Recording: Neon Genesis Evangelion - A Cruel Angels Thesis

Submitted Wed, 10/10/2012 - 17:58
by Deiselc

Your rating: None

I only spent a few days on this so its a bit rough but its ok. Also I dont like how youtube has compressed the audio, it sounds completely different but I have to live with that.


9 comments on A Cruel Angels Thesis

Hey man, nice cover! :D I

Hey man, nice cover! :D

I noticed you use a Line6 Spider. I think you can get a way higher quality recording if you use the headphones output of that amp and connect it with the Line-In of your soundcard. It'll give you a much better signal than your camera microphone. :)

Thanks Im glad you like it.

Thanks Im glad you like it.

Actually the amp is there but Im not using it for this recording. I used an Line 6 UX2 interface for this recording. But youtubes compression really changed the audio and made it sound alot more muffled than the original recording which is why it sounds a bit like I used a mic.

Wow, really? :o Do you use a

Wow, really? :o Do you use a DAW like Reaper, Ableton, Mixcraft, ... to record? If so, they probably have a multiband compressor VST plugin that you can use on the master track. I think that might solve your problem with Youtube's compression.

Also make sure to follow Youtube's encoding specifications.

And if you haven't already, check out FamilyJules7x's recording guide. It's REALLY helpful.

Yeah I use mixcraft to

Yeah I use mixcraft to record. And Ive never had a problem with youtubes compression before this video. Ive looked into the encoding but there should be no problems. Its really disappointing the audio got so messed up but I honestly dont have time to mess around trying to fix it.

And I havent watched any recording guides but I might check it out. :)

Multiband dynamic compression

Multiband dynamic compression won't help with Youtube's coding. Doesn't make sense--perceptual coding does worse with increased spectral density, not better. You should only use a multiband dynamic compressor if you want it to sound like you've used a multiband dynamic compressor, not to solve coding problems.

Make sure you don't mix up dynamic compression with data compression (i.e. perceptual coding). They're separate concepts.

Feeding Youtube the highest quality audio file possible will help, though. Tandem encodes add extra noise, which tends to push things well past masking thresholds, so you want to use the highest bitrate possible, much higher than necessary to sound good on its own. (That is, even if you can't tell the difference between 256kbit/s mp3 or 160kbit/s aac and the original, you should still feed Youtube a higher quality file than that if you want it to sound good.)

I'm not going to pretend I

I'm not going to pretend I fully understood that but I get the basic idea, higher bitrate equals better. Im much more specialized behind the camera. Sound is really not my strong point. I've basically learned everything by just playing around.

Thanks for all the advice guys but I do have to address n00bje that plugging my 150W amp into my computers Line In will fry my sound card in a matter of seconds, so I really dont recommend trying that.

Well, I'm not a really good

Well, I'm not a really good musician or anything, but I understand engineering stuff, so sometimes I can explain technical points.

"Compression" is confusing, because it refers to two things:

  1. Dynamic range compression. This is what a "compressor" does in a DAW. It makes quieter parts of the input louder.
  2. Data compression (perceptual coding). This is like MP3, AAC, etc. Youtube does this.

Data compression isn't usually called "compression" by audio people, because that's confusing. Instead, it helps to refer to data compression as "coding" (encoding and decoding), and to the data compressors as "coders" (encoders and decoders).

MP3/AAC etc. use "perceptual coding", which works by adding noise. It's possible to add noise to a recording that you're physically incapable of hearing, because your song covers up the noise. This is called "masking". Perceptual coders calculate "masking thresholds" to figure out how much noise they can add in each part of the frequency spectrum. It's never perfect, but the majority of the noise added is inaudible at a good bitrate.

In order to play an MP3, you have to decode it; Youtube has to do the same thing in order to re-encode it at the bitrate it wants to use. This is called "tandem encoding". (Note: there is no "bitrate peeling" in audio like there is for video.) Unfortunately, tandem encoding has to go through the process of adding noise *twice*. The first time added noise up to the masking thresholds; the second time adds noise again, but this time ALL of the noise added is audible. So the tandem encode sounds significantly worse.

In other words, encoding an MP3 from an MP3 makes it sound more than twice as bad as encoding once. You have to go for overkill on the bitrate to compensate for this.

P.S. Sorry if that was tl;dr

P.S. Sorry if that was tl;dr ;-)

No its cool. :) Your

No its cool. :) Your explanation actually made sense and will be pretty useful in the future if I ever have to work with audio professionally. Thank you.