pdarg
Max Output Level: -52.5 dBFS
- Total Posts : 2265
- Joined: 2004/03/26 17:52:53
- Status: offline
More questions about 64-bit recording
I am thinking about making the transition to 64-bit recording.
What happens at mixdown time? What bit rate would you mix this down to if you are going to do mastering processing later? Does Sonar X1 actually export 64-bit audio?
Also, people are telling me that switching to 64-bit recording puts many of the recording resources in RAM, and makes everything run very smoothly. Can someone confirm this? So, using a high bit rate actually frees up computing resources??
Thanks in advance.
|
SvenArne
Max Output Level: -48 dBFS
- Total Posts : 2719
- Joined: 2007/01/31 12:51:29
- Location: Trondheim, Norway
- Status: offline
Re:More questions about 64-bit recording
2011/12/06 15:45:54
(permalink)
Using the 64 bit application in a 64 bit environment makes you able to utilize more RAM. It has nothing to do with recording/rendering/exporting 64 bit audio files (which SONAR can do, but noone would want to...)
|
BenMMusTech
Max Output Level: -49 dBFS
- Total Posts : 2606
- Joined: 2011/05/23 16:59:57
- Location: Warragul, Victoria-Australia
- Status: offline
Re:More questions about 64-bit recording
2011/12/06 15:51:16
(permalink)
The OP has not made himself clear, are you talking about 64 bit operating system or the 64 bit mix engine.
Do a search there are whole debates on both these topics, if you still are confused ask again and watch the razors fly.
So my advice is do the search, do some research then as I say if you are still stuck and don't mind a dust up ask again.
Peace Ben
|
SvenArne
Max Output Level: -48 dBFS
- Total Posts : 2719
- Joined: 2007/01/31 12:51:29
- Location: Trondheim, Norway
- Status: offline
Re:More questions about 64-bit recording
2011/12/06 15:53:48
(permalink)
BenMMusTech
are you talking about 64 bit operating system or the 64 bit mix engine.
Neither, he's talking about working with 64-bit audio files!
|
BenMMusTech
Max Output Level: -49 dBFS
- Total Posts : 2606
- Joined: 2011/05/23 16:59:57
- Location: Warragul, Victoria-Australia
- Status: offline
Re:More questions about 64-bit recording
2011/12/06 16:51:42
(permalink)
Sorry I didn't read the OP properly, stick with 32 bit all the way, because not everything is 64 bit.
The theory and it's just a theory is that if you process at 32 bit and your audio file is 32 bit the whole system works better because there is no upsampling and no downsampling.
This is the final post on this matter for me, as I said look it up and there are many fights over this on this forum alone.
As I say what you are talking about is a theory that no one has proved or disproved but whenever I do a new project I set the audio recording to 32 bit, that way everything is 32 bit.
If I bounce down to master it tends to be 24 bit and nine times out of ten I will dither then, not when I bounce down to 16 bit, this is because it is the time based effects that suffer when you bounce down to a different bit depth.
In mastering because I never use time based effects, (someone correct me are linear phase time based effects) when I go from 24 bit to 16 bit we arn't losing the tails of our reverbs because we have kept everything pristine in the first place.
Ok someone explain this better than I can, it's like this all effects are process at 32 bit, but if we bounce down to 24 bit we are going to loose 8 bit's of the effects and this is where the problem is, so if we bounce down from 32 bit to 24 bit, to keep the tails of the time based effects we dither here rather than from 24 bit to 16 bit because if we don't the damage has already been done.
Once again someone please explain this better.
Peace Ben
|
Jeff Evans
Max Output Level: -24 dBFS
- Total Posts : 5139
- Joined: 2009/04/13 18:20:16
- Location: Ballarat, Australia
- Status: offline
Re:More questions about 64-bit recording
2011/12/06 17:04:44
(permalink)
As far as I know (but please correct me if I am wrong) there is no audio interface that actually works at 32 bit (64 bit?) only 24 bit. So setting record sessions at 32 bit won't achieve anything. You might as well set up a session at 24 bit. (24 bit does sound better than 16 bit and has a much lower digital noise floor)
All mixing can be easily done in 32 bit or on a 32 bit system. Having extra 8 bits over 24 bit is all you need and will minimise errors. The place where 64 bit comes into its own is on the mix buss where the extra precision is handy and known to be valuable. All exports can be done at 24 bit. The limiter or the very last thing in the mastering chain can be dithering down to 16 bit where it needs to be. (I use a PSP Xenon limiter and it has the dithering function built in, very handy) Working at higher resolutions (bit depths that is) than what I have described here will NOT produce any improved audible results. This may be different for sampling rates though as has been debated here many times. Although they say a sampling rate much higher than 48KHz is also not that audible either.
