Hi,
I have bought an M-Audio Delta44 to use with Cubase to increase recording clarity/headroom etc. I assumed the best setting then was to record at 96k/24 bit as per the spec of the card.
To maximise quality in clean tracks and resolution of reverb plugins (which still sound muddy) should I be using 32bit float?
thanks
Mark.
Comments
Hi Mark, from what you write I expect you to be more or
Hi Mark,
from what you write I expect you to be more or less a beginner in the audio world. I will write this from that point. As time goes and you learn more and find out about these things you will use your own ears and experience to change things.
So my suggestion is that you use 44.1kHz and 24 bit. There is rarely any really big difference going above 44.1kHz as sampling frequency. As for 24 bits of resolution I find that it saves a lot of trouble, not necessarily that it sounds very different from 16bits. The trouble you save is not having to be quite as exact with setting up levels.
Inside Cubase (at least the modern versions) are all 32 bit, but there is not much to think about there as that is done inside the program.
From the point of 44.1 / 24 you can start on making your recordings better. Positioning the mic and sound proofing your room is one of the things to work on.
Gunnar
I am an old-timer and record everything in 16-bit, 44.1kHz. 24
I am an old-timer and record everything in 16-bit, 44.1kHz. 24 and 32-bit is fine if you think eating 2 quarter pounders is better than eating 1 quarter pounder. Bit levels will not brighten the sound. It only means that you have more dots in the connect the dots picture. 32-bits means greater resolution and that's all. Good technique and equalization in your mix can brighten the sound, besides, CDs are still 16-bit 44.1kHz.
RemyRAD wrote: I am an old-timer and record everything in 16-bit
RemyRAD wrote: I am an old-timer and record everything in 16-bit, 44.1kHz. 24 and 32-bit is fine if you think eating 2 quarter pounders is better than eating 1 quarter pounder. Bit levels will not brighten the sound. It only means that you have more dots in the connect the dots picture. 32-bits means greater resolution and that's all. Good technique and equalization in your mix can brighten the sound, besides, CDs are still 16-bit 44.1kHz.
?? :-? ??
More dots? Huh?
Now, if we were talking sampling rates - maybe (but even still, it's not more dots, just more frequencies to plot with dots).
As for 24 bit or 32 bit, it simply means more volume resolution and headroom. When effects (specifically computer-based effects) work their voodoo on your audio, they perform mathematical computations. Considering the fact that these computations often take the numbers representing your digital music and lengthen them (multiply a number with lots of decimal places by another number with lots of decimal places) and you now see that the volume/signal has changed a tad (this is VERY oversimplified.)
Do this a few times, even at 24 or 32 bit and you'll start to hear these changes. Do it at 16 bit and you'll definitely hear these effects. Think about it, how do you remain at 16 bit when the number you are now working with doesn't fit into a 16 bit string???
Yep, you guessed it - truncate it. (Or round it off and cut it off).
BTW - the argument
CDs are still 16-bit 44.1kHz.
holds no value here.
With many of the dithering algorithms currently available, there are many CDs which have noise signals on par with 20 bit or even greater bit depths.
8-) J.
3dchris wrote: If your converters are 24 bit then setting nuendo
3dchris wrote: If your converters are 24 bit then setting nuendo/cubase to 32bit float is just a waste of space with no sonic adwantage. Please understand that this IS NOT an INTERNAL BIT RESOLUTION setting. It's a recording setting.
just my 2 cents
chris
Umm... Nope.
You're right that you're not recording at 32 bit from the converters, so therefore, you're recording 8 '0's at the end of the bit stream. (At least that's what I think you meant...)
But, what happens when you do anything to manipulate the sound? Ahh, that's where those 8 remaining bits come in. If you make a volume change, compression, or pretty much ANY thing else, you now perform a computation on your bit stream. The stream now becomes longer. 32 bit float allows you to take it all the way out to the point of insanity. It makes dealing with tons of effects a lot easier on the ears and the music.
