Mastering your tunes

  On 6/24/2020 at 12:46 PM, Nil said:

Believe it or not, I spent a lot of time trying to write the most constructive post I could, and not to eventually come across as a total asshole. I'd should have phrased it differently now that I've re-read it, mea culpa. 

It can also be read as my double failure to achieve something worth sharing.

I'm by no mean criticizing the tune itself (it's pretty cool actually), but if I were asked to master that tune I'd suggest the artist to improve the drum mix so that in the end the track sounds as good as it can. It'll always sound so much better when fixed in the mix, no matter who masters it, and in my opinion here it's an easy, identifiable and quick fix.

It was easy for me to identify issues as I haven't spend the (emotional) effort, nor the time / energy / involvement writing and producing the tune. And because I know how reliable are my monitors. Not because I'm on par with Beau Thomas and Mandy Parnell (guess what, I'm definitely not :D)

The total lack of context / reference confused me too. From my experience, and from what I've witnessed / been told / read etc... it's much more effective to pick a 15/30 seconds long part of the track that sums what the tune is about (sonically and musically), and to spend most time working on that part. Here I have no clue what the entire track can be like 

I'd rather be suggested modifications that could improve my track (and learn something / improve my skills in the process) rather than have my work mastered no matter what, for a sub-par result (or, if you prefer, a result that wouldn't meet my expectations). Actually it happened to one my tracks : the whole mini-LP mastering sessions went super smooth, then for the final track I honestly did a terrible, terrible, terrible job when mixing it. The mastering engineer did it best, but in the end suggested a few alterations. I came back at my place, reworked the tune, came back the next day, let the ME validate the revision (or not).  He did, and he did a great job mastering it.

Whenever possible, I'll now book a listening session with my mastering engineer a week or two before the mastering session, so that I have time to fix whatever issues (s)he hears, and to do so directly in the mix. It'll always sound so much better.

Expand  

Don't get me wrong, I absolutely like it that you criticized the mix from your perspective. And it was definitely constructive.

  On 6/23/2020 at 8:39 PM, Nil said:

Sorry but I struggle to hear a finished mix here, nor something reliably representative of the full tune : it's impossible to tell if the track will remain beat-driven or if melodies will be added later ? In both cases, the tune will require different treatment.

The main issue here are the drums IMO : they are quite saturated, yet too spiky / too dynamic, which prevents transparent and / or musical additional, competitive loudness. There's not much hierarchy in the drum kit, so everything stands in the foreground, upfront, and consequently EQing that mix is super challenging. I bet fixing that kick drum would already makes things a lot better though.

The raw esthetic isn't the issue here though,  if the drums were more erm contained (still talking about the mix, not the drum programming)  it would be a whole different story.

I gave it an honest try, twice... and trashed the 2 attempts. The masters were better than the mix, but I wasn't satisfied. I'm not sure the artist would have been either TBH.

What worked : 
- low-ratio compression gave it some much needed density (but didn't address peaks fully, wasn't the point here anyway) and overall movement
- along some low-end remodeling, a relatively narrowish dB boost at ~ 160 Hz (in the mids, as in M/S) made the bassline fuller and more pleasant (it's not solid enough in the mix (to my ears), especially for such a drill n bass, post-jungle track), but also brought up some nastiness in the kick / drums. 
- a super gentle / broad, stereo, 1dB cut ~ 900 Hz pocketed the snares and opened up the mix nicely
- it was relatively easy to make the tune wider without losing its focus / center solidity (and without compromising mono-compatibility either).

What didn't work :
- various degrees / sorts of saturations to tame peaks. It raised the RMS, but wasn't convincing enough.
- I gave both clipping and limiting a try, and nothing sounded good enough either. I prefer to limit/clip last, once I've got nice balance / density with EQ/comp, it's always worked better for me that.

Maybe I should have tried to raise the level first, and then to EQ etc... into the limiter / clipper.

I hope that post comes out in a positive, constructive way.

PS : the main loop was roughly at -18 LU (peaks at -1,8 dB FS), raised it to a bit more than -10 (peaks at -.5dB FS, true peak), tops. I suppose I could get much cleaner results if ignoring loudness imperatives.

