delerium – silence (sunmix)

To start with simply try dropping all the levels equally before rendering it to a (wav) file. It sounds like its peaking out and, having no-where else to go, its just distorting and becoming very noisy. It seems to be the top end thats suffering the most. Once you’ve rendered a nice clean wav file (depending upon your source sounds being clean too!) you can then use another program to boost the levels without them clipping.

I’m sure there’s people much more experienced at mastering who can offer you some better advice but thats what I’d try doing to start with.

Again, sort that out and this would be a track I’d listen to a lot! I think it may benefit from a nice, flowing breakdown in the middle somewhere just to break it up a touch and avoid the repetition of the beat from start to finish.

Good work though, very good work!

It still sounds a little grainy in the top end or it could just be the acapella itself. Is the acapella itself of a good quality? This really is a shame as, as I said before, I really like this. Overall, compared to the first version, it does sound fuller buts just let down by the ‘noise’.

Is it possible to post a completely dry version of this track, with no effects? It may be a little easier to pinpoint what is exactly going on. I sincerely hope someone else can point you in the right direction as I fear this is beyond the limits of my advice.

Hopefully if you can sort out these quality issues for any further work as this is a very promising start!

I ‘think’ I know what the problem is. I noticed something while playing your track in, of all things, WinAMP. On WinAMPs graphic EQ display Silence seemed to be missing a large proportion of the ‘top-end’ frequencies. So I ran silence through a spectrum anaylser to see what was going on. you seem to be missing everything above 16kHz. I am wondering if this is because you are using a bad MP3 compressor.

Below are two spectrum analysis graphs. The first is of Silence and the second is one of my tracks, for comparison, which has also been MP3 compressed.

[size size=150:1oqiczmk]Here is the spectrum analysis of Silence:[/size:1oqiczmk]

[size size=150:1oqiczmk]And here is a spectrum analysis of one of my tracks:[/size:1oqiczmk]

Please notice that (for some reason) the Hz markers are in slightly different places for each graph.

To make it a fair comparison the two tracks have been encoded at very similar rates. Although mine is a VBR with a maximum of 160kbps, yours is a CBR of 160kbps but that shouldn’t be making that much of a difference.

For all the frequencies below 16kHz notice the graphs are very similar however on yours there is a drastic drop off above 16kHz while mine drops a little at 16kHz before dropping off at around 18.5kHz.

A little drop off at the higher end of the spectrum is to be expected, after all thats what MP3 compression goes after first (I think!), the frequencies that would be out of the range of most peoples hearing. If I’m right (I probably aren’t but its worth a try anyway!) your MP3 encoder seems to be far, far too aggressive and is removing far too much. Unfortunately your MP3 encoder doesn’t tag the MP3 file with its own name (some do, some don’t) so I can’t see what you’re using.

Thats why at first I thought your levels were too high and being clipped, creating noise. Now I think the ‘noise’ may be due to the missing frequencies. What could be happening is that things like the vocals, the pads/strings, the snare/clap (can’t quite tell) are missing their upper frequencies which results in them sounding distorted/noisy. I also notice that the hi-hats have almost completely disappeared too.

Like I said, I may be wrong about it being the MP3 encoder but its worth trying, and easy to try, a different one. If that doesn’t cure the problem it may help if you can post a short uncompressed wav clip so we can have a listen to the original source… it depends if you think its worth the effort though!

Hope you understood all of that and I hope it helps!

Shades

[quote quote="sun":1bzgsq95]thanks , i checked my original output and indeed the high frequencies are very low, although if i "up" them in the mastering process the "strings" starts "screaming" a bit too much for my taste.

maybe it’s all about my choice in vst instruments, hehe.[/quote:1bzgsq95]

Hmm, not too sure it will be your VSTs as the upper frequencies of the vocals are missing too… unless you play the vocals through a VST or something?? Perhaps its something to do with your rendering options? If I choose the ‘best quality’ choice with some programs the final wav comes out sounding awful!!

Don’t try boosting frequencies, try backing everything else off! Like…

… then boost the volume back up without clipping.

Also I’ve just thought of something else…

Does your sound card have some sort of EQ or bass/treble controls? If so these should be as neutral as possible so as not to distort your impression of the over all sound of the final mix.

I once had a problem where the program I used to use had a ‘seperate’ EQ for playback but it got rendered into the final mix. So if I set up the EQ for how I like it (usually high bass, low mid and high treble) while composing, rendered the final wav and then listened to it in WinAMP (once again with the EQ how I like it, high bass, low mid and high treble) then the highs would be too high (and ‘scream’) and the bass would be too high and distort.

I’ve found its best to keep things as neutral as possible, with only minor EQ tweaks to certain sounds, so as not to affect the listeners own preference.

Just something to think about!

At least we appear to have got to the root of the problem. Just checked my version of the acapella (downloaded from another source) and it appears to be exactly the same. Will you try sorting it out or just taking it as something to watch out for in the future?

will no one use the sarah (mclocghin) version of the vocals?

I’ve heard so many re-sung versions

its no sweat though, I like this version. It sounds like it wants to break into a huge electro track