J.
from the cubase manual: About the record format When you record
from the cubase manual:
About the record format
When you record with effects you should consider setting the record
format (bit depth) to 32 Bit Float. This is done in the Project Setup dialog
on the Project menu. Note that this isn’t required in any way –
you can record with effects in 24 or 16 Bit format should you so like.
However, there are two advantages to 32 Bit Float format:
• With 32 Bit Float recording you don’t risk clipping (digital distortion)
in the recorded files.
This can of course be avoided with 24 or 16 Bit recording as well, but requires more
care with the levels.
• Cubase SX processes audio internally in 32 Bit Float format – recording
in the same format means the audio quality will be kept absolutely
pristine.
The reason is that the effect processing in the input channel (as well as any level or EQ
settings you make there) is done in 32 Bit Float format. If you record at 16 or 24 Bit,
the audio will be converted to this lower resolution when it’s written to file – with possible
signal degradation as a result.
Note also that it doesn’t matter at which actual resolution your audio
hardware works. Even if the signal from the audio hardware is in 16 Bit
resolution, the signal will be 32 Bit Float after the effects are added in
the input channel.
^^ I bet the majority of the people dont take the time to actual
^^ I bet the majority of the people dont take the time to actually read any manuals.
That is one of the first things I do when I get something, is read the manual briefly so I have an idea of what the f*ck I am looking at! :)
P.S. This applies to putting together furniture or any other thing that requires some assembly.
RemyRAD wrote: Good performers and good recording techniques alw
RemyRAD wrote: Good performers and good recording techniques always sound like they are 32-bit! With bad performances and bad technique it all sounds like crap! So what are you beginners talking about? It's not the gear, it's the people. The equipment is truly secondary.
uhh, excuse me...
First, it's interesting that you call us beginners when you've provided absolutely no information about who the hell you are!
second - your post is proposterous. true, a good recording and good musicians and agood engineer should sound good in almost any case. but, there are factors - such as digital resolution, which have a drect impact on the sound.
a native 16 bit recording that stays 16 bit from cradle to grave is easily identifiable.
j.
who the hell am I? Dear Jeremy et al., I'm sorry that you feel
who the hell am I?
Dear Jeremy et al.,
I'm sorry that you feel that some of my answers are preposterous. They're not. I guess you think that your 10-year tenure outweighs my over 35 year tenure professionally just because you had some money to buy some nice esoteric equipment? Just "who the hell am I and what are my
qualifications?" Really.... I'll try to be brief, it's difficult to be with as much experience as I have, so I won't be brief.
It appears that in this forum there are many newbies that really don't know what's going on. I am not trying to show you up, our moderator, just trying to help further competent, practical, perhaps even inexpensive, engineering. I realize that some simplified answers are necessary here (hear) for our beginners in this forum, in our audio community. Equipment junk is not the answers. I have done maintenance and construction from Fort Lauderdale, to Wally Cleaver's in Fredericksburg to Media Sound in New York City, which also includes Paul Wolff, from API/ToneLux to Bruce Kane at Sterling Productions Ltd.. I have been known to have an exceptional pair of H-EARS. I have done numerous recordings through the years of orchestral/choral music all over town and at the Kennedy Center, to operatic recording with the New Zealand symphony orchestra and Alessandra Marc (all with that low end equipment like API).
I grew up in Motown as a little kid, quite literally. How did I do that? My father used to play violin string tracks for Motown at United sound systems. He would take me down there from the time I was seven years of age. I knew then I would be working in this business the rest of my life.
He was a fairly accomplished violinist having been concertmaster for the New Orleans Symphony, St. Paul Chamber Orchestra, Syracuse New York Symphony and until his death with the Cleveland Symphony Orchestra.
My mother was also no slouch, a former Metropolitan opera star, a voice teacher from Michigan State University to Peabody Conservatory to the Catholic University currently, at 81.
I grew up listening to everything with a passion for all those knobs and dials. I entered the broadcast industry with my third class license, in only four days of study,at age 15. I had built my own radio station production facility, at home, from equipment my father had removed from his advertising agency production studio, at age 13. My console, was a Western Electric 23c, from 1948. My recorder was a Magnacord, from the 1950s. That was my first production facility I built without even knowing how to read schematics well.