Expand  

a lot of that made perfect sense, i think its not entirely useful to have such a short clip to reference as that whole track is like..6 mins long with a fair bit of dynamic range between the sections. plus, the rest of the ep was taken into consideration when mastering this no doubt, so all in all it was more for demonstration about dynamic range control between that quieter section and that loud bit which is probably the loudest part of the whole record. (whole track is here). (and it was mastered for vinyl so im sure there were some limitations goin on there too)

also, about the mix, yeah it wasnt great. the way the track was produced was a bit stupid, each drum sample hit had its own channel in ableton and there was a messy web of groups and sends. mixed down on an SSL AWS desk got it much tighter but still, yeah needed a proper sort out and frankly i didnt feel like putting another string of hours into what was a overwhelming project. some would say "it had that old skool charm of rough mixed" ..or something like that 

the track has some really quiet moments, as did the EP as a whole, so I do think the engineer did a really good job finding a middleground between making it more punchy and controlled without destroying the quieter, more gentle sections. Would be really interested to hear what a different engineer wouldve done to it. maybe in the future when i got the cash to burn it would be interesting sending one track to a bunch of different engineers and seein what they all do different.

  On 6/24/2020 at 4:25 AM, snack master said:

thanks for sharing @mause

no probs! I could have a dig thru my shit and post another if you guys are interested  maybe a full track this time

  On 6/24/2020 at 12:46 PM, Nil said:

Believe it or not, I spent a lot of time trying to write the most constructive post I could, and not to eventually come across as a total asshole. I'd should have phrased it differently now that I've re-read it, mea culpa. 

It can also be read as my double failure to achieve something worth sharing.

I'm by no mean criticizing the tune itself (it's pretty cool actually), but if I were asked to master that tune I'd suggest the artist to improve the drum mix so that in the end the track sounds as good as it can. It'll always sound so much better when fixed in the mix, no matter who masters it, and in my opinion here it's an easy, identifiable and quick fix.

It was easy for me to identify issues as I haven't spend the (emotional) effort, nor the time / energy / involvement writing and producing the tune. And because I know how reliable are my monitors. Not because I'm on par with Beau Thomas and Mandy Parnell (guess what, I'm definitely not :D)

The total lack of context / reference confused me too. From my experience, and from what I've witnessed / been told / read etc... it's much more effective to pick a 15/30 seconds long part of the track that sums what the tune is about (sonically and musically), and to spend most time working on that part. Here I have no clue what the entire track can be like 

I'd rather be suggested modifications that could improve my track (and learn something / improve my skills in the process) rather than have my work mastered no matter what, for a sub-par result (or, if you prefer, a result that wouldn't meet my expectations). Actually it happened to one my tracks : the whole mini-LP mastering sessions went super smooth, then for the final track I honestly did a terrible, terrible, terrible job when mixing it. The mastering engineer did it best, but in the end suggested a few alterations. I came back at my place, reworked the tune, came back the next day, let the ME validate the revision (or not).  He did, and he did a great job mastering it.

Whenever possible, I'll now book a listening session with my mastering engineer a week or two before the mastering session, so that I have time to fix whatever issues (s)he hears, and to do so directly in the mix. It'll always sound so much better.

Expand  

it all checked out mate, its very interesting to hear what a different engineer would do on the track. I think also having that feedback from the engineer is really constructive too, music's a learning game and it fuckin rewarding to learn another perspective on something you just spent so long exclusively hearing and deafening yourself to. Fresh ears and all that!

god i suck at watmm formatting posts, didnt mean to double post there

I'm glad it's all cool then  The ambient intro of your tunes is super lush by the way. If going for an old-school esthetic, I think some sorts of buss processing would definitely work : as if all the drums were coming out of the same sampler / compressor. Genuinely curious : how did you saturate those drums ? What keeps on puzzling me is that usually, saturation helps reducing crest factor between peaks and body of the sound while here it sounds like it make the drum buss more dynamic ?!

SSL AWS you say ? Nice !!! 

  On 6/24/2020 at 4:46 PM, Nil said:

I'm glad it's all cool then  The ambient intro of your tunes is super lush by the way. If going for an old-school esthetic, I think some sorts of buss processing would definitely work : as if all the drums were coming out of the same sampler / compressor. Genuinely curious : how did you saturate those drums ? What keeps on puzzling me is that usually, saturation helps reducing crest factor between peaks and body of the sound while here it sounds like it make the drum buss more dynamic ?!

