Is a soft signal of less quality than a loud signal?

N

nononsense!

New member
Let's say that +4 dBu average level is the best level for recording but that it also depends on
what level the equipment sounds best, but mostly it will be around or just lower than +4 dBu
average level (0VU).

But sometimes in soft passages the analog signal is on a weaker volume of -30 dBu. Does a
weaker volume of -30 dBu give a less accurate or less quality signal? I am not talking about
noiselevels. I mean the accuracy of the electric signal being described by the voltage.
Of course human beings can't hear any difference at first with only one track, but is there a
difference technically, in the quality of the electronic signal? Can this be proven?
 
Interesting question, I guess. I suppose the answer depends a lot on the integrity of the signal path. There is a chance that things will at least sound different at a low level like that. And your signal will get closer to whatever noisefloor you have to work with, which isn't a real good thing. And the tails of the signal that trail off will be closer to the least significant bit (or your lowest quantizing step), if recording digitally. Especially if recording 16-bit.

Summed up: I don't know, but it depends a lot on the quality/characteristics of your signal path.
 
Theoretically, no, there is no difference in the information contained in the quieter analog signal. An analog signal, by definition, has infinite detail, regardless of its amplitude.

However, there are practical limitations, the major one being self-noise of the entire recording chain. So if you record a signal at unnecessarily quiet levels, you will lose dynamic range.
 
"Quality" is in the ear of the beholder.

That being said - It's very rare that I hear a signal that's *too* soft...

Too *loud* on the other hand - All the time. Morning, noon and night.
 
mshilarious said:
Theoretically, no, there is no difference in the information contained in the quieter analog signal. An analog signal, by definition, has infinite detail, regardless of its amplitude.
Click to expand...

That would be a sort of relief for me if that was true. Can a electrotechnician verify this?

I am homerecording for about 20 years now and always misinterpreted the best recording
level. I always thought that +4 dBu 'the zero leds on the mixer' were to be taken as the
best PEAKlevel, but as it now seems it is to be taken as the best AVERAGE level.
I haven't got much of a problem with it but in fact I underrecorded everything.
 
Most analog gain circuits have an optimal gain setting where the signal will be operating in the most linear portion of the transfer function. This is especially true with non class A circuitry. Then there's tubes and transformers, which definetly sound different depending on how you drive them. In strict terms of accuracy though, and in a class A circuit, there should be little difference between using all or half of the linear portion of the curve.
 
You know...I had a practical experience with this just last night. I was testing out some new mics that I just got (from mshilarious actually). I was recording acoustic. I have an RME Fireface 800. I typically have my outputs and inputs set at -10. I plugged the mics in, turned up the gain on the FF until I got a signal and layed down a track.

It sounded dull....lifeless.....like poo. Not to mention, it wasn't a very "full" signal. I don't know how else to term it.

So...I opened up the FF settings and turned the inputs on +4 and turned the gain on the front of the FF up a little bit more.

What happened?

The recored acoustic sprung to life!!!!! So much so that it actually made me smile. I didn't even need any EQ to get it sounding good. Now...in a mix, it might sound different. But I was pretty damn happy with the acoustic sound I got.

This all being said...I'm not sure it even applies. :D
 
Well once we leave theory and deal with a real circuit, there is definitely a sweet spot. Too little gain will mean more noise, too much gain will mean more distortion.

As for the -10/+4 dilemma, it's possible the gear is taking a +4 signal and ramming it through a pad to produce -10. Or it could be taking a -10 signal and adding +12dB of low-quality (or digital) gain to get +4. More components than necessary usually doesn't help.
 
mshilarious said:
Theoretically, no, there is no difference in the information contained in the quieter analog signal. An analog signal, by definition, has infinite detail, regardless of its amplitude.
Click to expand...


That would be a sort of relief for me if that was true. Can an electrotechnician verify this?

I am homerecording for about 20 years now and always misinterpreted the best recording
level. I always thought that +4 dBu 'the zero leds on the mixer' were to be taken as the
best PEAKlevel, but as it now seems it is to be taken as the best AVERAGE level.
I haven't got much of a problem with it but in fact I underrecorded everything.
 
Well, I think you are missing the bigger picture. Sure you want your signal to be far enough away from the noise floor, but you only have so much headroom in your mixer, your A/D convertors or tape recorder, etc.
Depending on your mixer, you may not have a whole heckuvalotta headroom above +4. But if you are worried that your stuff doesn't sound crappy enough, then by all means, push the headroom limits of your signal path.
 
nononsense! said:
That would be a sort of relief for me if that was true. Can an electrotechnician verify this?
QUOTE]

Why do you need a concrete factual analysis of this? With so many variables in so many differnt circuits you might as well be asking if Strawberry is the best flavor of jelly.

Use your ears to find the "best" level to record at.
Click to expand...
 
And you didn't "under-record" everything. Lower is fine. And if you have lights instead of meters that are calibrated to VU, they're *VU* (ballistics taken into consideration). So that *IS* average level.
 
Zed10R said:
QUOTE]

Why do you need a concrete factual analysis from an electrotechnician? Use your ears to find the "best" level to record at.
Click to expand...


I think it is time after 20 years that I know for sure what the best level of a signal is, not one book about recording goes deeper into that.
So maybe an electrotechnician could be of help.

Most people's ears are certainly not that good to know what the best electronic level is.
 
Uh,....are we still talking about music here? If so, you are somewhere way out in left field. A music signal is all over the place; not one specific constant value. That is what makes music interesting--changes in level and frequency. If you tell us the specs of your mixer, we can tell you what is NOT an acceptable signal level.

Maybe you should go troll yourself at an electrotechnician forum.
 
You must log in or register to reply here.
Back
Top