ValhallaRoom 1.5.1 Released. New Electric Blue GUI

[@midnight]

Hey Sean, this is sort of a random question -

but can you explain why a plugin like EOS requires additional latency whereas ValhallaRoom seems to be running in zero latency? (It is truly zero latency, correct?)

[bmanic]

I don't agree here. Often the reverb becomes a part of the sound and is more like a effect and not a room emulation. It will be processed with the complete sound together. Sends for hardware reverbs or for reverbs which glue different sounds together. Some old rules with sends vs inserts are not set in stone

No. It's not a rule but good practice. Here's why:

First of all: there is NO difference in the sound if you run it as an insert or as a send. The sound, when you tweak the volume of the send channel, is identical (you can null it in case you don't believe me).

Second: keeping the reverbs on separate send channels makes it easy to automate the reverbs throughout the song. It's a very old and good trick to make clear differences between verses and choruses (or high-energy or low-energy parts) of a song. You can even accent certain lead or vocal lines simply by automating the reverb. Many reverbs do not let you smoothly automate the dry/wet knob and several have dual knobs making it even more difficult to do. It's also easier to find the reverb in large projects when they have their own channels (and thus usually their own name and color). This becomes crucial if several people are collaborating on a project.

Third: Reverb sends/returns, whatever you want to call it, are usually processed somehow. It is extremely rare NOT to process a reverb signal. Compression, Gate/Expander and EQ are the most common processing tricks but chorus and flanger can be used to a great effect as well. Then you can further send the reverb channel into a delay channel, or vice versa.

Fourth: You can send several elements into the same reverb, thus mixing the content within the reverb. You might argue that you can just duplicate the reverb to other channels but if the reverb has some random elements in it (noise, modulation etc) then it will NOT be the same and thus sound different.

The only advantage I see to having the reverb as an insert after whatever sound you need reverb on, is that you can then process both the original sound and reverb together. Then again, you can do this same thing if the reverb is a send. You just route both the original sound and the reverb channel to a bus where they are summed and processed. Again doing it this way has numerous advantages. Only disadvantage is that if you need tons of individual reverbs then it could get a bit cluttered (but ask yourself.. how often do you use more than 10 reverbs in a track?).

So, as you can see. There are very good reasons for using reverbs (and delays, choruses, flangers.. whatever needs additional processing) separated in their own channel. It's not recommended just because of some old "rule" but because it is good practice, giving the user a lot more flexibility and options.

Cheers!
bManic

[bmanic]

Yeah it used to be once when you only had one hardware reverb processor but these days its ok to use as a direct effect on a channel with automation on the wet/dry control and decay values.

It's interesting how many seem to think this way.. see my above post for explanation to why what you just said is one of those weird ".. in the old days the had to ... but not anymore.." myths.

In short: Processing the reverb send/return is why you have it separate. Using a reverb as an insert is only done when lazy. You gain nothing by having the reverb as an insert and loose a hell of a lot of flexibility.

Cheers!
bManic

[Funkybot's Evil Twin]

You gain nothing by having the reverb as an insert and loose a hell of a lot of flexibility.

I agree with you 99.9% of the time, but there have been times where I was working with a band and they wanted more "space" in between the vocalist and mic. What I did was threw a reverb set to only play the late reflections early in the vocal chain (with a very low mix setting), then processed the vocal chain (with the reverb inline) as normal. Once it hit the compressor, the early reflections really got exaggerated in a nice "singer way off the mic" sort of way. It didn't sound quite as authentic with the verb on a send. Anyway, it worked well in this case, but it's something I haven't done since.

[bmanic]

Yes, that kind of compression of source+reverb is done quite often. Again, it's something which is easy to do even when the reverb is a separate channel. You just group the source and the reverb and process from there.

If you want to do things quickly, then of course using a reverb as an insert is the way to go but during serious mixing, it makes no real sense.

Of course there are no real rules but I just want to bust the myth that they used reverbs as sends in the old days just because they didn't have that many reverb units.

Cheers!
bManic

[solidtrax]

Thanks Sean,

Sulaco is beautiful! We've made little showcase for it. Enjoy:
http://soundcloud.com/solidtrax/charlatan112-vroom109

[Fritze]

In short: Processing the reverb send/return is why you have it separate. Using a reverb as an insert is only done when lazy. You gain nothing by having the reverb as an insert and loose a hell of a lot of flexibility.

Cheers!
bManic

Yes, if you need or want that flexibility. Your examples are showing that very good. At the same time when doing so you're loosing simplicity. Obviously. I don't make a rule out of both but use what seems appropriate in the given case.

[zeep]

I wish all reverbs were like AA's, Tila, etc.. Separate DRY and WET. It just makes more sense.

I disagree. Separate wet and dry always annoys me. It's much easier to control with a single slider.

Besides, it's usually highly recommended to run a reverb as a send, not as an insert. You can't properly process the reverb if it's only as an insert.

Cheers!
bManic

Yes i remember my first reverb, Alesis Quadraverb. That was a good send reverb.

Lucky for us these days we can soundshape on one track. I still use sends, but not many.

When i work with specific sounds i automate the reverb on that channel. Some breaks i max the reverb time, other times i make it phonebooth short. In the EDM scene this is done a lot. If i use a send reverb for that then all the other sounds get swamped in reverb and away is the contrast.

None of my colleagues will call me lazy when i soundshape the heck out of a sound and keep my effects and automation tidy on their respective channels.

Anyway to each his own. I believe in no boundaries. The days when i "had to" work in specific ways went away when i sold my gear and went in the box.

Not to get in an argument but there are also reasons NOT to work like you say. It's not necessarily good practice.

