Advice for mastering

[Dad3353]

1 hour ago, Ricky 4000 said:

 

I dunno, Daddy. I've never mastered anything in my life.

 

I linked to Warren Huart as he is a pucker producer / engineer and mix engineer.

 

And if you have a look, he interviews some top mastering engineers on his YT channel.

 

I don't doubt that he's a fine mix engineer; he doesn't come across, to me, as a fine pedagogue, s'all, to judge by this sole video. 

[skankdelvar]

I'm still very much at the 'consciously incompetent stage' when it comes to this stuff but here's my tuppence-worth

 

* I usually (but not always) embed a reference track alongside a sub-master bus (as described above). I'm not trying to get my track to sound exactly like the reference. The reference is there to let me hear if my mix is too bassy, muddy, thin, toppy, dynamically flat or spiky compared to a professionally recorded track.

Another approach is to have several reference tracks and to try to get your track into a space where it stands comparison to any of them in terms of clarity. You're not trying to get your track to sound the same - it never will. The reference track is there to give you a yardstick, no more

 

In Reaper it's easy to use mute / solo buttons to switch instantly between your track and the reference track. Remember to keep the ref track and your track at about the same volume. Crank the ref track fader up and down to equalise their volume - not your sub-master fader

 

* Pretty much most of my EQ-ing, compression, etc is done at the track level. The tracks feed into buses (drums, bass, guitars, keys, vocals) where I balance the individual tracks against each other. The buses might have an eq to tweak certain frequencies if necessary and touch of very light compression to glue the tracks together. At this point I'd stick a soft clipper at the end of each bus fx chain to increase the volume (if necessary) before the bus goes to the (sub) master bus.

 

* At the sub-master bus I'd usually have an effects chain in this order: Volume trim to tame or boost the incoming signal > EQ (for very minor tweaks) > Compressor (light)> Limiter (to control final loudness) > LUFS meter (to measure loudness, check for clipping). The sub-master bus goes to an effects-free master output bus for rendering.

 

* The idea is that all the big EQ, compression, reverb etc moves are made at the track level, most of the loudness is done at the bus level and the sub-master is there to control the big picture rather than add anything much.

 

* I've found that (i) small EQ and compression moves are better than big moves (ii) too many plug-ins spoil the broth (iii) limit the number of instruments playing at any one time. Mixing becomes exponentially more difficult the more you've got to mix.

 

Now, after all this guff, I should point out that @Dad3353 has a far simpler approach than me and uses far fewer plug-ins than I do, and his entries in the BC Composition Challenge always sound ten times better than mine. In the end, just watch as many vids as you can, take notes and see where they agree on certain things. Then fire up your DAW and try those things out.

 

Recommended plug-in: LoudMax brickwall limiter for the sub-master or Master bus, available for download here. It's good, it's easy and it's free.

 

Edited July 24, 2022 by skankdelvar

[lowdown]

23 minutes ago, skankdelvar said:

 

 

* I usually (but not always) embed a reference track alongside a sub-master bus (as described above). I'm not trying to get my track to sound exactly like the reference. The reference is there to let me hear if my mix is too bassy, muddy, thin, toppy, dynamically flat or spiky compared to a professionally recorded track.

Another approach is to have several reference tracks and to try to get your track into a space where it stands comparison to any of them in terms of clarity. You're not trying to get your track to sound the same - it never will. The reference track is there to give you a yardstick, no more

 

In Reaper it's easy to use mute / solo buttons to switch instantly between your track and the reference track. Remember to keep the ref track and your track at about the same volume. Crank the ref track fader up and down to equalise their volume - not your sub-master fader

 

* Pretty much most of my EQ-ing, compression, etc is done at the track level. The tracks feed into buses (drums, bass, guitars, keys, vocals) where I balance the individual tracks against each other. The buses might have an eq to tweak certain frequencies if necessary and touch of very light compression to glue the tracks together. At this point I'd stick a soft clipper at the end of each bus fx chain to increase the volume (if necessary) before the bus goes to the (sub) master bus.

