ExposureRoom Home
  Log in Sign Up
Mark's Website
Mark Weiss
United States
Friends: 3
Focused on : 0
 
       
                                                             

Fireworks Display at Danbury Fair Mall, July 3, 2008

Not Rated YetNot Rated YetNot Rated YetNot Rated YetNot Rated Yet0votes
Views:787  Favorited:0 Reviews:0 Comments: 5
I recorded this fireworks display using multiple frame rates on the XDCam EX.
One was set to realtime record at 24fps, the other to overcranked 60fps. The
results were combined in post production. During the quiet parts of the intro, I mixed in the slow motion footage, so you'll se a lot of displays going off with no sound. This was edited this way on purpose.
Audio was captured using two large-diaphragm condenser mics, fed into the EX1.
No sound compression was used. The camera demonstrates superb dynamic range, which will exceed most computer sound cards. A premium low-noise sound card is highly recommended for playback of this display. Proper playback will require high power headphones and amp, or thousands of watts of stereo through speakers.
The dynamic range of the audio portion is 76dB! Proper level is set by adjusting volume until the crowd noise is at the typical level for people shouting, as heard from 50-75' away.
The finale has the well-known 1812 Overture mixed in.

Happy Independance Day!
Get Adobe Flash player
Play the Low Bandwidth Version Play the Medium Bandwidth Version Play the HD Version (High Bandwidth Required)
size: 30.55MBwxh: 640x360
Updated:  13 hours ago
Short
Duration:  00:04:23
Definition:  This Video is a High Definition (HD) Video
Dimensions:  1280 x 720
Size:  194.65MB
Share It
Digg It!  Add To StumbleUpon  Add to Del.icio.us  Post To MySpace  Share on Facebook
Embed Tags

See Recent Videos By Mark

Whitewood Rooster Test Video (h.264) Rooster Test Video

Comments



Shiv Kumar    July 05, 2008 02:56 AM
Hi Mark,

I think you've capture these fireworks really well. I'm not sure if we were missing anything, but what I see here was spectacular.

Thanks also for all the information you've provided. Makes the viewing a better experience.

I was a bit disappointed with the level of audio. It was way too low even with headphones on. I'm comparing with with virtually any other video or music I've listened to using the same PC and headphones.

Love the way you mixed in the 1812 Overture. Sadly, my speakers didn't blow out and nor did the needle (on my record player) jump a few tracks with those canon shots :) (If you know what I mean).

Mark Weiss    July 05, 2008 03:18 PM
Hi Shiv,

You should see the uncompressed 1920x1080 original footage. It's spectacular! XDCam really shines. The stills I pulled from frames of the video look cleaner and clearer than the stills my friend got from his expensive Nikon D-SLR.

The dynamic range of this recording is in excess of 76dB, as I carefully set the recording level so that no clipping or compression would result, then I normalized it for maximum peak levels. Unfortunately, 90% of the PC sound cards out there won't have a low enough noise floor to enjoy the full dynamics of this video's soundtrack, nor nearly enough amplifier power to enjoy it at the correct volume level. You'd need about 1500 watts in the average size livingroom through high efficiency speakers to reproduce the loudness of the fireworks as captured in this recording. It is more practical to achieve this with headphones and a good high gain headphone amp. (I'm talking about full-sized headphones, not Walkman-style earbuds.)
High dynamic range recordings are designed to faithfully capture the entire range of loudness, and as such, make big demands on the playback equipment.
Most commercial recordings are heavily-compressed and that makes the average levels sound louder, but loses the impact of the concussion.
I would think you would be glad your speakers didn't blow out!
But what you need is a sound system like the one at Basspig.com to play this back. :)

Shiv Kumar    July 05, 2008 04:37 PM
Hi Mark,

My reference to needle skipping tracks and speakers not blowing out was actually to do with the 1812 Overture:

Many years ago they re-recorded this piece with new cannon shots (bigger/louder). When the vinyl was released many people thought that that there was only one cannon shot because the sound was so loud (the ridge in the vinyl was so high) that the needle simply skipped to the next track and continued on without missing a beat.

