Login / Register
 
32-bit vs. 24-bit Recording
New Reply
Subscribe
drumkideric
Thread Starter
#1
30th January 2007
Old 30th January 2007
  #1
Lives for gear
 
drumkideric's Avatar
 
Joined: Apr 2006
Location: Nashville
Posts: 590

Thread Starter
drumkideric is offline
32-bit vs. 24-bit Recording

So I was just wondering today, would recording at 32-bit compared to 24-bit make a big difference at all. What are the technical reasons behind it?
Also, Where do you guys think digital audio is heading as far as bit-rate/sample rate wise? Do you think DSD will take over? Or can it even get that much better?
#2
30th January 2007
Old 30th January 2007
  #2
Lives for gear
 
Nu-tra's Avatar
 
Joined: Oct 2005
Location: Lost Angeles
Posts: 3,771
My Recordings/Credits

Nu-tra is online now
I'm happy with 96k
#3
30th January 2007
Old 30th January 2007
  #3
Lives for gear
 
Joined: Aug 2004
Location: Inside my brain...
Posts: 3,137

Lawrence is offline
There is no real benefit to 32-bit recording since there are no 32-bit converters. There _may_ be a tiny benefit to rendering processed 24-bit files / tracks / mixes as 32-bit float, if you use a 32-bit float native daw, since the end result won't have to be dithered or truncated to 24-bits.

As far as padding a 24-bit file during tracking to make it 32-bits... why? The daw is going to do that anyway as soon as the audio enters it... well in most native daws anyway. It's a waste of disk space.

DSD? You might wanna ask people who've actually heard it like Lynn or some others hanging around.
#4
30th January 2007
Old 30th January 2007
  #4
Gear interested
 
Joined: Dec 2004
Location: Washington, DC
Posts: 27

Architect is offline
I used to record everything at 32bits because the notion was if your recording at a higher resolution you get better sound quality, etc but after experimenting for a while I started really liking 24bit at lot better. It probably depends a lot on the type of music your recording, your recording style, etc. I can imagine classical recording engineer/producers wanting to record at the highest possible resolution available.
#5
30th January 2007
Old 30th January 2007
  #5
Lives for gear
 
zemlin's Avatar
 
Joined: Sep 2004
Location: Indianapolis, IN
Posts: 658

zemlin is offline
24 bit captures more dynamic range than the human ear can hear. What's the point of going any further?
__________________
Karl Zemlin - www.sonicartistry.net

I couldn't pick a pocket in a pile of dirty clothes - Chris Smither
#6
30th January 2007
Old 30th January 2007
  #6
Lives for gear
 
Protools Guy's Avatar
 
Joined: Nov 2006
Location: US of A
Posts: 1,286

Protools Guy is offline
Quote:
Originally Posted by zemlin View Post
24 bit captures more dynamic range than the human ear can hear. What's the point of going any further?
I couldn't have said it better myself...
#7
30th January 2007
Old 30th January 2007
  #7
Lives for gear
 
Joined: Aug 2005
Location: New England
Posts: 1,955

6strings is offline
Quote:
Originally Posted by Lawrence View Post
There is no real benefit to 32-bit recording since there are no 32-bit converters. [clip] As far as padding a 24-bit file during tracking to make it 32-bits... why? The daw is going to do that anyway as soon as the audio enters it... well in most native daws anyway.
The Cubase manual recommends 32 bit IF you are applying effects during tracking. Other than that there is no reason to record/convert 32 bit files.
#8
30th January 2007
Old 30th January 2007
  #8
Lives for gear
 
Joined: Feb 2006
Location: NYC
Posts: 831

Send a message via AIM to NesNeedsGear
NesNeedsGear is offline
Quote:
Originally Posted by zemlin View Post
24 bit captures more dynamic range than the human ear can hear. What's the point of going any further?
werd
#9
30th January 2007
Old 30th January 2007
  #9
Gear interested
 
Joined: Jan 2007
Posts: 12

vanamp is offline
Quote:
Originally Posted by zemlin View Post
24 bit captures more dynamic range than the human ear can hear. What's the point of going any further?

Couldn't you also say that 44.1 khz captures all frequencies humans can hear??

Then why are companies moving away from standard red book CD format?

MONEY$$$

I don't know I have heard so many different arguments on this
#10
30th January 2007
Old 30th January 2007
  #10
Gear Head
 
Joined: Dec 2006
Posts: 40

RWhite is offline
Just walked in here, but I assume that the advantage of 32 bit recording is for processing purposes only. That gives the processors some headroom to do all of the math on 24 bit files and not have to truncate to 24 bits when the actual result of the math needs 32 bits to express.
#11
30th January 2007
Old 30th January 2007
  #11
Lives for gear
 
Joined: Jul 2002
Posts: 3,881

Duardo is offline
Quote:
Also, Where do you guys think digital audio is heading as far as bit-rate/sample rate wise?
Well, it doesn't seem to be heading anywhere else in any kind of hurry...the common delivery format is still 16/44.1, and if anywhere it's migrating to lower-quality (or at least lower-bandwidth) formats. Unless that changes, I think that the formats we use for production are sufficient and even excessive.

