24bit?

[DrXparaMental]

Within the box is a term meaning mixing solely within the computer, using DAWS and/or software plugins only. Not routing to hardware mixers or hardware units such as Manley compressors or SSE desks etc.

Capturing audio signals with microphones is an art form of it;s own with a lot of dynamic range and possible problems to mess things up. vstis/softsynths etc. are either processed using wavetables or generated synthesis and much of the audio processing is controlled within the software, leaving it less prone to fluctuations compared to running it out into an amp and then micing it and recording it that way. Although there are times where this is desirable (if you get your mics positioned well, great pres, know how to engineer a little) but the quality of most vstis now and the extral controls gives more possibilities without the inconvenience of micing it all up, so it's not very widespread.


What I mean particularly are vocals, guitars, sax, etc to my ears, often sound a tad fuller and richer in 24 bit mode.

Thanks & Understood. At this point I am all in the box with respect to recording clips and improv sessions. The reasons I have become so specifically curious about bit rate is the following.


(a) General reason:
I hope to become a Live performer in more ways the one. I have been into improvisation for a good while now. About the last 10-12 years of the 30 total I have been playing. I started with floor controlled looping devices and progressively incorporated a V-Bass (midi synth module as I am by strictest definition an electric bassist) & VS1880 (multi track hardware recorder) into live improvisational performance. Every musical move I have made over the course of the last 4-5 months has been made with the exclusive purpose of forwarding my evolution in true quantum fashion as a musician. I have a added a decent laptop computer/external hardrive, a decent audio/midi interface & a decent 46 key midi controller. I am using Ableton Live 6.0 (of course), Reaktor 5 (don't know a damn thing about it yet) & Stylus RMX. To make a long story very short, it's like jamming with 100 times the capability I possessed last year at this time. And that's with the extremely small amount I already have come to somewhat understand about the new hard/soft ware.

(b) Specific Reason
One the absolute best developments over the course of the last year is meeting and teaming up with EM at http://www.whiterobotrecords.com/home.html

if you go/click on "musicbot", about half way down the page, there midst other treasure, lurk a few...well, I'll leave it up to you to determine what they are.

When we get together and "sketch" stuff via a virtual notebook approach to improv, he uses a basic G5 with a basic Pro Tools input and then masters this material via an Alesis Masterlink. I apologize as I don't know much about this rig (obviously) as it's not mine, but EM assures me the audio outcome, and my like of it, comes a great deal from the 24bit Pro Tools capability. It has a somewhat "warm" sound after all the normalized BS. I really don't know what this is due to but I am trying to get there because I really dig that analogesque warmth. One thing I will say is that damn Master Link is one of the slowest pieces of equipment on the planet. Certainly not built for those blokes like myself with an above average amount of ADD.

[dh187]

Dr. X,
I don't want to make you tweak about Live. It's awesome. I use it almost exclusively. However, the main reason your friends material sounds better coming out of ProTools is that the PT audio engine sounds better than the Ableton audio engine. This may cause a big flare-up in this thread, but I assure you, it's absolutely true. A lot of hard core Live users won't admit it, either out of wishful thinking or blissful ignorance, but I have A/B'd in a very controlled situation and there is no doubt about it. I don't know exactly where the differentiation occurs. For example, if I playback a file through the master output of Live and the exact same file through the master output of PT (with all variables the same including bit depth, sample rate, audio interface, buffer size, clock source, cabling, monitors, room...), the PT output is much more open, clear, more headroom, indubitably better. That's probably what you're hearing. On the other hand if I record audio into PT and the exact same audio into Live and play it back on a CD or quicktime or iTunes or whatever, it sounds the same. The word on the street is that the issue lay somewhere in the summing bus, that is the point at which individual tracks are combined into the stereo master track. I don't think anyone who doesn't have the software code knows the exact answer and those who do have the code are not exactly forthcoming regarding the shortcomings of their brainchild. Keep in mind that the main focus of Live is its workflow and making it processor and on the fly friendly. Corners have to be cut to make the instrument as creatively open as possible.
Also if you really want to get into bit depth and sample rate go read the white papers on those subjects at:
lavryengineering.com
The author of those theories knows a lot more than most of the hobbyists and regurgitators on this forum. Some mistakes have been made in this thread regarding particularly sample rate and its relationship to resolution.

mdb wrote: Your capturing more accurate representation of what you recorded because its able to sample the audio many more times a sec

That ain't true. I won't go into the specifics but rather yield to the experts. The world of digital audio has plenty of dark corners and pop philosiphy. Here's a little from me: use your own judgement and go with what sounds good to you.

