Why record higher than 44.1 khz?

  • Thread starter Thread starter c_olin
  • Start date Start date
C

c_olin

New member
I've noticed that most basic computer audio interfaces have the ability to record up to 96 khz. My question is, if you aren't planning on releasing your music in DVD-A format, what are the advantages to this? If the end result will be 44.1 khz, why record higher?

Possibly this belongs in the newbie forum.
 
If you are recording something like classical music with a very very very wide dynamic range (soft orchestral), you might want to use 96k.

Or if you're working with equipment that is locked on 48k (Soundblasters, ADATs, etc).

And if you're doing any video work, it's always 48k or you'll have sync problems with the frames.
 
because higher sample rates sound better

it's something that can, and has been, argued to death on this forum and others, but i have no qualms with saying that 96k sounds better than 44.1 or 48k. however there are trade-offs, as 96k will of course eat your hard drive and processing power alive. on top of that, you likely won't gain any benefit from tracking at 96k unless you have a really great room and front end.
 
  • Like
Reactions: apl
I'm all about certain processes at higher sample rates - But I've yet to hear a (good quality) converter that actually sounds "better" at higher rates. Unless you have a source, microphone, preamp, processing, amplifier, speaker, hearing capabilities (etc., etc., etc.) that go beyond 22kHz, there won't be a difference. It's been beaten to death by the best designers on the planet. They'll tell you straight up - If you can hear a difference in their converters between 44.1 and 96kHz, the converter is broken.

And not that I'm not arguing against higher resolution sample rates (although I wouldn't touch 96kHz unless the project was destined for video). There have been a few occasions that I've tracked projects at 88.2kHz (and usually regretted it) strictly out of the insistence of the client.
 
  • Like
Reactions: apl
I understand the reasons to use higher sample rates if the end result would be on a medium that supports higher sample rates.

My question was more along the lines of... If the recording is going to end up on a 44.1 khz CD... is there any reason to record at a higher rates? Intuitively it seems like there is no reason, but I know people who record at higher even if it will only be distributed as MP3.
 
  • Like
Reactions: apl
If I've got the drive space I usually record at 96khz, but usually I use 88.2khz. I don't know why, but my mind tells me that downsampling to 44.1khz for putting on CD by halving the number of samples will be more accurate than some weird factor between 96 and 44.1. I don't have any actual knowledge to back that up, it just seems nice to be working with factors of two. I don't have any reason behind recording at higher rates other than, "it might be a bit better" :)

But the most important thing is working at 24bit.
 
Last edited:
  • Like
Reactions: apl
I record always at 44.1. There may be some benefits at using higher rates, but I suspect that any such benefits are overwhelmed by other weaknesses in my signal path. I'll live with 44.1 and enjoy the lack of demand on disk space and CPU, which is of greater value to me than a potential increase in quality.
 
  • Like
Reactions: apl
I've read that different converters may have a different slope to their low-pass filter, which has an impact on how much aliasing gets into the 20-20k range of the spectrum. The result being that better-designed filters are better equipped to sound good at 44.1, where some converters may benefit from a higher sample rate to get the distortion farther above 20k.

Sounded reasonable, any truth to it?
 
  • Like
Reactions: apl
Why record higher than 44.1 khz?


............................Because I can.
 
  • Like
Reactions: apl
all I know and can stand by is this.......going over my archives I have found that in absolutely every case the music I recorded (tracked) in 44.1 sounds vastly better to MY ears than the same recordings done while experimenting with higher rates. (48-96)
I must also add that for some reason everything THAT I have recorded (tracked), especially in 48k, lacks something...it's just not as 'pleasing' to my ears.
Could be something wrong with my 'down-sampling' procedures....but I'm not too sure.
Anyways, I'll stick with the pleasure I get from tracking in 44.1, at 24bit, ofcourse!!
Cheers.
 
c_olin said:
I understand the reasons to use higher sample rates if the end result would be on a medium that supports higher sample rates.

My question was more along the lines of... If the recording is going to end up on a 44.1 khz CD... is there any reason to record at a higher rates? Intuitively it seems like there is no reason, but I know people who record at higher even if it will only be distributed as MP3.
Click to expand...
There are only two reasons I can think of to do so:

1.) if you happen to have an A/D converter that sounds so much better at the higher sample rate that even after the cost of downsamplling back to 44.1 you still wind up sounding better than if you just recorded at 44.1 and avoided the sample rate conversion. But it's important to remember that if this is the case, it's not because the higher sample rate is itself superior, but because that specific converter has design issues at the 44.1 rate.

(Massive used the term "broken"; I agree with the thrust of his point that it's a converter design/operation issue and not an issue of sample rate itself, I just think that can be a bit misleading to some to use "broken" as the operative description, as the converter may be operating without error according to spec but the problem is that the design/spec itself is wanting.)

