Hello,

The best is record at 24 bits, why?. Because with 16 bit you optain a theorical margin of 96 dB (Usually are worst), if youn don't record a top level (peaks at 0), you lost dinamic range and low level definition. Record with the peaks at 0 is very dangerous (Is easy optain a over).

If you record at 24 bits, you have a theorical margin of 144 dB (I don't know any machine with this S/N), you don't have the needed of use all of your bits for have a good S/N ratio, a more relaxed recording sesion.

Another thing is the edit, if you have a 16 bit file, and you edit this file (EQ, rev, etc.), low level artifacts appears in a audible range. With 24 bits file this artifacts is out audible range.

See this link:

http://www.digido.com/morebits.html

48 kHz is the studio standart, if you want a studio compatibility use 48 kHz.

Bye

P.D. Sorry for my english ;)

Uh, George, 24 bit audio on a Redbook CD? What are you talking about?

(I _KNOW_ you know better than this, so maybe I am just misunderstanding your point here.)


Peter

Hi Peter
We had a client with one of these. I copied this from the Alesis web site.


Think of it as a Complete Mastering House in a Box.


Presenting the new MasterLink ML-9600 two-track hard disc recorder - literally, the state-of-the-art in do-it-yourself mixing and mastering systems.

Only MasterLink lets you capture your mixes in stunning 24-bit, 96kHZ; edit your recording and apply finishing tools - such as high-resolution parametric EQ Compression, limiting and normalizing - all in the same system and then burn the finished recording in either industry-standard Redbook, or new high resolution CD24, fast becoming the new standard for archiving and transferring high-res. audio files to the mastering room.

But don't take our word for it. Visit your local Alesis dealer today and learn why top pros like George Massenburg, Roger Nichols, Elliot Scheiner, Al Schmitt, and Greg Ladanyi have traded in their DAT's for Masterlink. Or better yet, just get off your musically inclined little butt and buy one!

After all, it's only your career we're talking about here.

• Huge internal hard disk recorder...up to five hours of two-track audio capacity
• Burn CDs using Standard "Red Book" (16-bit/44.1kHz) and high resolution formats…up 24-bit, 96kHz
• Choose any combination of digital resolutions (16-, 20-, and 24-bit) and sample rates (44.1, 48, 88.2, and 96kHz) with full AIFF compatibility
• Organize song playlists with total control of fade-ins, fade-outs, track gain, start points, track cropping and more
• Stores 16 different playlists containing up to 99 songs each
• Onboard digital signal processing: compression, EQ, limiting, and normalization
• Uses inexpensive, readily available CD-Rs

I record 24 bit at 44.1 kHz. The audio just plain sound better recorded at 24-bit and it gives more detail to the plug-ins for rendering. Because (for the forseeable future) most audio is going to be on red-book CDs, I keep the sample rate at 44.1 kHz. Lots of people record at 48 kHz for the extra bandwidth, but I don't like having to resample the audio down to 48kHz. Resampling is pretty nasty stuff and you can definitely hear it.

yeah! figure a way to let us burn Masterlink compatiable CD's, and I wont have to buy one :)

As far as I can tell you can easily do this already (though not completely from inside Vegas, admittedly).

Alesis's CD24 "standard" seems to be nothing more than putting 24-bit AIFFs onto a data CD-ROM in the usual way, but with a slight restriction in what filenames are allowed. It seems that filenames are restricted to 8.3 format.

All you should need to do to make a Masterlink compatible CD24 should be to save the audio files as 24-bit AIFFs (easily done from Vegas or Sound Forge) and then burn them to a data CD using your usual software (Roxio EasyCDCreator, Ahead Nero, etc) making sure the names you give them are according to the 8.3 scheme. I haven't been able to test this myself (I don't know anyone with the cash for a Masterlink) but as far as I know it should work fine.

If it doesn't work doing it this simple way then you can use the rather more long-winded method supported by Alesis themselves. Download the free "CD24 ISO Image Builder" software from the Alesis website and use it to create an .ISO file containing the 24-bit AIFFs you produced earlier. Then use your normal CD-burning software to burn a data CD from this .ISO file.

Either way, it shouldn't present much of a problem to exchange files with the Masterlink in its CD24 format.

bgc:

If you record at 24 bit at 44.1 khz would that be about the same as recording at 24 bit at 48 khz? Vegas converts it to 16 bits at 44.1 khz. Aren't you going to have dither artifacts on conversion to Redbook at both 24@44.1 and 24@48?

jpresley

"you can definitly hear it"? 24/48 is pretty much standard for recording at the moment. Also there wouldn't be a 24/96 or even 192 if the re-sampling to 16/44 would introduce artifacts.

-Markus

Mark, thanks a TON for that info!
We are moving towards 24/96, and were trying to figure how we would get that format to the mastering lab. The two that I use both have alesis masterlinks, so we were planning on getting one. I will try this iso builder and see if it works at the lab on wednsday. If it does, I thank you immensely, as itll free up some money to go buy more converters :)

jpresley -
The thing is that during recording and mixing, Vegas doesn't convert the audio to 16-bit, 44.1kHz. If your sound card supports the format you choose for your project it will play back the audio at that sample rate and bit depth you specify for the project.
What Vegas does is convert all of the audio and dsp functions to floating point numbers (high resolution) for internal processing and busses. It only does any conversion at the very last step of the master bus where it has to convert it to the bit depth and sample rate you've specified for the project under properties (to send to the card).
Make sense?
This is different than the settings you choose for rendering. When you render a 16-bit, 44.1 kHz file to burn to a CD it does convert to 16-bit, 44.1kHz.
bgc

Hi Markus -
Really good sample rate conversion can sound decent, it's just that it's like dither and D/A and A/D processes. There's lots of art and science there and when it's done "OK" you can hear it (at least I can). I just prefer to not have to worry about what the sample rate converters are doing to my audio. How many taps in the filters, what type of interpolation, blah, blah, blah. In comparison dithering seems almost simple (though you can definitely hear that too).
bgc

I've been recording at 24bits and then making CD's. Does Vegas automatically convert them to 16/44? I haven't run any dithering...should I be? Would it be better to convert them before making the CD pre-masters? I just assumed it was doing it all automatically when I rendered the file for CD.
Just curious.

