Detecting a poor mix *before* using Har-Bal

[DaveEwer]

I recently pulled up an old track of mine and put in in Har-Bal. The spectrum is as shown here

It seems obvious to me that there's something very wrong here, especially on the bottom end, that should be fixed before I get to the mastering stage. I've noticed this wild bass response with quite a few of my mixes - especially the ones that were done completely with synthesized instruments.

Is this a problem of phase cancellation between instruments on the bottom end? Is there some other cause? I have noticed that levelling out the peaks in Har-Bal, while making the mix more even, also adds some mud, or lack of definition.

So what appears to be wrong with the mix?

Thanks.
-Dave Ewer

[har-bal]

I recently pulled up an old track of mine and put in in Har-Bal. The spectrum is as shown here

It seems obvious to me that there's something very wrong here, especially on the bottom end, that should be fixed before I get to the mastering stage. I've noticed this wild bass response with quite a few of my mixes - especially the ones that were done completely with synthesized instruments.

Is this a problem of phase cancellation between instruments on the bottom end? Is there some other cause? I have noticed that levelling out the peaks in Har-Bal, while making the mix more even, also adds some mud, or lack of definition.

So what appears to be wrong with the mix?

Thanks.
-Dave Ewer

Dave

It sounds as though the frequency range for each of the instuments in the track may be overlapping. If you go here http://www.har-bal.com/frequency.php you can see the range for the individual instruments.You can use high and low shelving on each of the tracks based on their range to trim off the excess. You would end up with a tighter mix with more definition. You wouldn't want a bass guitar and a kick drum taking using the same space. We had a discussion once on the use of harmonics in mixing.

You can find it hear and it will clear up a few things.
http://www.har-bal.com/ipw-web/bulletin ... =harmonics

As a matter of fact you were part of the discussion

Just out of curiousity....which instruments are you using?

Earle

[DaveEwer]

Yes Earle, I definitely remember the conversation, and I'm aware of the problem of overlapping instruments. With this graphic, I was more asking about whether or not the sinusoidal response in the bass region was being caused by cancellation. After all, though you can carve the EQ to give each instrument its own place in the mix, you aren't going to completely notch out one in favour of the other, right? I mean, if your bass guitar is boosted in the 100 Hz area, you won't notch the bass drum completely at 100 Hz. I'd assume that you'd notch the bass drum just enough to give the bass guitar prominence. Have I got this right?

The instuments in that particular mix are:

Synth strings
Bass
Piano
Glock

All instruments are from the Roland Dound Canvas.
There is no drums or percussion. No instrument is playing in the bass region other than the bass.

-Dave

[har-bal]

Yes Earle, I definitely remember the conversation, and I'm aware of the problem of overlapping instruments. With this graphic, I was more asking about whether or not the sinusoidal response in the bass region was being caused by cancellation. After all, though you can carve the EQ to give each instrument its own place in the mix, you aren't going to completely notch out one in favour of the other, right? I mean, if your bass guitar is boosted in the 100 Hz area, you won't notch the bass drum completely at 100 Hz. I'd assume that you'd notch the bass drum just enough to give the bass guitar prominence. Have I got this right?

The instuments in that particular mix are:

Synth strings
Bass
Piano
Glock

All instruments are from the Roland Dound Canvas.
There is no drums or percussion. No instrument is playing in the bass region other than the bass.

-Dave

Dave

You look at the fundamental frequency of each of those in the bass region and slightly boost possibly the second and third harmonic.

Earle

[HarBal]

Dave,

I had a look at your spectrum and I would say that I can't really tell if there is a problem in the bottom end or not. My gut feeling is that there laregly isn't.