At age 15, my parents divorced and we moved to Baltimore. I almost immediately went to work for WBJC as an on the air operator, 2 days a week, for jazz extravaganza with Vernon Welch and the Thursday night Opera theater with Duke Baugh. I also was able to take over the audio responsibilities of the Maryland ballet company from, Thomas Bray, my mentor and the father of vertical hold as we know it (from Bell Labs). I even designed and built a custom console for my Pikesville high school radio station from an old, salvaged, Presto tape recorder that had a 3 microphone input mixer, that I expanded to 5 inputs, with cueing, having removed all of the preemphasis and demphasis equalization circuits for a flat response. You know, all of this stuff had those nasty sounding vacuum tube things. By the time I was 17, I was a high school dropout but was in the business working for Flite 3 recordings in Baltimore. Craig Kenney, vice president, offered me a job as chief engineer but I told him that that position had too much responsibility for a 17-year-old. So I turned on my mentor to the job, who'd been let go from University of Maryland
after 23 years. I ended up being a production and young music engineer there. By the time I was 19, I ended up as a disc jockey for the number one progressive rocker in Baltimore, WKTK. By the time I was 21, I went to work as a production, on air and maintenance engineer for WFBR/WBKZ. That didn't last long as I got a phone call from my best high school friend, his father in the industrial film business, about making that recording studio we had always dreamed about in their new warehouse, Hallmark films and recordings. We didn't have much money but we did buy a brand-new MM 1200, 16 track off of the AES floor in New York. We
couldn't afford an API, like Flite3 had, so I custom designed and built a console, built around Op-Amp Labs modules. It sounded better than the factory built units as it was around all of my own gain staging. This, all in 1978, before you were born, remember? After that, I found myself designing building, maintaining and operating a studio for an international, syndicated advertising agency. Soon after that, I was awarded with a position, as a maintenance and design engineer, which also included on
air operations with the NBC radio and television stations in Washington DC where I spent over 20 years. In 1990, halfway through my over 20 years there, I built a mobile production studio the Audio Oasis through my smaller company Reel People recording services and consultants. It included an Ampex MM 1200 that had been left for dead as a parts machine. After procuring 8 more channels of electronics, building a new wiring harness for the transport (the schematic had numerous typographical errors that of course kept the machine from running properly. Ampex thought this was quite funny when they told me " you caught the typos!" This machine was quite a challenge to repair). My console was a used Sphere eclipse C. 40 mostly all discrete electronics (the important stuff like microphone preamplifiers, summing amplifiers
and output amplifies). I got real lucky, when NBC decided to get rid of their 2 custom, completely discrete (not a personality trait), Neve Consoles, beginning in 1996. These consoles, were so terribly intermittent in their pots and switches, that squirting stuff in them wasn't going to cut it. I believe I helped pioneer the use of sonic cleaners to bring these modules back to like new condition and they work reliably to this day, in this terrible environment called, a remote truck. This also included a great catch of vintage equipment like, LA 3A's, 1176s, nasty Neumann U67s, 87s, KM56s, 86s, etc. etc.. And the hits go on!
Why am I wasting my time here?? I am at home, recovering from serious brain surgery (not from rock-and-roll recording, sexual reassignment surgery or, my motorcycling, but it could have been).
I am a hopelessly practical engineer trying to help some of you newbies, with lots of passion and small budgets. I'm generally not impressed by recording school or university graduates. George Massenburg was a Hopkins dropout . You can read off numbers and specifications to me but they mean absolutely nothing.
Through all of this I have garnered, Grammy, Emmy and Soul Train music awards nominations. (I would've never even known about my Soul Train nomination, if I hadn't turned on the television, after coming home from from work at NBC and going into the kitchen to get something to eat. When I heard my live recording for
Yolanda Adams announced. I honestly thought somebody was in my house playing a joke on me! No joke! Incredible timing!)
I make beautiful recordings on broken 20-year-old, noisy, $300 Peavy PA boards, 'cause it's not what you got, it's what to do with it. Anymore questions?
