Getting the bass right.

steve@psbnoe.wanadoo


Hi, i'll try to explain this the best i can, when i've completed a mix and i check the spectrum in Harbal,  on the two responses (peak and average), peak seems to be pretty good (pretty much level across the range), but on the average it seems to be to high in the 70 - 300 area, so if i go back to the mix and lower the kick n bass and recheck it has lowered the average to a create a nice response but has obviously lowered the peaks to much, but on a commercial track both average and peak seem to have flat responses across the full range.How can i achieve this?
 
Cheers
Steve.

montezuma


Do it so it sounds good

mike_mccue


I didn't think it was hard to understand either...


:-)

I mean, if you already know that you have a bass drum hitting on 1/4 notes and the bass "instrument" hitting the 8's... what do you expect?

I'd expect bass build up and a field of mid range blur that buries just about anything you plan on floating above it.

The point I am trying to make is that this can easily be predicted by simply thinking about the arrangement.

I'm saying this should be considered way before the HarBal gets broken out.

It seems very easy to understand.

Here's a more subtle consideration; What are you using for monitors? Can you actually hear the 40hz bass parts you are writing? Sometimes people are surprised when they end up with a big fat broad bass tone at 80-120hz while they are searching to hear a 40hz signal that isn't coming out of the speakers?

best,
mike

bitflipper


Steve, I do exactly the same thing - use Har-Bal as a sanity check, and rather than EQing in Har-Bal I go back to the mix and fix it there. Over the years the discrepancies have become fewer and less severe, until by now the Har-Bal check mainly serves as a confidence booster.

Like most folks here, I also have the same small-room problem (10x13x7.5) that results in uneven bass response. Like many, I am entombed in 703, which helps a great deal with flutter echoes and stereo imaging, but like most I am physically unable to put in adequate bass trapping due to the layout of the room.

My best suggestion is to keep using Har-Bal to show you what you can't hear properly, and try to learn how to match what you hear to what you want to end up with. For me, it means ignoring an annoying peak at 143Hz and fighting the urge to EQ it away because I know it's an artifact of the room and not real.

Also, keep in mind that a perfectly flat graph isn't necessarily the ultimate goal. Not all commercial recordings follow the same curve. Many times I've tweaked a mix for that perfect EQ curve, only to go back and listen to an earlier version and realize that the new mix sucked in comparison! It can look right and sound wrong!

Here's a thought. You say the peak curve looks good but the average is too high in the 70-300Hz region. One possible reason is that you need to roll more lows off some of the midrange instruments such as acoustic guitar. A strummed acoustic guitar can add a lot of low frequencies to the mix, but they are almost subliminal because you only consciously hear the highs from the guitar. Another possible factor is an overcompressed bass guitar, raising the average RMS of the instrument. Try lower compression ratios or shorter release times on bass, which may help it punch through better without being too loud.

bitflipper


Steve, I do exactly the same thing - use Har-Bal as a sanity check, and rather than EQing in Har-Bal I go back to the mix and fix it there. Over the years the discrepancies have become fewer and less severe, until by now the Har-Bal check mainly serves as a confidence booster.

Like most folks here, I also have the same small-room problem (10x13x7.5) that results in uneven bass response. Like many, I am entombed in 703, which helps a great deal with flutter echoes and stereo imaging, but like most I am physically unable to put in adequate bass trapping due to the layout of the room.

My best suggestion is to keep using Har-Bal to show you what you can't hear properly, and try to learn how to match what you hear to what you want to end up with. For me, it means ignoring an annoying peak at 143Hz and fighting the urge to EQ it away because I know it's an artifact of the room and not real.

Also, keep in mind that a perfectly flat graph isn't necessarily the ultimate goal. Not all commercial recordings follow the same curve. Many times I've tweaked a mix for that perfect EQ curve, only to go back and listen to an earlier version and realize that the new mix sucked in comparison! It can look right and sound wrong!

Here's a thought. You say the peak curve looks good but the average is too high in the 70-300Hz region. One possible reason is that you need to roll more lows off some of the midrange instruments such as acoustic guitar. A strummed acoustic guitar can add a lot of low frequencies to the mix, but they are almost subliminal because you only consciously hear the highs from the guitar. Another possible factor is an overcompressed bass guitar, raising the average RMS of the instrument. Try lower compression ratios or shorter release times on bass, which may help it punch through better without being too loud.

batsbrew


i have found that it's nice to use har bal to finesse things, but it will not teach you how to choose and create sounds to mix.

could be, the frequencies that are bothering you, are REQUIRED to create the sound you want..

that's why the statement 'do it so it sounds good' is actually pretty close.


now, on most professional mixes, they've decided up front, that the sound is going to 'sound like ___________'


and the ME has made every track fit a nice little niche that makes it mix very smoothly and balanced.

it's a real art form, what the pro ME's do..
it's not an accident, or a standard that they work by, to get those 'pro' mixes.

plus, the use of extremely excellent equipment (especially on the mics, preamps, compressors and convertor sides) lets them capture the performances with absolutely stellar sound, which makes the mixing so much easier.

there are just too many variables for anyone here to say 'do this', or 'do that', and fix the bass issues.

in other words, i would not depend on Harbal to fix any mix issues.

batsbrew


i use the harbal on my 24 bit non dithered stereo mix file BEFORE i master (in wavelab)

typically, if i see issues in the harbal curve, i go back to the mix, and work on it there.


then, when i've dialed the mix in as good as it can be, i'll output another 24 bit stereo file, and put it in harbal, and only finesse it.

i have some curves i've made for reference, from my own material, as well as pro recordings, plus the 'rock' reference that comes with harbal.

between all of those, i MAY change mine.

but ultimately, using the A/B approach, i use my ears, and don't take what the reference curves tell me.


so, i never use harbal to FIX mix issues.

i use it to help me figure out whether or not i have mix issues, then i FIX them in sonar.

then, i check it again against harbal, and will slightly tweak using the harbal, and take THAT output file, and go to master with MBC and/or limiting, and nothing else.