Mixing Orchestra bottom end

[ManFromGlass]

I’m working on a partially orchestral film project that requires huge bass drums, low loud brass, lot’s of low strings. I’ve been checking out the Hollywood A-List guys scores for reference. I’m amazed at how good they sound in the car. I’ve been asking mixer friends how they would deal with all that low end. Here are tips I have been given but I’m looking for more because I have to mix it and I am not a great mixer.
(Bass drum hits and taikos/toms are driving the track)
- roll off anything below 20 Hz or higher on the bass drum
- same with other instruments if anything is down there
- doubling the low strings/brass with a low sine wave of the same melody might help
- Waves Maxx Bass might help the bass drum hits on smaller monitors
- no reverb on anything low or eq the low end out of the verb
- don’t be afraid to do extreme eq cutting if that helps
- and ???

[The Elf]

At any point in time only one instrument should be allowed to dominate that low end. Your job is to decide which one and make it happen.

I'd get busier with that low end filtering. Most sources will happily take a cut up to 150Hz (and many sources will take considerably higher) without losing their mojo, and I'd be looking to cut bass drums at around 30Hz to get rid of really low junk that will add nothing of any value to the mix. Use you ears - if you can hear the filtering it may be too much.

And think three times before adding *any* boost to anything below 150Hz. Does it really need it?

Others will do it differently - and that's fine!

[Drew Stephenson]

Might be worth looking at some ducking as well. If you've got low, sustained notes from strings or brass, sidechaining a compressor driven by your percussion hits can help them stand out without needing to crack the level.
Obviously you don't want to get into EDM-style 'pumping' so use your discretion...

[Rich Hanson]

In orchestral scores, if more than one low section are playing at the same time, they will generally be playing in unison or octaves, anything else tends to lead to mush.

[ManFromGlass]

Thanks all. An interesting side note -
I’ve been listening to some Batman cues. I have MP3s. They sound really good in the car even cranked at highway speeds. The car sound system is good but not high end. But under good headphones (Shure SRH 1840s) the cues sound pretty bad in the low end, mushy and not clear.
I’m wondering if my brain is filling in all the lost frequencies in the car and yet I swear I am hearing the fundamental of the monster bass drums. (which can’t be true)
Under headphones I can also hear the strings and low brass doubled with low synth, barely audible. Perhaps that is one of the secrets to a mix translating well to a non-optimal listening environment. Fascinating.

[RichardT]

That sounds like a real challenge. The bass frequencies are very hard to get right. Eddie Bazil has a course / ebook called ‘The Low End’ that might be helpful.

With heavy bass drums there is enormous potential for making the mix sound messy. I would cut away as much as possible without losing the power in the bass and the beater sound, which is much higher in frequency. So there may be an area in the middle where you can cut a lot away without bad effects.

Also be ruthless in cutting away frequencies higher than the ones you need for definition. It’s amazing how high up bass drum sounds can reach.

The other problem is the long decay time of taiko drums and the like. I use a transient enhancer (in fact, NI’s Transient Master) to emphasise the bang and shorten the boom on bass drums. That can be very effective.

Multiband compression can also help control the drum sound.

For other instruments playing in the bass register you may need quite a bit of fast-attack fast-release compression to bring up the sustain portion - the transient portion is going to be lost under the drums anyway.

[Martin Walker]

At any point in time only one instrument should be allowed to dominate that low end. Your job is to decide which one and make it happen.

Lots of good advice here, but if I could only remember one sentence from this thread, this would be the one

[ManFromGlass]

also
Boz has a plug called Transgressor 2. It’s great for shortening the ring out. I assume a gate could work too but I like Boz’s plug better.
I may try a really short ducker on the other low end instruments each time the drum gets hit. Perhaps this is a lazy approach but if it gets me there then I don’t mind.
I just remembered, Melda makes an autodynamic eq. If I can set up ducking with it I will have good control over the bottom of the other instruments leaving the higher frequencies untouched.

[Martin Walker]

also
Boz has a plug called Transgressor 2. It’s great for shortening the ring out. I assume a gate could work too but I like Boz’s plug better.
I may try a really short ducker on the other low end instruments each time the drum gets hit. Perhaps this is a lazy approach but if it gets me there then I don’t mind.
I just remembered, Melda makes an autodynamic eq. If I can set up ducking with it I will have good control over the bottom of the other instruments leaving the higher frequencies untouched.

I've got the Boz Transgressor 2 - it's really effective on problem sounds, plus a true creative tool.

On the subject of dynamic EQs, don't forget TDR Nova, whose basic version is not only really good, but free:

https://www.tokyodawn.net/tdr-nova/

Martin

[Drew Stephenson]

On the subject of dynamic EQs, don't forget TDR Nova, whose basic version is not only really good, but free:

https://www.tokyodawn.net/tdr-nova/

Martin

That's what I'd use.

[The Elf]

I'll just point out that many DAWs do feature these kinds of plug-ins, so check them out before you start spending money. It's a rare plug-in that I need to go seeking these days...

