What's new

Reverb on vocal - how to get it right?

I might add that there is very little discussion about this plugin on the internet. It has not generated much interest.

True dat. It was recommended to me by some engineers (in real conversation :) ) who use it quite a lot.
Personally, I toggle between this one and Precedence for similar tasks. Sometimes one is better, sometimes the other. Recently i had great results with an overdubbed alt saxophone choir (6 part voicing, block chords, with 2-3 doublings [i.e. 16 voices in total]) that I recorded with a Neumann in mono.

The trick is to use both the upmix gain and the brilliance knob with sensible ears; these make huge differences ("huge" in the sense of: these sliders can decide if the result sounds as if the source stands in a aluminium garage or if it's brilliant – at the same time quite dry but with a very defined sense of space.)
 
It will be part of a song, so it's voice over music.

All I received so far is a test run recorded on a cellphone.

This is probably going to be an ambience match issue. Check out the iZotope RX module called Ambience Match. There should be YouTube videos about it. If it describes your problem, see if you can borrow a copy of RX or get someone to help who has it. Unfortunately, processing in the trial version of RX can't be saved, so the trial version will be of no help.

Basically, you want the voice over, recorded in a different space, to work with the room sound of the music. A reverb plugin is not going to address that. Nor is a plugin that increases the "presence" of the voice-over track. Both may actually make the problem worse.
 
Last edited:
True dat. It was recommended to me by some engineers (in real conversation :) ) who use it quite a lot.
Personally, I toggle between this one and Precedence for similar tasks. Sometimes one is better, sometimes the other. Recently i had great results with an overdubbed alt saxophone choir (6 part voicing, block chords, with 2-3 doublings [i.e. 16 voices in total]) that I recorded with a Neumann in mono.

The trick is to use both the upmix gain and the brilliance knob with sensible ears; these make huge differences ("huge" in the sense of: these sliders can decide if the result sounds as if the source stands in a aluminium garage or if it's brilliant – at the same time quite dry but with a very defined sense of space.)

None of this explains why Schoeps used an inherently problematic recording in its own demo. As you may know, Schoeps mikes are widely used in sound recording for feature films and documentaries. There does not appear to be a single discussion about this plugin on forums that focus on location sound and post-production mixing. In one brief thread on Gearslutz, one person says that the plugin involves "shmeering" the sound. If that's true, it goes a long way to explaining why few people appear to be using it.

Hey, it may be the greatest thing since sliced bread, but the evidence isn't there, starting with Schoeps's demo.

In any event, I think that this discussion is on the wrong track. See the post just above.
 
Last edited:
Hey, it may be the greatest thing since sliced bread, but the evidence isn't there, starting with Schoeps's demo.

Honestly, i don't care that much about a specific YT demo. It wouldn't be the first bad product video since the existence of youtube.
However, you can download a 14 day demo from the PiA page, that's what i would recommend you anyway instead of buying based on a youtube video.

What you mention with the "smearing": yes, that's why i took the time and wrote some lines to its usage.
If you reread my post, the negative sound impression i mentioned could also be described as kind of "smeared" signal.
But as mentioned, adjusting the settings may change a lot.
However, as most tools, it's not the right one always, which is ok (since there is Precedence too).
 
It really depends on the voice and its recording, usually you can de-noise, de-verb it with Izotope RX or Era4 bundle and then compress and EQ, or EQ, compress, EQ and send it to 1 or 2 Reverb busses - try some Plate emulations if you have some, I personally always try UAD Emt140 on vocals first.
 
I looked further into Schoeps Upmix. It appears that the plugin was actually made by a Swiss company called Illusonic. Illusonic is best known for an audiophile signal processor that costs US$12,000 - $24,000 depending on number of channels. It’s for people who have lots of cash and very high-end hi-fi/surround sound systems. Illusonic has apparently also done signal processing work for Schoeps, Sennheiser and probably Logitech.

The purpose of upmixing is to convert mono to stereo or stereo to surround. Off the bat, I have reservations about applying upmixing to the spoken voice, which is inherently monaural because it comes from a single source, the mouth. That’s why the human voice is not recorded in stereo. There are people who argue for recording in stereo, but it’s a pretty fringe view.