If the instruments that occupy this space in the spectrum are not percusive and have sustained notes and the melody line they follow is confined to a few notes then it is quite possible that you'll see a spectrum that looks like that at the bottom end. Then if you try and correct it you'll be amplifying noise and other stuff that you shouldn't, which would certainly explain why you end up with mudiness. My suggestion would be to simply cut the one peak at around 75Hz by 5dB or so and give a broad boost from 100-300Hz to raise the peaks up in that region to the same general level as the mid range. Leave the peaks in tact. You might actually want to cut the mid a bit as well, but thqat would depend on the music and instrumentation, so you're the only one in a position to judge that one.

Let us know how you go.

Regards,


Paavo.

[DaveEwer]

Thanks for the help!
-Dave

[Jay]

You look at the fundamental frequency of each of those in the bass region and slightly boost possibly the second and third harmonic.

Earl, how do you find the fundamental on a complex sound (e.g., kick) or on a continuous part (e.g. a bass line). Of course, the fundamental is obvious on, say, a single bass note, but I don't see how to determine that when Har-Balizing a bass part for an entire song. And while a kick isn't the most complex percussion sound there is, it's still got enough frequency data to make it less than obvious (to me, anyway) -- I mean, I know it'd be something in the low freqs (rather than the slap freqs), but still not sure. Any guidance is appreciated.

Jay

[HarBal]

Jay,

I'm sure Earle will answer you but I'm wondering why exactly, it is important for you to know the fundamental frequencies of a particular part played by a particular instrument. In a sense, I wondering if your mind set is leaning too much toward "control".

Personally, I'd basically look at the spectrum and preserve what has significant power and just control any errant peaks. However, if the spectrum is of a single instrument only then you'll have to combine your knowledge of the part with what you see in the spectrum, otherwise you may end up misreading it. By that I mean, if there's one note that is played a lot more than the others, it will show up as a pronounced series of peaks in the average spectrum, not necessarily because it is too loud, but on average it appears in the track much more than the others. In cases like this your best bet is to judge balance by the peak spectrum trace rather than the average case.

Cheers,


Paavo.

[Jay]

Jay,

I'm sure Earle will answer you but I'm wondering why exactly, it is important for you to know the fundamental frequencies of a particular part played by a particular instrument. In a sense, I wondering if your mind set is leaning too much toward "control".

Hi Paavo,

I asked specifically because of what Earle said. It sounded as if what he was saying pertained to specific intruments rather than an entire mix. Did I misunderstand?

Jay

[har-bal]

You look at the fundamental frequency of each of those in the bass region and slightly boost possibly the second and third harmonic.

Earl, how do you find the fundamental on a complex sound (e.g., kick) or on a continuous part (e.g. a bass line). Of course, the fundamental is obvious on, say, a single bass note, but I don't see how to determine that when Har-Balizing a bass part for an entire song. And while a kick isn't the most complex percussion sound there is, it's still got enough frequency data to make it less than obvious (to me, anyway) -- I mean, I know it'd be something in the low freqs (rather than the slap freqs), but still not sure. Any guidance is appreciated.

Jay

Hello Jay

It is much more effective during the mixdown phase.

Finding the fundamental is very easy. You can't necessarily find it in Har-Bal just by looking though. Each instrument covers a certain frequency range. Look at the chart on our frequency page.

What would be the fundamental frequency of the violin......it is where is starts...in this case it would be 186hz if my old eyes are not deceiving me If you add 186 to it the second harmonic would be 372hz and so on.

You can find a frequency calculator here http://www.csgnetwork.com/harmonicscalc.html

Imagine what you can accomplish by manipulating the harmonics of the instruments during the mixdown phase.

Earle

[Jay]

Forgive me, Earle, I guess I'm not being clear. I understand fundamental frequencies with regard to *single* notes. You had said to "look at the fundamental frequency of each of those in the bass region and slightly boost possibly the second and third harmonic". What I'm asking about is how you determine the fundamental on a kick drum or on a *continuous* bass part (i.e., one containing multiple notes). I'm not understanding how you can determine the fundamental on either a percussion instrument or on a pitched instrument that is playling more than one note (like a 3-minute bass part). Can you clarify?

