Jump to content

Analyzing Waveforms of Heavy Hitting Movies


Recommended Posts

  • Replies 84
  • Created
  • Last Reply

No problem.  DVD and BD tracks are very similar.  You can see the ULF pulse clearly.  On the BEQ version, it is more insane.  Great effect.  terrific 'crack' at first, followed by the ULF rumble of one full cycle.  You can see the cones move quite far if your rolloff allows it, and it is a new experience compared to my old setup.  



Link to comment
Share on other sites

  • 1 month later...


Where is the activation for waveform on spectrum lab.



It's called amplitude bar on the 1st page of SL's settings or you can follow this post's directions and download the zip file at the bottom of the post and it will bring it up for you when you load the settings file: http://data-bass.ipbhost.com/index.php?/topic/435-bossobass-mini-gtg-thread/?p=8831

Link to comment
Share on other sites


Prometheus (2012)

Ship crashing into ALIEN ship LFE.1 dtsHDMA 7.1 English, monitored on Lucasfilm Ltd THX  sound system crossover/monitor 3417. SDDS Fader at  -2.9db. 


Where is the activation for waveform on spectrum lab.





That ^^ is such a frustrating scene to watch - you're expecting sofa-shaking depth and instead get some poor effort that sounds weak!

Link to comment
Share on other sites

  • 2 months later...

I've been thinking about amplifier burst testing lately.  The capacitor reservoir in amplifiers is different in every design and it makes a big impact on how the amp performs in sub duties for huge transients where it can give the amp more power than it can draw from the wall power outlet. 


There are no standards to what length amps are tested for burst power duration.  Some claim that they should be tested at up to 500 milliseconds for burst power while others use 1ms and sometimes less to show that an amp has a higher peak power despite having a small cap bank. 


With a timescale (horizontal) of 50ms per division, here are the different lengths of CEA bursts at a few different frequencies:


You can see how much more demanding a 5Hz burst would be because it is a longer duration which will drain a cap bank more than a shorter burst. 


Looking closely at the 5Hz burst, we can guess where the capacitor reservoir will start to discharge.  If we figure that we have a 120V/30A (3600W) power outlet and an amp that is about 60% efficient so that it can do a sustained tone at around 2205W, a 5 ohm speaker load at 5Hz would make for around 105V before running out of juice from the wall.  Everything above 105V would have to be covered by the cap bank.


This graph has a vertical scale of 50V per division and a horizontal scale of 50 milliseconds per division. 


Since we now know where to look for the transition from wall outlet power to cap reservoir (a bit over 2 divisions or 105V), we can zoom in on different CEA frequency bursts and see how much time the cap bank will have to pick up.




Single digits get pretty stressful on an amplifier no matter what topology it is. 


The question is what burst frequency is the most logical to use to test an amplifier's capabilities.  Stay tuned and I'll show some real world movie content shots of transient waveforms and their time duration zoomed in on to see if that helps pick a burst frequency that most closely matches the stresses that some of the coolest bass hits in movies inflict.  

Link to comment
Share on other sites

Hey Shred.

Personally I'd like to see the amp manufacturers start using the CEA-2010 type waveforms for their "burst" or "peak" ratings and reporting the results at the following bandwidths.


 10Hz, 31.5Hz, 100Hz, 1kHz, 16kHz, 20kHz.


I think that would just about cover it when reported along with a long term output rating. I'd like to see that be something reasonable too like 3 seconds sustained power or something. No need for more than that in any realistic scenario. Yeah I know 5Hz is tougher than 10Hz for the burst testing but there are a ton of amplifiers that are high passed above 5Hz or will protect with that signal and manufacturers still treat 20Hz as the end game for the most part. I think 10Hz is a good compromise without testing every bandwidth.

Link to comment
Share on other sites

This is continued from the pictures in post 40  http://data-bass.ipbhost.com/index.php?/topic/425-analyzing-waveforms-of-heavy-hitting-movies/?p=10248


Here is a scene from the movie Bourne Legacy where the cabin gets hit by a drone missile strike. 

There is an impressive transient when the missile hits the cabin that is an awesome effect especially coming from the peaceful quiet of the snowfall in the woods in the previous scene.  This effect can be missed if your amp doesn't have a sufficient cap reservoir.



Here is right after that where there is a sustained "boom" of the explosion.



For this scene it looks like an amp would have to supply power for transient peaks of around 10ms. More scenes to come...

Link to comment
Share on other sites

I appreciate what you are trying to do here.  Please forgive me for pointing out a few things:

  1. The strongest frequency in the spec is at 38 Hz or so.  A full wave-length at 38 Hz is 1/38 or 0.026 seconds long, or 26 ms.  The distance between zero-crossings for the wave is half the wave length or 14 ms.  That's basically what you are seeing in the time-domain data you posted.  I expect almost all pro-amps should handle that transient just fine as they are designed to provide plenty of power down to at least 40 Hz or so.
  2. The data you posted is (I assume) is directly from the soundtrack.  What the amp actually sees is something different due to crossover filters and any downstream signal processing including Linkwitz-Transform (LT) type signal shaping.
  3. The crossover will remove a lot of higher frequencies (that appear as finer oscillations in the time-domain data).
  4. The signal-shaping used in a ULF-capable, multiple-sealed HT sub system will drastically increase the emphasis of the lowest frequencies (the coarser oscillations in the time-domain data).
  5. The consequence of these two things is that time-domain data for the signal presented to the amp will look totally different, and for "full-bandwidth" effects, the oscillations that dominate will be very low in frequency and will have half-wavelengths much longer than 10 ms.  The half wave-length for 10 Hz is 50 ms.  At 5 Hz, it's 100 ms.
  6. If your amp depends on capacitance to sustain peak output with longer wavelengths, you'll need closer to 100 ms worth of storage for output at 5 Hz.  (One of your earlier pictures shows 32 ms at 5 Hz where I think you meant 64 ms.)

If you are able to capture the output from the signal shaping device in your signal chain, I think that this data would be very interesting for us to see what your amp actually sees.  Of course, the signal seen by each person's amp will depend on what kind of signal processing is used in the chain, but I'm guessing that for "full-bandwidth" effects, ULF will dominate in the signal going to most amps that are being used for ULF reproduction.  Hopefully my explanation makes sense.

Link to comment
Share on other sites



Pretty close.  I remember that the 5.1 was hotter overall, but not on this effect, apparently.




I used to use AVR processing/bass mgmt to study the tracks, but there are so many ways that AVRs handle the digital data with various xover and LFE cutoff points and slopes that I only look at the digital data now, no filters, only summation with proper channel levels.  While I know that isn't what a subwoofer amplifier sees, it is the most apples-apples way I could think of doing it.




Link to comment
Share on other sites

SME makes some of the points I was thinking to myself. It would be very useful to see a direct from disc waveform compared to the signal being sent to the amplifier after equalization, etc...He is right it will be much heavier emphasis on the sub 30Hz frequencies. As he states obviously that will be a bit different for each persons system but a comparison of just a couple of scenes from one or two of the heavy hitter movies would be great information.

Link to comment
Share on other sites


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

  • Create New...