post edited by Jeff Evans - 2011/12/06 17:10:50
Specs i5-2500K 3.5 Ghz - 8 Gb RAM - Win 7 64 bit - ATI Radeon HD6900 Series - RME PCI HDSP9632 - Steinberg Midex 8 Midi interface - Faderport 8- Studio One V4 - iMac 2.5Ghz Core i5 - Sierra 10.12.6 - Focusrite Clarett thunderbolt interface Poor minds talk about people, average minds talk about events, great minds talk about ideas -Eleanor Roosevelt
|
Chregg
Max Output Level: -51.5 dBFS
- Total Posts : 2382
- Joined: 2010/02/22 06:14:27
- Location: Perth, Scotland
- Status: offline
Re:More questions about 64-bit recording
2011/12/06 17:10:49
(permalink)
Ok someone explain this better than I can, it's like this all effects are process at 32 bit, but if we bounce down to 24 bit we are going to loose 8 bit's of the effects.................. 32 bit float is a 24 bit word, the 8 bit is the exponent, so its not a fixed integer !!! linear phase is a finite impulse response, that has a limited out put, moving average filters, do exactly that, average the signals amplitude, there can be a lag with depending on how many digital taps there are, which is why you get latency issues with linear phase, i dont think i want to call it a time based effect cuz that not the purpose of it, unlike a delay which is time based
|
BenMMusTech
Max Output Level: -49 dBFS
- Total Posts : 2606
- Joined: 2011/05/23 16:59:57
- Location: Warragul, Victoria-Australia
- Status: offline
Re:More questions about 64-bit recording
2011/12/06 17:20:15
(permalink)
Jeff Evans
As far as I know (but please correct me if I am wrong) there is no audio interface that actually works at 32 bit (64 bit?) only 24 bit. So setting record sessions at 32 bit won't achieve anything. You might as well set up a session at 24 bit. (24 bit does sound better than 16 bit and has a much lower digital noise floor)
All mixing can be easily done in 32 bit or on a 32 bit system. Having extra 8 bits over 24 bit is all you need and will minimise errors.
The place where 64 bit comes into its own is on the mix buss where the extra precision is handy and known to be valuable. All exports can be done at 24 bit. The limiter or the very last thing in the mastering chain can be dithering down to 16 bit where it needs to be. (I use a PSP Xenon limiter and it has the dithering function built in, very handy)
Working at higher resolutions (bit depths that is) than what I have described here will NOT produce any improved audible results. This may be different for sampling rates though as has been debated here many times. Although they say a sampling rate much higher than 48KHz is also not that audible either.
No you are right Jeff (hi) what happens is the 24 bit file is wrapped in a 32 bit file with the extra 8bit's just a bunch of zeroes. This is where the theory comes in , if you have a 32 bit audio file and are processing at 32 bit, the system has to do less processesing, ok remember, this is a theory, neither proven or disproven. have to disagree about the dithering thing if you have mixed at 64 bit and bounce to 24 bit how are we not losing data. In my opinion and Chregg seemed to have a handle on it better than me is that we dither from 64 bit to 24 bit because all the yummy data is in the 64 bit file. Once we get to 24 bit, we have lost all that extra information because it gets lost in the noise floor. Once againhelp me out and explain it better, this is probably why they won't let me into an RHD course or teach for that matter I can never explain myself properly. Ok Peace Ben
|
Bub 
Max Output Level: -3.5 dBFS
- Total Posts : 7196
- Joined: 2010/10/25 10:22:13
- Location: Sneaking up behind you!
- Status: offline
Re:More questions about 64-bit recording
2011/12/06 17:31:44
(permalink)
Here's my take on it. The higher you go, the more better off you are because if you need to bounce or apply effects, dither will not come in to play.
