| Author
|
When do we need more than 44100?
|
Kane
IsraTrance Junior Member
Started Topics :
23
Posts :
1772
Posted : Feb 2, 2009 23:24
|
Quote:
|
On 2009-02-01 18:18, x-rayz wrote:
Quote:
|
On 2008-08-08 00:31:08, Medea wrote:
I'm wondering, when does it make sense to use samplerates more than 44100?
Recording hardware synths? Or making final mixdown?
|
|
Both. Sample rate of recorded material should be at least twice times higher than the frequency u are recording.. So practically, u dont need to record in more than 8000-10000 sample rate, that is, if ur recording a voice :-D I always mixdown and rec at at least 88 kHz..
Upavas was making a good point but since I dont lowpass..
|
|
Twice as high as the highest frequency you're recording. Recording vocals at that sample rate would sound terrible.
  You believe in the users?
Yeah, sure. If I don't have a user, then who wrote me? |
|
|
|
The Andychrist
The Andychrist
Started Topics :
11
Posts :
150
Posted : Feb 3, 2009 00:27
|
Quote:
|
On 2009-02-02 23:24, Kane wrote:
Twice as high as the highest frequency you're recording. Recording vocals at that sample rate would sound terrible.
|
|
Hmmm, unless I am missing something, it would seem that recording vocals at that sample rate would sound fine for use in trance music.
Human voices are roughly in the range of 80 Hz to 1100 Hz (that is, E2 to C6) for normal male and female voices together. ( http://en.wikipedia.org/wiki/Vocal_range )
So twice the highest frequency involved is roughly 2200 Hz for the tonic of the highest voices. To account for nice overtones, I think we can safely say that 10000 Hz is a sufficiently high sample rate to record vocal harmonics up to 5000 Hz.
Coincidentally, telephone transmitters have a frequency-response range from 250 to 5000 Hz. Of course, telephone sound is not perfect, even for voice, but if the goal here is to use recorded voices in trance music ( which is debatable ), we probably will end up cutting many of the higher frequencies anyway for the sake of the mix, esssspecccccially if the vocalsssss are in englisssssssshhhhhh... =)
http://soundcloud.com/the-andychrist
http://unwashedtomato.com |
|
|
|
~d2~
Inactive User
Started Topics :
7
Posts :
751
Posted : Feb 3, 2009 00:40
|
Your first mistake was to look on wiki.
What you are talking about is the root or fundamental. Don't you want the harmonics as well?
Record your voice and run it through a low pass filter. It will soon become apparent. |
|
|
|
The Andychrist
The Andychrist
Started Topics :
11
Posts :
150
Posted : Feb 3, 2009 00:51
|
apologies, where I wrote "tonic", i should have written "fundamental".
If you look a few inches above your post, you will see:
Quote:
|
...To account for nice overtones, I think we can safely say that 10000 Hz is a sufficiently high sample rate to record vocal harmonics up to 5000 Hz.
|
|
this post is a section about producing trance music, no?
Try mixing english speech without a low pass filter into a track. Then you can play it out a big soundsystem and watch everyone wince in pain and get scared back 30 feet by the high frequencies. http://soundcloud.com/the-andychrist
http://unwashedtomato.com |
|
|
|
The Andychrist
The Andychrist
Started Topics :
11
Posts :
150
Posted : Feb 3, 2009 01:01
|
hey ~d2~,
I'm not trying to pick a fight here, but it is something I have noticed playing out.
Undoubtedly, the rules change for different types of music, and certainly with different languages. English however uses a lot of sibilance, which hurts when it is reproduced very loudly.
http://soundcloud.com/the-andychrist
http://unwashedtomato.com |
|
|
|
~d2~
Inactive User
Started Topics :
7
Posts :
751
Posted : Feb 3, 2009 01:09
|
Use a de-esser then.
 |
|
|
|
~d2~
Inactive User
Started Topics :
7
Posts :
751
Posted : Feb 3, 2009 01:15
|
Quote:
|
On 2009-02-03 00:51, The Andychrist wrote:
apologies, where I wrote "tonic", i should have written "fundamental".
If you look a few inches above your post, you will see:
Quote:
|
...To account for nice overtones, I think we can safely say that 10000 Hz is a sufficiently high sample rate to record vocal harmonics up to 5000 Hz.
|
|
this post is a section about producing trance music, no?
Try mixing english speech without a low pass filter into a track. Then you can play it out a big soundsystem and watch everyone wince in pain and get scared back 30 feet by the high frequencies.
|
|
Well I do live sound. Both recording and FOH. Every band I have worked with has been English speaking...not seen anyone wince in pain, ever.
The fact that it is a Psy-Trance forum doesn't alter good engineering practices. I really feel that you are just trolling about when you make remarks like that. I'm not looking for a fight though, so I'll give you the benefit of the doubt.
|
|
|
|
The Andychrist
The Andychrist
Started Topics :
11
Posts :
150
Posted : Feb 3, 2009 01:39
|
|
The Andychrist
The Andychrist
Started Topics :
11
Posts :
150
Posted : Feb 3, 2009 01:52
|
Live sound is very different from making trance music. I also ran Front of House for actual bands, but in that richly dynamic context, these matters can be treated differently.
