How to improve vocal track?

[Kim Lajoie]

aha, you as well.. - ubiquitous? Not at all! - Kim never seems to get tired of pretending it but actually the common pro advice is that it is in any case recommended to use a compressor with moderate settings before going D

Ok. This is how I see it.

There are two main reasons for using a compressor on vocals. The first reason many people use an analogue compressor pre-ADC is to get more digital resolution by recording a hotter signal. This is definitely a good idea if you're recording at 16 bit. However, if you're recording at 24 bit, you have a theoretical dynamic range of 144 decibels. Most gear in your chain will give you a usable dynamic range of around 100 decibels. At 24 bit, even if the quietest part of the performance is at -48 decibels, you are still recording with 16 bits of resolution - as if you were maxing out a regular 16 bit converter[1]. So long as you have a reasonably quiet signal chain, recording at 24 bit gives you enough recording resolution to not have to use an analogue compressor.

The other reason many people use an analogue compressor is for colour. Most analogue compressors have a sound that is quite different to digital compressors. However, an analogue compressor will not transform your recordings from crap to cool. At best, it will turn crap into slightly-less-crap. The advice I gave in my first post will have a much bigger effect on the quality of your recordings. Using an analogue compressor for colour is icing on the cake, and is most useful when you've already got the basics as good as you can get them.

The reason professionals use analogue compressors is mainly this:

1. They have compressors that sound fantastic (and cost appropriately).
2. They are experienced enough to know how to set it right first time to get a good sound when recording.
3. They already have all the basics of recording set up to near-perfrection, so tools like analogue compressors become a reasonable way to further improve the sound.


Again, that's just they way I see it. I'm happy to discuss differences of opinion.


Forever,




Kim.

[1] Not quite, but close enough for this discussion.

[mgc3003.com]

Jeez/Kim -- let me ask you this. Obviously neither case here is optimal, but what would you advise???

1) MIC to TUBEPREAMP to 16-bit SOUNDCARD

2) MIC to TUBEPREAMP to DIGITAL COMPRESSOR* to 16-bit SOUNDCARD.

(And I mean a compressor that operates digitally with an analog out.)

With option 1, you record a lower average volume... so you decrease the digital resolution as you pointed out in your post.

With option 2, your average volume is fine --- but passing through a digital compressor means that's *two* D/A conversions instead of one.

So if someone HAS to choose between those two options, what would be the best?

[Kim Lajoie]

Jeez/Kim -- let me ask you this. Obviously neither case here is optimal, but what would you advise???

1) MIC to TUBEPREAMP to 16-bit SOUNDCARD

2) MIC to TUBEPREAMP to DIGITAL COMPRESSOR* to 16-bit SOUNDCARD.

(And I mean a compressor that operates digitally with an analog out.)

Answer below...

With option 1, you record a lower average volume... so you decrease the digital resolution as you pointed out in your post.

Yes. My point though, was that if you're digitising at 24 bit, then a lower average level is not a big concern. So long as you stay above -48dB, you're still winning.

With option 2, your average volume is fine --- but passing through a digital compressor means that's *two* D/A conversions instead of one.

That's correct. Whether the signal degradation is significant enough to outweigh the benefits of the compressor is impossible to say without listening. Read below...

So if someone HAS to choose between those two options, what would be the best?

It's impossible to say, based on text alone. You would have to do some listening tests to see which combination has a sound that you like and is useful for you.

Without listening, my preference would be to have as clean a signal path as possible - don't use the compressor. You will already be recording with high bit depth (assuming you have the rest of the recording process well established), and recording without dynamics processing will give you more flexibility.

Of course, you might like the sound of your digital compressor, in which case you might want to use that while recording. I suggest comparing your digital compressor to software compression plugins to see if the reduced flexibility of using hardware is worth that sound.

Really, you have to try out some combinations and listen carefully. Consider sound. Consider flexibility. Consider convenience. It's your gear. Spend some time testing it out so you can make informed decisions on how to make best use of it.

And then make some music.