Remember, bit and sample rates don't have the same effect as different kinds of tape did. bit/sample rates are just methods of capture. They have no bearing on the actual sound you are recording ... until you start to apply effects, bounce, export.
I 'think' the theory behind all of this is, if all of your processes are a single bit depth, your PC will work smoother and faster because it doesn't have use extra resources converting between 32 and 64 bit and applying dither. For example, if you have all 64bit processed plug-in's, it may behoove you to record 64bit.
But ... It would take a lot of research to find how your plug-in's are processed. And some of them can be 64bit compatible plug-in's but process at 32bitF and vice versa. For example, Guitar Rig processes at 64bit, but the interface is 32bit. You can tell by enabling the Double Precision engine and look at the ticks on the Guitar Rig plugin in the FX Bin. Double ticks=64bit, single ticks=32bit. Guitar Rig shows up with single ticks (32bit) even though it's processing at 64bit.
It seems to me though, that even if you have a completely 64bit system, the extra size of the 64bit recorded tracks would slow you down more than your computer dealing with the internal processing of the mixture of 32/64 bit plug-in's.
I could be wrong about this, but I thought 64bit wave files were something proprietary used by Sony?
"I pulled the head off Elvis, filled Fred up to his pelvis, yaba daba do, the King is gone, and so are you."
|
pdarg
Max Output Level: -52.5 dBFS
- Total Posts : 2265
- Joined: 2004/03/26 17:52:53
- Status: offline
Re:More questions about 64-bit recording
2011/12/06 17:32:50
(permalink)
Okay - we're getting some answers.
Here's what I do now:
-Record at 24-bit
-Mix as needed
-Export as 32-bit float (which I understand does not require dithering, and does not have to re-eample bits from efx processing[?])
-Master 32-bit audio file, truncate, encode to MP3
Sounds like moving up to 64-bit may not be right for me(???)
|
Chregg
Max Output Level: -51.5 dBFS
- Total Posts : 2382
- Joined: 2010/02/22 06:14:27
- Location: Perth, Scotland
- Status: offline
Re:More questions about 64-bit recording
2011/12/06 17:34:33
(permalink)
64 bit float is a 53 bit word with an 11 bit exponent, so bouncing down to 24 bit, yeah your loosing data
|
Jim Roseberry
Max Output Level: 0 dBFS
- Total Posts : 9871
- Joined: 2004/03/23 11:34:51
- Location: Ohio
- Status: offline
Re:More questions about 64-bit recording
2011/12/06 17:43:49
(permalink)
Hi Phil,
A 64Bit DAW application... and 64Bit audio resolution are completely separate things.
IOW, You can run the 32Bit version of Sonar X1... and sum with 64Bit dual precision Float resolution.
Or... or you could run the 64Bit version of Sonar X1... and sum with 32Bit Float resolution.
Running a 64Bit DAW app allows you to take advantage of more than 4GB RAM.
Currently, A/D converters don't even offer a full 24Bits of audio resolution... so there would be no advantage to recording 64Bit audio files. That said, summing with 64Bit dual precision Float does make good sense. It makes rounding error a moot point. So, I could see a logical point to bouncing down your final mixes to a 64Bit Float audio file. You could then "master" the mixes at 64Bit dual precision Float resolution.
This would preserve maximum fidelity...
A fully 64Bit Sandybridge DAW can run substantial loads at ultra low latency settings.
If you make extensive use of soft-synths, you'll certainly appreciate the ability to use greater than 4GB RAM.
Hope that helps clarify...
|
Bub
Max Output Level: -3.5 dBFS
- Total Posts : 7196
- Joined: 2010/10/25 10:22:13
- Location: Sneaking up behind you!
- Status: offline
Re:More questions about 64-bit recording
2011/12/06 17:51:17
(permalink)
Wait until he asks about 192kHz. ;)
"I pulled the head off Elvis, filled Fred up to his pelvis, yaba daba do, the King is gone, and so are you."
|
pdarg
Max Output Level: -52.5 dBFS
- Total Posts : 2265
- Joined: 2004/03/26 17:52:53
- Status: offline
Re:More questions about 64-bit recording
2011/12/06 17:54:29
(permalink)
Hi Jim! Yes, it does clarify things quite a bit.
This means that I may need to get the newest version of WaveLab, since my current version (4.x) cannot handle 64-bit float.
