How To Benchmark Your Mix?

Hey everybody,

I know that mastering and sound engineering is an art just like making music.
But, you must agree that there is a lot of science in mastering, and I am looking to gain better skills in this regard.

I do not care much for the mastering articles and tutorials out there - they are boring me to death if I can be completely honest here.

So I am now attempting to benchmark my track against commercial tracks and I want to do it by two means:

  1. Listening to my track, and to the commercial track which I think sounds good
  2. Comparing the frequency curves of both the commercial track and my track at certain points.

Does anyone have an idea about how to implement the second option?
I wish that I could “freeze a copy” of the curve, so that I could have played the mp3 (as a sample in Renoise), freeze the spectrum, then play my pattern and see how close it is to that curve.

Thoughts?

Not sure about with Renoise alone, but in Izotope Ozone’s EQ, you can take a snapshot of a freq curve from another track and use it as an overlay to the eq

Yeah, thought about using it for the job.
I really hate the mastering stage.
I wish there was a plugin where you can input a benchmark MP3 and it will adjust your frequencies accordingly…

I thought Har-Bal did someting like that:
http://www.har-bal.com/

I honestly don’t see a point in changing your track’s EQ to be the same as another track… each track has different dynamics, therefore each track should have it’s own unique curve. All you’re going to end up doing is stifling your track’s dynamics.

Your thoughts are very appreciated!
Yes, Har-Bal looks like a very good direction indeed. By the short video demo, it looks like a standalone and not VST, but I guess it is not that bad.
So far, I am doing mastering only inside Renoise, but I am going to see if they have a demo.

Do you have any experience with it?

This is definitely nice and thanks for sharing.

You are absolutely right.
I do not want to make my track sound exactly like something else, but I want to have a reference - both by listening and by visuals.

I don’t think you would need Harbal as a VST inside Renoise.
At least if you want to cross reference with another similar type of song and then crank up the dynamics you can easily do that in Harbal. The only thing you can’t do there is the mixing, but you wanted a tool for mastering anyway.

Skip to the last line for the pearl of wisdom amongst my waffling.

I use the spectograph-wotsit to help visualize any problem areas in a mix. Particularly with individual instruments solo’d.
Its particularly useful to see which tracks are muddying up the bass end.
And to throw in my opinion on the matter of eq pattern cloning, now is probably the best time in the history of recordings not to sound commercial, cos the big labels aren’t doing that well nowadays are they?

OK… now then, a pint of the finest contintental lagers worth mixing/mastering tips.

With nearfield stereo monitor speakers, your head and speakers should make an equalateral triangle. You ears should be in the same plane as the speakers… or angle your speakers up or down to point at your ears. or raise your chair/stand up/sit down get speaker brackets… whatever.

Dont master with tired ears. Also dont mix down with tired ears.
Do your final mixdown on a day when you haven’t been composing. I appreciate volume helps get you into the groove. This tires your ears.
If you’re working on a project of numerous tracks you really ought to master them together. Not weeks apart in between writing for the project. This will make your life easier when making decisions for each track. Master tracks at the end.

When mixing/mastering, you’re doing it too loud if you cant converse with someone in the same room. Your ears will tire quickly and then you are wasting your time. You’re probably damaging your ears too.
The ears perception of bass, mids and trebles change the louder you monitor. I might have it the wrong way round- but im quite sure your ears become less sensitive to bass at higher volumes, -of course experience can correct this phenonomen.

When your at a concert/festival, look at the well paid sound engineer- he will be double checking things sound ok through his/her expensive soundproof monitoring headphones. Headphones are also great late at night.
Unless your vey lucky, your room will be full of reflections and standing waves. Louder volumes make louder reflections and at certain energy levels walls and furniture will start to resonate creating more noises to confuse you.
These noises and standing waves will create holes or increases of volume at certain frequencies in your room. (refer to headphones)
Bass traps and sound absorbing panels make a real difference to problem rooms.
Make one cheaply from a rockwool panel covered in cloth so its not itchy. A wooden frame around it will help it survive being moved.
Put it on the wall opposite your speakers.
(When tracking vocals, stand in front of this panel facing away from it- for double value for money).

The mastering stage generally requires a touch of eq, some compression/limiting and maybe narrow the stereo field in the low end. -although this narrowing is better done within the mix rather than across the final mix (-artifacts).
Dry minimalist mixes may benefit from a sprinkle of best quality Reverb across the Whole mix.
Also… reverbs can mud up the mix. Make sure you keep an ear on any reverbs’ bass end. (put a filter or eq on the reverbs buss)

I stear clear of using multiband compressors and enhancers across a whole mix, unless you have a very expensive wonder-master-finalizer-pro MkII © The rest of them just sound good to tired ears by exciting certain frequencies.
(EQing/multiband compression or filters can create a phase shift. you generally want very little phase shift across a mixdown)
Use multiband compressors/enhancers on buss groups or individual instruments where necessary.
(if you use on across the whole mix, being sparing with its settings will be alright)
If you have excessive stereo bass/kicks and your tune ever makes it onto vinyl the needle will jump out the groove… unless the cutting house adds even more mastering. Also a panned kick in a night club with a central sub-woofer may sound rather shit… you’ll hear the kicks attack to the left or right, and then a delayed whoompf in the middle.

If you need to massively change the EQ at mastering stage, you fucked up when mixing. Probably cos you monitored to loudly or your ears were tired. -Try mixing it again another day.

Once you have mastered, double check that it sounds ok through a teenagers’ mobile-phone loudspeaker setting.

