How to get more volume?

[mixsit]

MusicWater, please read the following:
dB dBu dBFS dBV
Especially the part that says "There is no decibel to dBFS converter

Notice - Comparing dBu and dBFS: There is really no fixed
world standard like e.g. −20 dBFS = +4 dBu = 0dBVU.
The digital peak scale is not equivalent to the analog RMS scale.

You can never match dBFS and dBu. "

Hope that helps...

In one of my earlier ('more nubee' :>) stages I actually asked RME (in so many words I don't remember how I phrased it..
'if there are these different analog I/O selections, which setting is (closest?) to 'true, 'neutral' (i.e. not attenuated or gained) to the converter level?'
Shame I don't remember which he said (one was.. 'a little closer') but there was definitely a bit of a chuckle' that came along with the answer to my question

 

[DM60]

miro - It's exactly what he's been saying all along. The guy with the -18dbu converters needs to keep his analog levels lower in order to hit -10dbfs, and when it comes back out to analog the levels will be lower. This puts you closer to the noise floor no matter which way you look at it. I don't know that it's some "secret of the masters", but it is there.

DM60 and ecc83 - Be careful here. First of all, digital gain does amplify the analog noise that came in with your signal, which really is usually the worst part of adding gain even in analog. I suppose it's obvious, but if your levels are down at -30dbfs, it means you're already way down in the analog noise floor. Then too there really is a noise floor in digital. It's really low, but not really that far away from good analog specs. If you do some absurd things, you will hear it. No matter how deep your mix engine is, you will never have any better dynamic range than what comes through the converters (without a whole lot of fucking around). Digital is a bit more forgiving than analog, but it's no excuse for poor gain staging.

Yes, I guess from my point is, it doesn't add noise (you have corrected that thought, so now it is amended to say "very little") as an analog amp would actually add depending on the quality etc. To your point, if it is there in the source, it will be there in the digital and it will get increased in the gain. I think that is pretty straight forward.

It should be understood that a good source signal is still regardless of its destination (analog or digital) is still best approach in recording/tracking.

Thanks for the additional information (as this was a learning moment for me).

 

[BroKen_H]

I learn a little more every time I come here. Someone once said, "The more I know, the more I realize I don't know." Feel like that sometimes.

 

[mixsit]

Yes, I guess from my point is, it doesn't add noise (you have corrected that thought, so now it is amended to say "very little") as an analog amp would actually add depending on the quality etc. To your point, if it is there in the source, it will be there in the digital and it will get increased in the gain. I think that is pretty straight forward.

It should be understood that a good source signal is still regardless of its destination (analog or digital) is still best approach in recording/tracking.

Thanks for the additional information (as this was a learning moment for me).

..as in your digital float point noise added with gain ITB is insignificant way way below the noise floor you've previously captured.

 

[miroslav]

miro - It's exactly what he's been saying all along. The guy with the -18dbu converters needs to keep his analog levels lower in order to hit -10dbfs, and when it comes back out to analog the levels will be lower. This puts you closer to the noise floor no matter which way you look at it. I don't know that it's some "secret of the masters", but it is there.

Not sure what you or he are saying.

The converter's headroom has nothing to do with changing the level of an incoming signal that is being converter (other than it sets up a ceiling/headroom for it).

You can change your dBFS reference (what YOU want call xxdBFS relative to incoming signal)...but that's different.
If you send the same +4dBu signal to two different converters, with two different headroom levels, you won't get two different levels in your DAW.

 

[Greg_L]

I learn a little more every time I come here.

Not from this thread, you didn't.

 

[mixsit]

I was ok up to here..

.. If you send the same +4dBu signal to two different converters, with two different headroom levels, you won't get two different levels in your DAW.

You would get two different recorded levels - if these 'two different converters use different i/o sensitivities.
RME three scales - I've always be thrown by their use of the term 'headroom, but actually maybe this (looking again), this thread's helped me see something maybe I didn't get previously- Their inclusions of 'headroom spec' are referencing to what might be typical' or expected'(?) with different 'systems' ..

