16th July 2012
|
#1 | | Lives for gear
Joined: Mar 2011 Location: Bogustown, Europe
Posts: 1,274
Thread Starter | Heavy lifting with a single gain plugin = distortion?
I've had projects in the past where I for some reasons have had a very low level sent to the master bus. No problem, I thought, I'm working in 24 bit so let's just put a gain plugin on the master and crank it up some 12 dB or so.
Well, I ended up with severe distortion. Why is this? Badly coded gain plugin or is this due to something else? Is this a bad idea per se?
Nothing seemed to be clipping, I didn't go past the zero; not even close.
I noticed this on an old session where the piano (software) sounded really awful, I was about to dismiss that piano library altogether when I noticed the gain plugin on the master, removed it and - hey presto!
Why is this? Is it a must to put a gain plugin on every track in order to increase the gain?
|
| |
19th July 2012
|
#2 | | Lives for gear
Joined: Mar 2011 Location: Bogustown, Europe
Posts: 1,274
Thread Starter |
Bump
|
| |
20th July 2012
|
#3 | | Lives for gear
Joined: Mar 2011 Location: Bogustown, Europe
Posts: 1,274
Thread Starter |
Bump
|
| |
20th July 2012
|
#4 | | Gear Guru
Joined: Mar 2005 Location: Long Beach, CA
Posts: 17,429
|
As always, when diagnosing a problem, simplify the test conditions and isolate the elements so you can test each element, one at a time, changing conditions until you start seeing the bad behavior you're trying to analyze.
Off the top of my head, from your description, I'd hazard a guess that there's actually some other element in the mix of variables that you're not taking into account.
However, maybe it's just a bad plug. But, let's face it, a plug that raises gain has got to be one of the simplest plugins to implement. (And many DAWs simply make them unnecessary with track and individual clip trims, which really help your flexibility.)
|
| |
20th July 2012
|
#5 | | Lives for gear
Joined: Mar 2011 Location: Bogustown, Europe
Posts: 1,274
Thread Starter | Quote:
Originally Posted by theblue1 As always, when diagnosing a problem, simplify the test conditions and isolate the elements so you can test each element, one at a time, changing conditions until you start seeing the bad behavior you're trying to analyze.
Off the top of my head, from your description, I'd hazard a guess that there's actually some other element in the mix of variables that you're not taking into account.
However, maybe it's just a bad plug. But, let's face it, a plug that raises gain has got to be one of the simplest plugins to implement. (And many DAWs simply make them unnecessary with track and individual clip trims, which really help your flexibility.) | I'll do a more scientific test tomorrow and post the results. Might be of interest to some. Unless it shows that I'm wrong ...
|
| |
20th July 2012
|
#6 | | Lives for gear
Joined: Apr 2008 Location: London UK
Posts: 686
|
Sounds like your output is hitting the amp too hard.
|
| |
21st July 2012
|
#7 | | Lives for gear
Joined: Mar 2011 Location: Bogustown, Europe
Posts: 1,274
Thread Starter | Quote:
Originally Posted by scanner Sounds like your output is hitting the amp too hard. | What are you talking about? It's a DAW, there's no amp.
|
| |
21st July 2012
|
#8 | | Lives for gear
Joined: Aug 2008 Location: Bradford UK
Posts: 553
|
Does scanner mean your monitor amp after your conversion?
It is easy to clip your converters too.
A common mistake I see is having a "control room" output and stereo output in cubase to the same physical outputs on the converters.
|
| |
21st July 2012
|
#9 | | Lives for gear
Joined: Apr 2008 Location: London UK
Posts: 686
| Quote:
Originally Posted by crosscutred Does scanner mean your monitor amp after your conversion?
. | That's the one....
Try slowly turning down the main output fader.... if the output is hitting the speaker/monitor amp too hard the distortion will clear up. You'll then know if the gain problem is at the output of your DAW or generated by the gain plugin ....which I doubt. Quote:
Originally Posted by crosscutred A common mistake I see is having a "control room" output and stereo output in cubase to the same physical outputs on the converters. | Good call...