However, if you got something sounding good, i’d fuck the rules/guides and go with it.
And Probably best you dont take it all too seriously cos the experts tend to be boring. :blink:

I may have missed one or two things.
Of course most guides and rules of thumb can be absorbed or ignored but…

------------don’t monitor to loudly------------ (apologies for not really answering the question)

The entire thing was worth reading even if just for the above comforting thought… :)

Hey, thanks for taking the time, even if it does not answer - it is nice to hear thoughts from the more experienced.
I do not think vinyl, and definitely not a club, I am still fighting to get a good 192kbps MP3 master…

I think I am able to get some above average (or at least above total crap…) results, but the thing is I dont feel I am doing it scientifically. EQ a little here, compress a little there, check the output wave in a wav editor, do it all over again about 5 times, add exciters, remove reverbs, and eventually, I am half satisfied.

So far, comparing the result with a commercial track is a good way for me - and Harbal looks perfect for the job - since I do not have golden ears, so what sounds good to me, doesnt necessarily have a commercially standard sound - I need the references.

Oh well. :)

Getting to know your speakers takes time. so yeah, listening to comercial tracks on them is good.
I lent mine to a friend whilst another friend who is refurbing his house lent me his… silly idea.

As an exercise/experiment suggestion.

(without using any mastering effects across your main outputs)

Take a loop from a tune you like…
Drop it in renoise, and set the tempo so its looping correctly.
Make your own pattern, (drums, bass, perhaps a riff too) and chain the two together… and mix your pattern it sits well with the loop.

It will give you an instant guide to the kind of levels youll need for individual parts within the context of the loops finished mix.

  • although there are text book levels for mixing rock music, i don’t think they work so well for electronica, But no harm in trying them, from a commercial stance theyll be similar. (im assuming your making dance music)

Your can then submix your pattern tracks to a buss with EQ, a compressor and a limiter on it once you’ve got it sounding somewhere near. The EQ will have quite a different effect depending on whether you put it before or after the compressor.
You can now try mastering your bit compared to the loop.

Also… another tip. The punchy part of a Kick is usaully somewhere around 80 to 90 hz.
Cutting some of that out of your bassline helps if your struggling to get your kick sounding right.
Conventional Kick and bass tend to clash. Look into sidechain compression for this issue, its better than just EQing the bass.

Have fun,
Chuffed to be of help.

Oh… i checked your tunes, i think they a rather cool, well mixed and if anything…
perhaps some of the reverb tails (on the bass line) boom to much? such as on hypocrisy? assuming i followed your links correctly from profile. Or maybe im mixing bass light on my friends speakers :confused:

Thanks buddy.
Yeah, You listened correctly :)
I tend to be trigger happy on the reverb, I know, working on that… its a sort of an addiction.
Specifically, Hyposcrisy was made before I knew anything about mastering (not that I consider myself knowing much now) - but I definitely knew less then.

Your loop suggestion may be good to try.
In fact, I did almost that - I cut a loop from a song I like, paster it in one pattern, then pasted one of my busy patterns after it and let the two play repeatedly while I adjust the EQs on my pattern.
Had to put command switches to turn the EQ off whenever I enter the benchmark pattern, but the process is managable.

Other ‘sonic map’ to use as a guide is Pink Noise. It has a pleasant yet ‘full’ intensity that you can theoretically listen to for a long time without fatigue.

For the record, I master using both loud and soft techniques. Both reveal important things.

EQ at the mix and mastering level should only be used reductively, especially in filtering out muddiness and harshness in the mids. Never to boost the tops. That’s the job of an exciter.

Foo - I noticd that the exciters are boosting the very extreme areas of the frequency range - which is not something I see on anything commercial. I am using X-cita.
Whats wrong here?

EDIT:
Also, by looking at Har-Bal’s video tutorial, it looks like it is acting as an equalizer, and is designed for final mastering. To make use of Har-Bal’s EQ, you may need to boost and not only cut - so Har-Bal is a bad thing also?

I wonder if it crossed their minds to take a small creative leap and call it Hair-Ball… :)

I wouldn’t use X-cita personally. Any one worthy would be ‘tunable’ and you can get great presence in the mids depending on the input mix.

Afaik Harbal could be great, but I’ve not had time to test it thoroughly myself. I know in my commercial mastering work I’d used nothing but LP10 reductively, with zero phase. Years of experience tells me that additive EQ in the tops is harsh, and it’s something I’d never do for clients.

And nothing beats analogue mastering for that matter. Why spend a heap of time being as anal as me about it? You could be out making music instead! Then, leave to your chosen professional to take to the last step. After all, if you’ve got boring songs, no amount of world class master can save you.

I see. I will try the others you recommended.
I am not sure I yet understand why is exciter better than EQ - I mean, how does exciter “excite” the sound, without boosting frequencies?

Well, nothing beats having a lot of money also, but I guess I will have to do with neither at the time… :)
Generally speaking, I tend to agree - leave the mastering to the engineer, yes.
But, I believe that all artists that make music, especially electronic music (since it already involves heavy use of the PC for sound generation) should be able to generate some sort of semi-professional mix. It should not necessarily be vinyl ready, etc, but it must not have too many mastering mistakes.

So in fact, this is my goal in this regard - have a master that does not sound like I know nothing about mastering, and that will sound good enough on most common retail speaker systems.

You’d maybe also like to try AAMS, which is kinda similar to Har-Bal.

http://www.curioza.com/

Appreciate it.
Screenshots look nice, I will most likely give it a try.

it’s pretty important to not only match eq curves but also the average rms powers per frequency range.

And how would you do that? What tool?

you could use har-bal for that

i would suggest to only play a bit with har-bal so you can learn what you have been doing ‘wrong’ in the first place. then move on to not use har-bal at all. because, well, it doesn’t auto sync to your own personal approach/sound… more like that of someone else. i wouldn’t want to depend on it to get a killer mix.
still, it’s a unique and handy tool to visualize and compare.