower level film mix... 

 

Avatar.  DTS-HD MA 5.1

-This is the scene where the tree falls  -12.04dBV from WCS

7fd0da0264b2bff996d974b7a6a14c58.png

c590cf29d751a301b90be2ab3327f941.png

 

Edge of Tomorrow.  DTS-HD MA 7.1

-The square wave intro  -7.04dBV from a WCS

22d52ec8cd42b98c3053219cabb13d58.png

feb915cc0d5dd5523f4d2931e7f7589a.png

 

 

Earth to Echo.  DTS-HD MA 5.1

-The ship at the end of the film  -2.03dBV from a WCS

7440a3de9bff7081fcd7ad9b0caedc0d.png

50a1c83206f103bc8b380fc809c9f271.png

 

Godzilla.  DTS-HD MA 7.1

-1 hour, 40 minutes in where Gozirra fights Mothra (or whatever that thing is supposed to be)  -5.78dBV from a WCS

dfcc5219171a8ed462508bb3c7e7ff5e.png

f53498796d678689ad911f0973343001.png

 

Guardians of the Galaxy.  DTS-HD MA 7.1

-Intro of movie where Chris Pratt says he is Starlord  -6.52dBV from a WCS

e7850ed32e17e4fbc3cfc0fb2190e971.png

673993238089aa4f5d47c65313a8dbb3.png

Link to comment
Share on other sites

  • Replies 84
  • Created
  • Last Reply

War of the Worlds.  DTS-HD MA 5.1

-Bridge collapse

I found it interesting how much weaker the DVD's dolby digital mix was than the DTS mix.  And I was really surprised to see that the DVD's DTS mix is actually hotter than the Bluray's master audio DTS track. 

-11.13dBV from a WCS (DVD DD 5.1)

-4.08dBV from a WCS (DVD DTS 5.1)

-6.27dBV from a WCS (Bluray DTS-HD MA 5.1)

67a7d24294a7bc340f26f40af9bff0de.png

ea38317ece7fad0639f9478586918e0e.png-5.1 DTS DVD

 

 

Here is a demo disc that maxmercy made for me that has almost 0dBFS of a 40Hz sine wave on all 7 channels re-directed and summed with the LFE.  This is an absolute worst case scenario and it isn't possible to have a hotter level out of the sub out than this.  It should be clarified that max encoded the levels down from 0dBFS by -0.1dB on all channels.  Usually digital audio discs are mastered down from 0dBFS anywhere from -0.1dB to -0.5dB because in cheaper DAC's a full 0dBFS signal will come out clipped. 

-0dBV from a WCS.

648209d68c7f2111c23763a4cc1f4c65.png

c29127c726bf895ca8dad0579b434979.png

 

 

I will add more waveforms for hot content as time permits. 

Link to comment
Share on other sites

How to Train Your Dragon.  Dolby True HD 5.1

-Dragon Crash  -1.44dBV from a WCS

57f19ed37eebfd7d4e385e3aed8cf94b.png

2773e997b6a6c10e89c0b81d0d0ecbce.png

 

Interstellar.  DTS-HD MA 5.1

-Right before the ship's automated warning is telling him to eject  -3.17dBV from a WCS

bc0e475c0eae5406ef1a59d9674c6621.png

65720cb6ca164e85ee0944f87b429180.png

 

Oblivion.  DTS-HD MA 7.1

-"Coming in hot"  -7.9dBV from a WCS

508c62f738e6eb066ebba00a2664d7da.png

9998ccfd9d86197d4d6cc626556721a7.png

 

Pacific Rim.  DTS-HD MA 7.1

-The opening battle  -4.68dBV from a WCS

180151b3bbaeb3bb15dc68027111c1f6.png

98b45d25d41e615b9b954c63c6ea09e4.png

 

Thor.  DTS-HD MA 7.1

-"Then GO!" in the frost giants battle scene  -4.48dBV from a WCS

16c239b902c3e2e06cb11713918e87ba.png

8199315ee9f6a47373d3d8186fe7d47a.png

 

 

Just a quick note:  The HTTYD dragon crash is only 1.44dBV from a WCS (worst case scenario). 

 

Earth to Echo is 2.03dBV from a WCS.

 

Interstellar is 3.17dBV from a WCS. 

 

While Avatar is down 12.04dBV from a WCS. 

 

I wouldn't be surprised as the loudness wars progress if a movie comes out that maxes out the headroom completely for a brief period.  It would be a powerful/potentially dangerous effect. 

 

Another thing interesting to me is that while some newer movies like EoT and Interstellar are using either clipped channels or aggressive limiting to create square waves it must be for the effect that square waves add to the mix because they don't seem to be maxing out the total headroom in all re-directed channels. 