" The 'standardized' studio levels do not result in a (often desired) full scale level, but take some additional digital headroom into consideration. The amount of headroom is different in different standards and again differently implemented by different manufacturers.
Because of this we decided to define the levels of the ADI-8 DS in a most compatible way.
The headroom of the ADI-8 DS is defined according to the chosen reference level.

Reference 0 dBFS @ Headroom
Lo Gain +19 dBu 15 dB
+4 dBu +13 dBu 9 dB
-10 dBV +2 dBV 12 dB

At +4 dBu a headroom of 9 dB offers a problem-free operation with most devices, and meets the latest EBU recommendations for Broadcast usage.
At -10 dBV 12 to 15 dB headroom are common practice, each mixing desk operating at -10 dBV is able to send and receive much higher levels.
Lo Gain allows to work with high levels, best suited for professional users who prefer to work balanced and at highest levels. "

( my add; and what I never could seem to get through my head before-
In this case "Lo Gain" is a six db pad which, simply moves the 0dBFS point up to +19dBu. This '15dB of headroom is still referencing to that same '+4' system.

The simple ver-
You hook up whatever line in kit you have
You set / run that front end -pre' or what ever) at/in it's nominal best gain setting' (job one.. "1st stage sets the signal-to-noise and quality (for the most part) of your entire system".
Then you match (or adjust' if needed) that to your converter sensitivity- to get to your dBFS target record level.
(If that happens to means an in-line pad' or what have you to get you there , fine)

 

[ashcat_lt]

Not sure what you or he are saying.

The converter's headroom has nothing to do with changing the level of an incoming signal that is being converter (other than it sets up a ceiling/headroom for it).

You can change your dBFS reference (what YOU want call xxdBFS relative to incoming signal)...but that's different.
If you send the same +4dBu signal to two different converters, with two different headroom levels, you won't get two different levels in your DAW.

Yep, but depending on your converter, it might show up as a different level in dbfs inside the machine. You don't usually get to set that, it's baked into the converter. You can change analog gain staging around it, but then you're not actually sending the converter a +4dbu signal anymore, are you?

..as in your digital float point noise added with gain ITB is insignificant way way below the noise floor you've previously captured.

You can get down there if you do really stupid things likes overcompress/distort really low level sources, but I think I said it doesn't matter how deep your mix engine is because the dynamic range is already limited by the ADC. That noise floor is not so far below the analog noise floor as to be completely irrelevant, especially if you're then doing stupid things like smashing the signal later.

To your point, if it is there in the source, it will be there in the digital and it will get increased in the gain. I think that is pretty straight forward.

I kinda figured you knew that, but...some people...

 

[miroslav]

I was ok up to here..

You would get two different recorded levels - if these 'two different converters use different i/o sensitivities.

Mmmmmm...I don't think a converter's I/O has "sensitivity" like a microphone.
It's a Line level I/O (not counting the all-in-one-boxes that include preamps).
You could maybe set it to -10 or +4 (which is what I have on my converters)....but once you choose either, you've established your reference.
So even with different brands/models....if the I/O is referenced to +4 on all of them, regardless of their headroom level, the same input signal would/should give the same dBFS reading.

Yep, but depending on your converter, it might show up as a different level in dBFS inside the machine. You don't usually get to set that, it's baked into the converter. You can change analog gain staging around it, but then you're not actually sending the converter a +4dbu signal anymore, are you?

Are you making assumptions or stating a fact...?

If you send a signal (any signal)....to two converters with different headroom specs...the signal will still remain the same level.

Will it show up as different dBFS meter readings....?...I doubt it. If that was the case, dBFS meter readings would be even more meaningless than they already are, seeing how everyone can choose what to reference their dBFS scale to.

