Old 07-04-2014, 05:49 PM   #1
saddle
Human being with feelings
 
saddle's Avatar
 
Join Date: Feb 2009
Posts: 155
Default Help with Hornet Auto Gain Pro

This might sound silly, but I'm having trouble getting this plugin to work for me on a vocal track. I understand how to use the plugin. Setting write automation and read it back. Tweak the controls... No problems with any of the setup. But I just seem to be having a hard time getting a combination of settings that really level the vocal track out.

I've tried peak and rms. RMS works best for this. I'm using internal as the source, to just level it to a constant signal.

Tried attack and release all over the place, it's just that I get too much overshoot in quiet parts, too loud in others.

What settings are you using? It's fairly dynamic vocal track. Maybe I need to so some manual automation first to level it out. I'm using Reaper x64 4.62. But that part is fine.

Any ideas?
Thanks,
Saddle
__________________
74 Takamine F-450s (Pre-litigation) 62 Martin D-28 : 2004 Fender Squier Tele
Rouge Acoustic Electric (work guitar.) Various other guitars and toys.
Reaper x64. Pro Tools 11.3.1
saddle is offline   Reply With Quote
Old 07-04-2014, 06:21 PM   #2
tut
Human being with feelings
 
Join Date: Feb 2012
Posts: 47
Default

Have you tried using JS:LOSER/WhiteNoise as the reference track? It seems to work pretty for me, give it a whirl. Good luck.
tut is offline   Reply With Quote
Old 07-04-2014, 06:31 PM   #3
saddle
Human being with feelings
 
saddle's Avatar
 
Join Date: Feb 2009
Posts: 155
Default

Quote:
Originally Posted by tut View Post
Have you tried using JS:LOSER/WhiteNoise as the reference track? It seems to work pretty for me, give it a whirl. Good luck.
I haven't. I figured the internal signal source would work ok. I'll see what happens with it.

Thanks
__________________
74 Takamine F-450s (Pre-litigation) 62 Martin D-28 : 2004 Fender Squier Tele
Rouge Acoustic Electric (work guitar.) Various other guitars and toys.
Reaper x64. Pro Tools 11.3.1
saddle is offline   Reply With Quote
Old 07-04-2014, 07:49 PM   #4
saddle
Human being with feelings
 
saddle's Avatar
 
Join Date: Feb 2009
Posts: 155
Default

Really strange if you ask me... but, with the external whitenoise, I'm getting automation that I would expect. But using the internal reference, it's all over the place.

tut, Have you ever used the internal source?
__________________
74 Takamine F-450s (Pre-litigation) 62 Martin D-28 : 2004 Fender Squier Tele
Rouge Acoustic Electric (work guitar.) Various other guitars and toys.
Reaper x64. Pro Tools 11.3.1
saddle is offline   Reply With Quote
Old 07-04-2014, 08:02 PM   #5
tut
Human being with feelings
 
Join Date: Feb 2012
Posts: 47
Default

No, I havent to be honest. Somebody turned me on to that trick when I first got AutoGain and I really didn't dive in any deeper. I'll have a mess with it later on and see if I can make any sense of it.
tut is offline   Reply With Quote
Old 07-05-2014, 12:54 AM   #6
Arran
Human being with feelings
 
Join Date: Aug 2012
Location: Bolton, UK
Posts: 234
Default

Quote:
Originally Posted by saddle View Post
Really strange if you ask me... but, with the external whitenoise, I'm getting automation that I would expect. But using the internal reference, it's all over the place.

tut, Have you ever used the internal source?
I find exactly the same thing, particularly with trying to use Peak settings. I've previously emailed the developer about this, and I know he checks these forums, so this should add a bit of weight to my findings. In the past he's sorted these bugs very quickly, if it is a bug - I think it may just be the difference in rms value between external noise/tone and the internal reference (presuming that even in Peak mode there is still some small degree of rms smoothing).
My normal vocal processing is: Autogain with external noise reference writing automation to the prefade track envelope to pre-level the track, then compression/EQ/fairy dust, then Autogain with an external reference of everything except vocals and drums to write the post fader automation, followed by some manual tweaking.
Arran is offline   Reply With Quote
Old 07-05-2014, 06:24 AM   #7
saddle
Human being with feelings
 