[Drew Stephenson]

I'll just point out that many DAWs do feature these kinds of plug-ins, so check them out before you start spending money. It's a rare plug-in that I need to go seeking these days...

Indeed, it's one of the weird gaps in Reaper's standard set though. Multiband compressor, yes, dynamic EQ no. Though you can do something clever with ReaEQ and ReaComp to make them act together as a dynamic EQ.
But it was easier to go to Tokyo Dawn and as it's a freemium model you get a good tool for nowt if you so choose.

[The Elf]

Multiband compressor, yes, dynamic EQ no.

I consider them much the same thing.

[Sam Inglis]

At any point in time only one instrument should be allowed to dominate that low end. Your job is to decide which one and make it happen.

In an orchestral context, surely that depends on the composition? If the composer has doubled a low part on more than one instrument, they've presumably done so for a good reason.

[The Elf]

At any point in time only one instrument should be allowed to dominate that low end. Your job is to decide which one and make it happen.

In an orchestral context, surely that depends on the composition? If the composer has doubled a low part on more than one instrument, they've presumably done so for a good reason.

If they've doubled it then they can effectively be considered one instrument. You could say that the composer has 28 parts all 'doubling' the low end - something has to give, however you cut it. Maybe you HPF one of the parts to keep the low end tidy, but allow the double to 'speak' in the upper frequencies. Or you leave both as they are and accept the hit on headroom.

Which example are you thinking of?

We can go looking for examples to break any generalisation. How about every every cello and bass simultaneously playing different notes? I'm sure it exists somewhere. Most rock bands have only one bassist, but I bet we can find an example of two or more.

My point still essentially stands.

And, as I said, 'others will do it differently'. So what's your approach? And how would you deal with the doubled part?

[Arpangel]

When you say "mixing" is it literally that? I had to mix an orchestra, it was a 24 track Samplitude project, it was great, as all the sections were close mic'd as well as a stereo pair on the whole band.
If it’s a stereo mix, that’s a real pain in these circumstances, I’d be wary of too much extreme EQ, you’ll find it’ll suck the life out of things, especially the bass, a bit of wayward dynamics isn’t such a bad thing.

[ManFromGlass]

Yes literally mixing.
Part of the problem is that I am using samples. Each library brings it’s own room sound. Some even have their own reverb built into the interface, which I turn off. So part of the challenge is not only dealing with each section’s low end but also room tone.
Close mic’d instruments will never sound like an orchestra in a hall from a listener sitting in the audience perspective, which is the sound I have in my head. I’ll never achieve that with samples but I’ll aim for it.

[The Elf]

Close mic’d instruments will never sound like an orchestra in a hall from a listener sitting in the audience perspective, which is the sound I have in my head. I’ll never achieve that with samples but I’ll aim for it.

Some of the spiffy (e.g. 8Dio and Spitfire) libraries have multiple mic's to blend with, from close to distant.

[tea for two]

Yes literally mixing.
Part of the problem is that I am using samples. Each library brings it’s own room sound. Some even have their own reverb built into the interface, which I turn off. So part of the challenge is not only dealing with each section’s low end but also room tone.
Close mic’d instruments will never sound like an orchestra in a hall from a listener sitting in the audience perspective, which is the sound I have in my head. I’ll never achieve that with samples but I’ll aim for it.

If it helps at all the visualisation of orchestra in a large hall with drums right at the back.

So heavy bass drums taikos should sound as distant thunder.

I’m working on a partially orchestral film project that requires huge bass drums, low loud brass, lot’s of low strings.

Here's a track with Japanese Kodo drummers and Orchestra starts with lowish strings, thereafter strong Brass for some kind of reference to what you mention
https://m.youtube.com/watch?v=NZfeUPKSPQA
Available on CD : One World One Voice.

[Drew Stephenson]

Multiband compressor, yes, dynamic EQ no.

I consider them much the same thing.

Indeed, I've been wracking my brains to remember why I decided I needed Nova rather than just using ReaXComp and I remember now it's because Nova allows an external sidechain but XComp doesn't.

[Wonks]

Obviously they aren’t exactly the same. A dynamic EQ will still apply the EQ if the signal isn’t loud enough to trigger the dynamics, whereas on a multi-band EQ with the same dynamic settings, the sound remains unchanged. But with thresholds set lower so the compressors are always working, they are reasonably similar.

[Sam Inglis]

If they've doubled it then they can effectively be considered one instrument. You could say that the composer has 28 parts all 'doubling' the low end - something has to give, however you cut it. Maybe you HPF one of the parts to keep the low end tidy, but allow the double to 'speak' in the upper frequencies. Or you leave both as they are and accept the hit on headroom.

Which example are you thinking of?

We can go looking for examples to break any generalisation. How about every every cello and bass simultaneously playing different notes? I'm sure it exists somewhere. Most rock bands have only one bassist, but I bet we can find an example of two or more.

My point still essentially stands.

And, as I said, 'others will do it differently'. So what's your approach? And how would you deal with the doubled part?