Thanks,
Jay

P.S. Take a look at the image below. It's the spectrum of a bass part which is a little over a minute long. Imagine a similar spectrum (shifted to the left) of a kick drum also. Were you comments directed towards something like this? I'm not getting how a fundamental can be determined with this kind of thing. Something tells me I've misunderstood you.
.

[har-bal]

Forgive me, Earle, I guess I'm not being clear. I understand fundamental frequencies with regard to *single* notes. You had said to "look at the fundamental frequency of each of those in the bass region and slightly boost possibly the second and third harmonic". What I'm asking about is how you determine the fundamental on a kick drum or on a *continuous* bass part (i.e., one containing multiple notes). I'm not understanding how you can determine the fundamental on either a percussion instrument or on a pitched instrument that is playling more than one note (like a 3-minute bass part). Can you clarify?

Thanks,
Jay

P.S. Take a look at the image below. It's the spectrum of a bass part which is a little over a minute long. Imagine a similar spectrum (shifted to the left) of a kick drum also. Were you comments directed towards something like this? I'm not getting how a fundamental can be determined with this kind of thing. Something tells me I've misunderstood you.
.

Jay

Even if a kick drum or bass guitar is continuous they still have their own frequency range.

Try this.

1. Load your kick drum track into har-bal and then close it.
2. Load you bass guitar track into har-bal.
3. Load your kick drum reference track.
4. If you have a synthesizer track you can do the same thing.

Can you see them both now?
Try low shelving possibly the bass guitar at its fundamental frequency and boosting it slightly at the second harmonic. Experiment....were they overlapping at the low end

The whole idea is insure each of the low instruments have their own space/definition.

The fundamental frequency of an instrument is usually its highest peak on the left side.

Earle

[HarBal]

Hi Earle,

I think Jay doesn't really have a problem understanding your advice but the terminology you have used. Strictly speaking, the term "fundamental frequency" only really applies to a single played note. A complete part will have many fundamental frequencies spanning the entire range of the melody all mixed into one. So, to be more correct, I think you should be referring to the fundamental frequency range of a part rather than the fundamental frequency. In Jay's example it looks as though the fundamental frequency range spans about 1.5 octaves from around 80Hz to 200Hz.

Regards,


Paavo.

[har-bal]

Hi Earle,

I think Jay doesn't really have a problem understanding your advice but the terminology you have used. Strictly speaking, the term "fundamental frequency" only really applies to a single played note. A complete part will have many fundamental frequencies spanning the entire range of the melody all mixed into one. So, to be more correct, I think you should be referring to the fundamental frequency range of a part rather than the fundamental frequency. In Jay's example it looks as though the fundamental frequency range spans about 1.5 octaves from around 80Hz to 200Hz.

Regards,


Paavo.

You are probably correct. Perhaps there any mixing engineers out there that may be able to explain this better. I just don't see this issue as being complex. All instruments have their own space. This issue really needs to be addressed in the mixdown stage.

Earle

[Jay]

You are probably correct. Perhaps there any mixing engineers out there that may be able to explain this better. I just don't see this issue as being complex. All instruments have their own space. This issue really needs to be addressed in the mixdown stage.

Earle

Yes, Earle, Paavo is correct. I understand about carving out space in the mix, and I do address this issue in the mixdown stage. I'm trying to understand your initial suggestion. So, I'll try one last time to clarify my question. A harmonic is a multiple of a single frequency, so If you don't have a single fundamental frequency, how can you determine the appropriate higher harmonic(s) to boost? E.g., given the spectrum I provided, the peak energy is roughly between 100-200 Hz, so how do you determine which freq in that range will be the fundament which you multiply to get the higher harmonics? I know I could just do trial and error (i.e., try 100 as the fundamental and boost, 200 and/or 300, etc, then, if that doesn't sound good, treat 110 as the fundamental, and so on), but I thought maybe you were suggesting a better approach.

Thanks as always for your advice.
Jay