What "NOT" To Do While Your Mixing ..

  • Thread starter Thread starter P-Loc
  • Start date Start date
P

P-Loc

New member
If u gonna send ur stuff 'out' to get mastered?..


anyone have advice or tips?
 
- Don't put a compressor on the mix buss (or if you must, then use VERY little!)

- Don't push your mixdown levels too hard - leaving 2-3db for the ME to work with is not a bad thing...

- Don't worry about fade-ins/outs - that part of the ME's job...

- Provide a clear time log for the ME, along with any notes you may have on each mix...

- Concentrate on getting the best mix you can - mastering works best if only a little polish is needed... remember - you can't polish a turd!
 
Blue Bear, Im interested in getting about 10 rap tracks mastered. The mixes are pretty good and most are already peaking at -0db and non of them are clipping. I would like you to email me at lilraydio@aol.com with prices and other info, if you have open slots. Until then, peace!
 
just kidding...

Bruce, I think you should contact your psychotherapist. I never heard of anyone before adressing himself as 'the ME'. Seems like a severe ego-problem to ME! LOL :D ;) :D

David.
 
Kinda like that Seinfeld episode where that guy (Jimmy?) always referred to himself in the 3rd person, eh?!?!

:D :D
 
dobro said:
What's a time log?
The start/end times of all the songs you want mastered/sequenced (assuming it's on DAT or analog tape!)
 
P-Loc said:
If u gonna send ur stuff 'out' to get mastered?..


anyone have advice or tips?

- When mixing try to make the songs as consistent as possible as far as EQ and levels from song to song. Create a CD of your mixes, and listen on a variety of systems. If you find that you have to reach for the EQ or volume knob on your stereo when going from one song to the other something is wrong.

- Keep frequencies balanced. It can be very difficult for example to fix a mix where the cymbals are bright and everything else is dull. Trying to enhance the high end in this case is going to make the cymbals too bright. The same thing applies when the bass is hot kick is low and vice versa.

- Don't normalize or provide any processing on the overall mix. Essentially you don't want to create something that can't be undone later. Even if it does sound better than the original, processing should be done as few times as possible in order to reduce the number of calculations perfomed on digital audio.

- Fix anything in the mix that may be problematic after it's mixed in stereo. For example if an individual track is hot (e.g. vocal or guitar) don't expect a mastering job to remix your material. Likewise don't expect mastering to add effects to individual tracks (I've had people ask me if I can add reverb to the vocal part during mastering).

- Try to keep the final mix in the highest resolution possible. In other words, if you mixed in 24 bit 96k try to provide that mix rather than creating an audio CD at 16 bit 44.1k and sending that to the mastering house.


If you have any further questions feel free to email me at:

info@masteringhouse.com
 
Re: Re: What "NOT" To Do While Your Mixing ..

masteringhouse said:
- When mixing try to make the songs as consistent as possible as far as EQ and levels from song to song. Create a CD of your mixes, and listen on a variety of systems. If you find that you have to reach for the EQ or volume knob on your stereo when going from one song to the other something is wrong.

masterh. -

i have a question about point #1 in your otherwise outstanding list of recommendations. I have asked my various mastering engineers this specific question, and they have invariably told me NOT to worry about balancing the volumes of songs relative to eachother - just worry about making really great sounding mixes of each song, and they'll worry about balancing them when they get them.

Do you have strong disagreement with this?
 
Re: Re: Re: What "NOT" To Do While Your Mixing ..

littledog said:
masterh. -

i have a question about point #1 in your otherwise outstanding list of recommendations. I have asked my various mastering engineers this specific question, and they have invariably told me NOT to worry about balancing the volumes of songs relative to eachother - just worry about making really great sounding mixes of each song, and they'll worry about balancing them when they get them.

Do you have strong disagreement with this?

Hey Littledog,

I don't have a strong disagreement with this statement, but I believe that great sounding mixes and balance in volume and EQ are not contradictory.