SSL AWS you say ? Nice !!! 

yeah it wouldve been ultra old skool to buss the drums like that, but with the amount of different channels going on it would be an absolute nightmare to keep them all under control

i'll have to have a look at the project again to give a proper answer to the saturation question but i think it was just a lot of units applying very slight software saturation in the chain, maybe on a parallel buss with some gating so the transients weren't totally smushed out  

yeah was super lucky to have access to it, did a lot of mixdowns on it fuckin lovely console. only regret is not using the bus comp more, that things a beaut!

Someone's welcome to take a punt at this if they fancy it, I'm happy with the track and the mix but it is what it is, i.e. bedroom/hobbyist level. I've always been shy of compression, just beginning to learn my way around it so my stuff's always on the quiet side.

Also, incredibly basic-arsed question, but when you're sending stuff to a M.E. do you generally send a finished track or do you send them the stems?

  On 6/24/2020 at 9:33 PM, Leon Sumbitches said:

Someone's welcome to take a punt at this if they fancy it, I'm happy with the track and the mix but it is what it is, i.e. bedroom/hobbyist level. I've always been shy of compression, just beginning to learn my way around it so my stuff's always on the quiet side.

Also, incredibly basic-arsed question, but when you're sending stuff to a M.E. do you generally send a finished track or do you send them the stems?

 

 

Your browser does not support the HTML5 audio tag.

 

 

 

 

 

 

Letter.wav 71 MB · 3 downloads

 

Expand  

this is the kind of process i'll do to a track before uploading to soundcloud, its not a pro master by any means but gets the whole track sitting around -12dB LUFs. i think i mightve overcooked the eq a bit there too..

usually, you'll just send a single bounce with a bit of headroom..i think some engineers might do it with stems but that would probably cost so much more to get

Have you tried to limit yourself to one instance of each processor ? Often I get much, much, much better results that way.

I've also found that most of the times, compression > EQ works perfectly : the comp to control dynamics, density and movement (as well as some subtle tone shaping if going for a colored), then EQ to rebalance / fix spectrum (also stereo image if working in M/S) and finally limiting / clipping. Saturation, when needed, can go anywhere in the chain, it really depends on what it's supposed to address / enhance : for instance, before a compressor to tame peaks and ease the compressor, or after an EQ so that they complement each other, or to shave peaks before a limiter.

-12LUFS is a healthy loudness target, if only it was standard! I can't wait for Soundcloud to implement loudness normalisation like most streaming services, I've had to push masters into absurd loudness just for SC way too often, to rival all the other squashed to death DIY masters on that platform  

  On 6/24/2020 at 10:47 PM, Nil said:

Have you tried to limit yourself to one instance of each processor ? Often I get much, much, much better results that way.

I've also found that most of the times, compression > EQ works perfectly : the comp to control dynamics, density and movement (as well as some subtle tone shaping if going for a colored), then EQ to rebalance / fix spectrum (also stereo image if working in M/S) and finally limiting / clipping. Saturation, when needed, can go anywhere in the chain, it really depends on what it's supposed to address / enhance : for instance, before a compressor to tame peaks and ease the compressor, or after an EQ so that they complement each other, or to shave peaks before a limiter.

-12LUFS is a healthy loudness target, if only it was standard! I can't wait for Soundcloud to implement loudness normalisation like most streaming services, I've had to push masters into absurd loudness just for SC way too often, to rival all the other squashed to death DIY masters on that platform  

Expand  

yeah i have tried limiting to one process but ive found doing incremental processing works quite well for me personally. I think aswell it depends on what processor your using, ableton's compressors tend to get out of control if you go too heavy on them hence having a couple in the chain to gradually build up with like that but with something like ozone or those real nice waves ones you could defo do it with just one. compression > eq is good sometimes too, ive done it on a couple tracks but as a producer ive got more experience using the eq first so its more comfortable, what is music production without comfort zones

interesting stuff about saturation there, i'll give it a go..i usually just ad tiny amounts of saturation to beef up the sound not usually to affect the compressor in that way..to control peaks ive always just ued the attack on the comp personally, lots of good bits to try in there tho will give em a go!

