HELP! A QUESTION CONCERNING Htz.

M

Maurice3000

New member
Seems like I’ve stumbled across this forum at the right moment … After being a strictly analogue user, I’ve finally taken the plunge - I’ve just bought my 1st digital 4-track (Fostex FD4). Although it’s ease of use is highly surprising to me, the one thing that is bugging me is the choice of recording modes:
>NORMAL MODE “32khz” “DAC (what's this?)”
>MASTERING MODE “44.1khz” “16-bit”
Primarily, I’ll be recording (mike into desk) steel strung & classical guitars. To get as natural a sound as possible, I was wondering what the best choice would be (?). I’ve recorded some tracks using the 32khz mode (sounds pretty good) but now I’m wondering if things would sound better in 44.1. Am I correct in thinking 32khz would have a warmer sound than 44.1 (more like analogue), or is it the other way round? I’ll be using the additional tracks & bouncing a few times, so what mode would also be more suitable for this?
As I’m in the middle of recording a few things at the mo, I don’t really want to 'experiment' with reformatting yet, but I’m willing to sacrifice these tracks if it means a better tone could be on the way. Although I’ve been a musician for many years, I’m a complete ignoramus when it comes to modern technology. It’s about time I learnt a few things, so I’d greatly appreciate a few pointers from anybody. Cheers.
 
44 KHz 16-Bit is the more common format for CD quality sound. If you unit supports that, I never use much else. That way if you get a good track early you can still use it without change in the final.

Many recorders support lower recording formats to save space. The higher the rate, the more space per second that is needed. Basically a quantity/quality decision.

Ed
 
>NORMAL MODE “32khz” “DAC (what's this?)”

It indicates how many samples per second your your machine is taking in the Analog to Digital conversion. The higher the better (sounding) but as Ed pointed out, the more samples per second, the more disk space consumed. The up side is that a higher sample rate is abetter representation of the analog signal since the gaps between samples are not recorded.

To address your question about warmth, if the signal going in is "warm" sounding, then in order to retain that warmth, a higher sample rate will sound beter then a low.

CD's are written at 44.1KHZ, 16bit so you should at least record at that. Otherwise, the machine will need to convert the sample rate up to 44.1 to burn the CD - this can introduce errors in the digital stream and can degrade the sound.

As you can probably see from any magazine ad's these days, 192KHZ is becoming the new sample rate standard though 96KHZ is more common.

Analog is effectivly an infinite sample rate since it does not exclude any information from the input signal. The question is; at what digital sample rate is it not noticable anymore?

Kevin.
 
Ed & Kevin … Cheers for the info! I’ll reformat as soon as I can.

Looks like I was looking in the wrong direction when it came to reproducing the ‘reality’ of the analogue signal. Although I now understand that the ‘higher’ the sample rate, the closer you get to the (effectively) ‘infinite’ sample rate of analogue, this has presented another question for me:

Why record at 96 or 192khz, if the final format (CD) is presented in 44.1khz.

Wouldn’t this be like an oil painter producing his work using 100 colours when his main source of presentation is a 25 colour format?

(Please excuse my ignorance of these matters).
 
Why record at 96 or 192khz, if the final format (CD) is presented in 44.1khz.
Click to expand...

Once the signal is converted to digital, all processing is done in the digital domain... eq, compression, reverb etc... It is sonically better to do the procesing on as much information as possible so that the processed signal is again as close a representation as you can get.

If you record at a high rate, and plan on getting the song mastered, it would be best to send the high rate info to the mastering house for processing so that they can work with that data and then do the rate conversion rather then the other way round.

Cheers
Kevin.
 
longsoughtfor said:
[B...Analog is effectivly an infinite sample rate since it does not exclude any information from the input signal. The question is; at what digital sample rate is it not noticable anymore?

Kevin. [/B]
Click to expand...

Isn't this a bit misleading? By some measures digital can have more accuracy than analog. We don't wan't to give the impression that a 96k record rate is way up on the priority list do we?
Wayne
 
Isn't this a bit misleading? By some measures digital can have more accuracy than analog. We don't wan't to give the impression that a 96k record rate is way up on the priority list do we?
Click to expand...

I don't think it is mis-leading. It was a rehtorical question really... maybe 48K is the best sample rate to use I'm not sure.

I do know that the main difference in music capture between analog and digital still remains that in digital you take a snapshot of the signal every so often and sometimes make assumtions about what happened when you weren't looking so it makes sense to shorten the time between snapshots while with an analog tape deck, you capture everything that goes in.

The original question was; should he use 32K or 44.1K to record with. I was trying to explain why a higher rate was prefered.

Kevin.
 
Very valid points.

Longsoughtfor has very valid points and I agree that recording at a higher bit rate is best. Try to think of digital recording like Longsoughtfor said "taking pictures" of the audio. When the computer is taking pictures of the audio coming into the recorder it is best to have as many pictures as possible to represent the original signal as close as possible. Keep in mind that when recording digital the missed information that is not captured due to recording a low sample rates is filled in (guessed at) by the computer using the information that has been captured already.

