The No.1 Website for Pro Audio
 Search This Thread  Search This Forum  Search Reviews  Search Gear Database  Search Gear for sale  Search Gearslutz Go Advanced
is worthwhile recording at 32 bits ? Effects Pedals, Units & Accessories
Old 30th August 2008
  #1
Lives for gear
 
Gabriel Sousa's Avatar
is worthwhile recording at 32 bits ?

hello,
all AD are 24bits

is worthwhile recording at 32 bits ?

thanks
Old 30th August 2008
  #2
Gear Guru
 
AllAboutTone's Avatar
 

Not I
Old 30th August 2008
  #3
Lives for gear
 
Martin Kantola's Avatar
 

No.
Old 30th August 2008
  #4
Lives for gear
 
haryy's Avatar
Wow, the area of great confusion (talking about me!heh)

I don't think i can answer your question i will only express some thoughts.

All daws that support 32bit file recording, have also 32bit mix engine at least.
Theoretically(?) recording at 24 bits and processing internally at 32, produces inferior result at the daw output compared to 32bit recording with 32bit processing, otherwise if they were the same, they would not have included the 32bit recording mode from the first place.

Besides this, does any one know if the daw produces dither noise at lower level when working with 32bit files compared to 24bit files, in a 32bit engine daw?? If the answer is yes, then i suppose that 32bit files are better than 24bit

Last edited by haryy; 30th August 2008 at 04:19 PM.. Reason: Wrong expression
Old 30th August 2008
  #5
PDC
Lives for gear
 

Quote:
Originally Posted by haryy View Post
Wow, the area of great confusion (talking about me!heh)

I don't think i can answer your question i will only express some thoughts.

All daws that support 32bit file recording, have also 32bit mix engine at least.
Theoretically(?) recording at 24 bits and processing internally at 32, produces inferior result at the daw output compared to 32bit recording with 32bit processing, otherwise if they were the same, they would not have included the 32bit recording mode from the first place.

Besides this, does any one know if the daw uses less dithering when working with 32bit files compared to 24bit files, in a 32bit engine daw?? If the answer is yes, then i suppose that 32bit files are better than 24bit
With all due respect, I don't think that is a proper application of the word "bits". Don't confuse the math processing ability of a mix engine with the bit depth of a sample. Two different things.

We will not have a true 32-bit converter. We don't really have true, perfect 24-bit converters now. The issue is heat and the fact that it really isn't necessary for music.
Old 30th August 2008
  #6
Lives for gear
 
haryy's Avatar
Quote:
Originally Posted by PDC View Post
With all due respect, I don't think that is a proper application of the word "bits". Don't confuse the math processing ability of a mix engine with the bit depth of a sample. Two different things.

We will not have a true 32-bit converter. We don't really have true, perfect 24-bit converters now. The issue is heat and the fact that it really isn't necessary for music.
Hello PDC, can you tell me what exactly was wrong and why? EDIT * (if it was the last sentence of the previous post i corrected it to what i think it makes better sense)

I know all converters are 24bit (yeah i know they're 20 or something but they all advertise them as 24 so how can you call them?heh) but if you choose 32bits at the daw and you press record, then what do you call that recording? A 32bit recording i suppose. Or a 32bit file of a 24bit recording is better for you? (just joking)
Old 30th August 2008
  #7
Lives for gear
 
staudio's Avatar
 

When recorded from a 24 bit converter the 32 bit file has a string of zeros after the first 24 bits. No additional data is captured because there is no data after the first 24 bits.

When being processed a 24 bit file accumulates more bits because of the mathematical algorithms being applied. This additional data is truncated by the processing system to whatever the mix engine is designed for, 32 bit floating point in many cases.