 

I also don't understand why anyone would choose to hold the bass levels down such an extreme degree as 15-20dB from a full reference output.  From what I understand, the farther from 0dBFS you get, the more distortion increases in digital audio. 

Link to comment
Share on other sites

Yeah, paste the -x.xxdBV from WCS on the waveform graph as well if you can.

 

This info is golden and should be attached to the MWB list by adding the graph to the linked post with the peak hold/average graph.

 

If you can come up with an easy way to factor in content, then this would be a perfect objective way to say which scenes are most intense from a playback stand point with little room left for debate. ;)

Link to comment
Share on other sites

Nicely done. I'm a sucka for graphs and graphics. Good examples of loud vs dynamic. Never understood all the fuss over that Interstellar scene. And I don't get why the DVD is superior to the bluray soundtrack in with so many movies. Master and Commander is another example. Ice Fields from Titan A.E. is a challenge. 

Link to comment
Share on other sites

Just a quick note:  The HTTYD dragon crash is only 1.44dBV from a WCS (worst case scenario). 

 

I also don't understand why anyone would choose to hold the bass levels down such an extreme degree as 15-20dB from a full reference output.  From what I understand, the farther from 0dBFS you get, the more distortion increases in digital audio. 

Don't your voltages from a WCS only relate to your particular Blu-ray player? For example, I can output 30V+ p-p (+24dBu is 34.7 p-p) and I'm sure the voltage difference from HTTYD and WCS would be different. Maybe it would be better to show as a percentage. People's crossover frequency, LFE low pass, and filter slopes will also factor in how they relate their output levels to your's.

 

I analyze all tracks of all movies I watch. Almost all tracks get very close to 0dBFS. 3ll3d00d should be able to confirm with his movies. However, you lose a lot of headroom when you use bass management. This is even more prevalent as when you go to a 7.1 track. As you saw with the OPPO, it is up to the manufacture how they will manage the bass and what tradeoffs they feel are important. Some desire to preserve sound quality and output levels over the majority of content (OPPO) while others will allow more headroom at the expense of loss of low level signal detail due to using fewer bits for output. 

 

Below is the first 00:03:15 of EOT showing that the waveforms reach close to 0dBFS for the center and LFE channels. The entire movie is mastered at 0 dBFS and at +1.4 dBFS when you take intersample peaks into consideration.

 

post-42-0-45660400-1432764413_thumb.png

 

Here are three methods of bass management on EOT. The first is using peak level normalization which will make sure the volume is still as loud as possible. The third is how a typical Blu-ray player or receiver will have to do bass management for a WCS. 

 

post-42-0-75098900-1432764781_thumb.png

Link to comment
Share on other sites

Don't your voltages from a WCS only relate to your particular Blu-ray player?

Yeah, like I said it's not important cause it's all relative to dBV and if my player isn't doing management like other pieces of equipment I plan on showing that too with this experiment. 

 

My computer's PSU shit itself so when I get it sorted I'll post some more stuff. 

Link to comment
Share on other sites

 

Here are three methods of bass management on EOT. The first is using peak level normalization which will make sure the volume is still as loud as possible. The third is how a typical Blu-ray player or receiver will have to do bass management for a WCS. 

 

attachicon.gifEdge of Tomorrow Bass Management.PNG

 

Please elaborate on the different bass management schemes shown (and implemented in gear).  Those waveforms are quite different in amplitude.

 

JSS

Link to comment
Share on other sites

I analyze all tracks of all movies I watch. Almost all tracks get very close to 0dBFS. 3ll3d00d should be able to confirm with his movies. 

I posted a list I pulled out of my library in http://data-bass.ipbhost.com/index.php?/topic/379-maxmercys-wcs-test-disc-beta-and-an-o-scope/?p=6359

 

It shows that films invariably max out the main channels but that it's less common to do so on the LFE channel. Obviously this is only the peak value for the track and it says nothing about whether those peaks are aligned or what the overall level throughout the film is like. 

Link to comment
Share on other sites

Please elaborate on the different bass management schemes shown (and implemented in gear).  Those waveforms are quite different in amplitude.

 

JSS

I shouldn't have called it bass management. They are just all channels combined but with different level of attenuation used. The first used peak level normalization to maximize volume without clipping. The third uses the -10.2 dB on LFE and -20.2 dB on other channels you recommended to me for use for spectrum lab analysis. The middle one is about half way in between with -4.8 dB on the LFE channel. It is similar to what is used by a receiver that did not take into consideration the extra two channel in 7.1 content and still used downmixing levels used for 5.1 channels. I was told by Roger Dressler that this happened for a while after 7.1 was introduced.