That sounds most likely.
You can have your mix peaking at mid scale, but if you output fader is at plus 10 you'll still overload you converters (if the metering is post fader, you'll see it peak, if pre fader, you wont.)
Same thing goes for bussing 2 sets of mixes to the output, as crosscutred said.
|
| |
21st July 2012
|
#10 | | Lives for gear
Joined: Mar 2011 Location: Bogustown, Europe
Posts: 1,274
Thread Starter |
But the master is sending the signal to conversion. And the master meter isn't any higher than normal. The gain plugin is obviously pre master output (ie what the master meter is showing). Am I missing something here?
|
| |
21st July 2012
|
#11 | | Lives for gear
Joined: Apr 2008 Location: London UK
Posts: 686
|
If you remove the gain plugin, then raise the master fader by the same amount as the gain plugin, does it have the same effect ?
|
| |
21st July 2012
|
#12 | | Lives for gear
Joined: Mar 2011 Location: Bogustown, Europe
Posts: 1,274
Thread Starter | Quote:
Originally Posted by scanner If you remove the gain plugin, then raise the master fader by the same amount as the gain plugin, does it have the same effect ? | It's not possible since the master fader starts at 0 and can only boost 6 dB or so. The gain plugin boosted maybe 20.
|
| |
21st July 2012
|
#13 | | Lives for gear
Joined: Apr 2008 Location: London UK
Posts: 686
|
Does a change in settings on the gain plugin register on the meter of your output fader ? ie. is you metering post fader ?
|
| |
21st July 2012
|
#14 | | Lives for gear
Joined: Mar 2011 Location: Bogustown, Europe
Posts: 1,274
Thread Starter | Quote:
Originally Posted by scanner Does a change in settings on the gain plugin register on the meter of your output fader ? ie. is you metering post fader ? | Yes. But it's certainly post insert so it shouldn't really make a difference, should it?
Edit: Appreciate the help
|
| |
21st July 2012
|
#15 | | Lives for gear
Joined: Aug 2008 Location: Bradford UK
Posts: 553
|
Do you have meters on the converter or it may have them in it's software. That will tell you if it is the converters.
|
| |
21st July 2012
|
#16 | | Lives for gear
Joined: Mar 2011 Location: Bogustown, Europe
Posts: 1,274
Thread Starter | Quote:
Originally Posted by crosscutred Do you have meters on the converter or it may have them in it's software. That will tell you if it is the converters. | Sure the converters will tell me if they're overloaded. This might have been the case though, since it was some time ago. I need to replicate it.
One thing though, the signal level going to the converters should be the same if the master bus meter in the DAW shows the same level, no? Should it really matter to the converters what my internal gain structure is (in this case in Ableton)?
|
| |
22nd July 2012
|
#17 | | Lives for gear
Joined: Aug 2008 Location: Bradford UK
Posts: 553
|
Not necessarily, if you are sending your control room out and mix out to the same set of converters the meters in the daw will only show the mix level, but you are actually doubling the power to the converters. Which will add 3db
|
| |
22nd July 2012
|
#18 | | Lives for gear
Joined: Mar 2011 Location: Bogustown, Europe
Posts: 1,274
Thread Starter | Quote:
Originally Posted by crosscutred Not necessarily, if you are sending your control room out and mix out to the same set of converters the meters in the daw will only show the mix level, but you are actually doubling the power to the converters. Which will add 3db | But still, that wouldn't change anything; if the meter on the 2bus peaks at -3, regardless of my gain structure, the signal level going to the converters would be the same. But maybe there was something else going on. I need to do a null test.
|
| |
23rd July 2012
|
#19 | | Lives for gear
Joined: Aug 2008 Location: Bradford UK
Posts: 553
|
I don't know ableton but on cubase if you send your main stereo bus out to, for example, out 1&2 on the converters.... and.... you have "control room" activated and that is sending to 1&2 then the meters in the daw will only tell you the level of the main stereo buss but you are actually sending it twice to the same set of converters.
If you either have the main stereo bus "not connected" or control room disabled then the meters in the daw will tell you what is going out, and you will still hear it.
