Stereo or Mono Tracking?

When recording to a DAW, it's best to nearly always track each instrument in mono... it's the only way you will have control over the stereo image and maintain space in your mix.

Some sounds you might well decide to keep in stereo, but this will most likely only be because you specifically want the stereo effect for that indiviual sound.

Very deep sounds (sub-bass) are mono anyway... even if you record them in stereo, they will still be mono.

One instrument you might well record often in stereo is a whole drum kit to preserve the stereo panning of the constituent sounds... unless you record each drum seperately, in which case record each sound in mono on their own track.

If a particular sound is going to be prominent in your mix... and you like the stereo sound coming from the FX, then record it in stereo... a good example of this is piano. However, if the piano is to be simply a supporting instrument in your mix, record it in mono.

Sorry if my reply is a bit rambling, but hopefully it will be of some use to you.

)

Dave

http://www.drstudio.demon.co.uk/
"Musicians fall into three categories: those with a physical block, those with a mental block and those with a wooden block". Unfortunately, I'm in the 1st category!

Nope you werent rambling at all...lol

Thanks for clearing that up for me.... My mixes did seem a bit cluttered at times...more space is exactly what I need...

Just a quicky! on a tecnical note..

if you want to be efficient with your mono samples
you could save them out as 22000hz as appose to 44000
reason is that when playing a mono sample, each Bit is maped x2 times
once for left and agan for right.. thus you still hear 44kz of sound.
where as with a stereo 44kh samples
each bit is only mappeds once for either left or right but not both.
so you get half 44 in the left and the other in the right so thats still only 22000 x2 times...

Chow for now..

Results not Excuses
MSN: fantom01@hotmail.com

Get optimised: Rev 9.1




QuinnX universal tranlator below:-
(lets not be misunderstood)

!!!!!! Emphasis not anger
!!?? Confusion or Stupidity
ABC are you getting the point
&*$% Anger (but rarely seen)


Fantom-X6 with Audio Track Expansion (OS 2.0)
Windows 2000
Intel P4 2g
512mb

Quinnx,

unless I've been horribly mistaken all these years I don't think that is correct at all.

Each sound is best played back at 44kHz irrespective of how many sounds, stereo or mono are played. Even if it is finally rendered at 44kHz, if you start off with 22kHz then each sound will be frequency restricted and mixing more of them together will not improve the fidelity.

The frequency of a 22kHz sample is limited to an 11kHz bandwidth even if it is upsampled to 192kHz... if the detail is not there to begin with, up sampling won't magically restore it.

As an analogy, take a picture with a digital camera at 800x600 resolution. Enlarge it to 1600x1200 and it contains no more detail than the 800x600 picture as each 2x2 block of the 1600x1200 picture is simply a 4 pixel representation of the original pixel in the smaller version... i.e. 4 pixels of the exact same colour.

)

Dave

http://www.drstudio.demon.co.uk/
"Musicians fall into three categories: those with a physical block, those with a mental block and those with a wooden block". Unfortunately, I'm in the 1st category!

the best comparison for this is
record a mono source at both 44 and 22 recording in mono
listen to them both
you will not get any extra benifit from 44

it has been explained somewhere that the human ear is sensative to 22kh thus stereo is 44
its 44 because each ear only hears every other bit

stereo= 2 mono pairs inter leaved
i you split them you get 2x 22khz mono pairs


also..
your analogy with the picture is slightly off..
since your starting with a mono picture you must start at its native resolution which IS its final resolution
so if its native resolution is 800x600 thats what it is
however if this is a 3d(stereo picture) each picture left and right
must also be 800x600 so final resoltuion (stereo) is 1600x1200
but each frame is still only 800x600
Now.. just because the original stereo image was 1600x1200
does not mean the same picture mono should also be 1600x1200
becase you are only using 1 frame 800x600
but if you require to send the same frame a right and left monitor each monitor will simply display the same picture 800x600
thus sound
left and right will hear 22hz


Results not Excuses
MSN: fantom01@hotmail.com

Get optimised: Rev 9.1




QuinnX universal tranlator below:-
(lets not be misunderstood)

!!!!!! Emphasis not anger
!!?? Confusion or Stupidity
ABC are you getting the point
&*$% Anger (but rarely seen)


Fantom-X6 with Audio Track Expansion (OS 2.0)
Windows 2000
Intel P4 2g
512mb

"record a mono source at both 44 and 22 recording in mono, listen to them both, you will not get any extra benifit from 44"

I'm sorry, Quinnx, but you are completely wrong.

It is true that human hearing only extends to around 20-22kHz. But the mistake you are making is in assuming that a 22kHz sample can play back a 22kHz tone... it can not. It is necessary for the sampling frequency to be at least twice the frequency of the material being sampled. This is called the "Nyquist Theroem". I'd recommend you google that phrase and have a read up on it. It will lead you to realise that a 22kHz sample can not reproduce a tone any higher that 11kHz.