Ms. Remy Ann David
703-532 -REEL (7335)
if anybody would like to call me to talk, you are certainly welcome to.
Well, Remy, I'm glad you think you have me figured out by visiti
Well, Remy, I'm glad you think you have me figured out by visiting my website.
First let me correct some of your errors:
1. I was born prior to 1978
2. I personally have been recording for over 20 years.
3. Wally Cleaver's place is by no means a great studio. I wouldn't even rank it as nice.
4. If you spent as much time reviewing my previous posts as you did surfing my web-page - you would see that I am the champion of the "little guy " on this site. I'm constantly the advocate for "enough equipment to get the job done "
5. you profess to advocate cheaper gear but you constantly Name-drop (APl, Neumann, etc.)
6. I didn't stumble across money and buy esoteric gear. I built my business, dollar for dollar from the ground up. If I were a newbie, I must be doing something right.
7. Everyone in this area has recorded at the Kennedy Center.
Now, before you come in here and trash people and then dispense bad advice, I would advise you to do your homework.
First: Modern CDs, though 16 bit, possess noise floors on par with 20 bit or greater recordings.
Second: If you are going to be performing any digital edits or effects (and, since this guy's using Cubase, that's a safe assumption) you are better off recording at higher bit rates. (It's called quantization. Somewhere in your 35 years of experience you must have heard of it.)
BTW... you'll notice that, to do higher bit rate recordings, this dude doesn't have to spend one more friggin cent.
So your argument about us ''Newbie" ''gear heads" recommending new toys to buy is just wrong.
I don't come in to your house and crap on your floor. Please don't do it in mine.
Jeremy Cucco
(And YOU may feel free to call me.
540-429-2335)
oh - BTW - Remy: 1st - so you understand that I'm not THAT much
oh - BTW - Remy:
1st - so you understand that I'm not THAT much of a jackass - i did miss your first several posts. So please accept my welcome. We don't have too many ladies here on the boards.
2nd - I think we can get along now that our pissing contest is over. My only beef was that you had a low post count, didn't introduce yourself and then referred to a few of the seasoned guys here (including myself) as newbies. this is considered very bad etiquette on the boards.
Your input is welcome here ( as is everyone's except jp22) but we do try to respect eachother.
Welcome...
Jeremy
(Edited to fix typos while posting from my PDA...)
You're right that you're not recording at 32 bit from the conver
You're right that you're not recording at 32 bit from the converters, so therefore, you're recording 8 '0's at the end of the bit stream. (At least that's what I think you meant...)
But, what happens when you do anything to manipulate the sound? Ahh, that's where those 8 remaining bits come in. If you make a volume change, compression, or pretty much ANY thing else, you now perform a computation on your bit stream. The stream now becomes longer. 32 bit float allows you to take it all the way out to the point of insanity. It makes dealing with tons of effects a lot easier on the ears and the music.
Cucco, with all due respect once again you're confusing 32 bit WAVE with 32 bit INTERNAL SOFTWARE OPERATION. They are not the same thing. Wave does not CHANGE or degrade when you make any eq, compression etc. in the software. Software resolution is FIXED at 32 bit regardless of the wave bit lenght recorded and it's the SOFTWARE OPERATION resolution that makes a difference.
just my 2 cents
chris
let's clear this thing up a bit.... 32bit is useless UNLESS you
let's clear this thing up a bit....
32bit is useless UNLESS you want to bounce tracks internally and capture the full resolution of the signal represented in the (32bit) mixer... the mixer will ALWAYS run at 32bits....
it's that simple....
and guys.... be nice.... 8-)
3dchris wrote: You're right that you're not recording at 32 bit
3dchris wrote:
You're right that you're not recording at 32 bit from the converters, so therefore, you're recording 8 '0's at the end of the bit stream. (At least that's what I think you meant...)But, what happens when you do anything to manipulate the sound? Ahh, that's where those 8 remaining bits come in. If you make a volume change, compression, or pretty much ANY thing else, you now perform a computation on your bit stream. The stream now becomes longer. 32 bit float allows you to take it all the way out to the point of insanity. It makes dealing with tons of effects a lot easier on the ears and the music.