Link to comment
Share on other sites

I shouldn't have called it bass management. They are just all channels combined but with different level of attenuation used. The first used peak level normalization to maximize volume without clipping. The third uses the -10.2 dB on LFE and -20.2 dB on other channels you recommended to me for use for spectrum lab analysis. The middle one is about half way in between with -4.8 dB on the LFE channel. It is similar to what is used by a receiver that did not take into consideration the extra two channel in 7.1 content and still used downmixing levels used for 5.1 channels. I was told by Roger Dressler that this happened for a while after 7.1 was introduced.

that is actually consistent with shredhead's captures from his oscope then isn't it? i.e. your 3rd graph hits about 0.45 which in the region of  ~7dB down from the peak and this is about the same headroom as indicated by his readings (of 2.56V vs 5.72V)

 

is the first one just the output from jriver with Adaptive Volume/Peak Level Normalize turned on? i.e. the user is left to set MV to their preference (on a track by track basis) 

Link to comment
Share on other sites

A quick note to all out there who use HTPC's.... My computer recently power cycled in the middle of watching a video.  I thought the power went out but just as soon as I realized it had not I started to smell the magic smoke escaping so I reached over to my computer which was in the middle of booting up and turned the power off to it as soon as possible.  I have a 120mm fan that is switchable on separate power from my computer's PSU so I turned it on and it was pumping out hot, stinky air.  No good. 

 

It was my PSU dying probably due to dried up electrolytic caps and going into wild oscillations.  Here are some measurements I took of the old faulty one vs. a new one with and without a load on the 12V rail. 

5fa9e88ea2f0ecfcd9ada1a0253090ae.png

 

It is normal to have some switching noise on a computer's PSU but you can see that this was far too much noise for any digital electronics.  The switching noise is at ~91kHz so if you were to zoom out, you would see a brick wall of PSU noise.  PSU noise seriously messes with digital electronics in unpredictable ways as I already mentioned in other threads about the iNuke series of amps with their power cycling problems. 

 

I'm pretty sure this PSU would have destroyed itself (and probably my MB as well) from oscillating had I not caught it in time so just a heads up to the other PC users out there: watch out for power cycling and abnormal heat from PC PSU's. 

Link to comment
Share on other sites

Here is another frustrating case of the DTS Master Audio 7.1 track being the most undynamic of all the mixes.  Apparently France gets the best mix on this one. :wacko:

 

Prometheus.  DTS-HD MA 7.1 (bluray), Dolby Digital 5.1 French language track (bluray), and the Dolby Digital English track (DVD)

-Ship crash suicide

570c0c36b1e7540001768550f2ae1c86.png

e16125f193f1c0a579635508e7099f87.png-DTS 7.1

a1d4efd38931144b598d3b8d4ed49139.png-French 5.1

Link to comment
Share on other sites

I've noticed that neither you nor your dad ever demo the Washington Monument scene from Olympus Has Fallen (tops on the list at that link).  I wouldn't say it's a good movie, or even a very fun one, but that scene is definitely a contender for most difficult sustained scene to reproduce.  I added the E2E scene right below the OHF scene at that link, just for reference.

 

The length of it is killer, plus that big 1-3Hz hit that's coupled with all the intensity 10-30Hz.  I'd love to see how it does both short and long term in this measurement, as well as in your guys' playback SL captures, if you could start to incorporate it into your rotation.

Link to comment
Share on other sites

Nube here's a SL cap of that scene which I use as a reference and also which made me realize the near field Raptor was shit. My 5-10hz was better but from 10-25hz I had a huge null. Back up front where they belong. Of course I'm not reading anything from 1-3hz but everything else is there.

 

post-35-0-88543100-1433088278_thumb.png

 

 

 

 

Link to comment
Share on other sites

I'll get on the Transformers 1 for ya max.  Is it at the end where they are fighting in the city? 

 

nube, I've been hearing about that scene for a while now but we don't have the movie and I've never seen or experienced it.  I'm pretty sure Bosso watched it at some point and he might have capped some SL action but I can't remember.  I'll keep an eye out for the movie in bargain bins. 

Link to comment
Share on other sites

Yup, right after Megatron shows up in the city.  He blasts Jazz with his arm cannon and it is the strongest ULF transient in the film.  For comparison, you could do the tank (wrongly named devastator) dying or Ironhide's flip, both of which had significant content in different registers.

 

Thanks!

 

JSS

Link to comment
Share on other sites

  • 2 weeks later...

Archived

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


×
×
  • Create New...