" Gain Staging Done Correctly ..... everyone is wrong "
- Thread starter BenIfin
- Start date
SillyOctpuss
Shredder
- Messages
- 1,899
This was posted yesterday in the calibrating input levels thread
Post in thread 'Calibrating Input Level for Plugins' https://thegearforum.com/threads/calibrating-input-level-for-plugins.816/post-276688
Post in thread 'Calibrating Input Level for Plugins' https://thegearforum.com/threads/calibrating-input-level-for-plugins.816/post-276688
![[Nathan]](/data/avatars/m/0/613.jpg?1726885163){kind=avatar}
[Nathan]
Roadie
- Messages
- 398
Skip to 10:45
1) Set the interface gain to 0
2) Adjust the plugin trim input based on the spreadsheet
(The above is literally all anyone has ever recommended and somehow the public has resisted)
3) If your signal isnt clipping increase the signal until its safe enough and still not clipping
4) Whatever you increased the signal by, offset that by lowering the plugin input trim (If interface went up by 2db, lower trim by 2db etc)
The funny thing is all anyone is seeing in this video is their saviour that "plugin accuracy doesnt exist". The dude is saying the same thing all of us have been saying, he's just bolting on some more steps haha. He has a bunch of interfaces and had to cherry pick a behringer one to even show an interface "can" have this problem. The reality is that most interfaces wont have any problems being left at 0 and he had to go out of his way to find one bad enough where it is a problem. To me its kind of obvious if your level is so insanely low that you would increase the gain on the interface and reduce the gain in the plugin, but its good to point it out I guess.
Either way, the video is doing extremely well because people are biting on the clickbait and reading into it what they want to see. If they actually followed the guys advice its the same as anyone has ever said about the topic.
1) Set the interface gain to 0
2) Adjust the plugin trim input based on the spreadsheet
(The above is literally all anyone has ever recommended and somehow the public has resisted)
3) If your signal isnt clipping increase the signal until its safe enough and still not clipping
4) Whatever you increased the signal by, offset that by lowering the plugin input trim (If interface went up by 2db, lower trim by 2db etc)
The funny thing is all anyone is seeing in this video is their saviour that "plugin accuracy doesnt exist". The dude is saying the same thing all of us have been saying, he's just bolting on some more steps haha. He has a bunch of interfaces and had to cherry pick a behringer one to even show an interface "can" have this problem. The reality is that most interfaces wont have any problems being left at 0 and he had to go out of his way to find one bad enough where it is a problem. To me its kind of obvious if your level is so insanely low that you would increase the gain on the interface and reduce the gain in the plugin, but its good to point it out I guess.
Either way, the video is doing extremely well because people are biting on the clickbait and reading into it what they want to see. If they actually followed the guys advice its the same as anyone has ever said about the topic.
After moving from a Behringer UMC to Audient ID series, I have absolutely no problems with input gain.
I leave it at zero and I'm sounding better than ever, doesn't matter if it's higher output D-activators, EMG or vintage style tele single coils.
Agreed that it might be a problem for someone, but how does this translate in a full mix where it's all buried away anyway?
I leave it at zero and I'm sounding better than ever, doesn't matter if it's higher output D-activators, EMG or vintage style tele single coils.
Agreed that it might be a problem for someone, but how does this translate in a full mix where it's all buried away anyway?
Yes, some time ago I had problems with a Zoom AMS-22 (very noisy, unuseable with hi-gain captures in NAM, leaving input gain knob at minimum), so I measured the input with a little more gain on the knob (just till the point of clipping) and adjusted accordingly in NAM. It improves a little, of course. Anyway, that especific interface has not too much headroom either, so I couldn´t save much more than a couple dBs, and the noise was still too much. The conclusion is that the Zoom AMS-22 is not a valid interface for this use.
Then, swapping to a Audient EVO4, the problem with the noise is not there (or it´s much lower). I can leave the knob at minimum and use a noise gate with not aggresive setting.
At the end, it´s all the same concept We´ve been aware of thanks to some of our mates here.
ivanmancic
Groupie
- Messages
- 26
Well, I have an Antelope Zen Go, which is certainly not a low quality interface, but following his advice did make a noticeable difference in reducing idle hiss in Helix Native.