So in my opinion the more bits the better! The only thing that I do question is how wide the frequency spectrum should be in any given recording. Human ears can only hear 20hz to 20khz but yet most if not all "pro" recordings are done at 96khz or at even 192khz! Hmmm, sounds a bit useless to me.

sonicpaint :D
 
sonicpaint – thanks very much for more interesting info!
----------------------------------------------------------------------------------
Unfortunately, the more I’m learning, the more (slight) confusion there is for me at the moment – What’s more important, a high frequency or a high bit rate? Does a high frequency give a naturally higher bit rate? What exactly is a bit? etc.

I think I’m right in saying that 96khz records at a rate of 96,000 samples per second, therefore, is (obviously) superior to 44.1. I haven’t heard DVD Audio yet, but taking the “taking pictures” analogy into account, I can fully understand why it should be more desirable on the ear than CD.

As sonicpaint said, we only hear up to around 20khz & this was why I originally thought that using my 32khz mode would be more natural on the ear. Wouldn’t a higher frequency spectrum record & present to us too much of the ‘high-end ring’ of a steel strung acoustic, for example? Wouldn’t this make it sound harsher & more unnatural on the ear compared to an analogue recording, or doesn’t it work like this? Is it a just a case of having to be more careful with the signal you put into the desk when using digital? I’m still very new to the digital world, so I’d be very grateful for some more input. Cheers.
 
Maurice3000 said:
sonicpaint – thanks very much for more interesting info!
----------------------------------------------------------------------------------
Unfortunately, the more I’m learning, the more (slight) confusion there is for me at the moment – What’s more important, a high frequency or a high bit rate? Does a high frequency give a naturally higher bit rate? What exactly is a bit? etc.
Click to expand...

Convert the term 'bit rate' to bit depth. That defines the dynamic range. 16 bit gets you 96db from full scale to converter noise. 24 bit expands the digital noise floor down to around -144db. (That normally exceeds the dynamic range of the analog parts, so you end up with around 100--110db.)

Sample rate defines the high frequency limit, approximately 1/2 of the sample rate. 96 for 48khs, 44 for 20khz.
In theory, the 44/48k rate does 20-20k fine. Increasing the rate is an attempt to get more fidelity.
Ususally if you feed a good 44k system sweets sounds, you get good sweets sound out, you're good to go.
:D
The other thing with the gaps and such, :) all I was saying was that, again, 'in theory', if there are guessing errors, it is by definition on the frequencies out side of the bandwidth.
But I learn as I go.
Peace
Wayne
 
Maurice3000 said:
sonicpaint – thanks very much for more interesting info!
----------------------------------------------------------------------------------
As sonicpaint said, we only hear up to around 20khz & this was why I originally thought that using my 32khz mode would be more natural on the ear. Wouldn’t a higher frequency spectrum record & present to us too much of the ‘high-end ring’ of a steel strung acoustic, for example? Wouldn’t this make it sound harsher & more unnatural on the ear compared to an analogue recording, or doesn’t it work like this? Is it a just a case of having to be more careful with the signal you put into the desk when using digital? I’m still very new to the digital world, so I’d be very grateful for some more input. Cheers.
Click to expand...

As far as making things sound thin and unnatural, you had it right when you guessed that it doesn't work that way.

Let say that your recording a guitar and the natural frequency response is from 150hz to 10khz and your recording at a frequency of at let's say 9khz then your recording would not contain any signal about the 9 khz. (this is just a hypothetical of course) So in fact when recording a lower frequency there is a high chance that the recording could be forced to sound strange do to the lost information during the recording process.

Hope he analogy is easier to understand.

sonicpaint :D
 
And it's not useless to record frequencies above the range of human hearing.
There has been a fair amount of research that indicates that high freq's, (as high as 40,000hz!) affects how we hear the audible freq's and how the equipment reproduces them.
It's kinda like.....having an amp with 1000 watts seems pointless but it's widely known that an amp like that sounds more effortless when it's putting out lower wattaged 'cause it doesn't have to work hard.
 
Wayne, Sonicpaint & Lt. Bob – Thanks very much for the extra information. Things are a lot clearer now you’ve cleared up my confusion with some of the terminology. Feel like I’m part of the 21st century now! CHEERS!!
 
Glad to see everyone help out!

It's nice to see that everyone was able to clear things up for you. That's what's great about this forum, everyone does what they can to share information they've learned.

sonicpaint
 
i was glad to read this thread because since i'm a self-taught producer with mostly a dance/Reason background (read: no real recording or production skills) i'm weak on the fundamentals.

that said, i don't think Maurice3000 needs to be the least bit freaked that all the cool kids are recording at 192khz.

use the fostex at 44.1khz and try to be patient and experiment to get the best results you can.

i'd guess that good performances, quality mics and pre's and thoughtful engineering are all more important than taking your sample rate to the max of available hw.

i wouldn't worry about it until it's time to replace your fostex, and even then i'd only make it one consideration of many.
 
You must log in or register to reply here.
Back
Top