 

[ashcat_lt]

Will it show up as different dBFS meter readings....?...I doubt it. If that was the case, dBFS meter readings would be even more meaningless than they already are, seeing how everyone can choose what to reference their dBFS scale to.

This is exactly the case, in fact. 0dbfs means as much as the ADC take in and the most the DAC can spit out. For some this is 30 something volts. For some it's less. All of the dbfs levels indicate the ratio of this the signal level to this number. For a given input voltage, a lower headroom number makes a smaller ratio.

 

[miroslav]

OK...let's back up the bus.

I still don't see what comparing apples to oranges gets anyone.
Regardless of what the dBFS meter is showing...the signal level that you input from your analog front end is the same level no matter which converter is used or it's headroom (assuming it's all set to the same reference, -10 dBV or +4 dBu)
That's what I'm saying.

The initial post that got this going, implied that there was some audio "secret" that the pros keep quiet about....and that using a converter with more headroom somehow will improve your digital audio quality.
That only might apply at the analog stage as you are hitting the ADC....and the analog signal level you are using. IOW, if you start to overload the ADC's headroom, then you get distortion.
Once it's in the DAW....the dBFS reading is pretty much meaningless AFA "audio quality".


This GS thread says it nice and simple: Utilizing converter headroom to increase loudness - Gearslutz.com

The only benefit that might happen, is at the analog stage. With a +24dBu converter VS a +18dBu, you can get that analog level up.
Thing is....once you convert to digital, the analog level is solely dependant on the listener's playback system.
So....the original dBFS reading is not going to matter directly. It will "scale up/down" to the playback device's capability.
If you don't use any analog process out from your DAW when you mix/master.....whatever extra level you got, won't matter in the digital domain.

 

[ashcat_lt]

Yeah, I never really got behind the idea that this was some secret, and I mostly agree that once it's in the box, the point is pretty much moot, except...

That link you (miro) posted talks at least some about going back out to analog for further processing. In that case, IF you actually have a bunch of headroom in some fancy piece of gear AND you actually want to use all of that, THEN you'll be missing out on some of it if your converter doesn't also have that headroom UNLESS you also apply some other intervention.

The debate about whether or not it matters is almost academic. It is sort of a seperate issue from the basic fact that a given voltage into one converter can, in fact, give you a different dbfs level than another and that a given dbfs level can give you a different output voltage depending on the DAC you're pushing it out, which is the only part of the thing that I was really commenting on.

 

[miroslav]

The debate about whether or not it matters is almost academic. It is sort of a seperate issue from the basic fact that a given voltage into one converter can, in fact, give you a different dbfs level than another and that a given dbfs level can give you a different output voltage depending on the DAC you're pushing it out, which is the only part of the thing that I was really commenting on.

Yeah...that part of the discussion, I wasn't 100% sure of as I've never used two converters with different headrooms to compare.
I said I doubted it would be different....but having thought about it on the way home, I will now say that it might depend on how/where it's being metered.

If you have two converters, (+24dBu headroom & +18dBu headroom)....metering the same analog input level on each of those converters' meters, could give different dBFS levels on the converters' meters.
However, I'm still doubtful that the dBFS levels would be different if metered in the DAW.

IOW...the analog signal going into each of the converters was the same. Now take each converted signal and put them up on tracks 1 & 2, and play them....and I'm betting the *DAW track meters* would register the same dBFS for both signals/tracks.
The DAW has no idea how much analog "headroom" the converter had when it converted the signal.

Now....if you play each of them back through their respective converter....it's possible that at each converter's meter, again, the dBFS might show differently, but that's where I say we are comparing apples to oranges since the reference for each of the converter's meters is different.
The reference for the DAW's meters is just the digital signal. I doubt that one signal would actually be louder than the other just because one converter had more headroom. They both received the same analog input level.
More headroom doesn't = more/better signal unless you give it more signal or unless you hit a lower headroom converter up its max, and then compare that to a converter that has more headroom....which again is apples to oranges.
That's the problem, IMO, with Music Water's perspective that got this discussion rolling.
Comparing dBu to dBFS is a mistake...and a unit's headroom level has nothing to do with the analog input level.