saddle's Avatar
 
Join Date: Feb 2009
Posts: 155
Default

Quote:
Originally Posted by Arran View Post
I find exactly the same thing, particularly with trying to use Peak settings. I've previously emailed the developer about this, and I know he checks these forums, so this should add a bit of weight to my findings. In the past he's sorted these bugs very quickly, if it is a bug - I think it may just be the difference in rms value between external noise/tone and the internal reference (presuming that even in Peak mode there is still some small degree of rms smoothing).
My normal vocal processing is: Autogain with external noise reference writing automation to the prefade track envelope to pre-level the track, then compression/EQ/fairy dust, then Autogain with an external reference of everything except vocals and drums to write the post fader automation, followed by some manual tweaking.
The developer has helped me with a couple of issues very fast before. Maybe he will respond.

I'll try leaving out the drums next time. Good tip, thanks!

What settings for attack and release have you found most helpful? Do you limit your db +- to less than 12?

Thanks
Saddle
__________________
74 Takamine F-450s (Pre-litigation) 62 Martin D-28 : 2004 Fender Squier Tele
Rouge Acoustic Electric (work guitar.) Various other guitars and toys.
Reaper x64. Pro Tools 11.3.1
saddle is offline   Reply With Quote
Old 07-05-2014, 12:23 PM   #8
Arran
Human being with feelings
 
Join Date: Aug 2012
Location: Bolton, UK
Posts: 234
Default

I tend to use it in Peak mode with the attack and release times around 300-600ms for both Main and Reference. Gain scaling is usually around 60.
I aim for around +- 3dB of gain riding in either direction, sometimes the RMS value of the vocal doesn't match the perceived level and I'll either edit the automation curve or use max gain adjustment to limit it.

There is a particular bit of weirdness that I've come across in that the automation points once written seem to precede the waveform, AutoGain doesn't have a look ahead feature, or report any latency, so it must be happening (if it is...) afterwards.
I've (hopefully) attached 3 screenshots: one is of a sine wave with abrupt level changes, one is an actual vocal with the points highlighted and inverted for easier reference, the final screenshot is the same vocal with the points shifted to align better.
In the sine wave pic the points precede the wave change by 200ms.
Attached Images
File Type: jpg AutoGTone.jpg (32.4 KB, 402 views)
File Type: jpg Points Normal.jpg (26.1 KB, 359 views)
File Type: jpg Points shifted.jpg (26.5 KB, 353 views)
Arran is offline   Reply With Quote
Old 07-06-2014, 11:07 AM   #9
HoRNet
Human being with feelings
 
Join Date: Feb 2011
Posts: 171
Default

Hi, I have to test the issue by myself yet but I think it's due to how reaper treats the automation probably. AutoGain pro writes the automation from the GUI thread because other hosts require that (they don't accept any automation from the audio thread) the desync it's probably due to the way audio and GUI thread are handled.

I'll check in other hosts and let you know!

Saverio
HoRNet is offline   Reply With Quote
Old 07-06-2014, 08:43 PM   #10
saddle
Human being with feelings
 
saddle's Avatar
 
Join Date: Feb 2009
Posts: 155
Default

Quote:
Originally Posted by Arran View Post
I tend to use it in Peak mode with the attack and release times around 300-600ms for both Main and Reference. Gain scaling is usually around 60.
I aim for around +- 3dB of gain riding in either direction, sometimes the RMS value of the vocal doesn't match the perceived level and I'll either edit the automation curve or use max gain adjustment to limit it.

There is a particular bit of weirdness that I've come across in that the automation points once written seem to precede the waveform, AutoGain doesn't have a look ahead feature, or report any latency, so it must be happening (if it is...) afterwards.
I've (hopefully) attached 3 screenshots: one is of a sine wave with abrupt level changes, one is an actual vocal with the points highlighted and inverted for easier reference, the final screenshot is the same vocal with the points shifted to align better.
In the sine wave pic the points precede the wave change by 200ms.
Arran, That is really odd. I ran a test last night. Using whitenoise as the source but with it set to RMS. I looked to see if the automation was leading the waveform as your pictures show. But the automation isn't abrupt enough to see a definite start point. I'll have to try again with a different vocal track.
__________________
74 Takamine F-450s (Pre-litigation) 62 Martin D-28 : 2004 Fender Squier Tele
Rouge Acoustic Electric (work guitar.) Various other guitars and toys.
Reaper x64. Pro Tools 11.3.1
saddle is offline   Reply With Quote
Old 07-07-2014, 06:31 AM   #11
amagalma
Human being with feelings
 