The only benefit to a 32 bit file is to capture a signal that has DSP applied to it with less data truncation. For example:

A signal is converted using 24 bits and then destructively processed by a plugin (DSP) before being recorded to disk as a WAV file. The plugin (DSP) adds additional bits to the word length of the signal, which will then be truncated to the word length of the destination file format. If the destination file format is 24 bit there will be more data truncation than if the file was 32 bit.
Old 30th August 2008
  #8
Lives for gear
 
haryy's Avatar
Great. If we assume that we'll perform lots of processing after we record a file in a 32bit engine daw, then which is going to have less dither noise in the end? A 24bit file or a 32bit one? Will they have the same level of noise?
Does it depend on the export bit file depth?
Old 30th August 2008
  #9
Lives for gear
 
staudio's Avatar
 

Quote:
Originally Posted by haryy View Post
Great. If we assume that we'll perform lots of processing after we record a file in a 32bit engine daw, then which is going to have less dither noise in the end? A 24bit file or a 32bit one? Will they have the same level of noise?
Does it depend on the export bit file depth?
Read these:

http://www.cadenzarecording.com/imag...tingdither.pdf

Digital Domain - Dither
Old 30th August 2008
  #10
Lives for gear
 
norman_nomad's Avatar
Quote:
Originally Posted by Gabriel Sousa View Post
hello,
all AD are 24bits

is worthwhile recording at 32 bits ?

thanks
bits allow for dynamic range... or the loudest vs the softest signal you're able to capture.

24 bits = 144 db of dynamic range. You'll never, in real life, capture this kind of volume change. Of course you'll never capture 32 bits either.

No 24 bit converter (that I know of) can actually capture this dynamic range because their self noise is louder than 144db range. (Even the best converters spec -120's A weighted noise floor). So you get more like 20 bits of actual dynamic range and the last 4 bits (way way down in the noise floor) is just noise.

Bits do no equal "quality" or "fidelity" they just allow for dynamic change. Mix engines process at 32 bits because it allows you to make drastic dynamic changes in a mix without truncating your 24 bit signal. In a 32 bit float mix engine you can drop all your faders down -70db and then add 70db at the master fader and loose no bits in the process.
Old 30th August 2008
  #11
Lives for gear
 
haryy's Avatar
I'll rephrase: why should i use 32bit files?
Old 31st August 2008
  #12
Lives for gear
 
Sk106's Avatar
 

Quote:
Originally Posted by haryy View Post
I'll rephrase: why should i use 32bit files?
I guess the foremost advantage of that is you will be able to apply much more processing before you get sound degradation - theoretically speaking.
A small advantage on the side is that with floating point you won't have to worry about internal overloads.

But then again, virtually all AD/DAs today are 24 bit, so what you'll be listening to is converted from 32. I know from steinberg (which is my platform of choice) that downsampling from 32 bit floating point to 24 for playback doesn't happen in the host software. It happens in the audio interface and/or the interface drivers. And here things are very different depending on what card it is. Some cards actually dithers down to 24, but God knows with what dithering algo. Some don't dither, they just downsample with a very raw approach. Some uses oversampling and filters, some don't.

So ... gah, I shouldn't have responded to haryy's post in the first place. Kept my big mouth shut
Old 31st August 2008
  #13
Gear Addict
 

Quote:
Originally Posted by Sk106 View Post
I guess the foremost advantage of that is you will be able to apply much more processing before you get sound degradation - theoretically speaking.
A small advantage on the side is that with floating point you won't have to worry about internal overloads.

But then again, virtually all AD/DAs today are 24 bit, so what you'll be listening to is converted from 32. I know from steinberg (which is my platform of choice) that downsampling from 32 bit floating point to 24 for playback doesn't happen in the host software. It happens in the audio interface or the interface drivers. And here things are very different depending on what card it is. Some cards actually dithers down to 24, but God knows with what dithering algo. Some don't dither, they just downsample with a very raw approach. Some uses oversampling and filters, some don't.

So ... gah, I shouldn't have responded to haryy's post in the first place. Kept my big mouth shut
nah, you shoulda answered... i thought it was quite a good one, and it showed up the problems with converters.
Old 31st August 2008
  #14
Gear Addict
 

Hmmmm. I'm not really feeling alot of these analogies. It seems people just never like to know or accept or want to think that the future holds things they just can't mentally grasp yet... Why ? I don't know, but it is the main reason only a few great inventors and world changers exist/existed out there. Their great minds never had tunnel vision and road blocks in them....

True and real 32 bit (or even higher) Converters, DAW's and Plug Ins, therefore recordings SHOULD come out eventually.. and they better for that matter or I will be let down by this crumbling world of crap, and "its good enough" theories.

"Oh no, man can never make it to the moon, oh no there is no reason to go to the moon or to discover more about our universe and space"... hog wash. They all thought the world was flat too, and probably at that time semi smart people theorized on why flat is better than round anyway, and flat is "all we ever need". Cordless phones ? WHY, we don't need them, our phones work great now.