Thanks!
|
drewfx1
Max Output Level: -9.5 dBFS
- Total Posts : 6585
- Joined: 2008/08/04 16:19:11
- Status: offline
Re:More questions about 64-bit recording
2011/12/06 18:11:17
(permalink)
Jeff Evans
The place where 64 bit comes into its own is on the mix buss where the extra precision is handy and known to be valuable.
In the real world there is no benefit. People think there is a difference because they read theoretical claims about it, but no one bothers to do testing to put the errors in context, in which case they would find out it will never be audible in the real world. But of course if you make a theoretical claim that something might be there, people will swear up and down that they can hear it.
 In order, then, to discover the limit of deepest tones, it is necessary not only to produce very violent agitations in the air but to give these the form of simple pendular vibrations. - Hermann von Helmholtz, predicting the role of the electric bassist in 1877.
|
keith
Max Output Level: -36.5 dBFS
- Total Posts : 3882
- Joined: 2003/12/10 09:49:35
- Status: offline
Re:More questions about 64-bit recording
2011/12/06 18:21:58
(permalink)
Chregg
64 bit float is a 53 bit word with an 11 bit exponent, so bouncing down to 24 bit, yeah your loosing data
Any "word length reduction" *should* include dither to remove truncation error. 64-bit float => 32-bit float, 32-bit float => 24-bit integer, etc. all "require" dither, even though in each of those cases you will practically not inroduce any audible aliasing in the output. In theory, simple truncation introduces rounding error, the rounding error is additive, and the rounding error manifests itself as aliasing. Dither introduces random noise at the highest bits to break up the rounding error... which effectively decorrelates the error from the signal. The bit depth of audio determines the dynamic range: each bit provides approximately -6db of range. 16bit X -6dB = -96dB for audio CDs, 24bit x -6dB = -144dB, etc. For floating point you need to factor in the exponent, resulting in a dynamic range of around -1500dB. 64-bit float even greater dynamic range, obviously. You can see that truncating 24-bit down to 16-bit, you're introducing quantization error into the audible range... for higher bit depths you just need to worry about being "correct", as truncating anything down to 24-bit integer without dither will introduce quantization noise in the inaudible range. There was a big lon thread about this on this forum about 6 years ago, and Ron Kuper added dither options when outputting 32-bit float to the 24-bit hardware, or when exporting 32-bit float to 24-bit stems... Sonar didn't have 64-bit float at the time... 
|
Jeff Evans
Max Output Level: -24 dBFS
- Total Posts : 5139
- Joined: 2009/04/13 18:20:16
- Location: Ballarat, Australia
- Status: offline
Re:More questions about 64-bit recording
2011/12/06 18:34:59
(permalink)
Hey drewfx 1 In the real world there is no benefit. People think there is a difference because they read theoretical claims about it, but no one bothers to do testing to put the errors in context, in which case they would find out it will never be audible in the real world.
I am totally with you with this. All of the stuff mentioned here I really doubt would be the slightest bit audible in a controlled AB test. I did read somwhere they did a test where they took a 64 bit file and did all the right things to dither it down to 16 bit (using the finest dithering processes) and compared that to a 64 bit file that just had 48 bits just truncated off and no one could tell the difference!
I think 24 bit is better than 16 bit though in the recording phase. That has been proven to be correct.
Specs i5-2500K 3.5 Ghz - 8 Gb RAM - Win 7 64 bit - ATI Radeon HD6900 Series - RME PCI HDSP9632 - Steinberg Midex 8 Midi interface - Faderport 8- Studio One V4 - iMac 2.5Ghz Core i5 - Sierra 10.12.6 - Focusrite Clarett thunderbolt interface Poor minds talk about people, average minds talk about events, great minds talk about ideas -Eleanor Roosevelt
|
Bub
Max Output Level: -3.5 dBFS
- Total Posts : 7196
- Joined: 2010/10/25 10:22:13
- Location: Sneaking up behind you!
- Status: offline
Re:More questions about 64-bit recording
2011/12/06 18:39:00
(permalink)
keith
Chregg
64 bit float is a 53 bit word with an 11 bit exponent, so bouncing down to 24 bit, yeah your loosing data
Any "word length reduction" *should* include dither to remove truncation error. 64-bit float => 32-bit float, 32-bit float => 24-bit integer, etc. all "require" dither
Question about this ^ ... if you have Sonar set to record 32bit wav tracks, and you have dithering turned 'off', does Sonar over ride this and dither anyway since there are no audio drivers that support 32bit or is that 32bit to 24bit conversion done by the audio driver?