Let's take an example to demonstrate a point about audio quality and volume differences. Say that on song 1 a drummer was more aggressive when hitting a snare than on song 2. When mixing, the engineer set the threshold of a compressor for the snare in song 1 properly in order to tame a few of the inconsistent hits. When mixing song 2 (in his haste to get the mixes completed for the band's CD release party in 2 weeks) he used the same compressor settings but raised the volume of the snare (post compressor) to compensate for the weaker hits. As a result a many of the inconsistent hits in song 2 were allowed to get by the compressor without hitting the threshold, thereby creating a greater dynamic range that has to be dealt with when producing the final mix. In order to prevent digital “overs” in mix 2, the overall level will need to be brought down relative to mix 1 due to some of the snare hits hitting close to 0 dbfs on the individual track.

If we looked at the 2 songs in our favorite digital workstation we would see major variations between the two waveforms as the snare peaks in song 2 would be taking up more of the dynamic range than in song 1. As we all know, volume levels are based more on average volume than peak volume. As a result, song 1 would sound louder than song 2 even though individually the mixes might sound good. Let’s say that the difference in the average level between the songs is about 12 db. In order to get the 2 songs to have the same apparent level, song 2 would have to be compressed by 12 db during mastering. Since digital audio represents levels in approximately 6db/bit, 2 bits were essentially “wasted” in song 2 in order to handle a few random snare hits. These 2 bits could have been put to better use representing the remainder of the mix rather than just getting thrown out later. Not using the full word length we have available in digital will make audio sound more “grainy” due to decreased resolution. Try making a recording at a low volume and another at normal volume using the full scale. Now normalize the recording done at the lower volume so that it becomes full scale and listen to the difference.

That said, if our fictitious engineer had listened to the 2 mixes together rather than individually, the difference in levels would have been easily heard and hopefully corrected.

Other examples might be a bass guitar track, kick that is all over the place between songs. Instead of having the mastering engineer roll off bass thereby compromising one or the other, fix it during the mixing stage. What I’m suggesting is to listen to the entire album and make it as cohesive as possible without doing any damage.

Again, the main concern should be a great mix. If there are slight differences in levels and EQs between songs that is to be expected, and is the reason we have the mastering stage in the production process (other than noise reduction, editing, final song arrangement, an objective ear, etc.). You shouldn’t fudge this and try to correct imbalances with normalizing, major amounts of compressing, and EQing on the stereo bus. Try to get it at the track level because this is what you are committing to as a mixing engineer. Once printed to stereo it is much more difficult to control things at the individual level.
 
Littledog, he doesnt mean the stereo mix buss level, so not talking about raising or lowering the song by messing with the two track, just saying to make sure the song's internal levels are similar relative to themselves. I think thats what he's saying, or at least, thats how I operate :)

On the PC its nice to make batch files to handle the other 11 snares after you get the first snare right, but it usually doesnt work out dynamics wise, so its best to use that for eq's and things or not at all, for example
 
Masterh.-

In the examples you describe i agree with your points. But here's my perspective:

Remember that in Pro Tools it is not trivial to go back and forth between two different sessions to compare levels since you can't have multiple sessions open at the same time. Also, I think maximizing bit usage is not nearly the issue in 24 bits as it was in 16 bits. If a track is only using 22 bits, it's no biggie.

In my work, not all CD's necessarily have the same style and instrumentation on every track. I've done plenty of projects where some tracks are a kick-ass all electric band, while others are acoustic "unplugged" arrangements. Some might not even have drums. There might even be a track of a capella vocals mixed in. To worry about whether those songs are balanced relative to eachother is in some way a waste of my (and my client's) time, since the mastering engineer can do instant comparisons once it's all in Sonic Solutions and do a better job balancing than I can.

I guess i'm assuming that I'm a good enough mixer not to have one song with an average level of -30dB while another song has an average level of -8dB. Or at least that I would notice that was the case at some point during the process.
 
littledog said:
Masterh.-

"Remember that in Pro Tools it is not trivial to go back and forth between two different sessions to compare levels since you can't have multiple sessions open at the same time."