Hell I wish the industry got their head out of their A$$es and standard CD players and CD's jumped up to true 24bit 96k formats, and all DAW's, Plug Ins, converters, etc... jumped up to 32bits. HELL screw that, CD's and their players all jumped to 32bit/192k, and all DAW's, Plug Ins and converters jumped up to 128bit/384k. MAYBE THEN THE MP3 FORMAT WILL BE TRUELY SEEN FOR WHAT IT IS. Audiophiles at homes and in the cars can again enjoy being miles ahead of the "others" who got cheaper formats, etc... And BAMMM , finally boost CD sells again cause there will be a real reason to buy them.. Nothing will sound even close let alone better.

People (yes even the great white paper written on this subject) can sit around and preach to me until their blue in the face about how we can't audibly hear past 96k.....and a dynamic range of 144db will never be heard or utilized, etc...and so it all is unnecessary to move forward in technology this way in the industry. BUT PLEASE everyone, think about this so the world can change FORWARD for the better: It's what you CAN'T hear in music/audio recordings that makes that music/audio so great and x better than y, etc.... Proof of point - I say it like this, neighboring frequencies effecting other frequencies. Inaudible frequencies can play a big part of how a audible frequency is heard and interoperated by the listener. (of course not to the limited minded person though, and they will deny this until their grave).

Kinda like subconscious frames in a video can serious alter ones outlook of how they FEEL about what they just saw. It was what was NOT seen that actually EFFECTED the feelings they felt about the video they just seen. Similar is music/audio, it can be what is NOT heard that can effect EITHER things that are heard, and/or your FEELING of the music/audio and how you interoperated it. Plus if you want to get real deep with it, who is to say ALL WE REALLY HEAR WITH IS ONLY OUR EARS ANYWAY ???? ooooooo, ever think of that ? Spooky huh ? But completely possible and can not be proven wrong or right, we just don't know and have to keep open minds about this. All these so called tests are only based on what we know about the ears and how the ear itself reacts to our brain... and that also is only by what we know so far about it. So how can we be so sure (as we always do and history proves us wrong about our old sureness) that we are 100% right about all this ?

Its also called residual effect. Kinda like this, think about a paper towel. If you put a 1 inch by 1 inch drop of water on that paper towel, the average limited minded person that lets say just happened to know nothing about paper towels and water would think, ok this will only effect a 1 inch by 1 inch spot on that paper towel. Well very quickly you will see that they were wrong and so would they. You would actually see immediately upon placing the water there a area on that paper towel get "effected" (in this case wet) that was much larger (about 3 to 4 times) than that the original source effect's size (the 1 inch by 1 inch water drop). NOW even more proof, WAIT for a hour and come back to the paper towel. You will find that now not only is the effected spot area much larger BUT the whole towel has been effected by "residual effect" and metaphoric "neighboring frequencies". (in a sense).

Problem is, is if none of us believe in good reasons to move forward in this industry with higher bit depth, sample rates, and quality in general, then the companies that are in charge will not even TRY to invent ways to make it possible, then affordable and possible, then affordable possible AND efficient. Which is how it has to happen. They read threads like this and go, "yea, no reason to invent that new sh*t yet, the people wont buy it, and even more funny don't think they need it"..."lets just keep making the SAME products with more bullsh*t features and with cheaper designs and charging them more because it is new, but no new R&D had to go into it"......

SCREW THAT. I am always for moving things forward to make things better with a positive open mind. That other type of thinking keeps things from progressing. Sure, your right about what you say, BUT point is, is, maybe there are ways around them limitations that we can't see now how to, but when we do, that IN TURN makes the new invention/things/true 32 bit recordings actually pay off and work... THEN we all say, "god how did we get by before with out that"... It just takes one mental breakthrough to allow the rest to come with ease.

