What Sample Rate Do You Work In?
- Thread starter Styvo
- Start date
jipchen
ForesterStudio
- Feb 27, 2008
- 189
- 0
- 16
It's interesting,
I pretty much feel the same way that working in 44.1 at 24bit is more than enough to give you a quality end product.Most times i prefer to work in this format,however for some reason,clients feel that a higher sample e.g 48k-88.2 will give them a better outcome.Today we took a client over to another studio to track a grand piano and the engineer loaded the session at 96k-24b.
On the way home my business partner was suggesting we do this on some upcomming recordings,i don't believe it will make a difference.
Hence i thought i'd find out how you guys felt about it,and if you did work in 88.2-96k did you feel that you had some sort of edge over recordings done at say 44.1 or 48k.
I pretty much feel the same way that working in 44.1 at 24bit is more than enough to give you a quality end product.Most times i prefer to work in this format,however for some reason,clients feel that a higher sample e.g 48k-88.2 will give them a better outcome.Today we took a client over to another studio to track a grand piano and the engineer loaded the session at 96k-24b.
On the way home my business partner was suggesting we do this on some upcomming recordings,i don't believe it will make a difference.
Hence i thought i'd find out how you guys felt about it,and if you did work in 88.2-96k did you feel that you had some sort of edge over recordings done at say 44.1 or 48k.
ze kink
THE BLACK WIZARDS
Was using 48/24 some time ago as an experiment, but I recently switched back to 44.1/24. I didn't notice anything to be better with 48.
Torniojaws
They call me Juha
I guess most people here work for CD releases, so there's no point in using non-native sample rates, because you'll have to downsample it to 44,1 kHz anyway for the CD.
HeadCrusher
Member
dumb question, but how do you select 32 bit summing in Nuendo?
EDIT
Upon reading, the summing is always 32 bit in Nuendo, correct?
John_C
formerly Skeksis268
am i right in saying that the digital 44.1 khz approximation to a sine wave at 22.05 khz with amplitude 1 would be:
1, -1, 1, -1, 1, -1 etc.....
EDIT: those numbers are amplitudes BTW
1, -1, 1, -1, 1, -1 etc.....
EDIT: those numbers are amplitudes BTW
Torniojaws
They call me Juha
No, you're not correct. There's 44100 samples taken per second (steps) and the height depends on the bit depth, eg. 24-bit -> 2^24 = 16777216 steps for peak-to-peak values. In other words, it's extremely close to an analog sine wave with that precision. The amplitudes would be more like (starting from origo):
0000 0000 0000 0000 0000 0000
0000 0000 0000 0000 0000 0010
0000 0000 0000 0000 0000 0100
0000 0000 0000 0000 0000 0101
...
1111 1111 1111 1111 1111 1111
...
0000 0000 0000 0000 0000 0010
0000 0000 0000 0000 0000 0000
1000 0000 0000 0000 0000 0010
1000 0000 0000 0000 0000 0100
1000 0000 0000 0000 0000 0101
...
0111 1111 1111 1111 1111 1111
To get the rough idea. Those are binary values of course. In decimals it's something like 0, 0.000014, 0.000020, 0.000026, ..., 1, 0.999994, 0.999988, ..., 0, -0.000014, -0.000020, ..., -0.999988, -0.999994, -1
All the values are just some random rough numbers to display the idea. The actual values would be calculated with Fourier and difference equations with programs like MatLab.
0000 0000 0000 0000 0000 0000
0000 0000 0000 0000 0000 0010
0000 0000 0000 0000 0000 0100
0000 0000 0000 0000 0000 0101
...
1111 1111 1111 1111 1111 1111
...
0000 0000 0000 0000 0000 0010
0000 0000 0000 0000 0000 0000
1000 0000 0000 0000 0000 0010
1000 0000 0000 0000 0000 0100
1000 0000 0000 0000 0000 0101
...
