automatic loudness correction that can handle split stems too?

[Cheesehead]

I think I agree with Mike.
After all, f you have any kind of dynamic processing over your mix bus its not going to react the same if you put the same plugin over the stems unless you have some sort of side-chaining going on.
I'm happy if my stems are in the ballpark in terms of level.
Its going to get re-mixed to a certain extent anyway if they're used.

[lae777]

Originally Posted by Cheesehead

I'm happy if my stems are in the ballpark in terms of level.
Its going to get re-mixed to a certain extent anyway if they're used.

This is exactly right.

What is the point of making the stems loudness compliant as well as the full mix?

If the stems are being used for foreign language dubbing then they will be re mixed.

If they are being used as separate entities (e.g. music stem for a trailer) they will be re mixed to spec anyway.

If they are being used to re version to another format (e.g. 5.1 to Atmos) they will be re mixed.

What possible reason is there for having the stems match the full mix loudness specs?

They just need to be close in relation to the full mix at loudness spec.

[seanmccoy]

Yeah, from a practical standpoint this is the only thing that makes sense to me. In many situations both the FX and Music stems may be very, very light, and would have to be cranked up to ridiculous levels to meet specs on their own.

[MIKEROPHONICS]

Dynamic control shouldn’t be achieved on a mixbus - it should be achieved by correct mixing at source level, maybe with some mastering help (refinement) at source group level,so that your stems are correct.

Having said that, deliverables are request lists made by idiots for idiots, often cut and pasted by those who do not understand. There are two approaches; a) give the fools what they ask for or b) explain what they should have and what will actually be useful.

The requirement for your stems to be compliant is meaningless as they are never transmitted. However, it is much easier to edit in and out of scenes if the levels off your stems bear some relationship to the mix - I.e don’t be putting the latest Waves Ultrasodamizer on your mix stem and expect to please anyone that has to reversion!





Sent from my iPhone using Tapatalk

[Cheesehead]

To be fair we have wandered a bit off topic here.
Alexy didn't suggest he wanted his stems to be loudness compliant, that would be bonkers.
He wanted to be able to use something like Nugen LM Correct to apply the same processing to a stem as to a full mix, in order to keep the stem at the same relative level as the level corrected mix.
I can totally sympathize with that and I think it would be a useful feature in some circumstances.

[its2loud]

“He wanted to be able to use something like Nugen LM Correct to apply the same processing to a stem as to a full mix, in order to keep the stem at the same relative level as the level corrected mix.”

Probably beating a dead horse here but this just isn’t technically the correct way to do this. I may be wrong but processes like LM Correct aren’t just adjusting overall gain level to make the output legal. I believe there are complex algorithms at work to make the mix compliant while keeping the adjustments relatively transparent.

If one’s mix is so out whack that they need LM Correct to make it to spec then I’m willing to speculate that that that person needs to start over with the mix. Not judging the OP here, just stating an opinion. Measure twice, cut once, deliver. With offline measuring and history graphs it’s not hard to know where to make the necessary adjments in the mix in order to make it deliverable. Stems and all.

[Farhoof]

Originally Posted by its2loud

I may be wrong but processes like LM Correct aren’t just adjusting overall gain level to make the output legal. I believe there are complex algorithms at work to make the mix compliant while keeping the adjustments relatively transparent.

That's only the DynApt extension. LM Correct itself is just measurement and normalizing, without dynamics.

[alexeymohr]

Wow this really got philosophical quickly! ;-)

Obviously we can wax poetic on the rationality or irrationality of a network or streaming service demanding that their split stems exactly equal the composite mix stem. But the fact is that many do.

Also indeed things like iZotope RX Loudness Control and LM-Correct are brute-force dumb normalizers, simply measuring for LKFS and then applying corrective global gain.

So it strikes me that given how straightforward that is, it really shouldn't be all that hard to have the plugin store the gain adjustment value that it determined and allow the same adjustment to be applied to subsequent clips. That's all this was about!

But thank you all for the discussion of the relevance of this; I always do wonder, when presented with a ridiculous spec, whether or not anyone is actually verifying these things upon receipt of the deliverables. That said, I also do my best to do exactly as any client asks, regardless of my opinion of how reasonable it is!

[its2loud]

Good to know about LM Correct an others about their dumbness. I have not met one spec that wasn’t ridiculous, time consuming, and technically incorrect in some regard. It’s only gonna get worse with the new streaming services coming up.

[MIKEROPHONICS]

Specs are like tech rejections, treat it like bridge or poker and their often poor opening gambit, to which you as the knowledgeable pro can easily outbid to their eventual advantage...

I love telling tech review peeps how stupid they are (nicely of course) and pointing out the absurdity of some cut and pasted tech spec...it’s all part of the game


Sent from my iPad using Tapatalk