Topic: Bits v. kHz

[javertim]

So, I have recently purchased an Edirol R-09HR and am anxiously awaiting its arrival.  I am trying to decide which format to record in, and having only recorded with MD and Hi-MD in the past, the whole 24 Bit/96 kHz thing confuses me.  What should I be more concerned with: bits or kHz?  What would sound better: 16 bits/96 kHz or 24 bits/44.1 kHz? ... Also, would it be pointless to make recordings in 24 bits and then burn the files to audio CD (as I believe "CD quality sound" is only 16 bit)? ... I would very much appreciate if someone would explain this to me in practical terms.

[F.O.Bean]

bits are WAY MORE IMPORTANT!!! I have been recording in 24/44.1k and LOVE IT!

[stantheman1976]

Bits will mean more when editing your recordings.  The higher the bit depth the better your recording will fare when altering it during post.  When amplifying a recording you will most definitely notice a difference between 16 and 24 bit.  With 24 you can amplify much more before noise becomes noticeable.

[javertim]

What happens if I record in 24-bit, then master, and then burn the file to an audio CD?  Will I loose something of the original 24-bit recording since CDs themselves are only 16-bit?  And what difference do kHz make in recordings?

[rhinowing]

What happens if I record in 24-bit, then master, and then burn the file to an audio CD?  Will I loose something of the original 24-bit recording since CDs themselves are only 16-bit?  And what difference do kHz make in recordings?

24-bit mastered to a CD will sound better than 16-bit.

[page]

What happens if I record in 24-bit, then master, and then burn the file to an audio CD?  Will I loose something of the original 24-bit recording since CDs themselves are only 16-bit?  And what difference do kHz make in recordings?

khz samples translate to frequencies recorded.

44.1khz tops out around 20,000, while 48k tops out around 23,000. Cymbals are found around the 14,000-18,000 range if I remember correctly.

[sunjan]

What happens if I record in 24-bit, then master, and then burn the file to an audio CD?  Will I loose something of the original 24-bit recording since CDs themselves are only 16-bit?  And what difference do kHz make in recordings?

khz samples translate to frequencies recorded.

44.1khz tops out around 20,000, while 48k tops out around 23,000. Cymbals are found around the 14,000-18,000 range if I remember correctly.

An important part of this discussion is also the playback equipment you use. If you record high frequencies, make an effort to use good playback speakers/headphones etc to enjoy the higher quality.

[DSatz]

javertim, if your "ultimate delivery medium" is a CD, then one thing is for certain: If you make your live recording at a sampling rate higher than 44.1 kHz, you will eventually need to convert it down. The processes available for this conversion range somewhat in quality, and can never improve the sound quality of recordings except by some kind of wacky accident; at best they can be qualitatively neutral while adding a little bit of noise.

Valid generalizations about which approach is best are therefore hard to make. However, I've noticed that this doesn't stop people from making broad statements; it doesn't even seem to slow them down. I view such advice with skepticism. It may be valid for the particular equipment or software that a given person is using, or it may just be what they'd like to believe they're hearing (or can hear), or it may be some of each. But if someone says "It sounds better if you start at 96 kHz" they can't possibly be speaking for all 96 kHz equipment, or all ways of converting 96 kHz recordings down to 44.1; that's simply not possible. And yet there is so much of that kind of careless talk. People only discredit themselves by stating that type of opinion.

There's one definite practical and (potentially) audible advantage to 24-bit live recording even when your eventual "delivery format" will be 16-bit. The advantage is that, with suitable equipment used correctly, you can set your recording levels conservatively. You can avoid accidents if an unexpectedly loud sound occurs, without giving up the full dynamic range that 16-bit recording offers.

What I mean is, if I'm recording live at 16/44.1 and I'm going to deliver a CD to the client, then I really will be trying to get the peak recorded levels to land somewhere between -1  dBFS and maybe -3 or -4 dBFS. If they're lower, people can't help imagining that they might have guessed more accurately themselves, while if they're higher, then there's distortion (overload). And that makes for a somewhat tense situation because, what if I'm wrong? I'll have to change the levels during the recording, keep careful notes, and compensate afterwards during the transfer to the final product. That's a fair amount of extra hassle and the dynamic range of the product won't be ideal.

With 24-bit live recording and 16-bit delivery, though, I can allow (say) 7 or 8 dB headroom. If the performers want to surprise me, they can go right ahead. In the end, either way I'll reset the gain and dither down to 16-bit, and come out smelling like a rose every time. I like that.

--best regards

[bhakti]

DSatz, i love you... i have never had a better explanation of all this dithering bizness...

so if you were using an Edirol R-09 for example, would you always record [in generally unideal situations where levels can't be checked constantly] in 24/44.1 or just use 16? love to hear what you have to say... i think you may have already said it, but i'm not entirely sure...

thanks again for the very useful explanation.... always happy when i see one of your posts because i know i will learn something....

+T brother...

