Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[0dbfs]

Hi,

I was hoping to gain some insight into how everyone deals with different plugins and their intended unity-gain / dbfs-ref-level design.

Please consider:

In an analog studio the outboard pieces are typically designed and implemented to use a unity-gain signal of +4 dBu.

In a DAW environment we have various plugins that are modeled, designed, imitated, or "whatever" to fill the function of the original units in a similar manner.

ie; waves ssl, UAD 1176, urs-neve-series, bomb-factory-la2a, fill-in-the-blank, etc.

Now, when we capture with unity at different dbfs reference points the audio being fed into the plug-in-daw-devices actually goes in and out at a different level which change the sonic characteristics and behaviour of the unit.

For instance, the analog 1176 would sound and behave differently when integrated into a unity=0dBu environment compared to a unity=+4dBu, or unity=+8dBu environment. How does that compare with various plugs to a signal captured at a dbfs ref of +4dBu = -12, -16, -18, -20 dbfs, etc..

I mean, who really cares. (apparently I do)

It seems to me that just about the whole world uses +4 as unity and the digital equivalent is not considered once in the digital realm and that it probably varies somewhat arbitrarily from plugin-manufacturer to plugin-manufacturer or even product to product.

I think it would be cool if the plugs had a dbfs-ref setting so that we could also set unity within our DAW in a similar way to what we do in the analog realm or to match any AD/DA conveters in each of our systems.

It is what it is in each specific case at the moment but does anyone have any comments?

Cheers,
j

[Tomas Danko]

0dbfs wrote on Wed, 29 October 2008 17:33

Hi, I was hoping to gain some insight into how everyone deals with different plugins and their intended unity-gain / dbfs-ref-level design. Please consider: In an analog studio the outboard pieces are typically designed and implemented to use a unity-gain signal of +4 dBu. In a DAW environment we have various plugins that are modeled, designed, imitated, or "whatever" to fill the function of the original units in a similar manner. ie; waves ssl, UAD 1176, urs-neve-series, bomb-factory-la2a, fill-in-the-blank, etc. Now, when we capture with unity at different dbfs reference points the audio being fed into the plug-in-daw-devices actually goes in and out at a different level which change the sonic characteristics and behaviour of the unit. For instance, the analog 1176 would sound and behave differently when integrated into a unity=0dBu environment compared to a unity=+4dBu, or unity=+8dBu environment. How does that compare with various plugs to a signal captured at a dbfs ref of +4dBu = -12, -16, -18, -20 dbfs, etc.. I mean, who really cares. (apparently I do) It seems to me that just about the whole world uses +4 as unity and the digital equivalent is not considered once in the digital realm and that it probably varies somewhat arbitrarily from plugin-manufacturer to plugin-manufacturer or even product to product. I think it would be cool if the plugs had a dbfs-ref setting so that we could also set unity within our DAW in a similar way to what we do in the analog realm or to match any AD/DA conveters in each of our systems. It is what it is in each specific case at the moment but does anyone have any comments? Cheers, j

As you know when talking digital level signal, 0 dBfs is the roof, even within a stated 32 bit floating point environment you can clip going into a plug-in. Most, if not all, plug-ins will sound better when given some headroom. Some are of course more sensitive than others to this.

Since I try and peak at -10 or preferably lower on the individual channels in my DAW, few plug-in instances will give me problems. If one instance will hit close to 0 dBfs on the output I'll either use it's internal make-up gain (ie if it's a compressor) or place a gain-adjuster in series afterwards.

I always make sure to have plenty of headroom if possible, and in worst-case-scenarios never hit full scale when entering or exiting a plug-in. The ones I tend to use the most, I've examined to find out how to make it sound as good as possible. I have some virtual synthesizers that will show that they do not clip internally as well as not clipping the channel fader but in reality it's crunching away big time unless I lower the internal level another 6 dB or more.

Regarding the UAD 1176 as well as LA2A plugs, you do have the input level control in the plug-in which allows you to hit it hard or soft to get the stuff you're used to when operating the original hardware units. IMHO, any plug-in that emulates a device that behaves differently (in a musical- or otherwise characteristically useful way, I might add) depending on how hard you hit it with the incoming signal should always have an internal pre-stage for that level instead of relying solely on the incoming one.

After all, full scale is full scale. So how else to do this inside a plug-in unless you have a larger dynamic scale that's above the incoming signal.