"I pulled the head off Elvis, filled Fred up to his pelvis, yaba daba do, the King is gone, and so are you."
|
Chregg
Max Output Level: -51.5 dBFS
- Total Posts : 2382
- Joined: 2010/02/22 06:14:27
- Location: Perth, Scotland
- Status: offline
Re:More questions about 64-bit recording
2011/12/06 18:49:45
(permalink)
i think wat happens bub is when the d/a coverter recieves the signal from sonar it truncates to 24 bit and any errors will be filtered out with a brick wall filter (considering the errors are not in the audible range) , some sound carfds like the m audio deltas have internal processing too, the its the converters that are at 24 bit depth
|
drewfx1
Max Output Level: -9.5 dBFS
- Total Posts : 6585
- Joined: 2008/08/04 16:19:11
- Status: offline
Re:More questions about 64-bit recording
2011/12/06 19:15:31
(permalink)
Jeff Evans
Hey drewfx 1 In the real world there is no benefit. People think there is a difference because they read theoretical claims about it, but no one bothers to do testing to put the errors in context, in which case they would find out it will never be audible in the real world.
I am totally with you with this. All of the stuff mentioned here I really doubt would be the slightest bit audible in a controlled AB test. I did read somwhere they did a test where they took a 64 bit file and did all the right things to dither it down to 16 bit (using the finest dithering processes) and compared that to a 64 bit file that just had 48 bits just truncated off and no one could tell the difference!
I think 24 bit is better than 16 bit though in the recording phase. That has been proven to be correct.
Yeah, 24bit recording gives you a lot of safety on the headroom side. But a lot of the 64bit double precision vs. 32bit single precision and dither arguments are based entirely on theory, not the real world. And just because something is there in theory, doesn't mean you can hear it. For the record, I used to think all that stuff was really important, based on stuff I had read. But no one ever gave objective measurements or put anything in context. So I bothered to do the math myself, and then confirmed it with null tests and listening tests. And the truth is people really overestimate all of this stuff. Much of it isn't ever remotely audible at all in the real world. And other stuff might be audible under very specific conditions.
In order, then, to discover the limit of deepest tones, it is necessary not only to produce very violent agitations in the air but to give these the form of simple pendular vibrations. - Hermann von Helmholtz, predicting the role of the electric bassist in 1877.
|
Jim Roseberry
Max Output Level: 0 dBFS
- Total Posts : 9871
- Joined: 2004/03/23 11:34:51
- Location: Ohio
- Status: offline
Re:More questions about 64-bit recording
2011/12/06 19:32:15
(permalink)
And other stuff might be audible under very specific conditions. And for what amounts to a miniscule amount of overhead, you can sum with 64bit dual precision float resolution... and put all thoughts of "what if" to rest... 
|
drewfx1
Max Output Level: -9.5 dBFS
- Total Posts : 6585
- Joined: 2008/08/04 16:19:11
- Status: offline
Re:More questions about 64-bit recording
2011/12/06 19:36:12
(permalink)
Jim Roseberry
And other stuff might be audible under very specific conditions.
And for what amounts to a miniscule amount of overhead, you can sum with 64bit dual precision float resolution... and put all thoughts of "what if" to rest...
Don't bother. There is no "What if?". It's never audible, and it's easily proven so.
In order, then, to discover the limit of deepest tones, it is necessary not only to produce very violent agitations in the air but to give these the form of simple pendular vibrations. - Hermann von Helmholtz, predicting the role of the electric bassist in 1877.
|
Lanceindastudio
Max Output Level: -29 dBFS
- Total Posts : 4604
- Joined: 2004/01/22 02:28:30
- Status: offline
Re:More questions about 64-bit recording
2011/12/06 19:38:45
(permalink)
Not really - In all honesty, you are not going to hear the difference- Lance
pdarg
Hi Jim! Yes, it does clarify things quite a bit.
This means that I may need to get the newest version of WaveLab, since my current version (4.x) cannot handle 64-bit float.
Thanks!