 

* At the sub-master bus I'd usually have an effects chain in this order: Volume trim to tame or boost the incoming signal > EQ (for very minor tweaks) > Compressor (light)> Limiter (to control final loudness) > LUFS meter (to measure loudness, check for clipping). The sub-master bus goes to an effects-free master output bus for rendering.

 

* The idea is that all the big EQ, compression, reverb etc moves are made at the track level, most of the loudness is done at the bus level and the sub-master is there to control the big picture rather than add anything much.

 

* I've found that (i) small EQ and compression moves are better than big moves (ii) too many plug-ins spoil the broth (iii) limit the number of instruments playing at any one time. Mixing becomes exponentially more difficult the more you've got to mix.

 

 

 

 

 

That's pretty much how I work. On paper, it probably looks a lot more complicated than it really is. But, in practise, it really isn't.

I tend to work with various templates for different projects, but the above, is always some kind of starting point when putting busses and FX chains together.

[ambient]

I ran a workshop for a group of music students on mastering recently. You're welcome to my Keynote slides and the handout I produced for them. 

 

 

Edited July 24, 2022 by ambient

[Velarian]

Unlike @skankdelvar I'm only at the threshold of moving from unconsciously incompetent to the conscious stage so please feel free to disregard everything I have to say. However, in my so far limited acquisition of knowledge I've learned a little which might be useful to consider.

 

I'm currently in the process of trying to figure out the best way to remaster some old band recordings from the early 80s. The original 1/4" masters are long gone and I only have cassette copies to work from so the quality isn't great. Obviously I'm only able to tweak the whole mix rather than having the luxury of working at individual track level. First of all I'd echo all the comments above regarding listening to a reference track for comparison. Our original recording were specifically influenced by what we were listening to at the time and comparing similar music from that time (albeit professionally and expensively recorded) was useful in revealing where our recordings were lacking. The pro recordings were clearer and more open soundings with deeper, more solid bass. Obviously a cassette with dolby noise reduction is never going to sound as clear as a modern/professional recording. However, I felt that it ought to be possible to improve the bass end of things which, on our recordings, sounded muddy.

 

In my search for information I came across a YouTube channel called "In The Mix" which has loads of information about mixing and mastering and in particular I found the tutorial below on using a free plugin called SPAN which is a frequency analyser. This can be used at an individual track level and across a whole mix so it was relevant for me. Comparing the frequency graph of the reference track and the recording gives a visual reference to what's different. I could hear that the bass sounded muddy and this plugin enabled me to identify that there was a pronounced hump at 137Hz and a corresponding drop-off below that. I've been playing with EQ and reducing that hump and adding a little boost to lower frequencies to improve the depth has made a useful difference. Now I need to figure out if there is a way to make it sound clearer at the top end when there isn't much substance to work with. Any suggestions would be gratefully received.

 

 

 

[lowdown]

21 minutes ago, Velarian said:

Unlike @skankdelvar I'm only at the threshold of moving from unconsciously incompetent to the conscious stage so please feel free to disregard everything I have to say. However, in my so far limited acquisition of knowledge I've learned a little which might be useful to consider.

 

I'm currently in the process of trying to figure out the best way to remaster some old band recordings from the early 80s. The original 1/4" masters are long gone and I only have cassette copies to work from so the quality isn't great. Obviously I'm only able to tweak the whole mix rather than having the luxury of working at individual track level. First of all I'd echo all the comments above regarding listening to a reference track for comparison. Our original recording were specifically influenced by what we were listening to at the time and comparing similar music from that time (albeit professionally and expensively recorded) was useful in revealing where our recordings were lacking. The pro recordings were clearer and more open soundings with deeper, more solid bass. Obviously a cassette with dolby noise reduction is never going to sound as clear as a modern/professional recording. However, I felt that it ought to be possible to improve the bass end of things which, on our recordings, sounded muddy.

