FreeG vs. Trim

[EMRR]

I find these little differences both fascinating and alarming. I can think of some analog mixing aspects that might be modeled in one, while not in another, that might account. If true, they seem like hidden toggle switches I'd rather see available in the GUI.

[bayswater]

This might explain why, when I was trying to do a null test between DP and Logic, I was often unable to get a bounced mix with varying fader levels done in DP to null against many other signal paths (although I could usually get Logic mixes to null out).

I suppose you'd need the full null test entirely in the digital domain to sort all this out, but how do we know that the effects of the DP fader and the Trim plugin (maybe even FreeG) are not a result of the response of the monitors used? Would it be because the effect of each is slightly different?

It would be odd if the DP fader and the Trim plugin had different effects -- wouldn't you expect them to use the same process for gain change?

If you did a null test, what would be the most revealing? I suppose we'd need some gain change standard we know we can rely on to compare the DP fader and Trim. What would you expect to happen with a Trim in the first insert set to -20db, and another in the second insert set to +20db? Or fader down 20 and Trim up 20, etc?

[Phil O]

So, I set up the following:
-Two mono audio tracks, each with the same audio file (12 second sweep from 20 to 20,000 Hz.
-Track 1 has trim plug; track 2 has invert phase plug.
-Set up memory cycle for a 12 second loop.

-All set to 0dB, Meter Bridge shows no output.
-Trim plug set to -6dB, Fader on track 2 set to -6dB,Meter Bridge shows no output.
Trim plug set to -10dB, Fader on track 2 set to -10dB,Meter Bridge shows no output.
Trim plug set to -20dB, Fader on track 2 set to -20dB,Meter Bridge shows no output.
Trim plug set to -30dB, Fader on track 2 set to -30dB,Meter Bridge shows no output.
Trim plug set to -40dB, Fader on track 2 set to -40dB,Meter Bridge shows no output.

Substituting FreeG for Trim yields the same results.

I think there's something else going on in your set-up. Sends, perhaps? Don't know. My tests show identical results for attenuation through Fader, Trim, or FreeG.

Tried the same with some music instead of audio sweep and got the same results. Attenuating the audio by trim or fader sounds identical to me.

Phil

[bayswater]

Phil, that tells us Trim and Fader do the same thing. Does it tell us they are both linear across the spectrum? I think that's what was at issue.

[Phil O]

That's why I used a full spectrum sweep. They have to be identical at all frequencies or else they wouldn't null for the entire sweep.

If they don't have a flat frequency response, then they all have IDENTICAL frequency responses. It would be insane for any one of them to NOT be designed with a flat response, but for all three to have an identical non-linear response...

Phil

[EMRR]

If it's nulling a sweep, pink, or white noise, then it's the same. I'd want to look in a more detailed metering app like Spectrafoo in addition to the simple peak meters in DP.

[Phil O]

If it's nulling a sweep, pink, or white noise, then it's the same. I'd want to look in a more detailed metering app like Spectrafoo in addition to the simple peak meters in DP.

Why? If it's nulling (and it is) then they're doing exactly the same thing. Right? It's not close to zero signal, it's zero - or as close to zero as minus infinity on the meter bridge is. The number just above minus infinity on that meter is -102dB and it's well below that. There's no indication on the meter at all. I'm sure the detailed SpectraFoo metering is equally adept at measuring no signal.
Or am I missing something?

Phil

[bayswater]

If it's nulling a sweep, pink, or white noise, then it's the same. I'd want to look in a more detailed metering app like Spectrafoo in addition to the simple peak meters in DP.

Why? If it's nulling (and it is) then they're doing exactly the same thing. Right?

Phil

Yes. Al doing the same thing and were assuming they're all doing the right thing.

[mhschmieder]

I can only skim this at the moment as work is intense today, but just wanted to say that the resulting files look noticeably different on my system when using FreeG vs. Trim but not when using Trim vs. the faders, in the lower attenuation ranges mentioned above.

So, it wasn't psychological (especially due to repeats and blind tests with someone else manning the controls) or due to the monitoring setup, but I can see that there might be a default preference on the second page of FreeG that I need to open and look at again, which might be deliberately non-transparent.