Due to the current RMS war in mastering studios, producers must take into account that their tracks will very likely end up compressed into a mutant -7 RMS block of love. This is why I choose to err on the side of safety.
Thanks for your comments though. I found them very informative.
http://soundcloud.com/the-andychrist
http://unwashedtomato.com |
|
|
|
The Andychrist
The Andychrist
Started Topics :
11
Posts :
150
Posted : Feb 3, 2009 02:25
|
|
~d2~
Inactive User
Started Topics :
7
Posts :
751
Posted : Feb 3, 2009 02:52
|
Quote:
|
On 2009-02-03 01:52, The Andychrist wrote:
Live sound is very different from making trance music. I also ran Front of House for actual bands, but in that richly dynamic context, these matters can be treated differently.
Due to the current RMS war in mastering studios, producers must take into account that their tracks will very likely end up compressed into a mutant -7 RMS block of love. This is why I choose to err on the side of safety.
Thanks for your comments though. I found them very informative.
|
|
I have yet to speak with a mastering engineer who would want to compress that much. It is not the fault of mastering houses. If you are the producer and you don't have the label pulling rank then it is your call.
If you want to low pass vocals go for it. It's not for me. I would be more inclined to high pass and I usually do.
It is worth noting that the filters used at 44.1 are quite steep. I don't know what you used on your samples. I will have a listen though if you want.
and don't worry about your tone early. I can see now that you genuinely are wanting to discuss these issues.
|
|
|
|
x-rayz
IsraTrance Junior Member
Started Topics :
11
Posts :
576
Posted : Feb 3, 2009 15:13
|
hey d2, Andy was onto the point I was making.. But I forgot to mention the fundamental, when I said frequency.. And I didnt know exact freqs of voice and telephony so I didnt want to get into that, it was just for example.. But indeed they took advantage of this in telephony, actualy vocoder was invented that way, they wanted to narrow the frequency range..
So if u use steep filter 44.1 should be enough.. If not, as in my case, I use 88.2
And while were still on steep filters, best one I was using before was fabfilter simplon with 8 pole filter, meaning 48 dB..
If anyone knows of better filter with 10 poles or brickwall eq with cutoff of 60 dB let me know..
http://www.facebook.com/xrayzproductions
http://www.myspace.com/xrayzproductions |
|
|
|
~d2~
Inactive User
Started Topics :
7
Posts :
751
Posted : Feb 3, 2009 15:28
|
Hi x-rayz
I was meaning the actual filter in the converter. At 44.1 the filter will have to be very steep. This is quite hard to design with out any side effects. Where as once you get to about 60kHz a very gentle sloping filter is used in the converter. This is much easier to design and therefore why higher sampling rates an sound better even though we won't hear much past 22kHz.
Andychrists example doesn't really have a vocal track. Its vocal edited to create an FX. A different thing.
The telephone system sounds the way it does because of the limited bandwidth.
But are you going to be working at different sampling rates in the same song? Very unlikely. So why would say we only need to sample vocals at 8-10kHz? Are you going to set the session to that?
Why do we need to filter? I don't know why people feel the need to (except high pass). |
|
|
|
x-rayz
IsraTrance Junior Member
Started Topics :
11
Posts :
576
Posted : Feb 3, 2009 15:41
|
Nah man, dont take me for exact words, it was just example, if u used only a vocal track than 10 kHz would be enough sample rate.. And of course u cant use different sample rates in the track..
How I do it, I run the track in the mix in 44.1 than I do the mixdown in 88.2 so I have more frequency range when I work on the mastering.. Its all in 32 bit until mastering is done when final product goes to standard 44.1 and 16 bit..
About the filters.. Low pass, not high pass, Upavas said it why it is used..
Quote:
|
On 2008-08-08 02:01, Upavas wrote:
make sure you lowpass your mix around 20khz when going for 44.1 so you avoid aliasing meaning the wavelength is too short to be correctly sampled, 21khz becomes 19khz e.g.
|
|
So again if anyone knows of steeper filter or brickwall eq with 10 poles/-60 dB let me know..
http://www.facebook.com/xrayzproductions
http://www.myspace.com/xrayzproductions |
|
|
|
The Andychrist
The Andychrist
Started Topics :
11
Posts :
150
Posted : Feb 3, 2009 20:59
|
Quote:
|
On 2009-02-03 15:28, ~d2~ wrote:
Andy's example doesn't really have a vocal track. Its vocal edited to create an FX. A different thing.
|
|
He's right x-rayz. my earlier comments don't really apply if you want to treat the vocals as, well, actual vocals.
Also, I checked out the fabfilter, and it has a -48db/octave mode. do you need one steeper than that?
http://soundcloud.com/the-andychrist
http://unwashedtomato.com |
|
|
|