 

In my search for information I came across a YouTube channel called "In The Mix" which has loads of information about mixing and mastering and in particular I found the tutorial below on using a free plugin called SPAN which is a frequency analyser. This can be used at an individual track level and across a whole mix so it was relevant for me. Comparing the frequency graph of the reference track and the recording gives a visual reference to what's different. I could hear that the bass sounded muddy and this plugin enabled me to identify that there was a pronounced hump at 137Hz and a corresponding drop-off below that. I've been playing with EQ and reducing that hump and adding a little boost to lower frequencies to improve the depth has made a useful difference. Now I need to figure out if there is a way to make it sound clearer at the top end when there isn't much substance to work with. Any suggestions would be gratefully received.

 

 

 

  

 

Yes, a very good channel for tips etc.

I actually find what he is doing pleasing on my ears. 

 

 

[Dad3353]

16 minutes ago, Velarian said:

...Any suggestions would be gratefully received.

 

I have been doing the exact same process, bringing reel-to-reel tapes from over half a century ago of bands I played with, rehearsing, into the digital age, playing them into Reaper for preservation and, if possible, treatment. These, too, are stereo mixes for the most part (some mono...), so nothing can be done on the 'stem' level. The frequency analyser, along with just listening, gives some clue as to what can be done, but there's also a lot that just cannot. My unwelcome suggestion would be similar to that posted previously above; that's to say:  know when to stop. Once into the upper ranges, no distinction can be usefully made between Vox, keys, guitars, cymbals, noise... and pursuing perfection becomes a hunt for the Dahu, or Chimera. If it's as good as it gets, it is sometimes (often...) as good as it gets, and that's all. There are moments when a doctor has to decide when to pull the plug on a terminally-ill patient. That moment will be retarded as long as possible, but, eventually, the time of passing will be pronounced and a sheet pulled over the departed. It's sad, but that's Life.
Just as an anecdote, and no more cheerful, I had all my weeks of painstaking transfers of tapes on an external USB Teradisk. One day : nothing. The disk was dead. I even sent the disk to the specialists, who could, for a very hefty fee, recover info from dead media. To no avail. I signed up to pay in case of success; they tried, but failed to recover this disk. All, then has been lost, and I must dig out again my crumbling tapes and start again. Lesson..? Back-up, then back-up the back-up, if the stuff is precious. Digital media are fickle, and go 'Poof..!' when they choose to, not when you would like them to.
Good luck with your audio archaeology; hope this helps in some way.

 

Douglas

[skankdelvar]

53 minutes ago, Velarian said:

Now I need to figure out if there is a way to make it sound clearer at the top end when there isn't much substance to work with. Any suggestions would be gratefully received.

 

If there isn't much to be working with at the high end then it's difficult to add high end without using something like an exciter. I've never had much luck with those things

 

One school of thought is that reducing muddiness and taming some of the low mids creates the perception of more high end. If this were a normal project I'd say:

* High pass to cut everything hard below 50hz but do it on the tracks, not the master bus. That way, you can put some back in on the kick and bass if it sounds too thin

* Narrow cut around 140hz to reduce boxiness
* Cut toms, cymbals, pads hard below 300-500hz

 

But this is a master so all you can do is fiddle with EQ across the whole thing.

 

There's lots of tutorials out there which cover muddiness. Some of them recommend putting a dynamic eq across the master bus and gain reducing the muddy frequencies, fast attack, fast release so that you can reduce mud build-up only when it occurs.

 

Easy to use, free 4-band dynamic EQ and spectrum analyser: Tokyo Dawn Nova

Edited July 25, 2022 by skankdelvar

[Velarian]

17 minutes ago, Dad3353 said:

 

