 |
|
 |
Atlantis
  Member
Topics: 84
Replies: 3227
Registered: 14.Jan.03 |
 |
Underneath all the flaming this was actually quite an interesting read. I just wish I would've been around at the time to take part, although I do know a lot more on the topic now.
Basically, I agree with MyztiQue. He hasn't had a lot of input, but master compression has it's advantages as well as its disadvantages, but the advantages do outshine the distadvantages. Compressing each instrument individually and then combining these into a final mix is often still not going to sound cohesive, and good master compression usually has the effect of glueing things together for a more consistent sound. And yes, radio mixes are a lot more compressed and limited (part of the reason I haven't listened to the radio in years), but not everyone has a $10,000 hi-fi home theatre setup to experience all those transients. I, for one, always apply master compression, and, as acecream pointed out too, when set right, the negative effects become practically inaudible.
But, the real reason I stumbled across this topic is because I wanted to point out something I've been discovering lately when limiting my masters. I read on this site (hmm, link not working for me for some reason) that an RMS volume of -12.0 dB is considered to be a good average volume of your final master, whereas my initial rule of thumb had always been never to limit with more than 2.5 dBs of attenuation. That's stupid of course, because it depends on the input sound how much you can limit it before it starts to sound squashed. Still, limiting with any more than 6.0 dB of attenuation is going into dangerous territory, and personally I even like to have a master RMS volume of around -14.5 dB.
So, what I initially wanted to bring up is this suggestion: After you've compressed/limited (although unfortunately most people just peak limit) your track, check the RMS volume (Tools->Statistics in Sony Sound Forge) and make sure the RMS of both the left and right channel averages to no less than -12.0 dB (i.e. make sure it is more negative). Styles typically suited for home listening as opposed to dance music (such as electronica, downtempo, etc.) are better left with an RMS of -13.2 dB to -14.5 dB, and classical music probably in the order of -18.1 dB or higher (more negative), although really you wouldn't want to limit a classical piece of music anyway.
If your track has both loud and quiet passages, resulting in the quiet parts skewing the overall RMS value to a greater negative number, select the loudest passage and check the statistics then. Again, make sure it isn't below -12.0 dB, as I've come across mixes with an RMS much lower than that on this site (closer to 0.0 dB), and it's something I sure don't like to hear. 
Oh, and one other thing: Don't normalise to 0.0 dB, but to -0.20 dB instead, especially when you're going to encode to OGG or MP3 afterwards. 
Atlantis [Atlantean Records .: extant music beyond existence]
Mixing engineer/multi-band professor |
|
Nifflas
Member
Topics: 80
Replies: 1599
Registered: 05.Mar.04 |
Hmmmz, it almost feels like I'm devloping a dislike for the clean nice mix, where everything is supposed to be mixed and compressed in specific ways. :/ Guess I'm going into insanity with my music sooner or later.
edit: removed some text from this reply. I realize Szartd did apoligize for his post, and I guess an apoligy is an apoligy...
I'd like to point out though, on internet, you never know the mentally state of the person you're insulting. My girlfriend left me a few months ago, my ferret died almost the same day, I broke most of my gear in panic. My SL-880 was alive for a while, until someone insulted me online like that. I totally lost controll once again, and smashed my $600 midi keyboard into pieces...
I still have some nasty things people have said to me, that really hurted me, playing in my head over and over and over again every day.
Fortunately, my music is much more interesting than my posts. |
|
Randor
 Administrator
Topics: 560
Replies: 2948
Registered: 27.Dec.02 |
Neat Atlantis 
Never paind attention to RMs values and checked some of my songs....now I'm bit confused about average vs. loudest since one of my latest tracks you said RMS volumes could be higher.....Loudest popint is -6.8db while average is around -15 dbs ... Those breaks (where peaks are) lose power if I reduce levels, tried to remaster track and notices that? Why? Should I do compression/limiting in mix?
Or just obey my ears do like this far; afterall I'm amateur 
|
|
Nifflas
Member
Topics: 80
Replies: 1599
Registered: 05.Mar.04 |
So... erm... perhaps this has been answered 10 times already here, without me being enough intelligent to understand the answer  Anyways, if I have exported one of my songs to .wav, I don't feel like compressing every instrument individually, nor compressing the whole mix, because it will simply ruin the sound I like - anyways, how do I get rid of those peaks, so the overall volume can get a few db louder, without clipping?  Is there any alternative ways?
Fortunately, my music is much more interesting than my posts. |
|
Randor
Administrator
Topics: 560
Replies: 2948
Registered: 27.Dec.02 |
| I presume there's two ways, compressing with multiband compressor so the tresholds are set to db level where only those peaks reach. Or use limiter and cut of 2-3 dbs wich usually doesn't affect to sound much |
|
Nifflas
Member
Topics: 80
Replies: 1599
Registered: 05.Mar.04 |
Ah I wish I had an audio editor which could mark the peaks, and things like that, so I could see it visually... That often helps in this situations...
Fortunately, my music is much more interesting than my posts. |
|
sonus
 Artist
Topics: 20
Replies: 501
Registered: 26.May.04 |
Atlantis,
You mentioned normalizing to -.2db is good when recompressing the wave to ogg or mp3. Why is this?
-Sonus |
|
Nygen Dale
Staff / Moderator
Topics: 68
Replies: 3617
Registered: 21.Jun.03 |
sonus wrote on 15 Nov. (22:15) :
Atlantis,
You mentioned normalizing to -.2db is good when recompressing the wave to ogg or mp3. Why is this?
-Sonus |
I believe when you process the file there might occur minor volume changes and if you normalize to 0.0db there's no room for possible volume rises and there's the danger of clipping. So I guess -0.2db is just a value which puts you on the safe side without being noticably lower in volume.
 |
|
Atlantis
Member
Topics: 84
Replies: 3227
Registered: 14.Jan.03 |
Randor wrote on 15 Nov. (17:04) :
Never paind attention to RMs values and checked some of my songs....now I'm bit confused about average vs. loudest since one of my latest tracks you said RMS volumes could be higher.....Loudest popint is -6.8db while average is around -15 dbs ...
|
When you say the loudest point is -6.8 dB, do you mean that that's your loudest peak volume? That is, you look at your wave file in an audio editor and the waveform doesn't touch the top of the screen but rather only goes to around half way? In that case, normalise the waveform to -0.20 dB and check the RMS again.
|
Those breaks (where peaks are) lose power if I reduce levels, tried to remaster track and notices that? Why? |
Can you explain this a bit more? I don't quite follow.
Atlantis [Atlantean Records .: extant music beyond existence]
Mixing engineer/multi-band professor |
|
Atlantis
Member
Topics: 84
Replies: 3227
Registered: 14.Jan.03 |
Ni wrote on 15 Nov. (16:31) :
Hmmmz, it almost feels like I'm devloping a dislike for the clean nice mix, where everything is supposed to be mixed and compressed in specific ways. :/ Guess I'm going into insanity with my music sooner or later.
Fortunately, my music is much more interesting than my posts. |
I can understand that coming from you. These are rules you don't really have to follow though, so long as you don't go in the opposite direction I guess. That is, you don't have to compress your final mix if you don't want to, but I'd say the real rule is to not over-compress and over-limit it just to end up with a wimpy loud RMS volume of -10 dB or any other number closer to -0.0 dB. There's a reason limiting to increase the RMS is good, but again it's not something you're required to do as a producer of course.
Atlantis [Atlantean Records .: extant music beyond existence]
Mixing engineer/multi-band professor |
|
|
|
|
|
Latest Topics
![[203.118.142.104]](pics/ip.gif){kind=small}