I wish I had two converters with different headroom specs to confirm the DAW metering thing....but I'm sure it would be like that.

So then the real question is.....where are we doing our metering?....at the converter or in the DAW after the conversion...or at the analog stage before the conversion?
In my case, I do all my critical metering at the analog stage, and I only glance at my converters' dBFS meters, but I never obsess over -18dBFS or -12dBFs or -6dBFS. I just make sure I've picked -10 or +4 at the converter, to match my analog source.

Which is why it's been said that once it's converted to digital....it kinda doesn't matter much. You got what you got AFA "quality"...and you can certainly raise/lower the digital level to your heart's content without issue (until you go back out to analog, or start hitting the 0dBFS mark.

The main use of metering in the DAW for me is to compare track levels between each other. I may even change my dBFS reference at times, but it still lets me compare level differences between tracks regardless of the dBFS reference setting. I also use the meters to look at track dynamics.
Funny...for someone who doesn't pay such critical attention to dBFS meters, I recently picked up two sets of metering plugs...the Brainworx Meter and the Waves Dorrough Meter plugs, not to mention that my DAW (Samplitude ProX) has its own rather robust metering options.
I just like to have "tools" at my displosal...even if I only use them occasionally.

If anyone has a valid, different angle on this stuff....I would be glad to hear it. Nothing confuses people more than dBFS metering, IMO...especially the people who never really worked a lot with analog gear and analog metering standards.

 

[ashcat_lt]

I've been talking about recorded levels in the DAW. For input levels, 0dbfs means "exactly the maximum voltage that the ADC can convert", and -12dbfs means "exactly one quarter of the maximum voltage that the ADC can convert." It is completely dependent on what I guess we're calling the headroom of the converter. If you stick a +18dbu(~17V) signal into a +18dbu converter, you get a 0dbfs signal in the DAW. If you stick a +24dbu(~35V) signal into a +24dbu converter, you get a 0dbfs. If you stick a +6dbu (~4.3V) signal into a +18dbu converter, you get -12dbfs in the DAW. If you stick that same signal into a +24dbu converter, you get 20 * log (4.3/35) ~ -18dbfs. In order to get the -12dbfs in the DAW on that +24dbu converter, you have to put in a quarter of its maximum voltage, or ~8.8V. (These numbers don't exactly total up because of rounding errors, but the point is made.)

I know you don't want to believe it, but it's the way it works. It's the same on the other end, too. Run a 0dbfs signal out of a +24dbu converter and you'll get 35V, if you try to push that through an analog mixer that clips at 17V without first attenuating...

 

[miroslav]

My point is....once the signal is converted, like say I send you an audio file...how does your DAW know what the hearoom level was of my converter in order to properly display on your DAW meter the correct dBFS level relative to my converter....???

The DAW sees no analog headroom or P-to-P info in that audio file....so we're back to the same thing, once it's converted, the headroom and P-to-P stuff is irrelevent, and the DAW metering has no connection to which converter you used.....right?

Metering at the converter is tied to the converter. Metering the audio file later in the DAW is not...I don't see how it could be.

 

[ashcat_lt]

My point is....once the signal is converted, like say I send you an audio file...how does your DAW know what the hearoom level was of my converter...

It neither knows nor cares. Nor does your DAW, for that matter.

...in order to properly display on your DAW meter the correct dBFS level relative to my converter....???

The DAW sees no analog headroom or P-to-P info in that audio file....so we're back to the same thing, once it's converted, the headroom and P-to-P stuff is irrelevent, and the DAW metering has no connection to which converter you used.....right?