My recommendation in this case would be that you bounce the first mix and include it as a reference in 2 tracks that you are currently mixing, muting and unmuting them as needed. Of course the reference should be the same style of music and instrumentation as you mentioned below. If not a previous track it may also be helpful to use a track from a CD the client likes. Just flying blind can waste more time than it takes to switch back and forth between sessions if things have to be remixed multiple times.

What I was referering to in the previous post was creating a CD to listen to the rough mixes on various systems. And checking balances. I always do this for a client as a take away when a session is done for the day so that they know what they're getting as a session progresses and know how the mixes translate for different systems. Just monitoring in one environment can be very misleading and usually ends up with the "I don't get it, it sounded great in the studio" syndrome and an unhappy client.


" Also, I think maximizing bit usage is not nearly the issue in 24 bits as it was in 16 bits. If a track is only using 22 bits, it's no biggie."

The question is how many bits are in use by the "average level" of the content. If you have spikes all over the place using all 24 bits but the average level is only using 10 and then has to be raised another 8 bits or more later to be competitve, maximizing bit usage is as important as maintaining a proper level in the analog world. That's one of the reasons people "slam" analog tape (other than getting a compressed sound). It reduces the amount of noise in the final product and makes the sound "punchier".

"In my work, not all CD's necessarily have the same style and instrumentation on every track. I've done plenty of projects where some tracks are a kick-ass all electric band, while others are acoustic "unplugged" arrangements..."

I agree totally and was going to mention this as a disclaimer previously but I didn't want to digress too much. The average level of an accoustic track should be less than a rocker. But not to the point where one has to strain to hear it when assembled on a final CD. Likewise one shouldn't go to the other extreme when mixing an acoustic track and make it too loud. This may also be a case for not using all 24 bits as described above. Discretion is key when trying to apply any rule of thumb. Much like the recommendations people give on EQ and compressor settings, try it, listen, and then decide what's appropriate for a given situation.

"I guess i'm assuming that I'm a good enough mixer not to have one song with an average level of -30dB while another song has an average level of -8dB. Or at least that I would notice that was the case at some point during the process."

I wish I could say the same for some of the projects I get in. Recently I had a session that was so all over the place in terms of level and EQ it sounded as if it was recorded in 12 different studios. When I approached the engineer about the inconsistency he said that he wanted it to "sound inconsistent". While I agree that an album shouldn't sound as if it came out of a blender, it shouldn't rip your head off in one song and be totally dull and lifeless in the next. One where the kick drum is sounding like a "pile driver" and then barely audible above the bass gutar in the next, and so on. You can imagine how difficult it is in a situation such as this to make it sound like "an album".
 
Apparently it's just a matter of degree then. Hopefully my work bears no resemblence to your extreme examples, and if it did, I would hope that the mastering engineers in Boston would have the good sense to let me know instead of telling me not to sweat the levels. After all these years of sending them my clients work, you'd think at least one of them would have the guts to tell me my stuff is total crap if it were true.
 
littledog, who do you go to in Boston?

Boston mastering engineers are NOT known for biting their tongue when theres a problem, so if you were screwing up, I promise you would have heard by now :)

I think youre ok
 
Littledog -

Judging from the concern and effort that you put into on the posts on this board, I'm certain that you produce great mixes. Send me some links sometime, I'd like to hear hear your finished work.

As far as bit resolution/24 bit I thought that the following article may be of interest:

http://mixonline.com/ar/audio_mixing_pro_tools/index.htm

See the following comment from Bob Brockman:

“You don't expect an artist who's recording their own stuff to be totally on top of the recording aspect of it. But I've found that one of the things that ends up happening in the case of Pro Tools is you have source material that may be very interesting to listen to, but which is often recorded at a really, really low level; it's 10-bit or 8-bit or even 6-bit. When you have a lower level recording at 24-bit, the lower the level gets, the lower the bit resolution. That's just the way PCM works. If you don't record a super-hot level, then you're not getting all the bits in the word resolution, and, consequently, it can sound pretty degraded."
 
There's an awfully lot of very well respected engineers who disgree with Bob Brockman about the need, both technically and musically, to record super hot at 24 bits.

I guess we'll have to agree to disagree on that one.
 
Back
Top