-12 seems to be a pretty good standard for streaming, i think its close to what sc uses but they change it every now and then i think..im sure theres a detailed specs on sc's algorithm somewhere..never really looked for it just made sure my tracks are somewhat consistently dynamic so they dont go super quiet

  On 6/24/2020 at 9:53 PM, mause said:

this is the kind of process i'll do to a track before uploading to soundcloud, its not a pro master by any means but gets the whole track sitting around -12dB LUFs. i think i mightve overcooked the eq a bit there too..

usually, you'll just send a single bounce with a bit of headroom..i think some engineers might do it with stems but that would probably cost so much more to get

 

Your browser does not support the HTML5 audio tag.

 

 

 

letter_mauster.wav 65.23 MB · 102 downloads

 

Expand  

Cheers for this man, great to get a look under the hood at someone else's approach!

@Leon Sumbitches : I gave it a quick try too. Whenever possible, I favor keeping the tone as close as possible to the mix (while fixing / enhancing what's needed / if needed). I'm not sure louder would make more sense here (and as you can see, streaming platforms would play it without altering its overall volume). I suppose the volume could be pushed further though, if really needed.

Chain was Tone Project Unisum (compressor) > Crave EQ > Leapwing StageOne (a bit of stereo enhancing > Pro-L2 > Airwindows Dark (dithering).

Man, that sounds great, it polishes it up and brings out the mix elements nicely- thank you for taking the time to give it some love. You know your trade!

I just listened to your tune (pretty nice one by the way) and (hopefully) reacted accordingly with the sound processors at my disposal 

Compression to emphasize / lock movement (no need for more than 2:1 ration, Unisum is a wonderful processor I must say), EQ to balance the spectrum and to make the music pop, a lil bit of stereo imaging to complement what was done previously and finally gentle limiting to shave the peaks. It wouldn't have fit here IMO to process the mix too much.

Glad you like it !

The one headache I always run into with mastering my own stuff is: I'll get the mix sounding awesome and huge, but then when I boost it in mastering to get it at a level comparable with other music, the bass frequencies distort and sound like shit.  Then I end up experimenting with the mix until I can make it loud with minimal distortion occurring.  I usually succeed eventually, but I'm sure compromises have been made to get it there most times.  Too much limiting softens things too much, and kills the dynamics, or makes it sound squishy and gross.  Anyway... this part always kills the fun.  Anyone else run into this problem time and time again?  

There are many albums by artists I like that are plagued by the unpleasant distort/squishy phenomena that I can't stomach on my own masters.  So... clearly there are pro mastering engineers who find this sort of thing acceptable.  I really can't imagine it being desired by the artist/producer.

Sometimes I just run the whole mix through an Alesis Microlimiter.

  On 7/27/2021 at 7:59 PM, Zephyr_Nova said:

The one headache I always run into with mastering my own stuff is: I'll get the mix sounding awesome and huge, but then when I boost it in mastering to get it at a level comparable with other music, the bass frequencies distort and sound like shit.  Then I end up experimenting with the mix until I can make it loud with minimal distortion occurring.  I usually succeed eventually, but I'm sure compromises have been made to get it there most times.  Too much limiting softens things too much, and kills the dynamics, or makes it sound squishy and gross.  Anyway... this part always kills the fun.  Anyone else run into this problem time and time again?  

There are many albums by artists I like that are plagued by the unpleasant distort/squishy phenomena that I can't stomach on my own masters.  So... clearly there are pro mastering engineers who find this sort of thing acceptable.  I really can't imagine it being desired by the artist/producer.

Expand  

Thats the importance of leaving headroom for the master, keep like..6db headroom then the processing doesnt totally dilloute the dynamic range

Even with the headroom it's still an issue (getting it as loud as everything else with all the meaty goodness clean and intact).

Mastering engineers are going as loud as the clients ask them too, simple as that.

@Zephyr_Nova which kind of loudness are you aiming for ? Loudness is so material dependant, some arrangements/mixes can be pushed while others will collapse when you go past a couple of dB of limiting. And limiting is just a part of the equation (but you already know that)

If your mix sounds awesome and huge (I’d love to hear), then there’s no reason it shouldn’t sound equally as good once mastered.

What’s your workflow ?

  On 7/28/2021 at 1:31 AM, mause said:

Thats the importance of leaving headroom for the master, keep like..6db headroom then the processing doesnt totally dilloute the dynamic range

As long as the mix isn’t clipped, it’s trivial to recreate headroom.