Forever,




Kim.

[Kim Lajoie]

Kim never seems to get tired of pretending it but ...

What are you talking about? When have I said this before? It's not like I'm on a mission to make people use clean recording chains.

If anything, my mission is to get people excited about Sonik Synth 2.

Forever,




Kim.

[mgc3003.com]

Jeez, thanks!
I got your point about 24 bit. In your response, did you realize that I'm still stuck in 16 bit right now?

My question was regarding 16 bit, specifically... Whether that added resolution would be so important to be worth the extra layer of A/D of the digital compressor, so as to maximize the resolution of the 16bit format.

(Forgive me if your same "impossible to tell" answer still stands. Just, from your response I think you might have missed that I meant specifically in a 16 bit environment.)
Thanks again!

[Kim Lajoie]

Jeez, thanks!
I got your point about 24 bit. In your response, did you realize that I'm still stuck in 16 bit right now?

My question was regarding 16 bit, specifically... Whether that added resolution would be so important to be worth the extra layer of A/D of the digital compressor, so as to maximize the resolution of the 16bit format.

(Forgive me if your same "impossible to tell" answer still stands. Just, from your response I think you might have missed that I meant specifically in a 16 bit environment.)
Thanks again!

Sorry, I didn't realise you were recording at 16 bit.

In this case, the compressor might make a significant difference. It's definitely worth considering. Of course, you'll have to listen for yourself and make up your own mind.

Two better options would be:

1. Go digital from the compressor onwards. That is, take a digital signal out of the compressor and feed that directly into the computer. You might not have this capability though.

2. Even better, instead of buying the digital compressor, use that money to buy a 24 bit audio interface for your computer.

If you're digitising at 16 bit, it's very important to get the signal as close to -0dB as possible (without clipping).

How come you're recording at 16 bit?

Forever,




Kim.

[mgc3003.com]

Thanks again for this advice. I agree, it makes good sense. I'm going to go back to this route until I can get a better compressor and soundcard. And computer.

I'm at 16bit because I've feared the CPU hit I'll take by switching to 24. Already I'm hitting the CPU limit enough that I have to bounce more than I'd like. And it's partly financial as well... I'm about to take up to a 50k *CUT* in salary, at least temporarily. I'm bailing on my game industry career which pays well but is miserable for me. And to be honest, I suspect as I'm adjusting to a new career and trying to get my salary back up --- I'll look back at purchasing a 24 bit card right now as being downright absurd. Hehe! I'm trying to get ready for that salary drop by living like I'm at that level now. Which basically means buying nothing. :-/

As soon as I get my new career and a decent salary again, my first purchases will be a new computer & 24 bit card! For sure.

[Kim Lajoie]

I'm at 16bit because I've feared the CPU hit I'll take by switching to 24.

That's a shame.

The CPU difference between 16 bit files and 24 bit files is almost nothing. All VST hosts process audio at 32 bit floating point (similar to 24 bit, but without resolution loss at extreme low or high levels). There is a 50% higher hard drive performance hit, but more modern hard drives are capable of streaming dozens of tracks anyway.

And still, most recording software programs will allow you to record at 16 bit any way, if you really want to.

Forever,




Kim.

[Evan]

I do not think 24bit will have any (significant if at all) CPU hit. It will have a hit on HD space however. Most ASIO drivers pack data to 32bit nowdays because it's more efficient for the PCI bus and the CPU. Same goes for most sequencers, the 16bit or 24bit integer data is usually converted to 32bit float as soon as it hits the sequencer's audio engine. Plus, in many sequencers you can easily mix 16bit and 24bit files, not everything has to be 24bit, only the recordings that matter.

[toine6]

I've got a 20 bit Darla Echo card for sale if you're interested, say $70 + shipping. Just thought I'd throw that in here.

[mman]

This is great. A lot of usefull things here.

I record at 16 bit too (ADAT). I can't afford new A/D converters and I'm not planing to chang the sound card neither.

Thank you very much, I think I can try some things here and see if it works for me.