I have been doing the exact same process, bringing reel-to-reel tapes from over half a century ago of bands I played with, rehearsing, into the digital age, playing them into Reaper for preservation and, if possible, treatment. These, too, are stereo mixes for the most part (some mono...), so nothing can be done on the 'stem' level. The frequency analyser, along with just listening, gives some clue as to what can be done, but there's also a lot that just cannot. My unwelcome suggestion would be similar to that posted previously above; that's to say:  know when to stop. Once into the upper ranges, no distinction can be usefully made between Vox, keys, guitars, cymbals, noise... and pursuing perfection becomes a hunt for the Dahu, or Chimera. If it's as good as it gets, it is sometimes (often...) as good as it gets, and that's all. There are moments when a doctor has to decide when to pull the plug on a terminally-ill patient. That moment will be retarded as long as possible, but, eventually, the time of passing will be pronounced and a sheet pulled over the departed. It's sad, but that's Life.
Just as an anecdote, and no more cheerful, I had all my weeks of painstaking transfers of tapes on an external USB Teradisk. One day : nothing. The disk was dead. I even sent the disk to the specialists, who could, for a very hefty fee, recover info from dead media. To no avail. I signed up to pay in case of success; they tried, but failed to recover this disk. All, then has been lost, and I must dig out again my crumbling tapes and start again. Lesson..? Back-up, then back-up the back-up, if the stuff is precious. Digital media are fickle, and go 'Poof..!' when they choose to, not when you would like them to.
Good luck with your audio archaeology; hope this helps in some way.

 

Douglas

Thanks, all good advice, especially the knowing when to stop and backup points.

[Velarian]

9 minutes ago, skankdelvar said:

If there isn't much to be working with at the high end then it's difficult to add high end without using something like an exciter. I've never had much luck with those things

 

One school of thought is that reducing muddiness and taming some of the low mids creates the perception of more high end.

Without knowing anything about it, I did wonder if an exciter might offer a solution. Something to investigate. However your other point about the perception is interesting and I think is valid. 

[ambient]

32 minutes ago, Dad3353 said:

 

I have been doing the exact same process, bringing reel-to-reel tapes from over half a century ago of bands I played with, rehearsing, into the digital age, playing them into Reaper for preservation and, if possible, treatment. These, too, are stereo mixes for the most part (some mono...), so nothing can be done on the 'stem' level. The frequency analyser, along with just listening, gives some clue as to what can be done, but there's also a lot that just cannot. My unwelcome suggestion would be similar to that posted previously above; that's to say:  know when to stop. Once into the upper ranges, no distinction can be usefully made between Vox, keys, guitars, cymbals, noise... and pursuing perfection becomes a hunt for the Dahu, or Chimera. If it's as good as it gets, it is sometimes (often...) as good as it gets, and that's all. There are moments when a doctor has to decide when to pull the plug on a terminally-ill patient. That moment will be retarded as long as possible, but, eventually, the time of passing will be pronounced and a sheet pulled over the departed. It's sad, but that's Life.
Just as an anecdote, and no more cheerful, I had all my weeks of painstaking transfers of tapes on an external USB Teradisk. One day : nothing. The disk was dead. I even sent the disk to the specialists, who could, for a very hefty fee, recover info from dead media. To no avail. I signed up to pay in case of success; they tried, but failed to recover this disk. All, then has been lost, and I must dig out again my crumbling tapes and start again. Lesson..? Back-up, then back-up the back-up, if the stuff is precious. Digital media are fickle, and go 'Poof..!' when they choose to, not when you would like them to.
Good luck with your audio archaeology; hope this helps in some way.

 

Douglas

That’s a very good point there about backing up. I had a panic moment toward the end of my PhD work when I thought I’d lost the stuff I’d recorded over three very long years. Luckily it wasn’t lost. I learned my lesson though, and bought separate hard drive. I use iCloud too.

 

it must be fascinating listening back to your old material. There’s something about recorded sound that I think is more nostalgia inducing than anything else.

[skankdelvar]

Had another think about this re-mastering @Velarian's old cassettes.

 

EQ's obviously the way to go and a dynamic EQ / multiband compressor is a refinement of this approach.

 

Then I remembered Kenny Gioia's tutorial on frequency splitting, a method which allows you to apply EQ, compression (and other effects) to defined frequency bands with a hard cut-off between each band. This uses Reaper's stock frequency splitter VST

 

The end result is a group of effected tracks broken out by frequency band (e.g. Lo track, mid track, hi track)and which you can balance with your track faders. A further refinement would be to automate each frequency track (or track effects) so they become controllable across the duration of the song rather than being a static EQ profile from beginning to end.