It is an easy mistake to make because the daw will have "control room" enabled by default and it makes sense to send the main mix to the converters so that's often the first thing to do when you set up a session.
You could eliminate this as a cause by rendering the mix then analysing/ listening to it in different software.
|
| |
23rd July 2012
|
#20 | | Lives for gear
Joined: Mar 2011 Location: Bogustown, Europe
Posts: 1,274
Thread Starter | Quote:
Originally Posted by crosscutred I don't know ableton but on cubase if you send your main stereo bus out to, for example, out 1&2 on the converters.... and.... you have "control room" activated and that is sending to 1&2 then the meters in the daw will only tell you the level of the main stereo buss but you are actually sending it twice to the same set of converters.
If you either have the main stereo bus "not connected" or control room disabled then the meters in the daw will tell you what is going out, and you will still hear it.
It is an easy mistake to make because the daw will have "control room" enabled by default and it makes sense to send the main mix to the converters so that's often the first thing to do when you set up a session.
You could eliminate this as a cause by rendering the mix then analysing/ listening to it in different software. | There's a cue mix that's set to 1-2 (same as main mix) by default. This can't be disabled AFAIK, you can turn down the volume, this lowers the metronome. The level doesn't go up if I turn the cue mix all the way up and send ir to the same outputs as my main mix.
|
| |
23rd July 2012
|
#21 | | Lives for gear
Joined: Aug 2008 Location: Bradford UK
Posts: 553
|
Try sending the main mix to somewhere else, or to nowhere and see if you can still hear it. If so that's your problem, or one of them.
Work on gainstaging your mixes to end up at right level as you go. It just takes practice, like getting really good at anything.
|
| |
23rd July 2012
|
#22 | | Lives for gear
Joined: Mar 2011 Location: Bogustown, Europe
Posts: 1,274
Thread Starter | Quote:
Originally Posted by crosscutred Try sending the main mix to somewhere else, or to nowhere and see if you can still hear it. If so that's your problem, or one of them.
Work on gainstaging your mixes to end up at right level as you go. It just takes practice, like getting really good at anything. | There's no sound going through the cue, only metronome. If I send the main mix to output 1, and the cue mix to channel 2 there's only signal going to the left channel (I'm on holiday and only have two output channels to play with ...). Don't know if you can send the mix through the cue or if it's only to send click to the musicians. I always do headphone mixes via auxes as I do on a console.
Gainstaging is not really a problem for me, I used to end up with a very low level towards the end of my mixes years ago though.
My point, however, during this whole conversation has been this:
I don't see how internal gainstaging can make a difference to the converters if the master level is the same!
|
| |
23rd July 2012
|
#23 | | Lives for gear
Joined: Aug 2008 Location: Bradford UK
Posts: 553
|
Well like I said, I don't know abelton, but I have explained twice how you can have a different level hitting the converters than the level of the mix buss in the daw.
As for gainstaging, if your mixes are coming out 20db lower than you want then you need to work on gainstaging. Especially if they are distorting when you bump them up.
If this is a past problem that you have moved on from and everything is cool now then don't worry about it.
Enjoy the music.
|
| |
25th July 2012
|
#24 | | Lives for gear
Joined: Mar 2011 Location: Bogustown, Europe
Posts: 1,274
Thread Starter | Quote:
Originally Posted by crosscutred Well like I said, I don't know abelton, but I have explained twice how you can have a different level hitting the converters than the level of the mix buss in the DAW. | Well the only thing you've said about it is the case where you send the main mix and a monitor mix to the same outputs.
And this is not the case, as I've mentioned. But most importantly, I still don't see what that has to do with using a massive amount of gain in one place as opposed to a little in many places (ie gain staging).
This has been my issue all along.
|
| |
25th July 2012
|
#25 | | Lives for gear
Joined: Aug 2008 Location: Bradford UK
Posts: 553
|
Ok, so does the rendered/exported mix still sound distorted when you are playing it on other software?
If yes, the answer lies with the gain plugin or the gain structure of the mix.