VV always outputs (to the soundcard) at the sample rate and bit depth of your project (set in properties). When you render the project to a wav file it uses the properties in set in Template (open a project, select Render as... as look at the dialog box). Click on the custom button and you'll see the attributes of various templates (you can add more).
If you select 16-bit as output Vegas (I'm almost 100% sure) just truncates the bits.
You will need to add dither on the output master bus if you want dither. VV doesn't automatically add it.

48K is the video standard. Stick with 44K1 for audio and leave out the resampling otherwise necessary to go to CD. Tracking in 24 bits is preferable is you system has the 'legs' for it !

geoff

Hello,

No, 48 kHz is the studio standard (Audio studio), since the first DAT machines. Read the AES suggestions about this.

Regards

48k has been on the first DAT machines but only the "pro" machines were able to, or enabled you to, record at 44.1. 48k was considered a bit too tinny... At the time the first ones were made it was also considered overkill because it was above the nyquist frequency of 44.1 was more than enough to reproduce the frequency of sound heard by the human ear (roughly 20,000Hz). You can argue until your eyes and ears bleed that sampling higher than that brings out harmonics and sub-frequencies lost in the conversion, because very low frequencies also have elements above the hearing range which people claim color the sound without our being able to directly hear them. Sympathetic frequencies. It still all comes down to what you think sounds better. Audio cd's are still 44.1 16bit stereo. Processing digital audio at higher bit-rates results in higher resolution of parts of the audio that are very low - fades, reverb tails, etc. The fidelity will be better at 24bit for this kind of processing. You can then dither down to 16bits for cd mastering.

I've only ever been asked for 48k when someone is doing video. That IS the standard for video. Obviously, everything is changing a great deal now and people just assume that more is better. I hesitate on that one.

I'm sure more learned, bookish technical types might correct me on some of this. It's been a while since I've looked at a textbook ! :-)

thanks.

Hello Charlesdem,

>At the time the first ones were made it was also considered overkill because it >was above the nyquist frequency of 44.1 was more than enough to reproduce the >frequency of sound heard by the human ear (roughly 20,000Hz).

I agree with mr.Nyquist.

>You can argue until your eyes and ears bleed that sampling higher than that brings >out harmonics and sub-frequencies lost in the conversion,

What sub-frecuencies?.

I don't know any people that earing more than 20 kHz (I don't ear more than 16 kHz)

>because very low frequencies also have elements above the hearing range which >people claim color the sound without our being able to directly hear them. >Sympathetic frequencies.

I don't agree, this is a oniric world, "color" is a very subjective word.

You can't listen anything over 20 kHz (I can't).

>It still all comes down to what you think sounds better.

"Better" is another subjective word (I can't measure), isn't objective, isn't real.

>I've only ever been asked for 48k when someone is doing video. That IS the >standard for video.

And audio.

>Obviously, everything is changing a great deal now and people just assume that >more is better. I hesitate on that one.

go to infinite -1, is better (More bandwith).

Regards

Regarding:
>>You can argue until your eyes and ears bleed that sampling higher than that brings >>out harmonics and sub-frequencies lost in the conversion,

>What sub-frecuencies?.

>I don't know any people that earing more than 20 kHz (I don't ear more than 16 kHz)

I don't know if this is what he was talking about, but your ear is a non-harmonic "instrument" that actually creates distortion tones - they're very audible - and tones/sounds in the audible range can be created by tones/sounds in the inaudible range. So it seems those analog/LP folks ain't so nutty.

that is what I was talking about, thanks.

There are frequencies outside the audible range that can be felt. No one will argue that a good solid 15Hz rumble can add a lot to a sound track by shaking the floor at appropriate moments. For that matter, my hearing tops out at about 12KHz (i have no middle ear bones, so we're not even sure why i hear at all ...) but a triangle wave at 74KHz makes my spine crawl and my shoulder muscles spasm. All things considered, i guess i'm glad that most sound systems won't reproduce anything this high.

Subharmonic frequencies (also called beat frequencies) are caused when two higher frequencies mix and cycle through cancellation and augmentation. For example, a 55,000Hz tone and a 55,440 tone mixed together will sound like an A (440Hz). These beat frequencies can add to the experience of listening to music, but may not be reproduced accurately by a digital system.

Hello,

You are wrong.

> a 55,000Hz tone and a 55,440 tone mixed together will sound like an A (440Hz).

The diferential tone (Is the correct name), is a defect of the human earing, if you don't listen the two tones (Your example), or one of the tones, you can't make the diferential tone.

The reply is NO, NO 440 Hz tone.

Regards

Ramallo, i'm referring to beat frequencies, which are not dependant on the human ear, defective or otherwise. Beat frequencies are a physical result of the interference of two similar sound waves of different frequencies mixing together and cyclically augmenting and cancelling each other.