Quote:
Do you think DSD will take over?
No.

Quote:
Or can it even get that much better?
I think it can get better...but I don't think that getting better is synonymous with more bits and higher sampling rates. I think 16/44.1 can still sound better than it does now...it's certainly getting better all the time...

But that's really hard to market.

-Duardo
#12
30th January 2007
Old 30th January 2007
  #12
Lives for gear
 
Tibbon's Avatar
 
Joined: Jul 2004
Location: OH/Columbus
Posts: 4,793

Send a message via AIM to Tibbon Send a message via Skype™ to Tibbon
Tibbon is offline
Your gear can barely render 20-bit recordings well, and 24-bits is already enough 'safety' room and overkill for your ears and your gear. Hell, most of these modern recordings (after mastering) would playback fine with an 8-bit system. I'm serious.

24-bit is 144dB of dynamic range. That means that if you had your speakers calibrated (I know that i'm going to use the wrong symbols on part of this..) so that 0dBfs is at 144dB at your listening location (a few meters), then a full strength signal would be the equavilant of a jet engine taking off. -144dBfs would be dead silence.

Your gear can't even make this happen. Even "24 bit" convertors are at best getting a -120dB noise floor (or 120dB dynamic range). Most preamps have around 100dB of dynamic range. Microphones are even worse.

Then you have to factor in as well for "zero" sound or -144dBfs to mean anything, then -144 needs to be at the noise floor of your room to be heard any difference between a signal at -140dBfs and -144dBfs. That's the NC on your room? I'm really guessing that most of the rooms are barely even meeting NC-30, let alone something really low like NC-10. That means that even moreso, your ROOM can't accomidate the dynamic range of a true 24-bit recording and replication, unless you want to turn up those speakers even louder.

Of course you can probably find speakers and amps that will go louder, but again there's that damn noise floor. With 144dB of gain you're seriously starting to push out random sounds and noise from just random interference and analogue noise in general. It's like pushing your camera to a 3200ISO speed. There will always be noise at that point. It's just physics.

And I can't remember who pointed this out in Tape Op, but someone did. For your console to be able to deal with that dynamic range of 32-bit recordings, you'd have to have MASSIVE (and by massive I mean making an SSL 9000 power supply stack look like a AA battery) set of power rails on the console. Trying to remember, but something about "20 log voltage difference" expressing that dynamic range. I think you'd get around 192dB dynamic range. Remember, this is on a Log NOT linear scale. That's giant.

I did find one link: http://blogs.msdn.com/audiofool/arch...22/713498.aspx

Basically a 32-bit system playing at full scale will be literally pushing the atmosphere back and forth nonstop. It will make you blow up. This seems like a poor idea to me.
__________________
David Fisher (aka tibbon)
#13
30th January 2007
Old 30th January 2007
  #13
Gear interested
 
Joined: Jan 2007
Posts: 12

vanamp is offline
Quote:
Originally Posted by Tibbon View Post
Your gear can barely render 20-bit recordings well, and 24-bits is already enough 'safety' room and overkill for your ears and your gear. Hell, most of these modern recordings (after mastering) would playback fine with an 8-bit system. I'm serious.

24-bit is 144dB of dynamic range. That means that if you had your speakers calibrated (I know that i'm going to use the wrong symbols on part of this..) so that 0dBfs is at 144dB at your listening location (a few meters), then a full strength signal would be the equavilant of a jet engine taking off. -144dBfs would be dead silence.

Your gear can't even make this happen. Even "24 bit" convertors are at best getting a -120dB noise floor (or 120dB dynamic range). Most preamps have around 100dB of dynamic range. Microphones are even worse.

Then you have to factor in as well for "zero" sound or -144dBfs to mean anything, then -144 needs to be at the noise floor of your room to be heard any difference between a signal at -140dBfs and -144dBfs. That's the NC on your room? I'm really guessing that most of the rooms are barely even meeting NC-30, let alone something really low like NC-10. That means that even moreso, your ROOM can't accomidate the dynamic range of a true 24-bit recording and replication, unless you want to turn up those speakers even louder.

Of course you can probably find speakers and amps that will go louder, but again there's that damn noise floor. With 144dB of gain you're seriously starting to push out random sounds and noise from just random interference and analogue noise in general. It's like pushing your camera to a 3200ISO speed. There will always be noise at that point. It's just physics.

And I can't remember who pointed this out in Tape Op, but someone did. For your console to be able to deal with that dynamic range of 32-bit recordings, you'd have to have MASSIVE (and by massive I mean making an SSL 9000 power supply stack look like a AA battery) set of power rails on the console. Trying to remember, but something about "20 log voltage difference" expressing that dynamic range. I think you'd get around 192dB dynamic range. Remember, this is on a Log NOT linear scale. That's giant.

I did find one link: http://blogs.msdn.com/audiofool/arch...22/713498.aspx

Basically a 32-bit system playing at full scale will be literally pushing the atmosphere back and forth nonstop. It will make you blow up. This seems like a poor idea to me.