Now with all that said, don't get me wrong, I think all the statements on this thread are coming from very intelligent and knowledgeable people that I respect very much. BUT we need to remember, sometimes it is that "intelligence" and "knowledge" that will block us from seeing the insane possibilities of the future of what we can do and accomplish. And then it takes a "considered crazy" super eccentric genius like Einstein to break us into the new way of thinking and seeing things possibilities and reasons, etc...
Old 31st August 2008
  #15
Lives for gear
 
haryy's Avatar
Quote:
Originally Posted by Sk106 View Post
I guess the foremost advantage of that is you will be able to apply much more processing before you get sound degradation - theoretically speaking.
A small advantage on the side is that with floating point you won't have to worry about internal overloads.
So, you say it's good to use the 32bit file system in place of the 24bit even in a 32bit engine daw?
Old 31st August 2008
  #16
Lives for gear
 
norman_nomad's Avatar
Quote:
Originally Posted by Infa View Post
It's what you CAN'T hear in music/audio recordings that makes that music/audio so great
LoL.

Thank you. Now I have a great comeback line the next time a client has a problem with a mix.

heh

Seriously though dude.... your hearts in the right place but you need to become familiar with some basic bit depth a sample rate definitions to understand why converters stopped at 24 bits and why there's limited to zero value in capturing audio at 32bits on the way in.

To lift from your paper towel analogy, asking for a 32 bit converter is like asking for 3 paper towels to sop up your drop of water when clearly 1 will do the job more than adequately.
Old 31st August 2008
  #17
Lives for gear
 
Sk106's Avatar
 

Quote:
Originally Posted by haryy View Post
So, you say it's good to use the 32bit file system in place of the 24bit even in a 32bit engine daw?
I used to feel alot more convinced about this, but not anymore. We're talking about so many factors that interacts to create too many unknown results here.

The theoretical analogy I was reaching for, is when you adjust the sound with eq and other fx. That results in an increase of the wordlength. A file may be recorded in 24 bit, but add a couple of bands of eq and you may be looking at an average wordlength of 30 bits. If you maintain 24 bits all the way, any bits above 24 will be brutally amputated - not rounded. This leads to an incremental decrease of audio fidelity. That's not just theory, you can hear that after plenty of processing. So, if you have a file recorded in 24 bits, and adjust the sound of it within a 32 bit floating point environment, then you will have extra wordlength to draw upon and the wordlength won't be amputated. Especially not with floating point math.

Then again, whatever happens in a 32 bit floating point host program will still play back amputated or dithered to 24 bits. So ...
Old 31st August 2008
  #18
Lives for gear
 
norman_nomad's Avatar
Quote:
Originally Posted by Sk106 View Post

Then again, whatever happens in a 32 bit floating point host program will still play back amputated or dithered to 24 bits. So ...
All of which you would never hear in the real world because even undithered truncated 24bit audio has quantization distortion buried far below the noise floor of your converters.
Old 31st August 2008
  #19
Lives for gear
 
Sk106's Avatar
 

Quote:
Originally Posted by norman_nomad View Post
All of which you would never hear in the real world because even undithered truncated 24bit audio has quantization distortion buried far below the noise floor of your converters.
Christ, I'm gonna throw up on this bit ****
Old 31st August 2008
  #20
Lives for gear
 
norman_nomad's Avatar
Quote:
Originally Posted by Sk106 View Post
Christ, I'm gonna throw up on this bit ****
It is a puke worthy subject. heh
Old 31st August 2008
  #21
Quote:
Originally Posted by Infa View Post
....HELL screw that, CD's and their players all jumped to 32bit/192k, and all DAW's, Plug Ins and converters jumped up to 128bit/384k.....Now with all that said, don't get me wrong, I think all the statements on this thread are coming from very intelligent and knowledgeable people that I respect very much. BUT we need to remember, sometimes it is that "intelligence" and "knowledge" that will block us from seeing the insane possibilities of the future of what we can do and accomplish. And then it takes a "considered crazy" super eccentric genius like Einstein to break us into the new way of thinking and seeing things possibilities and reasons, etc...
The thing about einstein is that he knew where to push to make that break. There's not much point in developing a giga-bit converter, if it's functionally no better than a 24bit one.