[jesQuick]

Dr. X,
I don't want to make you tweak about Live. It's awesome. I use it almost exclusively. However, the main reason your friends material sounds better coming out of ProTools is that the PT audio engine sounds better than the Ableton audio engine. This may cause a big flare-up in this thread, but I assure you, it's absolutely true. A lot of hard core Live users won't admit it, either out of wishful thinking or blissful ignorance, but I have A/B'd in a very controlled situation and there is no doubt about it. I don't know exactly where the differentiation occurs. For example, if I playback a file through the master output of Live and the exact same file through the master output of PT (with all variables the same including bit depth, sample rate, audio interface, buffer size, clock source, cabling, monitors, room...), the PT output is much more open, clear, more headroom, indubitably better. That's probably what you're hearing. On the other hand if I record audio into PT and the exact same audio into Live and play it back on a CD or quicktime or iTunes or whatever, it sounds the same. The word on the street is that the issue lay somewhere in the summing bus, that is the point at which individual tracks are combined into the stereo master track. I don't think anyone who doesn't have the software code knows the exact answer and those who do have the code are not exactly forthcoming regarding the shortcomings of their brainchild. Keep in mind that the main focus of Live is its workflow and making it processor and on the fly friendly. Corners have to be cut to make the instrument as creatively open as possible.
Also if you really want to get into bit depth and sample rate go read the white papers on those subjects at:
lavryengineering.com
The author of those theories knows a lot more than most of the hobbyists and regurgitators on this forum. Some mistakes have been made in this thread regarding particularly sample rate and its relationship to resolution.

mdb wrote: Your capturing more accurate representation of what you recorded because its able to sample the audio many more times a sec

That ain't true. I won't go into the specifics but rather yield to the experts. The world of digital audio has plenty of dark corners and pop philosiphy. Here's a little from me: use your own judgement and go with what sounds good to you.

Now hold it... Dont want to start a flame war, by no means... But we have already established, on several occasions, that there is no difference what so ever between the audio engines. That being PT, Logic whatever....

People have done extensive test and the outcome was always total phase cancellation... Am I right !?!?!?

The difference people experience is between the different AD/DA converters and monitors...

Best

-J

ps. Havent done tests myself, so this is based solely on numerous forum threads...

[eyeknow]

allow me to cliff note all "this".......

bits is very important.

sample rate is VERY important, but is confusing and very intense on the cpu. To confuse you more, certain types of instuments are more noticeable in this area than others (bass guitar being the MOST noticeable to my ears)

There are differences in audio quality between daw's and I'll take the fucking pepsi challenge on THAT any day! (although do NOT separate this from other points)

everyone's ears are different (that is INCLUSIVE of the other points)

And (and this is the mother of all mothers) your "state of mind" is going to play a SIGNIFICANT role in your studies (i.e. pot, booze, shrooms....girl/boy friend.....etc)

[DrXparaMental]

allow me to cliff note all "this".......

bits is very important.

sample rate is VERY important, but is confusing and very intense on the cpu. To confuse you more, certain types of instuments are more noticeable in this area than others (bass guitar being the MOST noticeable to my ears)

There are differences in audio quality between daw's and I'll take the fucking pepsi challenge on THAT any day! (although do NOT separate this from other points)

everyone's ears are different (that is INCLUSIVE of the other points)

And (and this is the mother of all mothers) your "state of mind" is going to play a SIGNIFICANT role in your studies (i.e. pot, booze, shrooms....girl/boy friend.....etc)

Immediate apologies from me. I really don't think I did a good job of being "clear" as far as WHEN this warmth comes into play. It does NOT happen as the music is recorded or at least I don't think so. If I take the same examples of raw audio that is yet non effected and non remastered, then do a basic algorithmic soft clipped -15.0 db normalization process on my own equipment (all in the box PC utilizing Pristine Sounds 2000) I get very similar results to his on the fly slower than slow alesis masterlink. I should have never brought this whole thing up because I am way too lost myself. I had just hoped that the end result "warmth" reference might achieve some kind of basic recognition via a helpful advanced member insight. That was foolish on my part. Again, I apologize.

[Spindrift]

The difference listening to a single track of 16 bit vs 24 bit is not that great.
16 bit has plenty of dynamic range and usually the limit there will be your converters rather than the bit depth of the file.

But when working with audio in your DAW there is other issues that makes 24 bit preferable.
The engine is working at 32 bit float usually, so when you record your 16 bit files they need to be either truncated to 16 bit or you could use dither.
Truncating is not ideal, and neither is dithering many individual tracks.
When played back they need to be upscaled again....and going back to CD format you again need to scale the audio down to 16.
32 float to 24 fixed on the other hand is a more painless conversion. A 32 bit float signal does not need to be truncated or dithered to scale down to 24 bit fixed.

So the reason 24 bit sound better when you use it for all your individual tracks does not have so much to do with dynamic range as with technical issues doing conversions back and forth while working with 16 bit.

CPU usage is not affected by bitdepth. If anything it might be a bit more work for the processor to be converting bitdepth of your files if you use 16 bit.
It was many years ago I felt hard disks limited my track count when working with 24 bit. With a good system and a regular IDE drive 32 simultaneous tracks at 44.1/24 should not be a problem.
If you use more than that and can't budget for a SCSI/RAID configuration it's not the whole world to use 16 bit, but usually I see very little reason for it.