I guess what I'm saying is that in the first instance, a congested or over-rich low end in an orchestral track may well be an arrangement issue, not a mixing issue.

[The Elf]

Obviously they aren’t exactly the same.

Take a look at Waves C6. You can apply static boost/cut *and* compression/expansion to a band. To me that's pretty much both an EQ and a frequency band compressor, or, at least, as close enough as makes no difference in use.

I think of it as dynamic EQ, because I tend not to use multi-band compression per se, but if there's a genuine difference it's not one I recognise.

[Sam Inglis]

I would say there are a couple of differences between a multi-band compressor and a dynamic EQ, at least in their traditional incarnations. One is that in a dynamic EQ you can have overlap between bands, which is not really possible with a conventional multi-band compressor. The other and probably more important one is that the crossovers in a multi-band compressor are active even when isn't actually doing any compression. A multi-band compressor is always applying a set of steep filters to the signal to chop it up into different frequency bands, regardless of whether compression is active in those bands. That will inevitably introduce phase shift, or pre-ringing in linear phase designs.

Thankfully there are modern plug-ins that cover the same ground as both, without the filtering artifacts -- I'm a big fan of FabFilter Pro-MB for example.

[ManFromGlass]

I guess what I'm saying is that in the first instance, a congested or over-rich low end in an orchestral track may well be an arrangement issue, not a mixing issue.
[/quote]
I would normally agree when it comes to my writing, but a Hollywood A-list composer would have an A-list arranger/orchestrator and mixer. From the scores I’ve been checking out for bottom end these folks really bring the magic and a huge bottom end that isn’t mud. I’ll muddle (hah) along in my own way but have picked up a few tips here that will help for sure!

[ITHertz]

Some good advice here!

From what I've read above, no-one has mentioned manual level automation. Yes, I know, it's time consuming however it might give better results than automated ducking, etc.

As an example, look for places where multiple LF instruments play simultaneously, pick the one you want to hear and turn the others down (momentarily). If the same combinations of instruments are playing in other spots the automation can be cut and pasted.

You can even take this to the "micro" level and duck transients and tails.

This sort of approach is pretty common in EDM production these days.

Cheers,

Chris

[sonics]

... huge bass drums, low loud brass, lot’s of low strings.

Those instruments have very different sound profiles.

I think RichardT gave some excellent advice in his reply.

My thoughts:
Of course the arrangement is key, but after that there are many tools to help control that low end. Manual or automated level control, multiband dynamics, dynamic EQ, modern tools like Soothe or Smooth Operator and many more.

I don't believe in any set of rules or set techniques. The project dictates the tools and methods.

Get the arrangement and samples sounding as good as you can and then tackle the issues, like a build-up of low frequencies.

The ambience used is very important indeed. I like my orchestra to sound "as one", but the synths can have a different space. Once again, no rules...

At any point in time only one instrument should be allowed to dominate that low end.

That's not right! You mean the basses have to stop playing every time the taikos hit?

[The Elf]

That's not right! You mean the basses have to stop playing every time the taikos hit?

This is about the mix, not the arrangement. If those two instruments are going to co-exist as-is then you have to allow headroom for it to happen; and that's going to dictate the level of the rest of the mix.

Or you compromise one and let the other dominate - done well the masking will hide the trick. This is the art of *mixing*.

Compromises, compromises...

[ManFromGlass]

Mr Elf - a question -
I have been rolling off lows and necessary highs on my individual tracks. Would it essentially be the same thing if I sent all those tracks without EQ to a bus (or Aux depending what your your DAW calls it) and then roll off the lows all at once?
To be clearer, I would gang similar tracks so in this case all the drums would get their own bus and EQ, low orchestra their own bus and EQ, synths their own etc.
I suspect the ganging would probably be a quicker way to go, but one would still need to do a bit of individual adjusting to say, a string patch with too much mid information.

[The Elf]

Certain DAWs call *everything* an 'aux'! Sigh...

It's hard to generalise because I don't know what your arrangement looks like. To give us something to start with, let's assume we have a basses, cellos, violas and violins mix, and they are pretty much all playing, all the time...

I can tell you that if the cellos and basses are in unison (or octaves) I would roll the extreme lows from the cellos. Let the basses dominate the low frequencies. The cellos will 'borrow' their low end from the basses and you should be able to get something that sounds natural. If the cellos do need to be heard in isolation at some point then allow the low end back with automation in those passages.

I don't think I would be looking to do any drastic EQ at Group level. The EQ I would suggest is about keeping that low end solid, and that requires something to be down there all the time, not just shaving low end off everything. You're making a targeted decision as to what is going to use your headroom - do it well and it will be as if no trickery was happening.

If there are any low drum hits then I might use a touch of compression side-chaining to the basses to preserve headroom.

I'm not much into rolling off highs, but that's just me.

I'll repeat - it is near-impossible for me to say 'always do this', because every moment of every tune is different, but guard and allocate those extreme low frequencies jealously, because that's your headroom-killer.