As norman_nomad pointed out, you don't understand the whole bit depth issue very well - a PERFECT 24bit converter can capture more dynamic range than it's possible for a human to hear - the threshold of the quietest sound, all the way up to pain levels. In fact, a well designed 16bit converter will still give you this, at least to the point where the noise floor limitations of the format are masked by the background noise present in the (even ultra quiet) listening environment. at the moment, the best 24bit converters are limited by the analogue stage, not the digital part.

what you should really be calling for is better designed converters in consumer level gear...that'd have far more significant effect than any sort of fanciful mega-bit depth converter.
Old 31st August 2008
  #22
Lives for gear
 
staudio's Avatar
 

Quote:
Originally Posted by psycho_monkey View Post
better designed converters in consumer level gear...that'd have far more significant effect than any sort of fanciful mega-bit depth converter.
I want an iPod with a Lavry DAC built in!!! I might even settle for a Duet DAC built into my iPod!!!
Old 31st August 2008
  #23
Lives for gear
 
Iggy Poop's Avatar
 

Excuse me if I sound uneducated on the subject (cuz I am) but if CD and mp3 files are 16 bit then why does it make such a difference to record at a higher bit rate? Is the whole thing about file degradation a proven fact or is it a theory? Again, I don't know that's why I'm asking.
Old 31st August 2008
  #24
Motown legend
 
Bob Olhsson's Avatar
 

Quote:
Originally Posted by Iggy Poop View Post
... Is the whole thing about file degradation a proven fact or is it a theory...
It's a proven fact.
Old 31st August 2008
  #25
Quote:
Originally Posted by Bob Olhsson View Post
It's a proven fact.
Yes indeed. Aren't there some 32bit DACs out there now?
jeff
Old 31st August 2008
  #26
Motown legend
 
Bob Olhsson's Avatar
 

My point is that digital signal processing degrades the signal and storing longer words after processing will degrade it less. There's no advantage to capturing an analog dignal using more than 24 bits.
Old 31st August 2008
  #27
Quote:
Originally Posted by Bob Olhsson View Post
My point is that digital signal processing degrades the signal and storing longer words after processing will degrade it less. There's no advantage to capturing an analog dignal using more than 24 bits.
Wouldn't the extra 8 bits give you not only increased dynamic range
but also finer gradation of amplitude as well? I don't know much
material that gonna swing 200db ;-) .. but maybe the finer
gradation of the dynamic range would be higher fidelity?

That assumes you're going into a 32 or 64 pipeline as well.

jeff
Old 1st September 2008
  #28
Gear Addict
 

Quote:
Originally Posted by norman_nomad View Post
LoL.
Thank you. Now I have a great comeback line the next time a client has a problem with a mix.

Seriously though dude.... your hearts in the right place but you need to become familiar with some basic bit depth a sample rate definitions to understand why converters stopped at 24 bits and why there's limited to zero value in capturing audio at 32bits on the way in.

To lift from your paper towel analogy, asking for a 32 bit converter is like asking for 3 paper towels to sop up your drop of water when clearly 1 will do the job more than adequately.
Quote:
Originally Posted by psycho_monkey View Post
The thing about einstein is that he knew where to push to make that break. There's not much point in developing a giga-bit converter, if it's functionally no better than a 24bit one.

As norman_nomad pointed out, you don't understand the whole bit depth issue very well - a PERFECT 24bit converter can capture more dynamic range than it's possible for a human to hear - the threshold of the quietest sound, all the way up to pain levels. In fact, a well designed 16bit converter will still give you this, at least to the point where the noise floor limitations of the format are masked by the background noise present in the (even ultra quiet) listening environment. at the moment, the best 24bit converters are limited by the analogue stage, not the digital part.

what you should really be calling for is better designed converters in consumer level gear...that'd have far more significant effect than any sort of fanciful mega-bit depth converter.
Points well taken guys. I do understand where you are coming from. And yes, I will admit you guys and ALOT of others are well beyond me as far as really understanding all this as far as the actual technical white paper is concerned.

Sometimes break throughs and serious revelation advancements in technologies like this come from some things and ideas so simple its why no one seen them before. It was TOO simple and obvious of a answer.

Maybe I should word it like this... Maybe the advancement should be more of a "back door" theory, coming in from the backside, and would be more of inventing a way for the other architecture (things besides bit depth) in a converter or DAW or plug in to actually make it so they DO benefit from 32 bit depth and higher rates, etc... (instead of change the water drop, change the paper towel(s) )

Or something like that. Just something we are not thinking of now, because we only know what we learn and know so far. Then we WOULD be able to hear a difference because the back end of things finally was figured out how to take advantage of that extra depth we "didn't need" before, but yet make it something we now seen we were missing. Basically, take advantage of something it can't utilize now/yet, to in turn make it something that actually audibly does make a difference to us.