Asus P8Z77-V LE PLUS Motherboard i7 3770k CPU 32 gigs RAM Presonus AudioBox iTwo Windows 10 64 bit, SONAR PLATINUM 64 bit Lots of plugins and softsynths and one shot samples, loops Gauge ECM-87, MCA SP-1, Alesis AM51 Presonus Eureka Mackie HR824's and matching subwoofer
|
Jim Roseberry
Max Output Level: 0 dBFS
- Total Posts : 9871
- Joined: 2004/03/23 11:34:51
- Location: Ohio
- Status: offline
Re:More questions about 64-bit recording
2011/12/06 19:51:15
(permalink)
Don't bother. There is no "What if?". It's never audible, and it's easily proven so. Yeah, and some folks will tell you that you can't hear the difference in timing between 5ms and 10ms... And that you can't hear a difference with audio recorded at higher sample rates. When the CD was invented, 16Bits was all the resolution we'd ever need. You can't hear the difference between different dither algorithms. (If you test, they're certainly not all equal.) Some would argue that you don't need dither (remember the original version of SAW and Bob's arguments?) I've heard it all over the past 20+ years. All things being equal (and the performance hit to sum at 64Bit is miniscule), it would be silly not to take advantage of the extra resolution.
|
keith
Max Output Level: -36.5 dBFS
- Total Posts : 3882
- Joined: 2003/12/10 09:49:35
- Status: offline
Re:More questions about 64-bit recording
2011/12/06 20:12:46
(permalink)
Bub
keith
Chregg
64 bit float is a 53 bit word with an 11 bit exponent, so bouncing down to 24 bit, yeah your loosing data
Any "word length reduction" *should* include dither to remove truncation error. 64-bit float => 32-bit float, 32-bit float => 24-bit integer, etc. all "require" dither
Question about this ^ ... if you have Sonar set to record 32bit wav tracks, and you have dithering turned 'off', does Sonar over ride this and dither anyway since there are no audio drivers that support 32bit or is that 32bit to 24bit conversion done by the audio driver?
Let's say you have a A/D unit that for practical purposes gives you somewhere in the 20-22 bit resolution range... as Jim mentioned, there really is no such thing as 24-bit A/D, at least as far as I know, though I'm not a gear head. so you have, let's say, 22 bits of info coming from the hardware, when that hits the software (either at the driver level, just before the driver, whererever) you need to turn that 22 bits of information into something readable by the host software via the driver... this is where you get 24-bit packed, 24-bit unpacked, big endiand/little endian, int vs. float, etc. I think some drivers support native 32-bit float data packets, but not sure which exactly. Regardless of how the data gets into the software and what format it's in, the software then needs to convert that to a format that will be used within the software itself -- for SONAR that's either 32-bit float or 64-bit float. Now, remember, you only have 22 bits of info coming in, even in the case where you're reading native 24-bit data units... so from the A/D's perspective, the downstream representation of that 22 bits is inconsequential -- you ultimately convert the 22 bits to whatever format is needed (e.g., 32-bit float), and you're always increasing resolution of the data representation in the A/D direction. E.g., 22 bits => 24 bit int, 22 bits => 32-bit float, 22 bits => 64-bit float, etc. Going the other way is a different story. The hardware is typically going to ask for 24-bit integers... so just like dithering down to 16-bit for CD, when going from higher resolution down to 24-bit ints you should apply dither. From a practical perspective, simply truncating your 32-bit float mix right into 24-bit integer hardware should produce no audible effect for 99.9% of humans in 99.9% of use cases. Not applying any dither to a 16-bit output will most definately produce audible nasties... You can do that test yourself.
|
pdarg
Max Output Level: -52.5 dBFS
- Total Posts : 2265
- Joined: 2004/03/26 17:52:53
- Status: offline
Re:More questions about 64-bit recording
2011/12/06 21:35:29
(permalink)
Oops . . . sorry to start a conflict. But I think that I understand now:
1) "64-bit recording" really means using the Sonar X1 64-bit version in a Windows 7 (64-bit) environment; THAT'S the key to releasing the RAM(?)
2) As far as exporting audio to 64-bit using the 64-bit precision switch within Sonar: the debate rages on; in any case, soundcards can only record at 24 (32?) bit, so in reality, I am proposing to record 24-bit audio using a 64-bit OS/application.