amagalma's Avatar
 
Join Date: Apr 2011
Posts: 3,451
Default

Quote:
Originally Posted by Arran View Post
There is a particular bit of weirdness that I've come across in that the automation points once written seem to precede the waveform, AutoGain doesn't have a look ahead feature, or report any latency, so it must be happening (if it is...) afterwards.
I've (hopefully) attached 3 screenshots: one is of a sine wave with abrupt level changes, one is an actual vocal with the points highlighted and inverted for easier reference, the final screenshot is the same vocal with the points shifted to align better.
In the sine wave pic the points precede the wave change by 200ms.
Same thing happens with MAutoVolume, so this could be a Reaper thing...
amagalma is offline   Reply With Quote
Old 07-07-2014, 09:32 AM   #12
Tod
Human being with feelings
 
Tod's Avatar
 
Join Date: Jan 2010
Location: Kalispell
Posts: 14,745
Default

Quote:
Originally Posted by Arran View Post
I've (hopefully) attached 3 screenshots: one is of a sine wave with abrupt level changes, one is an actual vocal with the points highlighted and inverted for easier reference, the final screenshot is the same vocal with the points shifted to align better.
In the sine wave pic the points precede the wave change by 200ms.
Wow, everything in those picts looks wrong. Even if you could move the envelopes ahead I don't think it would work out. I don't know, you'd have to create a single line to match the envelope and then overlay it to tell for sure.
Tod is offline   Reply With Quote
Old 07-07-2014, 02:07 PM   #13
HoRNet
Human being with feelings
 
Join Date: Feb 2011
Posts: 171
Default

Can anybody try waves Vocal Rider?

I've tested the shift issue with Reaper on OSX and it seems to be happening there too, but i don't see the same thing on ableton live

I'm going to do the same thing on Logic X and Studio One...

Saverio
hornetplugins.com
Attached Images
File Type: png reaper-osx.png (36.5 KB, 299 views)
File Type: png live-osx.png (26.7 KB, 329 views)
HoRNet is offline   Reply With Quote
Old 07-07-2014, 02:37 PM   #14
HoRNet
Human being with feelings
 
Join Date: Feb 2011
Posts: 171
Default

These are the results with studio one, it seems like we are in sync here, so i must conclud that is some sort of syncing issue in Reaper...

I will try to put a fix in place for the desync!

Saverio
hornetplugins.com
Attached Images
File Type: png studio-one-osx.png (24.1 KB, 327 views)
HoRNet is offline   Reply With Quote
Old 07-07-2014, 03:27 PM   #15
saddle
Human being with feelings
 
saddle's Avatar
 
Join Date: Feb 2009
Posts: 155
Default

Quote:
Originally Posted by HoRNet View Post
These are the results with studio one, it seems like we are in sync here, so i must conclud that is some sort of syncing issue in Reaper...

I will try to put a fix in place for the desync!

Saverio
hornetplugins.com
I'm wondering if the number of tracks and plugins total might be affect the results between what Arran and you have seen? Total PDC, or a plugin like Nebula. Thanks for looking at it!


Thanks
Saddle
__________________
74 Takamine F-450s (Pre-litigation) 62 Martin D-28 : 2004 Fender Squier Tele
Rouge Acoustic Electric (work guitar.) Various other guitars and toys.
Reaper x64. Pro Tools 11.3.1
saddle is offline   Reply With Quote
Old 07-07-2014, 04:11 PM   #16
HoRNet
Human being with feelings
 
Join Date: Feb 2011
Posts: 171
Default

Quote:
Originally Posted by saddle View Post
I'm wondering if the number of tracks and plugins total might be affect the results between what Arran and you have seen? Total PDC, or a plugin like Nebula. Thanks for looking at it!