To simplify just a bit to aid clarity:

The ADC divides the input voltage by its rail voltage, which gives it a number between 0 and 1*. This is the value that it sends to the computer, which gets stored on your hard drive and manipulated in your DAW. Your DAW then does that 20 log x thing to that number and shows you the answer on the meter.

Send the file to me. My DAW sees the same number between 0 and 1, does that 20 log x thing and shows me the same thing on my meter that you saw on yours.

Now I send it to an output. My DAC multiplies that number between 0 and 1 by it's rail, and the actual output voltage will only be the same as what you put into if the rail on my DAC is accidentally the same as that on your ADC.

Master a song to 0dbfs and load it on your phone. Your phone will try to get as loud as it can, not as loud as your computer can.

It works on the other end also, though. 1/17 is quite a bit bigger than 1/35.

Which is to say (the part of whatshisname's post that made sense) that if I tell you that I record to -10dbfs in my DAW, it doesn't tell you anything about the actual voltage I'm running on the analog path in unless I also include data about my converter.

Metering at the converter is tied to the converter. Metering the audio file later in the DAW is not...I don't see how it could be.

My converters (Fostex 2424LV, Tascam US1641, PodStudioUX1) don't have analog meters. The 2424LV has digital meters which read exactly the same as my DAW, but I know from experience that it runs a little hot.


*Anything bigger than 1 is clipped off somewhere along the line, of course. Also, of course, there are rounding errors introduced - the noise that I was talking to the "number guys" about.


Edit - Actually, now that I've looked back at my converter specs, let me get even more specific. If I take any line level source, and split it to both the Fostex and the Tascam, the Fostex will always be 4db hotter in DAW.

Edit again - Wait, let me get even more specific since I think you'll get a kick out of this story.

The Fostex machine feeds my studio computer. This is where I "prototype" my live rig for my most recent show. We rehearsed through it, tweaked things, played again, tweaked a little more and then decided it was fine and time to test it on the live machine. Saved the entire Reaper file to my USB stick, stuck it in the laptop, and opened it. All the same plugins live on both machines. Everything else the same. Unplugged pedalboard from Fostex, plugged into the Tascam and it immediately felt just a tad anemic and weak. I couldn't get it to hit the amp or work the edge of the overdrive the way I had gotten used to. I never really memorized the exact difference between the two, and didn't bother to go look it up at the time. I just turned up the amp sim's In knob until it did what I wanted it to do. Turned out to be right around 4db.

 

[mixsit]

.. You would get two different recorded levels - if these 'two different converters use different i/o sensitivities.

Mmmmmm...I don't think a converter's I/O has "sensitivity" like a microphone.
It's a Line level I/O (not counting the all-in-one-boxes that include preamps).
You could maybe set it to -10 or +4 (which is what I have on my converters)....but once you choose either, you've established your reference.
So even with different brands/models....if the I/O is referenced to +4 on all of them, regardless of their headroom level, the same input signal would/should give the same dBFS reading.

Mmm yes maybe 'sensitivity is the wrong term.

.. If you have two converters, (+24dBu headroom & +18dBu headroom)....metering the same analog input level on each of those converters' meters, could give different dBFS levels on the converters' meters.
However, I'm still doubtful that the dBFS levels would be different if metered in the DAW.

IOW...the analog signal going into each of the converters was the same. Now take each converted signal and put them up on tracks 1 & 2, and play them....and I'm betting the *DAW track meters* would register the same dBFS for both signals/tracks.
The DAW has no idea how much analog "headroom" the converter had when it converted the signal. ..

This is where I think the term 'headroom of a converter' may be the wrong way to look at them- 'Headroom' is typically a reference to some other level is it not?
As in the RME examples, there are I/O gain settings - designed in' or adjustable (I called this their 'sensitivities' - to your input or output voltages referenced to a particular dBFS point),
And 'headroom' (at least for RME) is a level referenced to nominal ( line level, +4dBU or other?