I think that Schoeps Upmix is actually designed for recordings that could properly have been done in stereo, but were done in mono. I’m not sure what to make of the Schoeps demo video (see post #20), which uses a mono recording made in a highly reverberant space that is problematic regardless of upmixing. The demo doesn’t give me a lot of confidence in the product, which in any event seems to have a very small user base. There are very few references to it on the internet. I can’t find a single discussion about using this plugin with voice over or dialogue, notwithstanding the fact that Schoeps mikes are widely used in high-end film production.

With respect to the substance of this thread, I remain of the view (see post #22) that neither a reverb plugin nor this one will address the issue, which on the face of it is matching narration recorded in one space to music recorded in another. I think that the RX Ambience Match module is likely to be more fruitful.
 
Last edited:
It retails at $999 USD?

Yes, the module that I’m talking about is only in RX Advanced, which is very expensive. That’s why I said earlier that you should try to work with someone who has a copy. Despite the expense, RX is very widely used. Just about any professional sound recordist or editor will have it, as will many semi-pros. You don’t need the latest RX either. This feature has been included for several years.

As I mentioned above, a trial copy is not a solution because you can’t save your work.

There are also manual ways to address this kind of issue, but I don’t think that throwing a reverb or upmixing plugin at the recording is going to do what you need.
 
Last edited:
Reverb is hard to get right and to understand. There are many factors and tweaks that are needed on a per basis situation. Always think of the "space" you want the vocals or instruments to sit in...if it's a narration, WHERE is the narration taking place? Hall, inside a cave, small room, car..etc etc...there are post reverbs that are made specifically to place sounds in a certain space.

However, there are a few things to learn and think about as well. Are you using pre delay? more often than not the reverb needs to be rolled off at roughly 250Hz. Generally the highs above 5k are also tamed as to not introduce sibilance and keep the vocal natural. You would be surprised at how proper EQ on a reverb will make the source sound...for better or worse.

Any mud or wonkiness in the vocal should be addressed first if possible of course. Always keep in mind what the objective is for reverb: it is creating false ambience where there is none, or to change the setting of the source. Once you have this figured out...then it's a matter of tweaking...no reverb that I am aware of has a set and forget preset (at least none that I've used).

Hope this helps a bit.
 
I did a narration thing recently, and this Valhalla Room patch worked nicely. Doesn't mean you'll like it, but maybe a decent start? You can cut and paste this directly into VRoom...

<ValhallaRoom pluginVersion="1.5.1" presetName="vi two" mix="1" predelay="0" decay="0.0065065063536167144775" HighCut="0.53020131587982177734" earlyLateMix="0.17100000381469726562" lateSize="0.5" lateCross="1" lateModRate="0.090909093618392944336" lateModDepth="0" RTBassMultiply="0.3333333432674407959" RTXover="0.090909093618392944336" RTHighMultiply="0.44444447755813598633" RTHighXover="0.53020131587982177734" earlySize="0.027927927672863006592" earlyCross="0.10000000149011611938" earlyModRate="0" earlyModDepth="0" earlySend="0" diffusion="1" type="0.083333335816860198975"/>
 
I did a narration thing recently, and this Valhalla Room patch worked nicely. Doesn't mean you'll like it, but maybe a decent start? You can cut and paste this directly into VRoom...

<ValhallaRoom pluginVersion="1.5.1" presetName="vi two" mix="1" predelay="0" decay="0.0065065063536167144775" HighCut="0.53020131587982177734" earlyLateMix="0.17100000381469726562" lateSize="0.5" lateCross="1" lateModRate="0.090909093618392944336" lateModDepth="0" RTBassMultiply="0.3333333432674407959" RTXover="0.090909093618392944336" RTHighMultiply="0.44444447755813598633" RTHighXover="0.53020131587982177734" earlySize="0.027927927672863006592" earlyCross="0.10000000149011611938" earlyModRate="0" earlyModDepth="0" earlySend="0" diffusion="1" type="0.083333335816860198975"/>

Thanks Andre, I'll try it out. :)
 
Top Bottom