Hi all, I registered here a few months ago but could never load the forum pages for some reason, until now. Probably a good thing, because I've learnt a great deal during that time in the world of engineering, saving me any embarrassing moments. Well, probably more to come. ;)
Anyway, I'm new to mastering, but not new to music production and mixing. I switched to doing only mixing earlier this year after having produced my own stuff for 7 years, in an attempt to follow my interest in mastering, only to find out later that was actually called "stem mastering". So as of recently I've started doing actual mastering on stereo mixdowns using Sony Sound Forge and Waves DSPs.
I know all you pros would probably frown at me for using a PC based setup, MOTU 828mkII FireWire interface and AKG K 240 S headphones, but that's all I can afford for now (was going to buy Dynaudio BM15As, but I was robbed before I could buy them :(). Besides, it's the engineer more than the equipment he uses that determines the final outcome I believe.
Anyway, to make this topic a little more interesting, I'd like some feedback on my current mastering procedure and quality of one of my masters from anyone with more experience than me if possible.
I don't know if I'm doing this right, but in my most recent master I'm using a total of 12 plugins in the plugin chain of Sony Sound Forge (mixing/mastering EQ (corrective cuts to certain instruments, wider boosts/cuts in certain frequency ranges)->multiband compression->multiband stereo imaging->corrective EQ (low shelf/hi-pass, low-pass)->multiband limiting). Is it usual to have a total of 31 EQ nodes to shape a mix? I started out only using 10 or so, but the more experience I got, the more problematic frequencies I started picking out, and the fact that none of the mixes I get to work on have actually been properly mixed down doesn't make this any easier.
After a couple of masters I was able to pick out 0.1 dB changes in EQ boosts or cuts too, so I often end up doing as little as a 0.1 dB boost at 220 Hz, Q 2.87 to add more 'body' to the snare, for example. It's quite frustrating because I end up spending hours on end shaping the mix properly. So, ideally I'd just like to know if this is usual. I understand that it's probably more mixing EQ I'm applying, but I have to correct any frequency problems first before multiband compressing the mix, right?
So, after the compressor I use multband stereo imaging usually to tighten the bottom-end inwards in order to create a wider and more defined high mid and top-end, and I follow this with the L3 (I usually aI'm for a final RMS of around -13.2 dB, and have actually never limited with more than -6.0 dB of attenuation. 8-))
But OK, a short sample of my latest work is available below. I'd be grateful if someone with more experience could give me some feedback (you'll have to excuse the artificial sounding reverb, as it was made in Renoise).
http://atlantis.plastiqueweb.com/temp/atlantis.ogg
http://atlantis.plastiqueweb.com/temp/atlantis-mastered.ogg
Thanks for reading this far and I hope to get a lot more questions answered during my stay here, and perhaps help out any others too. 8-)
Comments
Thanks for the reply. That fully clears up stem mastering now to
Thanks for the reply. That fully clears up stem mastering now too.
And yes, I know about the headphone issue, but I'm still a student and can't afford anything more for the moment. Because I still work with headphones, I haven't really considered adjusting the stereo field much yet, only doing some minor adjustments when really necessary. I just compare the sound I'm used to hearing in commercial tracks, and base my opinion on that. Not the best, I know, but I'm not doing this for a living yet or anything.
Well, I'm not mastering my own mixes. It's just that the tracks I get given come from amateur musicians, and the tracks haven't been mixed down properly in regards to EQ and volume balancing. For example, in the sample I posted in my previous post, one kick is playing at 55 Hz, another at 87 Hz, the snare at 220 Hz, the fundamentals of other instruments at 262 Hz and 523 Hz and so forth. And because often the frequency balance isn't dead on with, for example, the first kick not having enough presence, I boost at 55 Hz, while I use 87 Hz on the second kick and so forth. I just correct all of the mixing mistakes first.
Thanks for the RMS tip though. Didn't really think of it before but you're right. I just read on http://www.loudenessrace.net/ I believe it was that you shouldn't have a final RMS greater than -12.0 dB, though I even find that too loud.