Thanks
Saddle
No i don't think so unless Reaper changes the sync between audio and gui thread dynamically which is something really complicate.
But that won't impact the fix i have in mind, fortunately Reaper is one of the hosts that accept automation data from the audio thread, so i will simply write it from there and we should be fine

Saverio
hornetplugins.com
HoRNet is offline   Reply With Quote
Old 07-08-2014, 12:30 AM   #17
Arran
Human being with feelings
 
Join Date: Aug 2012
Location: Bolton, UK
Posts: 234
Default

Quote:
Can anybody try waves Vocal Rider?
I might be able to do that tonight. I downloaded MAutovolume yesterday and I can confirm that it exhibits the same sync issues.
MAutovolume responds much faster even at it's default settings than HorNet in Peak mode. In the brief amount of time I had to play with it, which was only using one vocal track, I felt that HorNet was smoother and more predictable, but I'd still like to see faster attack and release times and maybe an RMS size control.

Quote:
What settings for attack and release have you found most helpful? Do you limit your db +- to less than 12?

Thanks
Saddle
One thing I forgot to mention is the Hi and Low pass filtering - I always adjust these to only let through the frequencies in the reference track that I think are masking the vocals, so the LPF is normally wound all the way down to 3kHz and the HPF can be 300Hz or higher.
Arran is offline   Reply With Quote
Old 07-08-2014, 03:35 AM   #18
HoRNet
Human being with feelings
 
Join Date: Feb 2011
Posts: 171
Default

Quote:
Originally Posted by Arran View Post
One thing I forgot to mention is the Hi and Low pass filtering - I always adjust these to only let through the frequencies in the reference track that I think are masking the vocals, so the LPF is normally wound all the way down to 3kHz and the HPF can be 300Hz or higher.
This is almost what the original AutoGain does internally

Saverio
hornetplugins.com
HoRNet is offline   Reply With Quote
Old 07-08-2014, 04:32 AM   #19
Arran
Human being with feelings
 
Join Date: Aug 2012
Location: Bolton, UK
Posts: 234
Default

Quote:
Originally Posted by HoRNet View Post
This is almost what the original AutoGain does internally

Saverio
hornetplugins.com
It's taken all those extra controls, just to get me back where we started...


Quote:
Really strange if you ask me... but, with the external whitenoise, I'm getting automation that I would expect. But using the internal reference, it's all over the place.
Here's a LICECAP comparison. I multed a vocal to 2 tracks, put an instance of AutoGain on each, set the parameters the same, fed one with an external pink noise source, and normalised the reference levels.

Arran is offline   Reply With Quote
Old 07-08-2014, 07:00 AM   #20
saddle
Human being with feelings
 
saddle's Avatar
 
Join Date: Feb 2009
Posts: 155
Default

Quote:
Originally Posted by Arran View Post
It's taken all those extra controls, just to get me back where we started...


Here's a LICECAP comparison. I multed a vocal to 2 tracks, put an instance of AutoGain on each, set the parameters the same, fed one with an external pink noise source, and normalised the reference levels.
That is a striking difference in response. When I first looked at this I wondered if the HP and LP settings would help when using the internal reference. I haven't used them to this point. But they are disabled when set to internal.

+100 for more range (lower) in the Attack and Release controls. RMS size would be helpful.

Great plugin. Certainly a time saver!
__________________
74 Takamine F-450s (Pre-litigation) 62 Martin D-28 : 2004 Fender Squier Tele
Rouge Acoustic Electric (work guitar.) Various other guitars and toys.
Reaper x64. Pro Tools 11.3.1
saddle is offline   Reply With Quote
Old 07-08-2014, 09:33 AM   #21
HoRNet
Human being with feelings
 
Join Date: Feb 2011
Posts: 171
Default

A white noise reference is something completely different from the "internal" setting.

The "internal" setting sets a static reference to the value in dB set with the gain/level knob, each sample that comes out from the attack/release section is checked against that internal value and corresponding gain is applied. That gain is instantaneous so if the signal is at -20 and you set the reference level to -18, 2 dBs are applied, instantaneously.

If you use the external reference and put a white noise there, you are passing through another combination of attack and release control, the difference between the two main envelope followers is the used to compute the gain to be applied. This actually brings everything closer to a RMS processor. The envelope follower on the external sidechain is necessary because otherwise you would have a vey fluctuating reference signal and too fast gain changes that would render the effect almost useless.

Hope this clears up the differences!