0111 1111 1111 1111 1111 1111
To get the rough idea. Those are binary values of course. In decimals it's something like 0, 0.000014, 0.000020, 0.000026, ..., 1, 0.999994, 0.999988, ..., 0, -0.000014, -0.000020, ..., -0.999988, -0.999994, -1
All the values are just some random rough numbers to display the idea. The actual values would be calculated with Fourier and difference equations with programs like MatLab.
~BURNY~
Member
- Apr 20, 2005
- 5,091
- 67
- 48
Soundlurker
Member
- Nov 19, 2005
- 3,730
- 0
- 36
44,24 unless the drum tracks I received were recorded at 48/24 in which case I record the rest in 48 as well. When I'm done with the mix I use r8brain in very high quality mode to downsample and I feel it does a perfect job.
If you're wondering about something you can always try to find out how the pros are working. If I recall correctly, Andy uses 44khz,24bit, as well.
If you're wondering about something you can always try to find out how the pros are working. If I recall correctly, Andy uses 44khz,24bit, as well.
John_C
formerly Skeksis268
thanks, good explanation
however, i did say a 22.05khz sine wave
what you just described was an approximation to a sine wave with frequency 44100/(2*2^24)= 0.001314 Hz
(for anyone following along, that was the number of samples per second, divided by the total number of different displacements multiplied by 2 because a sine wave covers each displacement twice)
and "from origo"??? is that like "a priori"?
Torniojaws
They call me Juha
Oh I thought you meant you had a sinewave of 22 kHz recorded at 44,1 kHz
origo = the place where X and Y cross
origo = the place where X and Y cross
John_C
formerly Skeksis268
seeing as you seem to know a lot more about the technicalities of this than me.......
are these statements correct
1. A sine wave of frequency 22.05 khz has period 1/22050 = 4.54 x 10^-5 seconds
2. The spacing of the samples when recording at a sample rate of 44.1khz is 1/44100 = 2.27 x 10^-5 seconds
3. 4.54 x 10^-5 = 2 x 2.27 x 10^-5 Therefore 1 complete oscillation of a sine wave of frequency 22.05khz recorded at 44.1khz will consist of only 3 samples
i hope there's a mistake there somewhere, otherwise there are serious consequences for recording at 44.1khz IN THEORY
are these statements correct
1. A sine wave of frequency 22.05 khz has period 1/22050 = 4.54 x 10^-5 seconds
2. The spacing of the samples when recording at a sample rate of 44.1khz is 1/44100 = 2.27 x 10^-5 seconds
3. 4.54 x 10^-5 = 2 x 2.27 x 10^-5 Therefore 1 complete oscillation of a sine wave of frequency 22.05khz recorded at 44.1khz will consist of only 3 samples
i hope there's a mistake there somewhere, otherwise there are serious consequences for recording at 44.1khz IN THEORY
Torniojaws
They call me Juha
True
True
According to the Nyqvist-Shannon Theorem, that is enough.
But no human hears 22 kHz. Theoretically the ear can detect (barely) 20 kHz, while most people 12 years old and older will commonly only hear up to 16-18 kHz. In music, the "money shot" range is probably about 100 Hz - 10 kHz
Mikaël-ange
Member
- Nov 7, 2008
- 2,368
- 1
- 36
John_C
formerly Skeksis268
Ok, i get what the Nyquist-Shannon sampling theorem is, although i couldn't follow the arguments for why it's true
In my simplistic way of looking at things, the 3 samples i mentioned before if you join the dots you get a triangle wave. Which is a pretty shocking approximation of a sine wave
Let it burn...
Orgasm Donor
96k recording
44.1 mixing
Why? Because I can. lolz
No srsly. I record at 96 because the amount of samples taken a second gives a clearer representation of the waveform as it was in the analog domain. Then I just edit in 96 and convert to 44.1 (so the SRC to 44.1 has more samples to chose from)
44.1 mixing
Why? Because I can. lolz
No srsly. I record at 96 because the amount of samples taken a second gives a clearer representation of the waveform as it was in the analog domain. Then I just edit in 96 and convert to 44.1 (so the SRC to 44.1 has more samples to chose from)