I'll read this in more detail later on. There were no sends in my signal path for the tests I performed; just inserts. I deliberately simplified the signal path, and made sure all auxes were turned off and that their inserts were disabled, to remove all possibility of other stuff seeping into the signal path. Only one track was play-enabled for these tests.

A quick scan of these responses indicates that a null test detected no differences between FreeG and Trim, so I'm going to guess that I set a preference on the second page of its GUI a long time ago, and forgot about it. I might have done something to work with its pan controls, but in this case I'm talking mostly about centered mono tracks anyway.

[mhschmieder]

Just got home after a long day at the office, so I'll have to catch up on the details of what others wrote a bit later, but just wanted to mention that it occurred to me during my train ride that it might be a difference between MAS and AU (although I think this topic has been visited before).

If I get around to opening DP tonight, I'll check the "back page" of FreeG to see if there are any suspicious settings there. Pan Law surely wouldn't affect centered mono tracks, though!

[mhschmieder]

I have verified there's nothing unusual on the preferences page for FreeG. It's just about choice of meters and resolution, BBC vs. other standards, etc. Also Pan Law, but that's only for Stereo usage and defaults to 0 dB. Then again, maybe it's having a weird effect on mono tracks if they did a bad job of the mono version. But like I said earlier, I also got different results on a stereo track for a high-frequency synth pad vs. just the mono low frequency bass guitar track.

I've got to install a bunch of software now so can't read the null sum tests quite yet. I did look at the full signal path in the project though, and there's nothing weird (no aux sends, no pre-vs-post-fader stuff going on, no mix automation, etc.).

It's been quite awhile now since I first took notice of FreeG coloring the sound, but at the time it was more related to stereo synth tracks so I thought it had to do with Pan Law implementation details. It's only on this current project where I started using it heavily on mono tracks, and I had switched to VSL's PowerPan a few years ago for most contexts anyway.

[Phil O]

...the resulting files look noticeably different on my system when using FreeG vs. Trim but not when using Trim vs. the faders, in the lower attenuation ranges mentioned above...

Well it's good that Trim seems to line up with the faders. Sounds like maybe it's a FreeG problem. It just occured to me that I haven't updated FreeG in some time. I'm still using an old AU version (1.08) with 32 Lives. What version are you using? If you'd like, you can put together a small test project and send it to me. I'd be happy to test it on my system and see if I get the same results as you.

BTW, we're getting slammed by this winter storm at the moment. If I don't respond for a while it means the power went out. Wish us luck.

Phil

[EMRR]

If it's nulling a sweep, pink, or white noise, then it's the same. I'd want to look in a more detailed metering app like Spectrafoo in addition to the simple peak meters in DP.

Why? If it's nulling (and it is) then they're doing exactly the same thing. Right? It's not close to zero signal, it's zero - or as close to zero as minus infinity on the meter bridge is. The number just above minus infinity on that meter is -102dB and it's well below that. There's no indication on the meter at all. I'm sure the detailed SpectraFoo metering is equally adept at measuring no signal.
Or am I missing something?

Phil

In my experience you can detect a lot of changes below -102 dBFS, and the meter in DP is peak reading only which doesn't tell you much. Some real differences might show up in a spectrogram of noise below -102 dBFS and absolute noise floor which could account for observed differences. In short I'd really want to see the very bottom of the residual in detail to make a call. This could be a case of differences in noise shaping, which would be way down there and not show up on a peak meter.

[daniel.sneed]

Hope Motu will fix the 2dB trim-plug increment bug, in DP9.

This bug does not affect all setups.
Unfortunately, mine is affected...

[Phil O]

Some real differences might show up in a spectrogram of noise below -102 dBFS and absolute noise floor which could account for observed differences. In short I'd really want to see the very bottom of the residual in detail to make a call. This could be a case of differences in noise shaping, which would be way down there and not show up on a peak meter.

Do you think noise that low would be audible? MH is hearing discernible differences between the tracks and says they even look different. Sounds like there's more going on here than stuff below -102dB. Don't you think?

Phil