Because, here's my point that I must stand by to that gives me a HUGE hint that there is SOMETHING to what I say: if 16bit was already beyond what we can detect already, then please tell me why I can tell a difference in my 16bit recordings from my 24bit recordings when using identical audio source ? (and the difference is, 24bit sounds better). Same goes with 48k... Everyone knowledgeable in this field says 48k was/is beyond our audible detection, so 96k is just ridiculous overkill... well..... When I run my same tests I can hear a difference in my 96k recordings vs my 48k recordings.(and the difference is 96k sounds better). So WHY then if supposedly we can't detect past 16 bit 48k can I and others hear a difference when the ranges are upped past supposed "overkill" ?

I will tell you why, its the stuff you can't hear, effecting the stuff you can hear. YES it is minor, but it creates that magic "sheen" that no one can put a finger on. Almost something you feel vs something you hear.

Haven't you ever taken a audio clip/sound and at lets say 9k frequency on a precise EQ with a very narrow Q on, and lets say boosted or cut the frequency enough to hear a moderate difference (metaphor to lets say adding a moderate amount of Bit depth of what we cant detect a difference on)..... Well did you ever notice how that boost or cut you did at 9k somehow made a audio illusion (or maybe not a illusion) that made other frequencies FAR away sound different now ? That is called "neighboring frequencies effecting other frequencies". And whether it is a audio illusion or not doesn't matter. Point is, is, it caused you to interpret them other non-boosted/cut frequencies differently that were FAR away from the actual boosted or cut frequency.

I feel, that is what is happening when people say 16Bit is beyond our human ear detection, BUT YET me and several other people can (most of the time) blindly tell a difference between 16bit and 24bit recordings and the 24bit recordings sounds better. (most of the time)... and same goes for 48k vs 96k...

Don't you think as time goes on, the same would happen from supposed over kill 24bit recordings to 32 (or higher) bit recordings ? I do. And I think the reasons are just unexplainable now because we are limiting ourselves to our ideals of what we think we understand about the human ear. Especially if we approach it in a way I described above to make it so the architecture finally did take advantage of higher bit depths and sample rates so that it was a obvious difference even to the limited human ear ALONG with subconscious feelings of the music, etc...
Old 1st September 2008
  #29
Gear Addict
 

Quote:
Originally Posted by jmarkham View Post
Wouldn't the extra 8 bits give you not only increased dynamic range
but also finer gradation of amplitude as well? I don't know much
material that gonna swing 200db ;-) .. but maybe the finer
gradation of the dynamic range would be higher fidelity?

That assumes you're going into a 32 or 64 pipeline as well.

jeff
This is my point exactly. Using that same analysis, maybe it will help several other things besides dynamic range as well...

Very good point Jeff.
Old 1st September 2008
  #30
Motown legend
 
Bob Olhsson's Avatar
 

Quote:
Originally Posted by jmarkham View Post
Wouldn't the extra 8 bits give you not only increased dynamic range
but also finer gradation of amplitude as well? ...
Not really. The best converters can only put out around 21 or 22 bits. Digital signal processing turns that 21 to 22 bits into between 56 and 80 bits. At that point 32 bit or even 64 bit float files are worthwhile for intermediate storage of the processed audio.
Post Reply

Welcome to the Gearslutz Pro Audio Community!

Registration benefits include:
  • The ability to reply to and create new discussions
  • Access to members-only giveaways & competitions
  • Interact with VIP industry experts in our guest Q&As
  • Access to members-only sub forum discussions
  • Access to members-only Chat Room
  • Get INSTANT ACCESS to the world's best private pro audio Classifieds for only USD $20/year
  • Promote your eBay auctions and Reverb.com listings for free
  • Remove this message!
You need an account to post a reply. Create a username and password below and an account will be created and your post entered.


 
 
Slide to join now Processing…
Thread Tools
Search this Thread
Search this Thread:

Advanced Search
Similar Threads
Thread
Thread Starter / Forum
Replies
Professor / High end
124
Gabriel Sousa / So much gear, so little time
12
16/44.1 / Mastering forum
39
Stubacca / So much gear, so little time
35

Forum Jump
Forum Jump