"so final resolution (stereo) is 1600x1200"

I'm afraid you'll find the maximum resolution you can claim for that scenario is 2 lots of 800x600... so considering the horizontal orientation of binocular vision, the resolution "could" be considered to be 1600x600


)

Dave

http://www.drstudio.demon.co.uk/
"Musicians fall into three categories: those with a physical block, those with a mental block and those with a wooden block". Unfortunately, I'm in the 1st category!

okay try this...
you want to play a mono sample at 22 to the equvelent of 44
you simply route the same sound to both left and right total 44

thus on the fantom you can take a 22 sample and asign it twice
once for left and again for right
total 44

the same if you sample a mono source in stereo
your actualy makeing 2 mono captures
if you sample in stereo fast enough you manage to sample the same exact wav sound 2 times for that moment you sample
but your still recording a mono sound not stereo
t

Results not Excuses
MSN: fantom01@hotmail.com

Get optimised: Rev 9.1




QuinnX universal tranlator below:-
(lets not be misunderstood)

!!!!!! Emphasis not anger
!!?? Confusion or Stupidity
ABC are you getting the point
&*$% Anger (but rarely seen)


Fantom-X6 with Audio Track Expansion (OS 2.0)
Windows 2000
Intel P4 2g
512mb

I'm sorry to have to contradict you again, but stereo 22kHz is not equivalent to mono 44kHz. If the sample rate is restricted to 22kHz at any point the final audible result will be restricted to a bandwidth of 11kHz and not the full 20-22kHz range of human hearing.

)

Dave

http://www.drstudio.demon.co.uk/
"Musicians fall into three categories: those with a physical block, those with a mental block and those with a wooden block". Unfortunately, I'm in the 1st category!

of course its not (thats not what i said)

your still thinking linier...
think outside the box...


lets try be definative..

stereo consists of 2 mono pairs interleaved correct?

each mono pair seperatly is being played at the 22hz not 44 correct?

so mono bandwidth per channel is 22 not 44
correct?

so mono in its strictest sence is 22..

Results not Excuses
MSN: fantom01@hotmail.com

Get optimised: Rev 9.1




QuinnX universal tranlator below:-
(lets not be misunderstood)

!!!!!! Emphasis not anger
!!?? Confusion or Stupidity
ABC are you getting the point
&*$% Anger (but rarely seen)


Fantom-X6 with Audio Track Expansion (OS 2.0)
Windows 2000
Intel P4 2g
512mb

stereo consists of 2 mono pairs interleaved correct?

Yes.

each mono pair seperatly is being played at the 22hz not 44 correct?

No.

so mono bandwidth per channel is 22 not 44
correct?

No.

so mono in its strictest sence is 22..

No.

A stereo 44kHz sample has two 44kHz interleaved mono channels. At 16bit depth, one second of stereo 44kHz sampling requires 176kb of memory... that's 44100 x 2 channels x 2 bytes (16bit) per sample.

I'm afraid the idea of adding 2 copies at 22kHz to make a 44kHz sample does not work. It would only improve the sound quality if the samples in the 2nd channel were true original interpolated values between adjacent samples in the 1st channel.... and they would have to be the original values and not a copy otherwise a 44kHz representation could not be constructed due to the restrictions of the Nyquist Theorem.... and if all that were the case, it would in actual fact be a 44kHz mono sample that happened to have it's data organised in an awkward manner.

So I'm afraid it's nothing to do with thinking outside the box or thinking in a linear fashion.... it's physics.

)

Dave

http://www.drstudio.demon.co.uk/
"Musicians fall into three categories: those with a physical block, those with a mental block and those with a wooden block". Unfortunately, I'm in the 1st category!

Hi guys, I'm certainly not a recognized expert in this field but I have been earning a living for the past 20+ years as an electrical engineer specializing in RF/microwave data and communications equipment design. You both are correct in some of your points. Dave's explanations are technically correct and accurate.

The formula for sample size is: (# seconds) * (# channels) * (sampling rate) * (bit resolution) / 8 = file size.

A stereo example:
a 60 second, 44.1KHz, 16 bit stereo sample is 10.584Mb ((60*2*44,100*16)/8)=10,584kb.

A mono example:
a 60 second, 44.1KHz, 16 bit mono sample is 5.292Mb ((60*1*44,100*16)/8)=5,292kb.

When Quinnx originally wrote:
if you want to be efficient with your mono samples you could save them out as 22000hz as appose to 44000 reason is that when playing a mono sample, each Bit is maped x2 times once for left and agan for right.. thus you still hear 44kz of sound. where as with a stereo 44kh samples each bit is only mappeds once for either left or right but not both. so you get half 44 in the left and the other in the right so thats still only 22000 x2 times...

This is not true and the reason provided that when playing a mono sample, each Bit is maped x2 times once for left and agan for right.. thus you still hear 44kz of sound is not correct. Furthermore the last sentence: where as with a stereo 44kh samples each bit is only mappeds once for either left or right but not both. so you get half 44 in the left and the other in the right so thats still only 22000 x2 times... is also simply not correct.

***edit*** Also, when Quinnx wrote: the best comparison for this is record a mono source at both 44 and 22 recording in mono listen to them both you will not get any extra benifit from 44 this is incorrect also. The 22KHz recording in this example will not be able to faithfully record frequencies above approx 11KHz (Nyquist) whereas, the 44KHz recording will be able to faithfully record frequencies up to approx 22KHz. Also the 44KHz recording has twice the sampling rate (which, everything else being equal,) results in a lower SNR--but that is a different topic entirely!!

Bottom line: The higher the sampling rate, the higher the frequency that can be faithfully recorded AND the truer the recording will be to the original. A good analogy is the "stair-step jaggies" that are present when a low-resolution image is enlarged. The low-resolution digital image attempts to re-create curved lines with a series of straight line stair-step segments.

(don't get me started on engineering topics, I'll bore the heck out of everyone!)

more info is available at: http://music.northwestern.edu/links/projects/midi/pages/audvsmid.html


---------------------------
Learning to play like Sponge Bob !