Thanks for checking out the sample and giving me some feedback though. The "fuzzy bass" is actually caused by a distorted pad playing as well, but it's only for a moment in the song and emphasises an 'edgy' part to the piece. Besides, I can't reduce the top-end any further because this will make the hats fade too, and I'm really satisfied with the current sparkle on them. 18795 Hz was another problem zone with one of the sitar notes having a rich overtone here, but I reduced it as much as I could I without affecting anything else too much.
But, this is after about two months of mastering work. I hope to improve a lot more in the future. :)
I thought that I should add to the good advice already given. I
I thought that I should add to the good advice already given. I may be wrong, but it sounds at lot like you're approaching mastering as a mixing engineer. You've mentioned a lot of the mix elements that you're working on, which is not a bad thing, but mastering is something that is often more broad than mixing. A lot of mix engineers listen to each individual instrument when mastering as opposed to looking at the bigger picture. In other words, fixing a snare drum is really best fix at the time of mixing and not the mastering. Mastering is typically where you are looking at the overall big picture. You'll likely realize even better results if you're not trying change a mix at the master stage but rather enhancing the mix that was given to you. Although, having a good mix to start with is very helpful.
HTH,
Erik
Ahh, zen and the art of mastering. This is why it takes enginee
Ahh, zen and the art of mastering.
This is why it takes engineers awhile to get the hang of mastering. It's not the same mind set. You have to unlearn certain things that you learned as an engineer. It took me many years to do this. You definately have to listen differently, if you need more than a handful of things to achieve what you need, then my guess is Erik is 100% right.
A great example of the downfalls of fixing a mix in mastering wo
A great example of the downfalls of fixing a mix in mastering would be doing everything possible to pull out a vocal and make it sound so much better than the original mix. Creative compression and wonderful EQ is applied to the vocal range of the mix. The vocal sounds great now... right? Well, a great sounding vocal was created at the expense mangling of every other part of the overall arrangement. I bet the guitars now sound awful and the drums have some odd ring to them. Now that the vocal is perfect, what to do next? Try to fix what was broken when creating the perfect vocal? My point is, fixing a mix during the master stage is a slippery slope and one can easily get stuck working on a single master for numerous hours. Something got broken, then it gets fixed and the fix ruins something else. Now we're stuck in a never ending vicious cycle of adding more and more unnecessary processing at the demise of the original mix.
-Erik
soundfreely wrote: one can easily get stuck working on a single
soundfreely wrote: one can easily get stuck working on a single master for numerous hours. Something got broken, then it gets fixed and the fix ruins something else. Now we're stuck in a never ending vicious cycle of adding more and more unnecessary processing at the demise of the original mix.
-Erik
and at this point you loose your perception & it's starts to sound better :?
Ed
In short - "Stem mastering" is mastering from the stems of a mix
In short - "Stem mastering" is mastering from the stems of a mix - Instead of being presented with a stereo mix, you might get a drum mix, guitars mix, vocal mix and effects mix, all at unity gain. It lets you tweak in a more "microscopic" fasion. Some mix engineers like to have it done that way, some don't.
But as long as we're getting into semantics, the "actual" term for the preparation of a production master disc that will wind up being the final version before a glass master is made is called "premastering." Although, the "accepted" term has been truncated to "mastering" without much argument.
I agree on the engineer's ears and years over the gear, but there is also absolutely no substitute for a quality monitoring chain. The best engineer and /or gear in the world is worthless without a quality monitoring chain to hear it properly. It's essentially impossible to make stereo field adjustments efficiently using headphones. An "open air" listening field is an absolute requirement. The same sound needs to hit both ears at different times. With headphones, that can't happen.
The only time I could ever imagine anyone using 12 plugins, 31 EQ nodes (or even 10 for the most part) and "hours" on anything is if you're trying to master your own mixes. One of several reasons why it's generally a very bad idea.
Try to never aim for a particular RMS level - Every track will "ask" for a particular level. Some want to be loud, some don't. Aim for a particular "SOUND" and worry about volume later. This is why most commercial releases sound so damn... Deep breaths... Deep breaths... Sorry. Almost lost it there.
Anyway, your files sound fine for the most part. The "fuzzy bass" that comes in around 0:35 thru the end has an awful lot of high-end energy in it, but that can easily be chalked up to the mix and artistic license.