The speed limit on the highway is 65 but I still want a corvette.

But seriously,

Whats the deal with Tape then. Is it just they way the sound hits the tape that people like?

I don't have enough experience with 2" tape to really hear a difference but its still around for a reason...right?
#14
30th January 2007
Old 30th January 2007
  #14
Lives for gear
 
Tibbon's Avatar
 
Joined: Jul 2004
Location: OH/Columbus
Posts: 4,793

Send a message via AIM to Tibbon Send a message via Skype™ to Tibbon
Tibbon is offline
Tape

There will always be raging debates between the analogue and digital forms of things IMHO.

Tape sounds different because it IS different. It's doing something totally differently than a A/D convertor is.

Tape vs A/D
Film vs CMOS/CCD
Film vs video
traditional steering vs fly-by-wire

I think that some people in photography actually have figured it out right. Capturing something on film is an emulsion process. It's capturing light onto a physical object via a chemical/physical process. It's called a photograph or photographic negative. A digital camera does not take a photo. When a digital camera grabs it, then it is corrected termed an Image. It captures an image via recording electronic values via a grid of sensors hooked to A/D convertors. If you look at it then way as photo vs image then (using correct terms), then you can see that it's comparing Apples to Oranges. Yea, they both get the same thing done (just as Apples and Oranges both feed you fruit!), and capture a moment in time, but they are totally different processes and not really comparable. Film looks like film. Digital looks like digital. If you like the look of film, but need the digital workflow, then scan your negatives and store them for later darkroom use. I actually have a "photography" company that's called "AFDF Imagography" since we work almost only in digital, and it's not really taking a photo (photography) if you're taking an image. So it's Imagography instead.

If you need the sound of tape, then use tape. It sounds different than digital, because it is different. Two different circuits rarely yield the exact same result, and a Rosetta 800 looks NOTHING at all inside (on or paper) like a Studer A80.

But back to your question, there are subtile differences between analogue and digtal recordings. In the end, it all gets the same thing done. I don't think one is better than the other. Digital was pretty bad at first, but so was analogue at first. Digital has came into it's own. It's a decision in the recording process just like choosing what key (C vs C#) tempo (120 vs 121 bpm), or preamp. There's the chance that the end user wouldn't know the difference, but it's still a critical choice.

The reason I write this is that i wish the analogue vs digital debate would stop. They are different. Period. Digital can emulate analogue, but it's still a fundamentally different process. Pick one, and make music.
#15
30th January 2007
Old 30th January 2007
  #15
Gear interested
 
Joined: Jan 2007
Posts: 12

vanamp is offline
Quote:
Originally Posted by Tibbon View Post
. Pick one, and make music.
#16
30th January 2007
Old 30th January 2007
  #16
Lives for gear
 
bcgood's Avatar
 
Joined: Jun 2005
Posts: 6,157

bcgood is offline
Quote:
Originally Posted by Tibbon View Post
Pick one, and make music.
Or both...

bcgood
#17
30th January 2007
Old 30th January 2007
  #17
Lives for gear
 
zemlin's Avatar
 
Joined: Sep 2004
Location: Indianapolis, IN
Posts: 658

zemlin is offline
Quote:
Originally Posted by pegleg View Post
I guess when you make this statement, you're assuming digital zero is at or under the threshold of pain? Since the threshold of pain is considered between 120 and 140 dB SPL, then, yes, 24 bit recording (approx 144dB dynamic range) would leave its quietest few dB below the threshold of hearing.
As highlighted by Tibbon above, there are so many other limitations on the analog side of the system, I don't need to make this assumption. With current technology, it isn't possible to build a playback system with that much dynamic range (so I've been told - not in a position to debate the details).
#18
30th January 2007
Old 30th January 2007
  #18
Lives for gear
 
Joined: Aug 2005
Location: New England
Posts: 1,955

6strings is offline
Quote:
Originally Posted by RWhite View Post
Just walked in here, but I assume that the advantage of 32 bit recording is for processing purposes only. That gives the processors some headroom to do all of the math on 24 bit files and not have to truncate to 24 bits when the actual result of the math needs 32 bits to express.
#19
30th January 2007
Old 30th January 2007
  #19
Lives for gear
 
A27Hull's Avatar
 
Joined: May 2005
Location: Tennesse Valley, AL
Posts: 705

A27Hull is offline
Quote:
Originally Posted by vanamp View Post
Whats the deal with Tape then. Is it just they way the sound hits the tape that people like?
I wrote some nonsense earlier, so I apologize to any who read my post. I dont know enough about tape/tape machines to comment.

guess I'll be hitting the books now!

Last edited by A27Hull; 1st February 2007 at 02:48 AM.. Reason: stupidity
New Reply Submit Thread to Facebook Facebook  Submit Thread to Twitter Twitter  Submit Thread to LinkedIn LinkedIn  Submit Thread to Google+ Google+ 
 
Thread Tools
Search this Thread
Search this Thread:

Advanced Search

Forum Jump

SEO by vBSEO ©2011, Crawlability, Inc.