Cucco, with all due respect once again you're confusing 32 bit WAVE with 32 bit INTERNAL SOFTWARE OPERATION. They are not the same thing. Wave does not CHANGE or degrade when you make any eq, compression etc. in the software. Software resolution is FIXED at 32 bit regardless of the wave bit lenght recorded and it's the SOFTWARE OPERATION resolution that makes a difference.
just my 2 cents
chris
No, not true.
Yes, if the software is a 32 bit engine, this is the case. The software/mixer will operate at 32 bits.
However, if you record a file IN 32 bit (float or fixed) you ARE in fact recording a 32 bit wave file. However, since most converters out there aren't recording past 24 bit, the last 8 bits are not used. That is of course, until you process the WAV file.
If you were correct, then answer me this one VERY important question.
If I'm NOT in fact capturing 32 bits of information in a 32 bit WAV file, why then is it twice the size (exactly) of a 16 bit WAV file? Why? Cuz there ARE 32 bits there and you ARE using them.
This is basic math and science folks.
Yes, your mixer hopefully runs at 32 bits, regardless of your recording resolution.
But, when you record a file at 32 bit, you ARE in fact recording a 32 bit wave file.
I can't make this anymore clear.
And Iznogood - I'm trying to be nice - 8-)
J 8)
If you were correct, then answer me this one VERY important ques
If you were correct, then answer me this one VERY important question.
If I'm NOT in fact capturing 32 bits of information in a 32 bit WAV file, why then is it twice the size (exactly) of a 16 bit WAV file? Why? Cuz there ARE 32 bits there and you ARE using them.
This is basic math and science folks.
Yes, your mixer hopefully runs at 32 bits, regardless of your recording resolution.
But, when you record a file at 32 bit, you ARE in fact recording a 32 bit wave file.
I can't make this anymore clear.
Cucco, I don't think I was clear enough in my previous post. I did not say that the wave file we record IS NOT 32 bit. Of course IT IS 32 bit. My point was totally different. What I was saying is that IT DOES NOT MATTER if the wave is 32 bit or 24 bit. It won't make any difference in the sound we're getting regardless of how many operations we do in the software (eqs, fx etc.). The only difference according to cubase manual is that recording in 32bit gives a bit more headroom. That's all to it. So if you don't care about the HD space then by all means us 32bit (as I do) but let's make it clear: 8 extra bits are totally WASTED and it won't make a difference during mixing.
thx,
chris
Chris - I agree we were talking different things, but, if the C
Chris -
I agree we were talking different things, but, if the Cubase manual states that, it is incorrect.
simply put - when you record a 32 bit wav, you are originally capturing 24 actual bits and 8 zeros. But, compression/eq/etc are mathematical computations. What happens when you multiply 2 numbers together? The numbers physically get bigger (more numbers). Well, if you stay in 24 bit, you must round these new numbers off. This is called a "rounding error" and is a real concern.
If, however, you record in 32 bit, these computations often keep the number within the possibilty of a 32 bit representation. If they go beyond that capability, the rounding error will be far less aggregious.
The important part is to stay 32 bit until final mixdown/dither.
Does this help clear it up a bit?
j... 8-)
When I recorded with Cubase VST 32 I always used the "True Tape"
When I recorded with Cubase VST 32 I always used the "True Tape" 32 option.
It sounded better. I had the best luck running 24 bits at 48K (since my motu could only go that high).
I still can't justify taking my studio to 96K yet... one of these days I'm sure. Some of my gear can sample at 24/192. If I had the money I'd record their all the time. Wait no, if I had the money I'd have a Studer 827 gold edition 2" 24 with an optional 8track "ultimate analog" head stack.
