Do plugins "work better" when the average digital level is higher?

I heard it asserted today that plugins work better if you first normalize the file to say
-10dbFS before using the plugins. What do others have to say?

Tim
 
Some will certainly react differently -- A compressor set with a threshold of -6 isn't going to do much with a source of -18...

Of course, I'd rather be working with the source of -18dBFS and just lower the threshold...
 
Depends on the plugin

Emulation plugins that attempt to duplicate hardware have a "Line Level" level above which they will begin to add modeled distortion artifacts as the hardware does

So feeding an emulation of say an 1176 compressor with a -18dBFS RMS signal will sound "cleaner" than feeding that same compressor a -6dBFS RMS which will cause some distortion (which may or may not be a good thing depending on the sound you are after)
Also if you have plugin emulations on a lot of tracks in your mix and are feeding them all a virtual "Hot" signal you will end up with a lot of added distortion in your mix which could be a problem

If it's not an emulation plugin then it should perform the same (just louder or quieter) regardless of the levels you send it

Make sure you know what your plugins are doing with the signal and then you will know what are the appropriate levels are to get the result you want
 
Some will certainly react differently -- A compressor set with a threshold of -6 isn't going to do much with a source of -18...

Of course, I'd rather be working with the source of -18dBFS and just lower the threshold...

Is that bitrate dependant? What's your take on that? Worth noting? I'm revisiting that old argument that plug-ins sound better at higher bit-rates...
 
Assuming you mean bit depth (word length), then no, it really doesn't apply. Even if you're working on 8-bit material, the software is almost undoubtedly working in 32-BFP (or even up to 80-bit in the case of Samplitude and perhaps a few others).

Unless you mean sample rate -- There are a few plugs I've worked with that sound a bit different (well, "hardly" different, but perhaps a whisker different in the very top end) at higher rates or upsampled. But most of those allow for upsampling on the fly anyway.

Put me squarely in the "eh" category there...
 
You know that in theory it should make no difference at all as all you do is adjust the gain structure, but I am sure that for some reason if the level is on the low side it does not sound as good through the plugins as a healthy level signal. Maybe it's my old school side showing up.

Alan.
 
Thanks for the replies. I'd never heard of this. I guess I'd always assumed (with the exception of level dependent ones) the plugins just crunched the numbers and didnt care what levels they represented. Seems I'd assumed right. Thanks for the confirmation.
Cheers Tim
 
Assuming you mean bit depth (word length), then no, it really doesn't apply. Even if you're working on 8-bit material, the software is almost undoubtedly working in 32-BFP (or even up to 80-bit in the case of Samplitude and perhaps a few others).

Unless you mean sample rate -- There are a few plugs I've worked with that sound a bit different (well, "hardly" different, but perhaps a whisker different in the very top end) at higher rates or upsampled. But most of those allow for upsampling on the fly anyway.

Put me squarely in the "eh" category there...

I meant bit rate. Sorry for the terminology screw up there. That answered my question :)
 
Some will certainly react differently -- A compressor set with a threshold of -6 isn't going to do much with a source of -18...

Of course, I'd rather be working with the source of -18dBFS and just lower the threshold...

I record pretty low (peaks around -12dBFS) so my threshold is usually super low too (sometimes around -20dBFS). I've seen a lot of garbage floating around about not setting the threshold that low. But, it really depends on how low the source material is. If you're peaking at -3 dBFS, then -20dBFS will crush the sound usually, but threshold is dependent on the material's volume.
 
I would suggest not, bear in mind some analogue modelled plug ins expect levels to be as they were in the
analogue world, check your manuals, use the refercence -18dBFS = 0vu / +4dBu.

In any event observe gain structure principles and work at 24 bit for best results.

It will plugin dependent in any event.
 
Plugins work the same no matter what you throw into them . . . they don't care about levels. But their effectiveness can be compromised if the signal is unsuitable for their operation.
 
use the refercence -18dBFS = 0vu / +4dBu.
Using this as a reference would only have relevance if your particular converters were calibrated to those numbers. The reference can change by a few dB in either direction.

My converters are calibrated at -15 dbFS = OdBvu for example.
 
Using this as a reference would only have relevance if your particular converters were calibrated to those numbers. The reference can change by a few dB in either direction.

My converters are calibrated at -15 dbFS = OdBvu for example.

Hmm, might be missing it, but I don't see the relevance. Once in the box we're all on the same scale.

Tim said:
I heard it asserted today that plugins work better if you first normalize the file to say
-10dbFS before using the plugins. What do others have to say?
-10 average or peak?
Actually it seems if you mean average levels around -10 for a lot of sources that's rather hot already. If -24--18 average is a good nominal target range for tracking (it is for me), your peaks are often already -10 or greater.
 
I don't see the relevance. Once in the box we're all on the same scale.
Yep. there is no relevance. I was referring to the numbers Barry had posted as being arbitrary because there is no standard for the numbers he posted in the digital scale verses the analog scale even if i did somehow apply to analog modeled plugs.
 
Back
Top