Hi, big fan and user since first versions here. I was wondering when we will get the possibility of automatically “translating” the loudness ( Crescendo and diminuendo) of an imported vocal. So when a vocalist goes LOUDER the automation on the parameter track fro that gets added. One could first define the general LUFS of the whole vocal, then take the median as 0 and the louder and softer parts could get automation on the fly. If this already is possible I must have been blinded by all the fantastic possibilities 
Help appreciated!
Not fast but possible in 2 steps:
1/ DAW: Generate level automation in an exported CC1 midi file.
2/ Use a script (like mine “ExtractVelocityFromMidi”) to import the corresponding loudness.
Not easy to find the DAW option to do that (but I found it on Cubase).
This could be much simpler if I found a way to directly read a wav file and get the volume. But for now I can do it with these steps.
1 Like
Hi jfa , thanks for chiming in so quickly. Your work-around does sound interesting but it presupposes level automation in my DAW which I did not intend to have to do. I was hoping for a way to just naturally “follow” the singer’s intonation and variation of loudness and loudness of secondary and tertiary syllables, instead of having to add that manually. If I had to do that in the DAW I could just as well do it directly in SynthV, right? Secondly it sounds like you have a proprietary script there that I can find no mention of , maybe it is posted here somewhere? I am just fishing here, my guess is that maybe Cubase can automatically assign CC1 values to audio when it translates to midi, gonna see if Logic Pro does that as well when it does its audio-to-midi translation but I have never seen it done Any further tips gladly !! Cheers!
Ps. Dreamtonics ( don’t know if you ever read the forums) : this feature of automatically assigning loudness based on a median LUFS over the song would be a GREAT step towards excellency!!
My first point was to automate (with internal DAW function) the audio volume of a CC1 in a new midi track. Not at all to create it. That’s what I was checking with my DAW but not so easy to find (but it’s done).
A script made to be shared see my repository: jfa synthv-scripts
The only way to inform them is to contact them directly using their support mail.
Thanks. I got that wrong a bit. I’ll go check on Logic Pro now if it has that functionality. Thanks also for sharing thosescripts : I have been trying to do some myself but javascript ones don’t work no more in Logic Pro and will have to adapt to LUA. Yeah… I have tried before to get an answer from DT but that didn’t happen so… We’ll see in a couple of days what version 2.0 promises Cheers!
check out Melodyne as i think the MIDI export of the audio also have velocity/level information set.
Thanks fossile. I have Melodyne but was not aware it can save that info to midi. I wonder then if SynthV is able to read that fro the midi… then I’d be set. I will experiment.
Edit : I just checked and Melodyne does save loudness ( or velocity ) info but when I import it into SynthV and give it to Hayden the dynamics are nowhere near what I sung ( Very low to very hard) so somehow that info is NOT used as far as I can tell so far. Maybe I’m just missing something… anybody had more success getting the AI voice to really follow the performance of the “donor”?
The limitation comes from SynthV which only imports notes and pitchbend.
Do not confuse extracting midi notes from audio (SynthV already does this) and exporting velocities.
You have to find a way to extract the velocities to automation events (CC1 in my example as I do with Cubase). And then thanks to the script, read the CC1 controller and assign the loudness.
Thanks jfa, it seems that finally my braincell has sprung into action but I haven’t found a way to extract those to automation events in Logic Pro- I will search some more, maybe there’s a way with the Midi transform in Logic. I will then try your script. It seems a good idea to await the SynthV 2.0 demo in a the following days to see if they have anything planned for this, if not will write in. Thanks all!
You’re welcome! I’m still trying to find a way to extract loudness data from an audio file directly in a script or with additional software.
Velocity to CC
There is an outboard utility included with Garritan Personal Orchestra called “GPO Standard MIDI File Translator”. Essentially, it translates the traditional volume control for notes, namely note velocity, and substitutes it with CC#1 messages, as GPO uses the Mod Wheel (for most instruments) to control volume contour. MFX Velocity to CC does the same sort of thing, but as an in-host plugin (including real-time play!). Also, you could map other controllers (like Var1 and Var2 in the GPO player) to certain velocity ranges for other expression techniques you may dream up. Also works in any other synths you might find a use for!
I don’t know this software but the link is blocked by my bitdefender.
On my side, I just released a script (LoudnessFromAudio) using another external software.
Audiowaveform : audiowaveform
This tool is able to generate a json data. With my script, I automate the process to run this external tool and I recreate the loudness with the json.
The lack of this is only tested on Windows.
It is necessary to generate a version of this tool for your platform and then adapt the calling process inside the script. Too much to do I think.
1 Like
i have all the tencrazy plugins… just that he let the site cert expire so the HTTPS warning
On another isolated computer I tried to access the tencrazy website and they offer many interesting tools. The Velocity to CC tool is a Windows dll plugin compatible with some DAWs (not a standard VST2 or VST3 plugin so I can’t try it). But the main issue is Windows only, so out of scope (MAC user with Logic Pro).
My solution is not perfect but already works fine and is fast. The “audiowaveform” can be built for multiple platforms (like for linux, MAC etc.). But I don’t have MAC to build it.
I also hope that a future version of SynthV (v2?) will be able to extract velocities directly with midi.
2 Likes