:D
Cucco, I agree with you on the rounding thing but what I'm tryin
Cucco,
I agree with you on the rounding thing but what I'm trying to say it's not the recorded 32bit wave itself that is being truncated. It is the SUM of all the waves. In other words the more eqs, fx etc you use the FINAL wave is gonna lose it's definition while the original waves (either 23 or 32bit) will remain the same. The question here is: will 32bit waves allow the mixdown to sound better? Maybe yes and maybe not. I never tested this. Did you?
chris
3dchris wrote: Cucco, I agree with you on the rounding thing but
3dchris wrote: Cucco,
I agree with you on the rounding thing but what I'm trying to say it's not the recorded 32bit wave itself that is being truncated. It is the SUM of all the waves. In other words the more eqs, fx etc you use the FINAL wave is gonna lose it's definition while the original waves (either 23 or 32bit) will remain the same. The question here is: will 32bit waves allow the mixdown to sound better? Maybe yes and maybe not. I never tested this. Did you?chris
Okay, well, the elements of your logic are in fact correct.
If you record a wave at 24 bit - from beginning to end, it should not change in quality or sound until bused down.
Here's the issues with that though -
1. If you only do "real-time" effects, (ie, monitor the effects but don't actually apply them until final mix-down) then you're fine. However, if you use any effects that are off-line, you have now caused a truncation and rounding-error.
2. If you internally bounce prior to the final bounce, you have now created a rounding error.
3. If your program doesn't actually use all 32 bits UNLESS your waves are 32 bits, then you're screwed. (In other words, if you record all in 24 bit, but then say you want to bounce down to a 32 bit file, some DAWs - not most newer versions though - will still only work within the 24 bit constraints that you've set up with your project to begin with.
I find the above 3 scenarios to be too limiting. Therefore, I record 32 bit. It doesn't take up that much more HD space or processing power, so in the end, I feel as though it's worth it.
J. 8-)
3dchris wrote: Cucco, Yes You are correct on this, however the q
3dchris wrote: Cucco,
Yes You are correct on this, however the question is... isn't a CD just 16 bit anyway? :)hehe... :)
chris
Well, yes. But what tells us the depth of a CD? It's noise signature? Well, in that case, most dithering algorithms (or I should say, some good ones, such as UV-22 or POW-R) produce CDs with signal to noise ratios on par with 20 bit systems.
So I would say, no. CDs are easily 20 bit mediums at todays technology.
(Oh, and Iznogood is absolutely right. Send your ME the highest res file possible. Let them do the dither down to 16 anyway...)
J. 8-)
Hmm. I think I have to chime in here as there seems to be a bi
Hmm.
I think I have to chime in here as there seems to be a bit of misunderstanding going on. Dithering reduces the SN figure, but moves the noise to places where the ear is less offended.
So a few statements here that I know to be correct, if you want to I can delve deeper into the issues. And please forgive me if I sound like the typical professor here, I find it difficult to write short in a foreign language, Swedish beeing my native tongue.
1) First, I tend to consider 16 and 24 bit digital audio as represented as a signal between -1 to +1. A full scale signal is exactly that, from -1 to +1. The level there is called 0dB FS. In this representation there is absolutely no headroom above odB FS. You get clipping without warning there. Painful.
2) Bit depth says how detailed the information is. The best measure to work with is generally the SN figure. SN is short for Signal to Noise ratio. The SN signal tells us how far from the maximum signal the noise floor can be. Mathematically this works out as 6.02 x number of bits + 1.76 (all in db). So, truncating we get about 6x16 = 96dB SN from a 16 bit signal and 6x24 = 144 dB from a 24 bit signal.
3) Now 32 bit float is something different. There is still 24 bits reserved for the signal so we get a SN of 144dB (this is a simplification). But it also allows you to represent signals that are much larger than -1 to +1. In decibels this actually turns out at about +700dBFS. This enormous headroom without loosing SN is what makes us lazy people love float.
3) If you look at real world converters and mics and such though, very few has more than 120dB of SN. In fact most mics may end up somewhere around 80dB and few sound card has more than about 110dB of dynamic range and hence cannot have larger SN. This is more like 20 bits, the rest of the bits are simply random noise. So basically when the people sell you 24 bit converters they know they are trying to fool you. There are even a few so called 24 bit systems that in reality does not even deliver 16 bits. My testing of the Edirol R1 points to it having a true SN ratio corresponding to about 15 bits. Very useful still, but far from the stated 24 bits. Now marketing is not really lying to you as you really get 24 bits, but they are not telling the full truth either as several of the bits are simply random noise.