Second: keeping the reverbs on separate send channels makes it easy to automate the reverbs throughout the song.... This becomes crucial if several people are collaborating on a project.

Well i disagree with the easy to find part. When i work in a project with 30 to 40 tracks, on the 6th channel or so and i'm tweaking and automating it for breaks / attenuation etc, then it's a chore to keep scrolling all the way down to the send channel, tweak automation, then scroll all the way up again, edit some more etc.
Also, 9 out of 10 times when i collab on a project or get sent parts, there are no send/return bounces. All the tracks have effect on them and i get those wet and dry. when a specific effect needs to be separate then it's put on a send.

Third: Reverb sends/returns, whatever you want to call it, are usually processed somehow. It is extremely rare NOT to process a reverb signal. Compression, Gate/Expander and EQ are the most common processing tricks but chorus and flanger can be used to a great effect as well. Then you can further send the reverb channel into a delay channel, or vice versa.

Live offers ability to split audio on a single channel to multiple channels and process all separate. This works very quick too. And most DAW's like Logic & Live offer multiple automating lanes visibility in one channel which is very convenient and at a glance you can see them all and edit them close by. That's better than scrolling back and forth to a send like i said above.

Fourth: You can send several elements into the same reverb, thus mixing the content within the reverb. You might argue that you can just duplicate the reverb to other channels but if the reverb has some random elements in it (noise, modulation etc) then it will NOT be the same and thus sound different.

True and handy. I do this too.

The only advantage I see to having the reverb as an insert after whatever sound you need reverb on, is that you can then process both the original sound and reverb together. Then again, you can do this same thing if the reverb is a send. You just route both the original sound and the reverb channel to a bus where they are summed and processed. Again doing it this way has numerous advantages. Only disadvantage is that if you need tons of individual reverbs then it could get a bit cluttered (but ask yourself.. how often do you use more than 10 reverbs in a track?).

Very often! 10 is not much at all.

I understand your valid points, but to keep it brief there's reasons enough to stray from the 'normal' path and work different. I work a lot faster than way and in Live there's so many routing possibilities i hardly ever 'need' to use a send.


*sorry if i'm typing too much. *

[bmanic]

When i work with specific sounds i automate the reverb on that channel. Some breaks i max the reverb time, other times i make it phonebooth short. In the EDM scene this is done a lot. If i use a send reverb for that then all the other sounds get swamped in reverb and away is the contrast!

I think you have misunderstood me. A send/return reverb does NOT automatically mean that you send multiple channels into a single reverb!! It just means that instead of using it as an insert, you give the reverb it's own channel and route whatever single instrument you want into it.

Besides, if you split the reverb on a separate channel on the same track (Reaper can do this too, up to 64 channels per track) then you have basically created "a send/return" channel.

Basically my point was: You have absolutely zero to gain by using it as an insert and everything to gain by using it as a send/return channel, flexibility wise.

Perhaps it is more common in traditional pop/rock/dance/whatever mixes than when the producer him/herself mixes and produces at the same time?

Btw, 40-60 tracks is a pretty small mix in today's standards, at least in my experience. The norm is around 100 tracks and the reverbs are extremely easy to find as long as you color them, name them and keep them all organized in one place (like in a folder for instance).

Cheers!
bManic

[@midnight]

I use 3 reverbs sends:

1 short @ 500-600ms
1 medium @ 2-3 seconds
1 long @ 7-8 seconds

and then I use a delay send also, and I send that to the reverbs too.

I only use an insert reverb if it's like a spring reverb on a guitar or something. But I usually don't do that.

[bmanic]

What kind of music do you produce? By today's typical rock/pop/dance it's rather few reverbs.

Of course it depends on the room/ambience amount and quality of whatever sound sources you are using but part of the "thick" sound of modern mixes, even the seemingly dry ones with sparse arrangements, comes from using a lot of subtle reverbs. A lot can mean anything between 5-20 plugin reverbs/hardware boxes.

Cheers!
bManic

[A.M. Gold]

Anyone here like creating hybrid algo/convo reverbs by using multiple busses?

[@midnight]

Anyone here like creating hybrid algo/convo reverbs by using multiple busses?

Ozone's new reverb uses convolution for the early reflections and algorithms for the tails. Makes a ton of sense to me, also, not so hot on the CPU though....

[mkdr]

What kind of music do you produce? By today's typical rock/pop/dance it's rather few reverbs.

Of course it depends on the room/ambience amount and quality of whatever sound sources you are using but part of the "thick" sound of modern mixes, even the seemingly dry ones with sparse arrangements, comes from using a lot of subtle reverbs. A lot can mean anything between 5-20 plugin reverbs/hardware boxes.

Cheers!
bManic

There's really no rules to this. I know plenty of producers who use just one reverb on modern pop/dance productions.. And with channel counts closer to 10 than 100. Over complicating things just for the sake of doing it like everyone else does serves no purpose. Just like over complicating the mixing process for doing a send for everything that could just as easily be done with an insert..

I use sends for "master effects". The verbs and delays and modulations that go on several channels(this is the case for roughly 90% of channels that need those effects). But the rest is easier, and more sane to do with inserts.

I do agree that the myth is wrong and there's no quality benefit of doing it with inserts.



I'm loving the ValhallaRoom btw. Best verb i've ever used

[hibidy]

In short: Processing the reverb send/return is why you have it separate. Using a reverb as an insert is only done when lazy. You gain nothing by having the reverb as an insert and loose a hell of a lot of flexibility.

Cheers!
bManic

Can't agree with this completely, though I do prefer bus/send for my verb.