2.) if you are working with audio for digital video, which typically operates at 48kHz, you might just want to do everything at 48khz instead of 44.1khz. But that still does not justify 96k outside of #1, above.

Which reminds me...
Tim O'Brien said:
Or if you're working with equipment that is locked on 48k (Soundblasters, ADATs, etc).
Click to expand...
I have never seen either a SB card or an ADAT machine that was "locked" on 48kHz. Either they are 44.1 only or they are user switchable between the two rates...unless maybe you know of some exceptions I have not seen?

Now, if one has an ADAT tape that was originally recorded by someone else at 48kHz, I could see the argument for leaving everything at 48k when you transfer that stuff to a DAW, and not do the downsample until the end.

G.
 
I've read about the differences about this for over 5 years and could copy and paste 96 gigs of text on the subject but it will just waste memory and processing power as does tracking above your target rate for cd's of 44.1k or for video 48k IMHO. Personally I hear no differences tracking at higher rates.
 
  • Like
Reactions: apl
yes, indeed....
So Glen, I guess that the actual 'down-sampling' process, be it by software or hardware, is an extremely important issue to consider, and not necessarily just the rate at which we record?....maybe they go hand-in-hand?
 
On the digital signal theory side, there is no advantage to tracking at rates different than your final product.
 
gee....I hope I get my 1000 posts up on this forum!!....cheers!
 
superspit said:
yes, indeed....
So Glen, I guess that the actual 'down-sampling' process, be it by software or hardware, is an extremely important issue to consider, and not necessarily just the rate at which we record?....maybe they go hand-in-hand?
Click to expand...
Well, IMHO, I'd hesitate to use the phrase "extremely important" in anything involving this issue; in the grand scheme of things I personally feel that this whole sample rate issue is towards the bottom of a long, multi-page list of things that really matter when it comes to making a good-sounding recording. This is a huge mountain made out of a tiny molehill of a variable, IMHO.

That said, when it comes to gear, everything comes down to signal chain, and every link in that chain has to be considered in the overall equation. If link A increases quality by 10% (I'm using very generic terms and numbers here just for illustration, don't take them as factual values) but link B decreases quality by 10%, it doesn't make a whole lot of sense to have both of those links in the chain (unless there's a desired color change involved.)

So, yeah, in that respect, SRC quality is important, yes. *If* one happens to have a converter that does indeed seem to sound better at a higher sample rate for whatever reason, but the down conversion for whatever reason largely negates that advantage, then it's just one step forward and one step back, and one has wasted their time/energy/resources for nothing.

Funny how Carl Perkins, Frank Sinatra, Elvis Presley, Bruce Springsteen or U2 never seemed to lose any sleep over any of this stuff though... :D

G.
 
  • Like
Reactions: apl
SouthSIDE Glen said:
Well, IMHO, I'd hesitate to use the phrase "extremely important" in anything involving this issue; in the grand scheme of things I personally feel that this whole sample rate issue is towards the bottom of a long, multi-page list of things that really matter when it comes to making a good-sounding recording. This is a huge mountain made out of a tiny molehill of a variable, IMHO.

That said, when it comes to gear, everything comes down to signal chain, and every link in that chain has to be considered in the overall equation. If link A increases quality by 10% (I'm using very generic terms and numbers here just for illustration, don't take them as factual values) but link B decreases quality by 10%, it doesn't make a whole lot of sense to have both of those links in the chain (unless there's a desired color change involved.)

So, yeah, in that respect, SRC quality is important, yes. *If* one happens to have a converter that does indeed seem to sound better at a higher sample rate for whatever reason, but the down conversion for whatever reason largely negates that advantage, then it's just one step forward and one step back, and one has wasted their time/energy/resources for nothing.

Funny how Carl Perkins, Frank Sinatra, Elvis Presley, Bruce Springsteen or U2 never seemed to lose any sleep over any of this stuff though... :D

G.
Click to expand...

and there is my 1K post!.....thanx mate!:)

As a performer, I shan't be losing any sleep, but as a sound-recordist, mmmmmm...I just might over this stuff!!! ...lol!

Cheers....
 
  • Like
Reactions: apl
SouthSIDE Glen said:
Which reminds me...I have never seen either a SB card or an ADAT machine that was "locked" on 48kHz. Either they are 44.1 only or they are user switchable between the two rates...unless maybe you know of some exceptions I have not seen?
G.
Click to expand...
The older SB cards were locked at 48k. You could record at 44.1k, but it was downsampling on the fly. That coupled with the fact that it only had 16 bit converters, when they advertised 24 bit were a couple of main reasons
that these cards sucked for recording use. (it allowed you to record and play back 24 bit files, but the converters were 16 bit)
 
BTW, the Metallica 'black album' was mastered from a 44.1k 16 bit DAT tape. It's sold 16 million records so far...
 
You must log in or register to reply here.
Back
Top