how to proffesional engineers get it so loud?

  • Thread starter Thread starter idahoe
  • Start date Start date
Massive, I've read quite a lot of posts from you in which you recommend to record at -18dBFS in order to not overload the preamps. And I know I already asked you this following question before, but I can't remember the answer or where we had that discussion, so here goes again:

What would be the problem in using the right gains in a preamp that puts out +4dBu "studio level", but running the adc/soundcard at "consumer level", i.e. -10dBV in order to not overload the pre but at the same time get healthy levels into the computer so that more (around 2, that is) bits are stored on disk? Okay, the problem will rear its head in mixing, when you'll have to turn the tracks down to prevent the mixbus from overloading. But I'd rather go safe and record the highest resolution possible even if I have to turn the tracks down eventually (and lose the resolution in the process). But what if not? Then I'd have wasted "good bits", wouldn't I? Or am I just going crazy?
 
Ethan Winer said:
You can also do this manually in a program like Sound Forge. If there are only a few loud peaks that prevent raising the entire track level enough, select just the one portion that goes too high (or too low) being sure to set your editor program to bound the selection on zero crossings. Then lower the volume of just that small portion.
Click to expand...

I used to do this years ago, cause I figured it would be a nice way to control the volume without causing too much distortion, but in the end, I'm just too lazy. If you've got to do this for a complete album full of songs, you're gonna be editing your ass off for hours. And I'm not one who usually shies away from doing editing work. Now, I'm just using it on individual tracks, if there are only a few transients, that are way off, I would fix them this way. But to be able to use it on complete mixes is quite nice. So maybe it's worth it to invest in this plugin. I'll give it a try, the demo runs for 30 days. Thanks for bringing that up, Ethan.

Edit: Well, I just tried it out and yes, it does quite gracefully what I did when I still used the technique, but in no comparable amount of time. Very nice. The only thing that bugs me is the crossover distortion when moving from an unaltered to an altered cycle. Depending on the material, this can be quite drastic. And this, I guess, is the advantage when doing it by hand. If the signal just bends too sharply, you can always pull the adjacent cycles down too.
 
Last edited:
TragikRemix said:
what would you classify 'hot' as? i think im tracking to hot, my signals are peaking just below clipping. i think that might be too hot, and i cant get a decent mix.

on the other hand, if i make something using only midi, the levels are sorta pre-adjusted and those mixes sound great.
Click to expand...
That's *SO* hot it's ridiculous. Look at the preamp - Don't even pay attention to the digital meters (again, no VU meters, use -18dBFS or so. Safe bet there).

And yes - Sounds that are actually *produced* digitally aren't running through any sort of outboard gear. They aren't going to overload a preamp or a converter, so they can come "into existence" at a hotter level.

Of course, they're probably going to have to be turned down considerably before mixing with other signals...

nessbass said:
Massive, I've read quite a lot of posts from you in which you recommend to record at -18dBFS in order to not overload the preamps. And I know I already asked you this following question before, but I can't remember the answer or where we had that discussion, so here goes again:

What would be the problem in using the right gains in a preamp that puts out +4dBu "studio level", but running the adc/soundcard at "consumer level", i.e. -10dBV in order to not overload the pre but at the same time get healthy levels into the computer so that more (around 2, that is) bits are stored on disk? Okay, the problem will rear its head in mixing, when you'll have to turn the tracks down to prevent the mixbus from overloading. But I'd rather go safe and record the highest resolution possible even if I have to turn the tracks down eventually (and lose the resolution in the process). But what if not? Then I'd have wasted "good bits", wouldn't I? Or am I just going crazy?
Click to expand...

Those bits are there so you don't have to use them. They're there for headroom - They're there so you can use the rest of your gear at a nominal level without compromising the sound. Even in 16-bit this was standard operating procedure. In 24-bit, it's nonsensical to do anything else.

And definitely - Don't be pushing a balanced signal into an unbalanced input just to get more gain... That's out there... I'd rather have signals that peak at -30 than to add even a small amount of "fake" gain. It adds nothing to the signal. THERE IS NO NEED TO RECORD HOT.

I'm sure I sound like a broken record, but you can record with *PEAKS* at -47dBFS and *STILL* have higher resolution than a compact disc.

I still just don't get it... :(
 
This is kind of funny as I have a project going on, and the guys friend comes in to "help" mix. He has an 001 PT setup. Anyways....he has given me some wave file of bass tracks for me to import into a few songs. Man! They were way hot, short of clipping. I told him there is no need to run his signal that hot. I think you can get pretty darn good resoloution with todays gear at a much lower level. Uh..... He does not believe me. I always, at least I try to run with my Alesis DH24 around -15db as the manual says that is equal to 0 db analog. I get nice sounding tracks and don't worry about clips or in most cases the need to compress/limit on the way in. The guy I have been sending projects to for mastering is very happy with the levels of the final mixes. He likes headroom also.
 
thanks massive, im using either the pres in the Digi002 or on an ADA8000, so theres no meters, except the digital ones. the ada only has a clip indicator.
 
jmorris said:
This is kind of funny as I have a project going on, and the guys friend comes in to "help" mix. He has an 001 PT setup. Anyways....he has given me some wave file of bass tracks for me to import into a few songs. Man! They were way hot, short of clipping. I told him there is no need to run his signal that hot. I think you can get pretty darn good resoloution with todays gear at a much lower level. Uh..... He does not believe me. I always, at least I try to run with my Alesis DH24 around -15db as the manual says that is equal to 0 db analog. I get nice sounding tracks and don't worry about clips or in most cases the need to compress/limit on the way in. The guy I have been sending projects to for mastering is very happy with the levels of the final mixes. He likes headroom also.
Click to expand...