Zen Go does require +22dB boost to get to Helix Native calibrated level (which is what I normally use), but to get to near clipping level it should be set to +34, and then -12dB in Helix Native. If I would set it to 0dB, the difference would be even more prominent.
MirrorProfiles
Rock Star
- Messages
- 3,242
This should really have been a clue to record a hotter signal, most interfaces don't have their instrument inputs set up to produce such a quiet signal (they're typically in that 12dBu area). When the signal is this far below 0dBFS its going to have issues, no one is going to suggest otherwise.
I believe Antelope also have digitally stepped gain controls? You'll probably find the difference in setting the preamp to +22dB a lot more drastic than the additional 12dB (which is already likely well below the noise floor of your guitars and in use shouldn't make much difference on a higher end interface). Nothing wrong with going hotter still if you have headroom, but if you use different guitars and pedals with different levels, you'll find yourself adjusting settings a lot
- Messages
- 6,623
There's no hard rule imo, except for not wanting to clip (unless you want that lol). Listen with the ears and hands. I wish I could say I had a set formula for input level in Helix Native but I don't. It can vary from 0 to +4 db. Motu M4 is pushing some too.
The models in Native react differently so playing around with input gain can be useful.
The models in Native react differently so playing around with input gain can be useful.
What he said is clear enough. Does anyone have info. on where I can learn more about this subject - how some of the other technical details work - like in James Freeman's thread - using a sine wave, obtaining RMS value of that sine wave, understanding how that converts to an equivalent voltage, dbus etc.
I like how this Voldemort guy broke it down in the first video - does anyone know of other good videos that do this for some of the other topics mentioned above?
Edit: you know what, maybe I should just read James's thread in its entirety - great info in there! But if anyone does have any good articles or videos, please do share.
I like how this Voldemort guy broke it down in the first video - does anyone know of other good videos that do this for some of the other topics mentioned above?
Edit: you know what, maybe I should just read James's thread in its entirety - great info in there! But if anyone does have any good articles or videos, please do share.
Last edited:
deathbyguitar
Roadie
- Messages
- 158
This is part of the reason that I like having hardware amp modelers instead of using plugins. Impedance/gain staging/blah blah blah is already built in so there's nothing to mess with in that regard.
- Messages
- 6,623
But it's really not that big a deal with software either. Not sure why it's gotten so much discussion tbh. Set the interface input at a healthy level and then adjust the software accordingly.
ivanmancic
Groupie
- Messages
- 26
Yes, it's digital, which is very convenient for this use case.
Additional 12dB did make an audible difference, but maybe I'm a bit anal about noise in general, not having messed with real amps in over 20 years :)
This video represent an almost perfect example of the straw man argument, as in refuting an argument different from the one actually under discussion. Good advice have been given on how to achieve good results in most cases - although simplified, i.e. taking into account that a lot of modern consumer level interfaces have high Z inputs with max input level @ minimum gain in the 12-13 dBu range combined with decent SNR @ the minimum gain setting. Thus, in most cases, leaving the input gain on your interface at zero and then adjusting the gain level in the box to the appropriate level your plugin expects is good and simple advice. He is arguing AS IF this recommendation is given as an "this is correct in all circumstances", which no one with a slight understanding of the issue has been arguing. Of course, if your interface have specs that differs significantly from the ones being discussed, you should take the specs of your interface into account. Adjusting the recommendations to your specific circumstances could very well mean applying gain on your interface and then bringing it back down in the box before hitting your plugin, i.e. to deal with signal to noise issues if necessary.
That being said, I kind of agree with him that the signal to noise ratio has been somewhat neglected lately (perhaps due to simplification for the average guitarist). But I dislike his take on this, with him unnecessarily creating conflict without substance, with what I guess is getting clicks as the motive.
That being said, I kind of agree with him that the signal to noise ratio has been somewhat neglected lately (perhaps due to simplification for the average guitarist). But I dislike his take on this, with him unnecessarily creating conflict without substance, with what I guess is getting clicks as the motive.
ian_dissonance
Roadie
- Messages
- 561
If your interface isn’t noisy and your Hi-Z is already at the right level you don’t gain (ha) anything by getting louder at the input and attenuating at the plugin besides bigger wave forms to look at.