..snip The reference for the DAW's meters is just the digital signal. I doubt that one signal would actually be louder than the other just because one converter had more headroom. They both received the same analog input level.
...snip
I wish I had two converters with different headroom specs to confirm the DAW metering thing....but I'm sure it would be like that.

So then the real question is.....where are we doing our metering?....at the converter or in the DAW after the conversion...or at the analog stage before the conversion?
In my case, I do all my critical metering at the analog stage, and I only glance at my converters' dBFS meters, but I never obsess over -18dBFS or -12dBFs or -6dBFS. I just make sure I've picked -10 or +4 at the converter, to match my analog source. ..

With a given input voltage, if you use a different input gains scaling (.. I still don't know what to call it, but refer to the '+4' vs 'Lo Gain' as in the RME for example again) you very definitely end up with that difference showing up in the track's recorded level.

FWIW ..nothing I never look at the converter's meters. Set up the preamp's gain, analog chain, compressor what have you, then it's either adjust an output fader, or if needed insert an in-line pad (or a different range on the converters in some cases), to get to into the nominal target dBFS range on the DAW meters.

 

[ashcat_lt]

The fun part is that there is no standard in any of this. Not for the actual values or for the nomenclature used. The Fostex manual says "Reference -12 db", the Tascam says "Headroom: 16db", others I've seen will give a maximum output level either as dbu, dbV, or even just voltage.

Generally, if you keep your analog chain happy, the digital will just fall into line. It's when you start pushing the extremes on either side (digital or analog) on either end (noise floor or headroom) that you will start to run into issues in the interfacing.

 

[mixsit]

The fun part is that there is no standard in any of this. Not for the actual values or for the nomenclature used. The Fostex manual says "Reference -12 db", the Tascam says "Headroom: 16db", others I've seen will give a maximum output level either as dbu, dbV, or even just voltage.

Yeah. I was just looking for an example of -what I though was fairly common in the pro converter world - wasn't hard to find; Apogee AD-16x,
http://www.google.com/url?sa=t&rct=...xVn3vAeadzcWTWOk0C9xfiQ&bvm=bv.72676100,d.cGU
- fully adjustable analog scaling for input voltage vs dBFS' levels +6 to +24dBU. (pg 6 I think
One converter, trim input voltages to pretty much any record ref level you like.

Generally, if you keep your analog chain happy, the digital will just fall into line. It's when you start pushing the extremes on either side (digital or analog) on either end (noise floor or headroom) that you will start to run into issues in the interfacing.

I've mentioned this before; Peak + RMS meters, 24dB scale, when the levels pop up into view in the DAW, good to goo! :>)

 

[ecc83]

Mention has been made of converters working at "line level".

They don't, well not the ADC chips themselves. Because they are powered from +5V or less the in/out levels for FS are usually limited to ~4.5V rms or +15dBu.

So, if you want a max out of +24dBu you need to bolt on an extra 9dB of gain and since a good quality chip might have a noise floor better than 110dB below FS the design of the input and output amplifiers needs care if the noise performance is not to be compromised. There are of course top end converters with -120dBFS noise floors and much of their cost is probably in the quality of the analogue front and back ends.

Signal levels have to be Jodrelled about in all sorts of ways to walk the noise v clipping line. Many people would think if asked I am sure that mixers (good ones!) run internal levels at 0dBu or 0dBV. Not a bit of it. Best NOISE with adequate headroom comes from an internal level of around -2dBu (615mV) for studio recording work but a broadcast mixer, where noise is not such an issue but clipping is, might run way down at -16dBu or 123mV.

Ashcat: I understand that analogue noise will come up with digital gain. Twas ever thus!
My point was that applying 20dB of such gain to a low signal will do little other harm.

In practice our signal to noise ratios are set totally by the analogue parts of the system (assuming 24bits) . This limit could be imposed by poor amps around the ADC chip but more likely by the outside world, ambient noise, most of us have rooms/10-20s, noisier than our mics/preamps!
And how "quiet" is even a very good synth?

Dave.