Then again years later they rerecorded the 1812 Overture with newer bigger cannon shots but this time for CD. Now CD’s have a dynamic range capability of 90+db (unlike tape which is about 45-50db). The problem with this is (as you know) is that the low level sounds are really low and the high level sounds are REALY high. So people bought these CDs and sat back in the easy chairs to listen to this track. Well, they pumped up the volume really high because they couldn’t hear the low passages they way they liked. As the tracked reached the crescendo and the first cannon shot fired, they heard only one. But this time because their speakers blow out.

Like you seem to be, I’m a major audio buff. Years ago, I used to design and build audio system (where I designed the circuitry for all the electronics) for friends and discotheques. I had the opportunity to build some really fine amplifiers with matching pro-level speakers.
I understand what you say about the way you’ve recorded the audio but still, I am using pro head phones and have a decent (better than average) sound card but still can’t hear the cannon shots loud enough :)

Mark Weiss    July 06, 2008 12:48 AM
Hi Shiv,

Obviously my notoriety has not extended to your circle (yet) or you would have known about my Telarc days :-)

You are referring to the UHQR pressing that sold for $59 in the Telarc catalog. I bought the original pressing (and my tonearm/stylus handled it) back in '83 and it was a blast. Two years later, I bought it on CD. 7 of the cannon blasts are BADLY clipped in the recording. I heard the clipping for years, and suspected something with the recording, but it wasn't until I got my first CD ripper in 1995 that I was able to actually LOOK at the waveform and locate the clipped cannon shots. The loudest blast is over 9dB clipped off, based on extrapolation of the rise time. I rebuilt the peaks by hand a couple years ago and built a version of the final movement of the 1812 that has 9dB more dynamic range than the CD, due to my pushing the overal master level down by that much to make room for reconstructed peaks. Now instead of 2,000W to reproduce the peaks, you need a little over 18,000W to reproduce the cannon shots. :-)

I designed my current sound system in 1982, although it was in various seminal stages of development since 1977. It took its final footprint in 1984, with the contruction of two large folded horns (later converted to twin vented mid-bass cabinets).

My earlier sound systems were all tubes. I designed and built all the vacuum tube electronics for that system. In fact, I'd been designing valve/tube amps for decades prior. My current speakers are the 11th generation. I designed many earlier types and sold them off to friends, finally ending up with the configuration I have now.

I designed a state-variable filter network 2-way crossover back in '82, while, at the same time, Linkwitz & Riley were inventing the same exact thing in parallel and it became a reference filter topology.

Oh BTW, the cannon shots in the fireworks recording are not full level. Only the fireworks actually reach 0dBFS (in fact the loudest part is about 0.2dB clipped due to my pushing the levels up in post, in an attempt to get the loudness maximized without creating objectionable distortion and loss of impact.) The 1812 piece was mixed in at -6dBFS so as not to drown out the fireworks.

Played back on the MotU 896, through the 15kW sound system here, it can go louder than life with ease. However, my Turtle Beach Montego DDL compresses the peaks and has an audible noise floor, which prevents me from playing at the proper volume while still having a blanket of silence underneath the crowd noise. The MotU is the only thing quiet enough to reproduce it all flawlessly. Of course I could burn it to a CD and play that on my Denon CD player and enjoy a clean background too. But the peaks are transients that measure in the microseconds in terms of rise time--they won't be fully audible if the transient response of the transducers and the amp driving them is not darned fast.

Shiv Kumar    July 06, 2008 08:44 AM
Hi Mark,

Well, you amaze me with all the stuff you've done and the knowledge you have!



Menus

Theme

Privacy Policy  |  Terms Of Service  |  Contact Us  |  Support  |  Help/FAQs  |  News