 

Example below uses an acoustic guitar but the principle is applicable to a whole song:

 

 

[Dad3353]

1 minute ago, skankdelvar said:

Had another think about this re-mastering @Velarian's old cassettes.

 

EQ's obviously the way to go and a dynamic EQ / multiband compressor is a refinement of this approach.

 

Then I remembered Kenny Gioia's tutorial on frequency splitting, a method which allows you to apply EQ, compression (and other effects) to defined frequency bands with a hard cut-off between each band. This uses Reaper's stock frequency splitter VST

 

The end result is a group of effected tracks broken out by frequency band (e.g. Lo track, mid track, hi track)and which you can balance with your track faders. A further refinement would be to automate each frequency track (or track effects) so they become controllable across the duration of the song rather than being a static EQ profile from beginning to end.

 

Example below uses an acoustic guitar but the principle is applicable to a whole song:

 

Potential for a huge Rabbit hole down which to bolt, but an excellent concept. Thanks for sharing; I'll don a parachute and jump in when re-doing my old tapes. Good Stuff. 

[skankdelvar]

11 minutes ago, Dad3353 said:

a huge Rabbit hole down which to bolt

 

And that's before we start to think about even more recherche refinements (mid-side to push the bass to the centre, stereo-widening the high frequencies, that sort of thing).

 

Stock Reaper widener MDA Pseudo Stereo has option to widen using a comb filter rather than the usual Haas effect. Works really well and doesn't negatively affect mono picture like Haas does

[lowdown]

On 24/07/2022 at 10:36, skankdelvar said:

 

 

Recommended plug-in: LoudMax brickwall limiter for the sub-master or Master bus, available for download here. It's good, it's easy and it's free.

 

 

I have just given that a run out. It's low on resources and does the job well..!

 

👍

 

[Dad3353]

8 minutes ago, lowdown said:

 

I have just given that a run out. It's low on resources and does the job well..!

 

👍

 

 

I'm very wary of Loudness in general, and brick-walling even more so (except, of course, 'Unhalfbricking'..!). Possibly a result of the genres I'm most familiar with (no 'Disco' and very little 'EDM'...); rather more airy-fairy flutes and bongos for me, so... Still, if it's Good, it's Good. 

Edited July 25, 2022 by Dad3353

[skankdelvar]

33 minutes ago, Dad3353 said:

 

I'm very wary of Loudness in general, and brick-walling even more so (except, of course, 'Unhalfbricking'..!). Possibly a result of the genres I'm most familiar with (no 'Disco' and very little 'EDM'...); rather more airy-fairy flutes and bongos for me, so... Still, if it's Good, it's Good. 

 

I used to run everything into a limiter on the master bus and crank it to the max. Then I realised that one can retain more dynamics and reduce artifacts by getting it louder before it goes into the limiter. For me, the limiter's there to catch a few overs rather than to crush everything flat.

 

That said, I'm still consciously incompetent. There's so much I need to improve before I'll be anywhere near happy.

[Dad3353]

2 minutes ago, skankdelvar said:

 

I used to run everything into a limiter on the master bus and crank it to the max. Then I realised that one can retain more dynamics and reduce artifacts by getting it louder before it goes into the limiter. For me, the limiter's there to catch a few overs rather than to crush everything flat.

 

That said, I'm still consciously incompetent. There's so much I need to improve before I'll be anywhere near happy.

 

Yes; I think that the concept of Limiting appied much more in the Bad Old Analogue Tape days, when saturation was a potential problem, and in the Bad Old Days of 8-bit Digital, where any clipping resulted in horrific distortion. With modern 24-bit stuff, there's (to me...) no need to raise levels above the dreaded Noise Floor, as there is (to me...) none there. I record, and treat, at much lower levels than in the Dark Ages, so I'm always well away from any limiter's trigger point. When 'Normalising' stems, I aim for -1db, and never need to increase beyond that. It's been a very long time since I suffered clipping of any sort. That's just me, of course; those 'pushing the envelope' may well need such devices. No, I'm not looking at @Leonard Smalls (although sometimes I wonder  ...).

