This is a very interesting thread. I’ve been happy with my mixes for some time now (save for a few exceptions) so I am struggling with mentally “picturing” (hearing?) what spending a lot of time on mastering type of activities is going to buy me.

Don’t get me wrong: I know that my mixes could sound better. But when the difference in quality could be attributed to anything from the player (me) to the signal chain getting to my computer to automation levels or any number of other, non-mastering related things then I tend to focus my activities on getting the mix to sound right.

At that point, I typically slap on a “warming” plug-in (Magneto), a limiter (typically the stock Cubase limit with the Light Peak Limiting preset), some EQ (typically to add presence in the upper frequency ranges), and (lately) FerricTDS and SlickHDR as final steps.

I do plan to investigate the statement that a maximizer should be put in the post-fader slot. But before I do my own A/B comparisons, can someone explain why this is a necessity?

Maximiser necessity - maybe
A maximiser in NOT necessary per se, but it can up the general levels of a mix to be more in line with what others in your genre are doing.

It needs to be applied judiciously as it is quite aggressive (with obvious artefacts) when the set maximum is reached, and pushing up levels to produce too many opportunities for such aggression can really ruin a mix. Better to adjust the mix for those tracks that contain large peaks, to softly round them down with compression or automation. Doing that well may produce less need for maximisation, or at least less aggressive settings.

While a brickwall limiter is mainly concerned with the very maximum level, a maximiser works over an adjustable band of levels, which can dramatically alter the dynamics of a mix. Until used to what it does, you may want to do a few try-then-listen cycles, and with others if possible, so that you are happy with the results.

Aggressive maximisation can destroy subtleties in a mix, so some genres can take more while gentle or sparse music may need more care.

Post fader necessity - absolutely!
It needs to be post-fader because it needs to be tied to the absolute maximum level possible for the output file or DAC for best effect, mainly because it allows NOTHING over its maximum.

If you have a level control after it, you will either:
a) Lower the level, in which case you have just wasted some dynamic range because nothing will appear above the maximiser’s maximum output level, or
b) Increase the level, in which case you have guaranteed clipping for EVERYTHING that exceeds the file’s or DAC’s maximum possible level.

Basically, it’s a lose all round if you adjust a level after a maximiser.

Maximiser ride alongs
Because of this absolute nature of a maximiser, it is the natural candidate in which to include any sample-rate conversion, bit-rate truncation and dithering, all of which are best applied as the very final processes of rendering.

Another setting that a maximiser may have is to handle inter-sample clipping. This is where while the samples either side may be less than the maximum, the resulting audio waveform between them would have exceeded the maximum. If the facility is switched on, the maximiser will calculate the real maximum and compress the peak more.

If outsourcing mastering - don’t use, but know what you want
Note that if you intend passing your mixes on to a mastering engineer, do NOT use any maximisation, sample-rate conversion, bit-rate truncation or dithering, as any of these are best applied by them AFTER they have done their magic.

However, clearly state what sample-rate and bit-level you require. Also, indicate how aggressive you want them to do any maximisation. I would be good for you to try out some settings yourself so you are in the ballpark of what you like for your material. It is one area that can give you a real shock if you are not on the same page.

This is the part I’ve never really understood. Do you mean there’s a window of frequencies to which it is applied? Or does it work with all frequencies but at different “strengths” depending on the frequency? If the former, is it fixed or does it vary based on some RMS algorithm to find the most prominent frequencies at any given time? Etc.

With the iZotope Maximiser, you can set at which level (for all frequencies) it starts doing its stuff (Threshold), as well as the maximum level to which it works (Margin). The lower the Threshold, the greater the maximisation.

As for the internal mechanics of how it exactly does its stuff, other than being some sort of compressor-expander-brickwall limiter, I don’t know. There are several different algorithms, some of which are very CPU-intensive.

Well, the Cubase Maximizer plugin doesn’t give you any control over things, which I’m not a huge fan of at least not for something like this. So I started looking around and found out that Ferris TDS is considered to be a maximizer based on the way it operates. So I’m using one without knowing I’m using one. Heh. I still need to insert it post fader, however.

I wrote this long reply, and it didn’t get posted for some reason. Well, it was probably more than you wanted to know anyway.

I master right in Cubase right in the project. The PC is powerful enough enough that this works. I can, if I want, go tweak something in the mix that affects the mastering. If I’m making a CD, I use Wavelab, although no effects, just the organization and administration.

On my master bus, I first put a Cubase compressor, set for mild mastering compression. I tweak it to ensure that I don’t compress by more than 2-3 decibels. This is the first place where the instruments interact dynamically. In fact, I do this during mixing. Gives it a little cohesiveness.

On the second slot, I put an amp simulator to get a little warm distortion. I am using TLS_Saturated Driver which is free.

Then I go to Ozone. I used to use the Cubase plugins for this. What I am trying to do is: 1) equalize the whole thing 2) stereo widening, and maybe harmonic excitation 3) loudness 4) dithering (because I’m recording at 24 bit).

For equalizing, you could just use the Cubase 4-way equalizer. The aim is to compare your mix to your favorite reference mixes. Those are your favorite albums/recordings in your genre. For the stuff I do, which is fusion-y jazz stuff, I find that the low mids are always muddy and need to be reduced, and the highs (7000) need to be raised a couple of dB. Probably a consequence of using guitars and keyboards.

