conquistadore
Forum Moderator
Joined: 03 Aug 2004
Posts: 541
|
| Metering and recording - Analogue and Digital |
|
|
This is one area which is unfortunately not much adressed to in today's recording world. So lets get this thread running and see how much we can educate ourselves..
*Recording levels on your DAW - headroom, definition of "0db"
*System calibration
*Metering systems
Last edited by conquistadore on Fri Nov 24, 2006 12:10 pm; edited 2 times in total |
Sat Nov 11, 2006 3:39 pm |
|
|
conquistadore
Forum Moderator
Joined: 03 Aug 2004
Posts: 541
|
| Recording levels |
|
|
Ok, for whoever who feels a bit lost, lets start of with the recording process. How do you record your signals in the digital domain? Do you record till just before your signal clips or you use some kind of 'reference' level? |
Sat Nov 11, 2006 3:47 pm |
|
|
|
|
amitbarde
Forum Moderator
Joined: 24 May 2006
Posts: 79
|
| here it is! |
|
|
Agree...metering is probably one of the most important topics when it comes to recording in a studio environment or for that matter anywhere in the field of sound...be it live, studio....etc.....
Well for those who read this:
I'll tell you what the "0dB" stands for.....and why do engineers insist that the fader must be at 0dB when gain for the incomming signal is being set.
Loosely speaking 0dB, also known as "unity gain", is the point where the incomming signal is not being cut or boost in terms of it's strength. The signal level a meter shows you when the fader is at the unity position is the signal that is being output from that particular instrument/device.
I hope this does make sense to people who are reading it. If it does sound a little confusing please do let me know.
Will put up more later. |
Mon Nov 13, 2006 10:34 pm |
|
|
conquistadore
Forum Moderator
Joined: 03 Aug 2004
Posts: 541
|
| 0dB |
|
|
Yes, in analogue gear, 0dB is usually the point where the sound sounds best. Because, the signal is neither too high to 'distort' or too low to get lost in noise. That is why its always best to record, when using analogue equipment, with the signal dancing around 0dB.
But when we come to the digital domain things get different. If you notice, 0dB in the digital domain is the highest point a signal can go, anything higher than that is heard as terrible digital distortion - called clipping. While in the analogue world, we have a reference level (the 0dB explained in the previous para), in the digital world there is no defined 'ideal' level as such. So that is why we find many people recording as "hot" a signal they can in the digital world. But do we really need to record such hot signals?
Not really, it is believed that its best to have a reference of anywhere between -15dBFS (dBFS it dB in the digital world) and -20/-22dBFS to record. Recording signals that dance around -20dBFS gives enough 'headroom' for sudden peaks and transients, and reflects the way the analogue world works (analogue gear usually have a headroom of 18 to 24dB above 0dB and greater). |
Tue Nov 14, 2006 10:02 am |
|
|
|
|
amitbarde
Forum Moderator
Joined: 24 May 2006
Posts: 79
|
| Re: 0dB |
|
|
quote:
Originally posted by conquistadore
Yes, in analogue gear, 0dB is usually the point where the sound sounds best. Because, the signal is neither too high to 'distort' or too low to get lost in noise. That is why its always best to record, when using analogue equipment, with the signal dancing around 0dB.
Thank you conquistadore! That's a very important point you've put forth. It is extremely important to know the differences between analogue and digital audio. Speaking of which i think it would be useful if a parallel discussion on the topic was started (ofcourse barring the arguments about which of the two mediums is better ) |
Tue Nov 14, 2006 10:16 am |
|
|
amitbarde
Forum Moderator
Joined: 24 May 2006
Posts: 79
|
| no posts? |
|
|
Looks like no one's paying attention to this topic. It would be great if some one else could contribute to this topic too. It's already been bumped down to some extent. Anyone with more information please do post! |
Fri Nov 17, 2006 7:58 am |
|
|
|
|
wailingalleycat
Forum Manager
Joined: 07 Oct 2003
Posts: 762
Location: Jersey C.I (UK) |
| Headroom & Material |
|
|
i find alot of the time the amount of headroom i leave is very dependant on what i am recording. for example a heavily compressed rock guitar will have little dynamic range, so you can safely let the input level dance up to -4db or so without much worry, especially as guitars are noisy instruments, so you need to get optimum S/N ratio throughout the signal path. however if i was recording an opera or possibly jazz singer, i would leave much more headroom, perhaps a peak of -16 during average programme level would be reasonable in a noise free environment. if i was particular concerned about Signal to Noise, i might look at using a bit more input gain, but placing a compressor in the record path, with a fairly high ratio and high threshold, just to prevent any anomalies jumping through and ruining the take. _________________ If In Doubt...Hit It With a Hammer, If Still In Doubt... Find a Bigger Hammer. |
Mon Nov 20, 2006 2:02 pm |
|
|
conquistadore
Forum Moderator
Joined: 03 Aug 2004
Posts: 541
|
| Recording and SNR |
|
|
Hi wailing!
Dont you think 4dB headroom is less for a guitar? Yes it may not clip during the recording, but you are sure to equalize at some point during the mix down stage and that is sure to eat up all your headroom and clip your signal.
And talking about signal to noise ratio, if you are recording in a 24bit system it doesnt really matter that much because with 24 bits you get a theoretical headroom of 144dB (which maybe a bit lower practically) which is a lot! So if you record something at -10dBFS you would end up having a dynamic range of 134dB (theoretical), which is more than you would need, because noise right down at 130dB or even a 100dB would not be
that
noticable. Recording a hotter signal and issues like signal to noise ratios would enter only if your input signal chain and analogue to digital convertors were noisy, which is not the case today as even the cheaper AD convertors have good signal to noise specs.
So it would be better dont you think if you record safer at around -18 to -20dBFS and have a more than comfortable headroom during mixdown and not worry about your EQ or other processors clipping your signal (or even the vocalist, guitarist or drummer getting a bit excited while recording!).
People may find signals recorded at -20dBFS too 'soft', but that depends on how your monitoring system is calibrated, and thats when we can bring in the issue of calibration and monitoring systems. |
Mon Nov 20, 2006 3:32 pm |
|
|
|
|
wailingalleycat
Forum Manager
Joined: 07 Oct 2003
Posts: 762
Location: Jersey C.I (UK) |
| Clarifying |
|
|
sorry Varun,
didnt make that clear enough, im talking about a maximum peak level of -4db, but like i say thats only in very special circumstances of a heavily compressed instrument with limited dynamic range. most notably if recording in a budget home studio, signal to noise is at a premium, so its always a good idea to get a good level from the beginning. if i wanted to add extensive EQ etc. it would be easy enough to drop the level down in the DAW with destructive processing or whatever to be certain of avoiding clipping.
also bare in mind the bit quantisation that will occur when recording at very low levels. ill try find a diagram or something to explain this. _________________ If In Doubt...Hit It With a Hammer, If Still In Doubt... Find a Bigger Hammer. |
Mon Nov 20, 2006 4:46 pm |
|
|

|
|
Goto page 1, 2, 3, 4 Next
All times are GMT. The time now is Thu Mar 11, 2010 6:37 pm
|
|
|
|
| |