I'm adding a few points to Alesis on my respectometer for that.

We can add your friend to the "doesn't get it" list... :D
 
i understand that dBVU, dBFS and the digital signal are all different.

im interested in dBVU, which i cant read because there is no vu meter (duh)
i should check my manual to see the equivalence/conversion (?)
 
If there are no meters, you can pretty safely use -18dBFS or so. Some (converters) are calibrated a little hotter, some are a little cooler. But that's generally a nice place to be.

Lower isn't a problem either. But in almost any case, -18 is going to be close. It's never too quiet, and only rarely too hot. And when it's too hot, it's still reasonable.
 
ness,

> The only thing that bugs me is the crossover distortion when moving from an unaltered to an altered cycle. <

The result should be identical whether you do it manually or with Peak Slammer. Either way, the cycles on either side of the region have a different slope. But in practice I regularly achieve 4 to 6 dB increase in level with no audible artifacts. YMMV as they say.

--Ethan
 
That somewhat reminds me of the perversion of a mastering plug-in I wrote (a non-limiter limiter). I called it "pow", which is short for "power". It alters every cycle exponentially. For each point in the input sample:

output = input ^ k

where ^ is the power operator and k is a value between 0 and infinity, and a value of 1 is unity gain output.

It's a neat effect between about .8 ant 1.2. Beyond those limits, it sounds... wrong.... :) Regardless, there's nothing quite like the look on your clients' faces when you tell them that their tracks are too hot, so you're going to take the square root! :D :D :D

Anyway, if anybody wants to write a fun plug-in, that's a fun one to write.
 
dgatwood said:
Regardless, there's nothing quite like the look on your clients' faces when you tell them that their tracks are too hot, so you're going to take the square root! :D :D :D
Click to expand...
Hahaha!!!!!!

I believe that when you record digitally at 24-bit, you can keep the input much lower than if you were recording at 16-bit. If I remember correctly, this is because there is more data in a 24-bit recording. The more data the computer has, the more accurately it can work with it. Please correct me if I'm wrong.
 
though I will admit to being one of the people trying to get to 0dB this makes perfect sense... because a 24bit signal has a 144dB dynamic range... and pushing preamps always adds noise and distortion... so by recording softer you get a cleaner signal... even a 16bit signal has 96dB to work with...
 
IronFlippy said:
Hahaha!!!!!!

I believe that when you record digitally at 24-bit, you can keep the input much lower than if you were recording at 16-bit. If I remember correctly, this is because there is more data in a 24-bit recording. The more data the computer has, the more accurately it can work with it. Please correct me if I'm wrong.
Click to expand...

Yes. Sorry, I should have said "You should see the look on your clients' faces when you tell them their tracks have too much dynamic range, so you're going to take the signed square root (k=0.5)."

To be more precise, yes, it's the signed square root. Otherwise, you'd have the negative portions of your waveform behave very strangely.

For positive input values:

output = input ^ k

For negative input values:

output = 0 – ( |input| ^k)

where || indicates absolute value.
 
THANKS!! This is so helpful! I'm new to recording and I had no idea, I totally get the concept.

Here is something I still dont quite understand though.

I don't understand the difference between DB, DbFU, and DBFS

Bottom line:
When looking at my input meters where should they read? Between -20. and -18 Db?

Should the input on my recording software be set the same?

I'm useing the Line 6 Toneport with the included Gearbox software.

In the Gearbox software there is a knob that "sets the level sent digitally to my recording program."

Is this the one I should go by?

Here is why I ask:

When I set this knob to -18.0DB the two meters read at anywhere from -45.0 Db to -35.0 Db

That is such a huge difference!! I dont understand it. Should I ignore the meters and go strictly by the knob that "sets the level sent digitally to my recording program?"

Thanks guys, any clearity would be appriciated.
 
PaulKarate said:
Here is why I ask:

When I set this knob to -18.0DB the two meters read at anywhere from -45.0 Db to -35.0 Db

That is such a huge difference!! I dont understand it. Should I ignore the meters and go strictly by the knob that "sets the level sent digitally to my recording program?"

Thanks guys, any clearity would be appriciated.
Click to expand...
I'm thinking that "this knob" is giving you dBVU - and the meters are giving you dBFS. If you try to get the signal to 0dBVU, you'll be hitting 18dB higher than -35dBFS - or (presto!) right around -18dBFS.

Unless something else is out of whack somewhere by around 18dB...
 
So if I understand correctly this knob should be set to 0.0 Db right?
Again, this knob controls what is being sent out to my Recording Software.

If I do that (set it to 0.0Db) the meters will read at about -18. Db. right?

This is where you say We should be at when recording tracks?


Thanks.
 
You must log in or register to reply here.
Back
Top