ProChannel Gain Questions

I'm just wondering what's going on... trying to understand...
 
I understand gain structure pretty well... I've been dealing with it for enough decades... ;-)
 
Why is it that if I have a clip whose gain is not clipping frequently presenting too much gain to the ProChannel plugins, but not if I bypass the gain...
 
I mean, even the first device... I have a peak of near -0.6 and when feeding into either the PC76 or the CA2a it immediately will have gain overs feeding into it... I would expect the input to handle gain coming into it below clipping...
 
Please... it would be very comforting if someone could explain the gain issue here? Thanks...
 
Keni
 

Yup please take us through step by step. Tell us what you do and say what you see :)

First off in Sonar a signal over 0 can be processed because its either 32 bit FP or 64 bits FP. Its very hard to actually clip the internal routing of Sonar.
 
My question is how do you know its clipping? For me its the clip indicator light going into the red and staying there. The other thing is why not just lower the volume fader? If this is too low for the compressor to work with adjust its input gain on it.    

Thanks guys...

For new it's the LED indicators as well and they are flashing with occasional red peaks. I find it hard to read this light peak in the PC itself but it's very visible in the inspector or CV...

I'm not understanding how signal below peaking can be too hot to a fixed input supposedly unity gain...? I'm forced to lower the track gain so the feed into the PC doesn't cause overs. I understand how the output gain of these devices might overload the input of the next device in line, but how does the original signal being free of overs be too much for the first device?

Dan... Yeah I've been quiet... Much of X3 is better tho I still liked pre-X better to work in... The X generation has it's advantages and living in the past is not something I can do...

I still dislike Lanes and all the problems they've have introduced... The new comping has some good points but so far I'm still faster before lanes got in my way...

I'm totally frustrated with the zoom issues and as they are Ware and fixing it (?), I'm remaining calm to see what they do with the coming fix...

Complacent? No, but my life is crumbling here as I've lost my heat already, my other bills all in arrears, and my house in jeopardy of foreclosure only days from now... I'm trying to play with X3 to keep my mind off things I can do nothing about but not very focused as the nagging thought of being homeless, broke, and 62 up in the mountains.... Survival is cramping my ability to care too much about anything right now...

Thanks for noticing!

Keni

That's the point... If there is no gain control on the front end, it should not be possible to be over going in? Some compressors use an input control to decide threshold, but even there if there is so little signal coming in that such a control is raised yet it's not high enuf to even trigger the compression, how can the LED be showing an over?

Right now the only thing I can do to deal with such is to lower the gain control to lower the feed into the device... This works, but I'm trying to understand how the gain is happening at each of these stages... This part was much more simple in the analog days! ;-)

Keni

I noticed on the input nob on the PC76 that its default appears to be -.2 dB. If one centers it that may be the problem. I don't know for sure though. 

useful? http://forum.cakewalk.com/FindPost/2880052

Yeah, the ProChannel meter lights are a bit of a flawed design IMO. They could really do with an unambiguous "This is now over 0db" indication. The way it gradually fades into red, and as far as I can tell, reaches red well before 0db, is not all that useful.

Well, hard to say without seeing the project, but one possible scenario is that the the gain coming in is very high, but the compressor's threshold is still above the gain.

John T
Well, hard to say without seeing the project, but one possible scenario is that the the gain coming in is very high, but the compressor's threshold is still above the gain.

Thanks John...

I could understand that once it gets increased by a secondary device, but if it's currently -0.6 peak it should be low enough to feed the input of the first device without overs... In the analog world it could then be amplified to greater levels but in the digital realm it can never be over zero (possible ISD)...

Now if the fixed threshold style device isn't triggering and raising the input on the device is required to get it to trigger level? I don't understand how that can be yet I've run into it a number of times...

As a recorded clip cannot be over zero, when it gets sent into a device that does not raise gain to trigger the hers hold but instead allows adjustment of the point to trigger (threshold) it should not be possible to have an over at a unity input... Should it? Now in the analog world, increasing gain after compression could be used to actually amplify a signal louder than it was... But when recording to tape, you might have to reduce signal gain before sending it to tape which has a ceiling far more limited than the actual range of possible volume... Which in turn can again be amplified on playback...

In the digital world it never truly is any louder than zero tho apparent volume such as mastering deals in can be louder or lower relative to this fact...

So if the clip (tape track) is limited in this fashion there is no way to ever have volume louder than zero, how can it be too hot for a unity gain input?

I'm more confused about these LEDs now than before! ;-)

Thanks...
Keni

The Sounding Off page at the back of this month's Sound On Sound has some good stuff to say about this kind of design and why it's unhelpful. I hope lots of DAW developers have read it.
 

With X3 one could use the pop out EQ and see what it says. If its not doing its job than place it anywhere along the chain you want to monitor and this would let you know exactly what is going on. 

John T
Well, as explained in that thread that lawp linked to above, it shades from yellow to orange to red as it approaches 0db. And there's a whole lot of "deep orange" in there that looks pretty red to most people. It's ok, for what it does, but I think there should be a separate clip light, that lights up *only* on overs.

It's kind of a weird design, for a piece of software; it's the kind of thing you'd do on budget hardware to cut down on the amount of LEDs you need for each unit, but makes no sense at all when an infinite number of "LEDs" costs the same as one.

Indeed, it's not clear to me why it's even an LED. There's enough room for a little strip meter.

Thanks...

I'm gonna tackle that thread next...

So there is an interim color that's been confusing me. That's the orange that looks slightly red when the CV shows no clipping at all I guess...

Very true... I'm with you and would prefer a little meter or at least a virtual LED stack with an actual over indicator separate from the almost indicators! I find the current indicators impossible to read well and obviously needed...

Anyone for mass feature requests? ;-)

Keni

gswitz
Keni, keep in mind that some of the plugs become over driven before you clip the signal. They model some of the gear with being able to run it too hot as a feature.

Ears should help choose the right setting.

If unsure, you could apply the effect with gain and threshold adjusted the same amount and bounce to tracks. Then a/b the results.

So in that case, a colored pc overdrive light might indicate that the plug is running hot, but not that you are clipping.

Thanks gswitz...

I always use my ears as well... But I remember that my ears don't always notice everything hence the need for metering and the likes. I heard no unusual distortion or artifacts, but if the meter says it's there I give it the benefit of the doubt.

To date I have always readjusted my gain to remove overs wherever I see them regardless of hearing them...

I'm just trying to understand what's going on here..

Yes, I remember that assorted hardwares are more or less sensitive... And I've subscribed to the theory that if properly modeled, the digital should show the same traits... So I'm believing that's the case... For example, the Console Emulator is far more sensitive than the other modules... Hence Craig's note regarding the input stage of it... I don't remember any of the 3 consoles modeled being that sensitive to input gain tho...

Keni