For stereo widening and harmonic excitation, Cubase has some plugins that I never really got to work well, but I do like what you get in Ozone. This is to taste, and basically, when you like it, dial it back a notch to be conservative.

For loudness, there are two things. Multi-band compression brings the volume up by frequency band, and can really have an effect on the equalization. This is a difficult one, and you have to balance your tweaks here with what you’ve done in the equalization stage. Some would say just skip this and effect the changes elsewhere. Cubase has MBC, and I would just say it’s difficult to understand and control. But I’d say that about Ozone too. I’m still learning.

The second loudness thing is a maximizer. Cubase has one, which I think works pretty well. The ones in Ozone are more sophisticated, I think. You really can’t tell what Cubase is doing, other than what you hear, but you have a better shot at seeing/hearing what Ozone is doing. If your mix had interesting dynamics beforehand, you should be able to make it louder (almost as loud as your reference recordings) without killing those dynamics. If your recording was very flat from beginning to end, neither will help.

Dithering is the standard last process, and you can read about it on the internets. Cubase gives you a plugin. You should be recording at 24 bits.

To check the final recording, I listen on all systems available. All the headphones you own, all the cars you own, laptops you own, oh, also your monitors. Always comparing with the reference recordings you love so much. I sit in my car and take notes.

To hear my results, listen to my album in my signature.

That has happened to me a few times. Very frustrating! If I know I am going to write a tome, or find that my post is becoming one, I will cut it all to Notepad, so I can at least see it all, but also to have an offline copy in case of loss while saving.

Sometimes when I have lost something, I just think “Who cares anyway”, and go do something else instead!

Yes, excellent idea! I really hadn’t intended it to be so long, neither the first time nor the second!

Pretty much what I do too these days. Having spent $$$ in the past on ‘pro-mastering’ and generally been unimpressed with the results (and I’ve been one who buys into 'Emporers New Clothes") I subsequently concluded that ‘mastering,’ generally is over-hyped and probably one of those things that is so often made out to be far more than it is or actually needs to be. Actually there was one occasion where I ended up RE-mastering some tracks myself because I thought the pro job had actually made the mixes sound worse!

And it certainly doesn’t make financial sense for most of us, certainly not me anyway, who are never likely to make any real kind of return on the investment anyway, let alone cover costs. So I’m all for DIY mastering, keeping it simple and real. And of course in this day and age of single digital downloads much of what mastering was originally about is no longer relevant to most anyway, though I myself must admit I did master my last album for CD distribution as well as digital. Listening on different systems is useful, particularly if one of those ‘systems’ includes earbuds!

Generally, I reckon if you’re happy with your mix then you may well not want to mess with it too much. It’s easy to get seduced by some mastering FX plugins but I so often found that under later review I usually preferred my original ‘untainted’ mix and therefore now tend just to focus primarily on optimising levels.

I always mix into some low-level compression and if it’s a single track (as opposed to album track) then I also do like to add a small amount of tape-sat across the whole mix (Magneto II these days), but apart from that… if there’s an issue I’ll always fix it in the mix. If I know the tracks will be part of an album (CD) then I often layout the finished mixes in a separate mastering project and mess with relative levels, tape sat etc there instead.


I just want to add that cubase’s mastering template is even better when you add the channel strips they do a great job. Great sound quality.
Especially the compressor I almost exclusively use it instead of a plugin insert… Always comparing my mixes to major label releases.

  1. Welcome back!

  2. How does Magneto II compare vs. Ferric TDS, assuming you’ve used the latter?

Larry, haven’t used Ferric TDS. Originally I used the old Magneto on most mixes, then later tried out the C7 channel strips tape sat and now of course Magneto II. One thing I find is differences generally are so subtle that for me it really doesn’t matter. I know that when I’m reviewing mixes in my car I’m not gonna know or even care which tape sat I used! :sunglasses:

As I think I mentioned in another thread my goal is to reduce my plugin footprint and the more good useable stuff that is bundled with Cubase itself the better. And the included FX/plugins etc generally are plenty good enough I reckon. Certainly with the releases over the last few years I’ve really been able to do away with a good number of 3rd-party plugins. I’m all for keeping it ‘lean and clean’ these days. :sunglasses:

Yeah for signal processing I don’t use too many 3rd party. The ones I use are from BlueCat (rarely), SoundToys, NI, Variety of Sound. And I will be starting to use FabFilter’s EQ in place of the BlueCat Triple EQ and the Cubase EQ VSTs as well as 4-band in channel EQ.

Right, Ian, I could have added that I’m not actually selling anything! I remember a thread over at Gearslutz where a guy put up the before and after of something he sent to a mastering house, which was no better than before, and the guys over there giving their best to make it “night and day” as they always do. As usual, the original poster didn’t have experienced enough ears, that was the problem!

Is Magneto II something that comes with Cubase 7? I am still on 6.0.7. I had the original Magneto with Cubase 4.5, but it sometimes gave me strange results.

I still find myself reaching for Ferric TDS, even though I like Magneto II a lot.

Magneto II comes with 7.5. Ferric TDS is from Variety of Sound, and it’s free.