[Michael-SW]

...
24 bit definately gives you more headroom and I think it gives the hobbyist more leeway for errors etc. In 16 bit, you had to try and get your levels hot but without clipping for optimal sound, 24 bit gives you more of a comfort zone to record lower and still have it sound good and run less risk of clipping.

Perfectly said. THIS is the very reason you should work in 24 bits. Unless you have your levels perfectly adjusted, 16 bits might very well end up as only 14 (or lower) bits when you freeze, record, render etc. With 24 bits, you have a lot of margin when you finally convert to 16 bits.

[noisetonepause]

Now hold it... Dont want to start a flame war, by no means... But we have already established, on several occasions, that there is no difference what so ever between the audio engines. That being PT, Logic whatever....

People have done extensive test and the outcome was always total phase cancellation... Am I right !?!?!?

The difference people experience is between the different AD/DA converters and monitors...

Best

-J

ps. Havent done tests myself, so this is based solely on numerous forum threads...

Funny, I have the opposite impression from a number of forum threads... there are differences once you mix down.

[eyeknow]

can we all at least agree that it costs too much no matter what?

[DrXparaMental]

Now hold it... Dont want to start a flame war, by no means... But we have already established, on several occasions, that there is no difference what so ever between the audio engines. That being PT, Logic whatever....

People have done extensive test and the outcome was always total phase cancellation... Am I right !?!?!?

The difference people experience is between the different AD/DA converters and monitors...

Best

-J

ps. Havent done tests myself, so this is based solely on numerous forum threads...

Funny, I have the opposite impression from a number of forum threads... there are differences once you mix down.

It's really very tough to say from my layman's listening standpoint. The ears are good, but the brain has much catching up to do. I hear a distinct warmth with relation to bass response when I listen to the pro tools analog audio input feeds as well as after mix down. The fact is, this initial warmth can be destroyed all too easy somehow with rash or "quick" normalization. It seems as though a PROPER mix down must be key in getting said warmth to arrive on location via a burn to CDR post 16bit degradation. I have no clue what this process is although I am beginning to believe it has something to do with correctly compressing the audio.


I was paying very close attention today when we recorded at White Robot and straight from the rack mounted Pro Tools sound card there is a measurably greater degree of warmth than my e-mu's (I have an 1820m & 1616m) will render. Why is this when they both use the exact same 24bit converters??? This (seems) to be with respect to their ability to convert external analog audio input to digital and what results in my ears where the audible end results. What the heck is this plain and simple?

I would LOVE the opportunity to do an actual controlled side by side comparison between the e-mu & pro tools sound cards, but I am tapped for this year so NO Pro Tools for this strapped young lad. It seems as though this learning curve is both fascinating & frustrating in that I sometimes literally cancel out knowledge that just the day before I would have sworn to be valid.

[Pantytec]

You're probably going to experiance dropouts switching from 16 to 24 if you have less than a gig of ram. It depends what else is going on... fat hog CPU monsters like Massive, a shit ton of plugs, modulations in Live, etc.

I would be shocked if you can "just switch" to 24bit without any sideffects.

There is no free lunch, turd.

[DrXparaMental]

You're probably going to experiance dropouts switching from 16 to 24 if you have less than a gig of ram. It depends what else is going on... fat hog CPU monsters like Massive, a shit ton of plugs, modulations in Live, etc.

I would be shocked if you can "just switch" to 24bit without any sideffects.

There is no free lunch, turd.

There's one in every crowd.

2.0 core2duo/2 gig Ram (no problems here)

But the real issue is that THANKS to the GOOD folks here, I know that bit rate choice has ZERO to do with CPU usage.

[DrXparaMental]

You're probably going to experiance dropouts switching from 16 to 24 if you have less than a gig of ram. It depends what else is going on... fat hog CPU monsters like Massive, a shit ton of plugs, modulations in Live, etc.

I would be shocked if you can "just switch" to 24bit without any sideffects.

There is no free lunch, turd.

There's one in every crowd.

2.0 core2duo/2 gig Ram (no problems here)

But the real issue is that THANKS to the GOOD folks here, I know that bit rate choice has ZERO to do with CPU usage.

[Michael-SW]

You're probably going to experiance dropouts switching from 16 to 24 if you have less than a gig of ram. It depends what else is going on... fat hog CPU monsters like Massive, a shit ton of plugs, modulations in Live, etc.

I would be shocked if you can "just switch" to 24bit without any sideffects.

There is no free lunch, turd.

This IS an almost free lunch. Since everything internal in Live is 32 bit anyway (effects, synths etc), it doesn't really matter what you render/record in. Files will be larger and your HD will have to work a bit harder but your CPU won't have to do any more processing.

[timothyallan]

All I know is that I get over 25 bits of pre-processed word length when I truncate my 24bit wav's. If I convert them to a 320mp3, BUT dither them using an 8->24bit dithering algorythm, I can totally tell the difference, as can people in a club. They subconsiously will react, even if your AD/DA's were sync'd by a proper midi splitter during the process. So always use 24bit, because people can hear the difference.