Please correct my misconceptions - or confirm that "by George . . . I think he's got it!"
:)
|
SvenArne
Max Output Level: -48 dBFS
- Total Posts : 2719
- Joined: 2007/01/31 12:51:29
- Location: Trondheim, Norway
- Status: offline
Re:More questions about 64-bit recording
2011/12/07 05:39:41
(permalink)
keith
Not applying any dither to a 16-bit output will most definately produce audible nasties... You can do that test yourself.
Funny thing, when I was young and dumb (around 2006) my band and I did a whole album and accidentally omitted dithering when going from 24 to 16 bit. And you CAN hear the truncation artifacts, but only in the very end of the tracks when the tails are fading into absolute silence (plus you'll need headphones turned WAY UP in a silent room). In the end, we decided it wasn't worth the trouble to remaster them. Certainly, no one outside the band ever noticed... Sven
|
drewfx1
Max Output Level: -9.5 dBFS
- Total Posts : 6585
- Joined: 2008/08/04 16:19:11
- Status: offline
Re:More questions about 64-bit recording
2011/12/07 10:08:01
(permalink)
SvenArne
keith
Not applying any dither to a 16-bit output will most definately produce audible nasties... You can do that test yourself.
Funny thing, when I was young and dumb (around 2006) my band and I did a whole album and accidentally omitted dithering when going from 24 to 16 bit. And you CAN hear the truncation artifacts, but only in the very end of the tracks when the tails are fading into absolute silence (plus you'll need headphones turned WAY UP in a silent room).
If you can only hear the truncation if you turn up the volume specifically in order to hear it, that means you can't really hear it. The fact that it's definitely there and would be audible if you turned up the volume means you are can only hear it hypothetically. For this reason, people shouldn't claim something is audible because they can hear it when they turn up the volume - that's cheating. Having said that, if something is perhaps borderline audible, or audible only under very specific circumstances, it does make a certain amount of sense to think about it.
In order, then, to discover the limit of deepest tones, it is necessary not only to produce very violent agitations in the air but to give these the form of simple pendular vibrations. - Hermann von Helmholtz, predicting the role of the electric bassist in 1877.
|
keith
Max Output Level: -36.5 dBFS
- Total Posts : 3882
- Joined: 2003/12/10 09:49:35
- Status: offline
Re:More questions about 64-bit recording
2011/12/07 10:57:11
(permalink)
drewfx1
SvenArne
keith
Not applying any dither to a 16-bit output will most definately produce audible nasties... You can do that test yourself.
Funny thing, when I was young and dumb (around 2006) my band and I did a whole album and accidentally omitted dithering when going from 24 to 16 bit. And you CAN hear the truncation artifacts, but only in the very end of the tracks when the tails are fading into absolute silence (plus you'll need headphones turned WAY UP in a silent room).
If you can only hear the truncation if you turn up the volume specifically in order to hear it, that means you can't really hear it. The fact that it's definitely there and would be audible if you turned up the volume means you are can only hear it hypothetically.
For this reason, people shouldn't claim something is audible because they can hear it when they turn up the volume - that's cheating.
Having said that, if something is perhaps borderline audible, or audible only under very specific circumstances, it does make a certain amount of sense to think about it.
It's not "cheating" to turn up the volume. The aliasing is there in the audible range whether or not the volume is on 1 or 11. The fact that you can't hear it wen the volume is low just means that your hearing is not quite as acute as someone else's. If the aliasing were at -130dB, then that would pracitically be inaudible for just about everybody... unless the aliasing were to be enhanced into the audible range via downstream processing (e.g., extreme compression or whatever). Which brings up another point: if you're going to allow nasties, don't allow them to be enhanced later. It comes down to application: truncating a thrash metal tune probably won't make too much difference to practicaly anybody, truncating a solo cello piece with very delicate reverb tails is a different story.
|
pdarg
Max Output Level: -52.5 dBFS
- Total Posts : 2265
- Joined: 2004/03/26 17:52:53
- Status: offline
Re:More questions about 64-bit recording
2011/12/07 11:13:43
(permalink)
And, one last question:
If one is exporting to 32-bit, is it wise to uncheck the 64-bit precision processing option, since that sends the mix into 64-bit and then back into 32-bit?
|