Hey eMD,

First of all, there are two types of main reverbs; Convolution and Algorithmic. Convolution Reverbs - in this case, I assume you use VSTs to achieve reverberation - use real sound samples, recorded from live environments, known as "Impulse Responses". Since they use real sound samples, they create more "realistic tail" and feeling of reverb. Effect process of convolution reverbs basically mix your signal with filtered signals that you want to achieve. So you can think that for one sound, you basically process two sounds. Therefore, it has a huge impact on your CPU as well as your DAWs performance.

As you can guess from the name, Algorithmic Reverbs create paramaters to mock "realistic tail" to generate reverbration. All of DAWs come from one or more algorithmic reverbs. In my opinion, algorithmic reverbs sound okay when you apply on ambiance sounds, FX etc. If you apply on solo instruments or real instruments, it does not give that realistic feeling.

I'd recommend you to use convolution reverb on synths, snares etc. where you can use algorithmic reverbs to give little bit more space to additional/extra sounds on your track.

If you're an Ableton user, they do great job with their custom reverbs. If you'd like to use another VST for reverb, you can try Arts Acoustic for algorithmic reverb (I am sure that lot of producer friends in this forum use it) and Liquid Sonics Reverbrate for convolution.

Personally, I only have maximum 4 reverb (two algorithmic, two convolution) bus track to achieve reverbration in my tracks. That way, I minimize CPU usage and do not create muddy, non-realistic mix.

Hey eMD,

First of all, there are two types of main reverbs; Convolution and Algorithmic. Convolution Reverbs - in this case, I assume you use VSTs to achieve reverberation - use real sound samples, recorded from live environments, known as "Impulse Responses". Since they use real sound samples, they create more "realistic tail" and feeling of reverb. Effect process of convolution reverbs basically mix your signal with filtered signals that you want to achieve. So you can think that for one sound, you basically process two sounds. Therefore, it has a huge impact on your CPU as well as your DAWs performance.

As you can guess from the name, Algorithmic Reverbs create paramaters to mock "realistic tail" to generate reverbration. All of DAWs come from one or more algorithmic reverbs. In my opinion, algorithmic reverbs sound okay when you apply on ambiance sounds, FX etc. If you apply on solo instruments or real instruments, it does not give that realistic feeling.

I'd recommend you to use convolution reverb on synths, snares etc. where you can use algorithmic reverbs to give little bit more space to additional/extra sounds on your track.

If you're an Ableton user, they do great job with their custom reverbs. If you'd like to use another VST for reverb, you can try Arts Acoustic for algorithmic reverb (I am sure that lot of producer friends in this forum use it) and Liquid Sonics Reverbrate for convolution.

Personally, I only have maximum 4 reverb (two algorithmic, two convolution) bus track to achieve reverbration in my tracks. That way, I minimize CPU usage and do not create muddy, non-realistic mix.

Thanks for the response! I was wondering where Valhalla Vintage Verb falls in this?

I do use Ableton, my follow up question would be do most people use reverb on returns or on buses? I find I have more control when using buses. What are some pros and cons?

Thanks for the response! I was wondering where Valhalla Vintage Verb falls in this?

I do use Ableton, my follow up question would be do most people use reverb on returns or on buses? I find I have more control when using buses. What are some pros and cons?

Reveerb as a send can be handy and use less cpu than lots of insert reeverbs, and also you can mix and process it on its own, add eq, sidechain, etc.
Reeverb as an insert can give you the advantage of tweaking each reeverb settings for the differnt channels, instead of haveing the same settings for every channel, as the send reeverb would do. Because maybe you dont want the same decay time for your synth lead and for your snare.

I'm trying to bring my mixes to the next level, I understand the basics of setting my faders and using EQ as tools to level out my mix but I find it really hard to grasp the concept of placing my sounds in a room, more specifically bigger rooms. From my understanding, using reverb not only places a sound in the mix further in the back but also the reverb should create a "room" that the sound bounces in.

To place your sounds "in a room" you need to think like they are in some space. Use EQ to cut unnecessary highs, especially for some distant sounds with long reverbs. Panning might be useful too.
Don't add reverb to the sound if that reverb is masked in the mix. You can use the noise gate to get rid of masked reverb tails.

You can use reverb creatively and try to compress, filter, distort it and resample and layer with your samples. In case you have some complex arrangement.

Quote from: triartmusic on January 12, 2016, 01:17:28 pm

Hey eMD,

First of all, there are two types of main reverbs; Convolution and Algorithmic. Convolution Reverbs - in this case, I assume you use VSTs to achieve reverberation - use real sound samples, recorded from live environments, known as "Impulse Responses". Since they use real sound samples, they create more "realistic tail" and feeling of reverb. Effect process of convolution reverbs basically mix your signal with filtered signals that you want to achieve. So you can think that for one sound, you basically process two sounds. Therefore, it has a huge impact on your CPU as well as your DAWs performance.

As you can guess from the name, Algorithmic Reverbs create paramaters to mock "realistic tail" to generate reverbration. All of DAWs come from one or more algorithmic reverbs. In my opinion, algorithmic reverbs sound okay when you apply on ambiance sounds, FX etc. If you apply on solo instruments or real instruments, it does not give that realistic feeling.

I'd recommend you to use convolution reverb on synths, snares etc. where you can use algorithmic reverbs to give little bit more space to additional/extra sounds on your track.

If you're an Ableton user, they do great job with their custom reverbs. If you'd like to use another VST for reverb, you can try Arts Acoustic for algorithmic reverb (I am sure that lot of producer friends in this forum use it) and Liquid Sonics Reverbrate for convolution.

Personally, I only have maximum 4 reverb (two algorithmic, two convolution) bus track to achieve reverbration in my tracks. That way, I minimize CPU usage and do not create muddy, non-realistic mix.

Thanks for the response! I was wondering where Valhalla Vintage Verb falls in this?

I do use Ableton, my follow up question would be do most people use reverb on returns or on buses? I find I have more control when using buses. What are some pros and cons?

I agree with Luis and Xan. It's just matter of your personal taste. If you have similar sounds let's say two violins, you don't want to apply same reverb settings to each channel. Instead, you can create a bus setting for reverb and send a signal to reverb channel to save some CPU.

As I mentioned in my previous post, after I design my sounds, I create two short, two long reverb bus channels and only use them in all track to prevent high CPU usage as well as muddy mix.

Quote from: eMD on January 12, 2016, 08:06:03 am

I'm trying to bring my mixes to the next level, I understand the basics of setting my faders and using EQ as tools to level out my mix but I find it really hard to grasp the concept of placing my sounds in a room, more specifically bigger rooms. From my understanding, using reverb not only places a sound in the mix further in the back but also the reverb should create a "room" that the sound bounces in.

To place your sounds "in a room" you need to think like they are in some space. Use EQ to cut unnecessary highs, especially for some distant sounds with long reverbs. Panning might be useful too.
Don't add reverb to the sound if that reverb is masked in the mix. You can use the noise gate to get rid of masked reverb tails.

You can use reverb creatively and try to compress, filter, distort it and resample and layer with your samples. In case you have some complex arrangement.

That's an interesting point. "Don't add reverb to the sound if that reverb is masked in the mix". That makes a lot of sense. Could you give some common examples of this?  I did a track once that had a low mid pad. The reverb on this pad made the bass sound as if it had a reverb on it. It never occurred to me to just remove the reverb from the pad. Is this kind of what you mean?  Please expand on this :-)

Cheers,

Rob