4) If you dither a signal, what you do is adding noise. Pleasing to the ear noise, but still. This reduces the SN. Typical triangular dither changes the equation from 6.02 x bits + 1.76dB to around 6.02 x bits - 3dB. In effect dithering eats about 4.8 dB of the SN.
5) any time you move from one representation to another, say from 32 bit float to 24 bit audio you have to handle the signal some way. This handling can create problems. For example, there is no way to directly represent a large float signal, say -2 to +2 on the limited range of -1 to +1 of a 16 or 24 bit signal. There will be clipping there. And the conversion process may add a few bits of noise. If you are running 24 bits, the added noise is far from the signal so you will not really hear it.
Anyway, a bit of food for the thinking man.
Gunnar
Excellent post Gunnar!!! One small edit or note though.. 3) I
Excellent post Gunnar!!!
One small edit or note though..
3) If you look at real world converters and mics and such though, very few has more than 120dB of SN. In fact most mics may end up somewhere around 80dB and few sound card has more than about 110dB of dynamic range and hence cannot have larger SN. This is more like 20 bits, the rest of the bits are simply random noise. So basically when the people sell you 24 bit converters they know they are trying to fool you.
This is not due to the limitations of the A/D chipset - it is rather due to the limitations of electrical circuits in general, which inherntly have many limitations.
A 24 bit converter can be a true 24 bit converter in that the bits that are there are quantifiably 24 of them per sample. However, some of them are noise, based on the analog circuitry. This is why many converters sound SSOOO drastically different. Otherwise, converting audio into numbers is a VERY simple (and accurate) process.
J.
I opened a can of worms here didn't I? I guess I should've aske
I opened a can of worms here didn't I?
I guess I should've asked if I have 96k/24 capability, is there any reason NOT to use it :)
Anyways, as it doesn't matter about the quality of the gear, but the performers etc, I'm selling my DAW and digging out the Fostex 4-track and couple of cassettes. :wink:
thanks for your input
Mark.
markgo wrote: I opened a can of worms here didn't I? I guess I
markgo wrote: I opened a can of worms here didn't I?
I guess I should've asked if I have 96k/24 capability, is there any reason NOT to use it :)
24/96 is good to have but you dont have to use it. Well I would use 24bit but not at a 96k sampling rate. Unless your project is going to SACD or DVD dont bother wasting drive space recording in 96k. Plus the higher the sample rate the more work your CPU has to do during recording and playback, and if you system isnt Hercules you can run into some problems. But it is good to have that type of capability incase you ever need to use it.
As it happens I am having HD problems. Gonna post a seperate th
As it happens I am having HD problems. Gonna post a seperate thread.
Disk was handling the load fine, ATA100 Seagate, AMD3200+ 1GB RAM M-Audio 44. Running 10 tracks.
But it died last week, power seems to shut down to it when opening several files, gone all of a sudden, no drop outs beforehand.
Took the opportunity to buy a SATA disk, but getting crackles from the disk buffer.
Hi Mark, from what you write I expect you to be more or
Hi Mark,
from what you write I expect you to be more or less a beginner in the audio world. I will write this from that point. As time goes and you learn more and find out about these things you will use your own ears and experience to change things.
So my suggestion is that you use 44.1kHz and 24 bit. There is rarely any really big difference going above 44.1kHz as sampling frequency. As for 24 bits of resolution I find that it saves a lot of trouble, not necessarily that it sounds very different from 16bits. The trouble you save is not having to be quite as exact with setting up levels.
Inside Cubase (at least the modern versions) are all 32 bit, but there is not much to think about there as that is done inside the program.
From the point of 44.1 / 24 you can start on making your recordings better. Positioning the mic and sound proofing your room is one of the things to work on.
Gunnar