Hi,
please add an auto leveler like Waverider or Vocalrider to Nuendo and make it available in the strip. Preferably with AI or something. It’s trendy right now
Thanks
Hi,
please add an auto leveler like Waverider or Vocalrider to Nuendo and make it available in the strip. Preferably with AI or something. It’s trendy right now
Thanks
Just use those plug-ins. Or are they terrible?
+1 Agreed.
It would be nice to have one, but it can be achieved using native tools.
I was very impressed with Izotope Nectar 4’s new Auto Level module as it automatically attenuates silent parts. Works way better than Waverider or Vocalrider.
But you can get better results by combining AI Detect Silence and Squasher
Try Detect Silence, we have covered it here
None of VST levelers are good enough, because of none look ahead possibility. Also possibility of VST plugins to write level automation is very limited. Actually best is Izotope RX leveler, which is RX internal only, because it needs to read whole file several times. Internal process which will read clips loudness, then change clip volume to regarding clip loudness, and write volume automation to compensate clip dynamics will very usefull.
It has the same issue as it is basically a compressor with a makeup gain and like other compressors with a makeup gain it is not good wih silence parts where it tends to boosts gain… also it’s algorithm has not been updated from last 10 years so it is not very intelligent in my opinion. Especially for long clips.
As mentioned above Nuendo has better tools by which you can quickly get extremely good results with precise fine grain control to achieve the desired gain riding.
Waves Vocalrider does…
Fredo
I´m absolutely not a fan of all automatic reatime vocal riders.
Here is a new solution with way better and perfect results -
Noiseworks Gainaimpro uses ARA and it has Volume Rider/Sibilance/ Breath Controls:
Looks like a great tool for this.
AFAIK all supplied Steinberg dynamics plugins are look-ahead by design. You can switch off look ahead on some of them by activating the ‘live’ button.
For proper dialogue leveling is necessary to have estimated 3 seconds look ahead period. (100 or 200 miliseconds are usable for limiters only). I tested it on off speed applications. Good solution is combinig VU period (400 ms) with 3 sec period. Best solutions is if application is able to get overall loudness of region before it starts leveling. Which of mentioned VST solutions have 3 sec look ahead possibility ?
Martin
What is the purpose of that and how does that change the quality / characteristics of the auto levelling? I’m guessing you must be needing loudness levelling or levelling according to RMS values. I think I must have missed some of the details of this topic (I will read again).
Looks like a good solution to the OP’s request.
It is perfect to get smooth correcting of dynamic peaks of dialogue. It’s like you are watching waveform and pull fader down before large peaks, and slow pull fader up when dialogue is low. With that long look ahead period it didnt listen as compressed.
Yes, as GainAimPro uses ARA, I think this is only solution for Nuendo now.
I think the idea with these plugins is to basically ‘replace’ the manual riding of a fader. When we ride levels manually we don’t react as quickly as a compressor / limiter does (for the most part) so these plugins would be closer to manual. By running the plugin you would hypothetically not have to automate levels manually on whatever track you apply it.
Personally I never found that it was faster than me just doing it manually but maybe I just have to test the latest versions out there.
APU Loudness Compressor does a maximum of 3s of lookahead (you have to increase the lookahead budget in “Settings”.
It is a great tool all around, very cheap and with a very active developer. Highly recommend.
Yes, there is 3 sec. Good tip.
However … if inserted it causes 3 sec delay on start.
Now testing in DOP mode.
I’m afraid that will be the case with anything you put a lookahead of 3sec. Can’t go against physics!
While you are not using live inputs it have to be possible to fill buffers faster than realtime.
However I know that actually it is not possible.