I’m hearing my monitors pop on the low frequencies in some of my mixes, it’s not clipping in the waveform, what am I doing wrong? I’m guessing I need compression or a low freq filter. 

I stay away from visual analyser as much as possible, though have you noticed anything weird running these mixes through SPAN or similar piece if software ? 

  On 7/28/2021 at 10:12 AM, Nil said:

Mastering engineers are going as loud as the clients ask them too, simple as that.

@Zephyr_Nova which kind of loudness are you aiming for ? Loudness is so material dependant, some arrangements/mixes can be pushed while others will collapse when you go past a couple of dB of limiting. And limiting is just a part of the equation (but you already know that)

If your mix sounds awesome and huge (I’d love to hear), then there’s no reason it shouldn’t sound equally as good once mastered.

What’s your workflow ?

Expand  

Thanks man.

I'm aiming for roughly the same loudness as all the other tracks I've released in this project, which is just loud enough that they don't sound wimpy in a Spotify playlist with commercial pop stuff (so, loud but not obnoxious).  This one may share an EP with other cover songs I've done.  The song in question is a Justin Timberlake cover, of all things.  (Th other covers so far are a Mr. Bungle song and an Ennio Morricone song.)  Workflow is Ableton for recording/mixing, T-Racks for mastering.  I've already messed with the mix a bit to accommodate the mastering problem, so I'll make sure I didn't totally ruin it before posting.

  On 7/28/2021 at 10:12 AM, Nil said:

Mastering engineers are going as loud as the clients ask them too, simple as that.

@Zephyr_Nova which kind of loudness are you aiming for ? Loudness is so material dependant, some arrangements/mixes can be pushed while others will collapse when you go past a couple of dB of limiting. And limiting is just a part of the equation (but you already know that)

If your mix sounds awesome and huge (I’d love to hear), then there’s no reason it shouldn’t sound equally as good once mastered.

What’s your workflow ?

Expand  

Here is the unmastered mix:

And here's a previous cover I did that sort of set the goal post for what I'm aiming for in terms of a master:

https://themunicipality.bandcamp.com/track/retrovertigo

Thanks for your input - much appreciated.

Amazing tune ! I haven’t listened to it on my monitors yet, will do ASAP.

@Nil yeah good recomendations about mastering , I would advise the same

(not being a professional mastering engineeer but your talk make sense to me)

about loudness war : yeah also tired of  trying to push it hard so that it don't seem weak for the listeners in a random playlist of their own)

but I've made my path in thinking, people who are not dumb (say, not deaf) would think about just raise thier listening volume up if they need to get more volume from a song they like.

I mean, what's the point mastering hard / limiting hard, and then so that everybody won't turn up their system more than 20% of max volume because otherwise it's too loud when listening.... tha'ts what headroom is for , is'nt it

well, ok, i'm old school, now with most of people listening on phones (edit : i mean smartphones, or portable usb speakers), with no much output power, wou might need to smash it hard so it gets a bit of something...

but I let it to the listeners, if you want some quality and high volume you just need a decent sound system (at least a pair of speakers, not a phone or laptop speakers)

in a way, that's an industrial / commercial disease, so to say...

also it reminds me of this interesting "labelling" project about loudness in music, 

https://www.kvraudio.com/product/tt-dynamic-range-meter-by-pleasurize-music-foundation

I used to use this as a reference.

like, education of listeners is the key to proper sounding musical outputs

2nd edit : not going to say "everyone spend a lot of money on speakers and amp and so on.... "we don't need lots of money to this:

just listen on a pair of speakers or a dedicated headphone, even cheap ones.

we don't  need trying to master on the hardest rms as we can. if someone like your music, he / she will love it however the rms level is, he / she would just pull the volume up. if he / she don't like it, he won't like it more with more rms level, they would just skip the track so...

  loudness war has sto stop, everyone should be able to choose their own listening loudness, I mean.

ha ha ha I might sound  so "old-school-wisdom-talker" haha ... but still, that's the whole point, I know most of you here will understand

3rd edit : also i don't mean mastering is useless. adapted mastering is usefull to get a decent listen on various output device,

and to bring glue / shine on your products. but the process of mastering have to be ... mastered.