[Leonard Smalls]

18 minutes ago, Dad3353 said:

No, I'm not looking at @Leonard Smalls (although sometimes I wonder  ...).

 

I'm fighting against my old BBC training, where 1KHz reference tone on a digital recording was set at -18dBU...

So I tend to record all instruments at peak of around -6, then smear a bit of neutron 3 on it so RMS level is between -15 and -10.

Then at the master I use either a Shadow Hills Compressor plus Waves limiter set at absolute peak of -0.5, or recently I've been using Ozone 9 with a Greg Colibri preset such as "high detail wide", then turn down compression slightly as this would otherwise give a dynamic range of only about 3dB! However, when producing a masterpiece such as this month's CompoCompo piece you want it to be as loud and unpleasant as possible - mainly to benefit those of us of a more delicate persuasion (eh Doug? 😁)

[skankdelvar]

26 minutes ago, Dad3353 said:

when saturation was a potential problem

 

Nowadays people (including myself on occasions) slap a saturation plug-in across a track. There are some benefits to this, one of them being an increase in perceived loudness without increasing gain. Other people think that it helps to add 'vintage mojo' but I'm not so sure.

[PaulThePlug]

Real Interesting Read... Is this for a TDK C90 Normal or the Chrome? Dolby NR button in... Yes?

Edited July 25, 2022 by PaulThePlug

[WinterMute]

This is a can of worms...

 

I'm not a Mastering Engineer, I'm a tracking and mixing engineer and these are fundamentally different things.

 

I do "master" my mixes, but only to ensure that the loudness profile is consistent with the delivery format, which avoids the algorithms smashing the dynamics flat on Apple Music or Spotify. All of the streaming services have different standards when it comes to loudness, and thats a bit of a pain, but I've found the mastering to -14LUFS seems to work reasonably well for most things, with a peak of around -2db. For CD mastering I go with -12LUFS ad 0bd peak. I use the Nugen Visualiser package that includes the very useful True Peak limier and the Mastercheck pro plug-ins.

 

In order to hit those loudness targets, I will use a limiter in the output stage, which is paralleled to a feed from my summing mixer that passes through an AMEK 9098 comp/lim. I use the UAD2 Oxford limiter for the digital path.

 

I'm of the opinion that if you're doing a lot of work in the mastering stage on dynamic control and EQ, then you haven't finished mixing yet.

Edited July 25, 2022 by WinterMute
usual typos

[skankdelvar]

Just in case anyone wants to try to replicate that thing we used to do where you played a cassette encoded with Dolby B but switched the decoder off to make everything sound brighter:

 

From Wikipedia: "The Dolby B system is effective from approximately 1 kHz upwards; the noise reduction that is provided is 3 dB at 600 Hz, 6 dB at 1.2 kHz, 8 dB at 2.4 kHz, and 10 dB at 5 kHz".

 

Flipping that on its head one could fire up the DAW and try boosting one's track by 3db at 600hz, 6 dB at 1.2 kHz, 8 dB at 2.4 kHz, and 10 dB at 5 kHz.

 

It might sound weird 

[wishface]

On 24/07/2022 at 11:40, ambient said:

I ran a workshop for a group of music students on mastering recently. You're welcome to my Keynote slides and the handout I produced for them. 

 

 

Sounds great thanks

[meterman]

I’d love to be able to do my own mastering, I feel like I should understand how to do it. My first ‘proper’ job was as a tape op, then I trained as a recording engineer but packed it in after a couple of years as the round-the-clock hours were doing me in. But I ought to have a few trial goes at it, as I always get my tunes mastered by a pro if they’re being released, and the costs mount up over the year.

 

My hearing is pretty bad though (around 30% in one ear and 55% in the other) so I’m not sure if I’d be up to the task. Might try on headphones. It just seems more scientific than I have the brains for 😂