Places to look; is the gain plug is before a master compressor, limiter or other mix buss processing on the main buss? is the gain plug simply making existing distortion audible? is there a fault with the gain plug programming? (unlikely but you can discount that by trying another brand of gain plug)
If no, then the answer lies with your monitoring.
Places to look; is the level from the DAW the same as level to the converters? are the converters clipping the analogue monitor setup at any stage? (amp, speakers etc..)
In the end the problem seems to stem from "using a massive amount of gain in one place as opposed to a little in many places (ie gain staging)" which can be avoided by gain staging your mix carefully from source to output.
I find that if there is a problem like this I can usually fix it by dropping all the faders and building the mix levels from the ground up again. If I have been working on a mix for longer than a couple of hours this is a useful tool to reassess your decision making regarding placing the elements in the mix anyway. One of the big problems of having instant recall on a mix (ie ITB) is that it's easy to become too attached to decisions made weeks or months ago that sounded good then but other elements have been added since which can skew the balance of the piece, and require revisiting earlier decisions.
I wouldn't bother with a null test, what information would that give you?
We await with bated breath to find out what the issue is, it may help us all to know if a particular gain plugin causes distortion, I expect the software designers would like to know too.
|
| |
25th July 2012
|
#26 | | Lives for gear
Joined: Apr 2010 Location: North Carolina
Posts: 638
| Quote:
Originally Posted by Sotsirc Well the only thing you've said about it is the case where you send the main mix and a monitor mix to the same outputs.
And this is not the case, as I've mentioned. But most importantly, I still don't see what that has to do with using a massive amount of gain in one place as opposed to a little in many places (ie gain staging).
This has been my issue all along. | Try the gain plug in on a different channel and see if the action is repeated.
Does Abelton allow Pre/post switches on a insert?
How does Abelton route the output of a plugin?
What 'gain' plug in is this?
Does Ableton have a gain control built into each track/channel? If so it may be possible that the plugin is controlling both its own gain and the track gain through a coding mistake. Or maybe the plugin uses a Midi CC control number to boost or cut - and that is also sending info to an internal gain control that also uses Midi CC control numbers.
|
| |
25th July 2012
|
#27 | | Lives for gear
Joined: Mar 2011 Location: Bogustown, Europe
Posts: 1,274
Thread Starter | Quote:
Originally Posted by crosscutred
I wouldn't bother with a null test, what information would that give you? | Well if the plugin is causing any type of distorsion, the files wouldn't null. It would tell me wether or not my brain is playing tricks on me. I've heard distorsion when using some limiters even without gain reduction.
But yes, your notion that the meter on my 2bus isn't necessarily the same level that hits my converters is something that I take to heart.
|
| |
25th July 2012
|
#28 | | Lives for gear
Joined: Aug 2008 Location: Bradford UK
Posts: 553
|
Which files would you be null testing against each other?
|
| |
25th July 2012
|
#29 | | Lives for gear
Joined: Apr 2011 Location: Vermont
Posts: 2,261
|
make a low level sine wave and render the file, then add your gain plug-in to boost it "a lot" and render that file...
then look at them with an analyzer...
they should be the same spectrally, only one will look louder...
__________________
"it's a long way to the top if you wanna rock -n- roll" -Jesus
|
| |
25th July 2012
|
#30 | | Lives for gear
Joined: May 2006
Posts: 3,152
| Quote:
Originally Posted by Sotsirc There's a cue mix that's set to 1-2 (same as main mix) by default. This can't be disabled AFAIK, you can turn down the volume, this lowers the metronome. The level doesn't go up if I turn the cue mix all the way up and send ir to the same outputs as my main mix. | The cue outputs can be set to something other than what the main outs are using. Setting them to some unused physical out will let you test whether it's the cue outs that are giving the problem. In Live it's easy to send things to the physical outs from tracks other than the master bus; the cue output is one such thing. Each audio / instrument track has an output destination which can be set to "External" and routed to the main outs, which will overload them as there's no metering internal to Live that registers the sum of what's hitting that output.
-synthoid
|
| | | |