[Trance] The Return (mp3)

[tvdude]

Hey everyone!  Here is a new song from me (finally)!  The stereo is nice, and the mix seems pretty good to, what do you think?  Here's the link
http://213.129.73.14/members/my_tracks/audio/DT_ACF961B4605971.mp3

Hope to hear from you.

[bvanoudtshoorn]

I like! I like the way you first get tricked into thinking the beat is where it isn't. And yeah, the stereo is pretty intense. There's a lot happening in both channels, but it tends to be pretty symmetrical, so there's no sense of one ear overbalancing the other. I know that I tend to favour my right ear over the left; I don't think you've done anything like that here. It takes a while before you introduce any chord changes, but when they come, they're good, and make a nice change.

Overall, top stuff as always.

[Sam_Zen]

Indeed, as always.
I like the way you build it up to a more massive sound.

Some small tech thing : I noticed a sharp tick at the end of the song, so I zoomed in on what could be the cause of it.
So I found quite a strong DC-shift in the negative value instead of zero. The jump back to zero in the last bytes made the tick.
This change in DC bias starts at about 5:36.43, where the concluding sound starts. But it was rather easy to fix.

[tvdude]

Thank you for the comments guys.  

Sam, I remastered the song straight from modplug, brought the end value of the song back up to zero, loaded it into audacity, checked the end value, which is at zero, and i am still getting a pretty bad click.  Any suggestions?  This is the only song I've made that I've run into this problem.  The click doesn't play in windows media player, but it gets played online and everywhere else.  This is perplexing me.

[Sam_Zen]

You can see here the head and trail of the offset-area. The start is the moment, that the 'dry' sound stops,
and only the echo continues to sound. A clear DC-shift.



At the end you can see why you checked the last byte as a zero value, but before that, there's a sharp shift from offset to zero.

I wonder, if this area has the length of the bytes of silence, often added to begin and end of a MP3 file.
Maybe WMP anticipates on this and stops before that area.

I suppose you exported the song as a WAV, imported it into Audacity, and converted and saved it as a MP3 ?
The part with the DC offset shouldn't be there of course, but I guess it's also the conversion.
So I did some testing.

I cut off the tick-area (~968 bytes), so no zero-value as last byte, and saved as WAV.
To check some WAV2MP3 conversions again.

Both 'Lamedrop' and 'Cool' came up with the extra silence bytes again, so with the tick.
'WinLame' did too, but it also has an option called 'nogap encoding', to skip the added silences.
Nevertheless also there the offset caused a jump to zero, so a tick.

As an extra, I did WAV2OGG, knowing that OGG doesn't add silence bytes, but still the last bytes produced a tick.
because of the offset.

[Saga Musix]

MP3 files consist of frames which always have the same length, this is why they are always padded at the beginning and at the end. LAME has the ability to encode "gapless" which still leaves these paddings, but the special LAME MP3 headers tell the audio program to skip a certain amount of bytes. Bad web players may ignore this, hence the click.

[LPChip]

If you make a selection from the last second or so where this negative offset is taken place, and you add a fade-out to it, you'll correct the problem without getting any sound.

The line will then curl up from below the mid section to the center. Since there's no oppposite waveform, there won't be sound, although it might be possible that a noisefloor dissappears.

[älskling]

A good idea might be to check which sample has the DC offset and then fix that sample.

[LPChip]

It can also be possible that your echo uses a flanger or chorus by using an offset (this is a known used technique). In that case, changing the echo would also fix it.

[Saga Musix]

Uuuh, Flanger / Chorus uses a time-based offset (i.e. horizontal) and not a DC offset (vertical)...

[Sam_Zen]

A good idea might be to check which sample has the DC offset and then fix that sample.

One can repair something, but this is exactly the point. What's the source of this offset in the first place ?

[tvdude]

Heya Guys, sorry I've been away so long.  I did do a fade out to zero db, but I still have a problem with a vst or something, as I am still getting some noise, and a click at the end (although it is somewhat quiter).  I'll have to track it down in the vst; i'm sure that's where it is originating.  Thanks for all your replies, it is much appreciated.

[aGIANTpupafish]

ignoring the technical issues, i'm not one for those :p

loved the build up here, and the intro where u confused me briefly with that change of beat, but in a good, fun, exciting way. like an unexpected present :p

all in all some nice trancey sounds here tvdude.  enjoyed it muchly.