Let me know if I can help you further

Saverio
hornetplugins.com
HoRNet is offline   Reply With Quote
Old 07-08-2014, 11:52 AM   #22
Arran
Human being with feelings
 
Join Date: Aug 2012
Location: Bolton, UK
Posts: 234
Default

Thanks, I'd figured that the internal reference was a static digital value, a digital DC if you like. What I was hoping to show was how it compares to an external reference. Given that the internal ref was provided to negate the need for an external noise source, with the faff of an extra track and setting up sends etc, I don't think it performs as well and could be improved by making it behave more like a noise source.
Arran is offline   Reply With Quote
Old 07-08-2014, 02:19 PM   #23
HoRNet
Human being with feelings
 
Join Date: Feb 2011
Posts: 171
Default

Quote:
Originally Posted by Arran View Post
Thanks, I'd figured that the internal reference was a static digital value, a digital DC if you like. What I was hoping to show was how it compares to an external reference. Given that the internal ref was provided to negate the need for an external noise source, with the faff of an extra track and setting up sends etc, I don't think it performs as well and could be improved by making it behave more like a noise source.
Yes, that's something to consider, I'll try to change the behavior of that and fix the issue with reaper as soon as possible!

Saverio
HoRNet is offline   Reply With Quote
Old 07-08-2014, 03:33 PM   #24
Arran
Human being with feelings
 
Join Date: Aug 2012
Location: Bolton, UK
Posts: 234
Default

Awesome! Looking forward to seeing a good plugin turn great.
Arran is offline   Reply With Quote
Old 07-12-2014, 01:32 AM   #25
Arran
Human being with feelings
 
Join Date: Aug 2012
Location: Bolton, UK
Posts: 234
Default

From this thread about the changes in 4.72
http://forum.cockos.com/showthread.php?t=142613

Quote:
Originally Posted by nicholas View Post
In short, raising the thread priority will cause audio to be processed ahead of other threads, such as graphics.
Is this going to have any bearing on the latency issue, or it's potential fix?
Arran is offline   Reply With Quote
Old 07-12-2014, 03:49 AM   #26
HoRNet
Human being with feelings
 
Join Date: Feb 2011
Posts: 171
Default

Quote:
Originally Posted by Arran View Post
From this thread about the changes in 4.72
http://forum.cockos.com/showthread.php?t=142613



Is this going to have any bearing on the latency issue, or it's potential fix?
Yes it could be but it sounds strange to me since before writing automation I acquire a lock so thread should be synced at that point....
HoRNet is offline   Reply With Quote
Old 07-24-2014, 08:46 AM   #27
HoRNet
Human being with feelings
 
Join Date: Feb 2011
Posts: 171
Default

Just a quick update on the issue.

We have released today an updated version of the AutoGain Pro that addresses mainly the automation accuracy and predictability and the behavior of the internal reference (it's smoother now)

Unfortunately i cannot do anything about the automation desync because it's because of Reapers' "anticipative fx processing" option, the only way to bring back in sync everything is to disable that option while autogain pro is writing the automation.

Saverio
hornetplugins.com
HoRNet is offline   Reply With Quote
Old 07-25-2014, 03:11 AM   #28
Arran
Human being with feelings
 
Join Date: Aug 2012
Location: Bolton, UK
Posts: 234
Default

Quote:
Originally Posted by HoRNet View Post
Just a quick update on the issue.

We have released today an updated version of the AutoGain Pro that addresses mainly the automation accuracy and predictability and the behavior of the internal reference (it's smoother now)

Unfortunately i cannot do anything about the automation desync because it's because of Reapers' "anticipative fx processing" option, the only way to bring back in sync everything is to disable that option while autogain pro is writing the automation.

Saverio
hornetplugins.com

Great, behaves much better with the internal reference now. Even with 'Anticipative FX' enabled the sync difference appears to be reduced to around 80ms instead of around 190ms.
I did a quick test on bass and I think it's inevitable that I'll still end up sliding the points around, this is with anticipative fx enabled, so a degree of 'Look ahead' processing:



Can we have faster attack times for the next release please?
And of course, many thanks for this update!
Arran is offline   Reply With Quote
Reply

Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump


All times are GMT -7. The time now is 06:04 AM.


Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2024, vBulletin Solutions Inc.