[sygdwm]

24/44.1 > run conservatively > raise gain in post > dither to 16 bit

[javertim]

Wow. Thank you everyone for your input! I think I'll go ahead and record in 24/44.1 so that I do allow myself the headroom that you are talking about, DSatz. ... As far as what I use to listen to my music, I generally use either my Sennheiser HD-280-Pros or my Shure se420s, although I've been thinking about upgrading the former pair to an open pair (maybe the Sennheiser HD600s).

[Javier Cinakowski]

I generally use either my Sennheiser HD-280-Pros or my Shure se420s, although I've been thinking about upgrading the former pair to an open pair (maybe the Sennheiser HD600s).

Try the DT880 by Beyerdynamic.  They are open air and more acurate than the Senns...  Though many folks prefer the coloration of the Senns...

[evilchris]

24/44.1 > run conservatively > raise gain in post > dither to 16 bit

ditto.

[sml42]

I mostly record at 24/44.1, for various reasons:

- my ultimate target is a CD
- I don't have state-of-the art PC, so the extra cpu overhead of downsampling hurts. Much easier for me to stick to one samplerate
- I currently run a MT2496, which doesn't seamlessly autosplit. 24/96 gives a file break approx every hour, unacceptable to me (24/44.1 is usually ok)

My previous recorder (JB3) only did 16 bit. The difference going to 24 bit was like night and day.

best regards,
stephen

[yug du nord]

My belief (although it may be wrong) has been that if your goal is to end up with a CD, that 16/44.1 is best.  Nearly everyone claims that 24/?? is better...  and that might be true...  if it's not going onto a CD.  I've read somewhere at some point that there is no dithering scheme that is as accurate as just recording in 16bit.  I understand about the headroom that 24bit gives you, and the advantages of 24bit or higher for post work, but I've always thought that a "true" 16/44.1 recording is technically superior to a 24/?? recording that's dithered down.  Like I said...  I might be completely wrong on this issue...  but DSatz might have given part of my belief some creditability.  Maybe...  DSatz... does this make any sense, or am I just talking out of my rear-end?  I've never recorded in 24bit....  so I'm not the best person to make these claims.

[DSatz]

Hi, uncleyug. Let's try to reason through this carefully with a practical example.

Let's say I set up a pair of microphones to record a concert. I've been to a rehearsal beforehand, so I know what to expect as far as the sound levels are concerned--and as usual, I've allowed 3 or 4 dB of headroom for last-minute enthusiasm on the part of the musicians. From then on, the dynamic range of the recording is up to whatever the musicians decide to do. Whether they use that 3 or 4 dB at the top or not, the recording might be sonically first-rate or not, depending on all the usual variables.

However, 3 - 4 dB is easily audible as a difference in loudness--and even a 1 dB difference or less in playback levels can have a real impact on people's opinions about the quality of a recording. Hi-fi salespeople have known this for years: If you can even sneak the level up just 1/2 or 3/4 of a dB on the pair of loudspeakers you want the customer to buy, people start hearing "more detail" and "more emotion" and "better imaging" and "more soundstage depth"--all sorts of qualitative differences, rather than noticing, "Gee, it sounds a tiny bit louder than it did a moment ago."

So if there's 3 or 4 dB difference between my recording and (say) one that the client got from another engineer a month before, then--well, my recording might be worlds better sounding if the client would only make the comparison with the volume levels set scrupulously equal in playback. But the average musician who's not an engineer will simply listen to one recording, listen to the other, and make a snap judgment: My recording just doesn't have the "life" and "energy" of the other (louder) recording.

By the way, this kind of situation has led many good people to firm but mistaken conclusions about all kinds of things in audio. And those people really have heard what they say they've heard--but the uncontrolled conditions make it quite impossible for them to know why things sounded as they did.

I mean--on one level I really can't argue with the client; he pops in the other guy's CD, and he pops in mine, and the other guy's CD sounds better. If there's something I can do to make my work sound better in such a comparison, the client naturally assumes that I should already have done it--what am I waiting for?

So if I'm in one of those situations where I absolutely must make things sound as good as possible in a comparison, I will recopy my recording so that it peaks at -1 dB or even -1/2 dB below digital full scale. And the moment that kind of signal processing comes into the picture, it's better if I'm starting from a 20- or 24-bit live recording.

Think it through: If I start from a 16-bit live original, then copy that in order to raise the levels by 3.5 dB, then in the hypothetical best case, the noise floor of the result would be 3.5 dB higher than before. But if I'm starting from an 18- or 20- or 24-bit live recording, then the noise floor of the recording doesn't increase at all when I boost the gain by 3.5 dB. The resulting CD will in fact have 3.5 dB wider dynamic range than if I boost a 16-bit recording by the same amount.

If you want to make the comparison more realistic by considering dither, then you can simply consider the dither itself to be the digital noise floor of the recording. The one which starts from a dithered 16-bit original will still be 3.5 dB noisier than the one that starts from dithered 18, 20 or 24 bits.

What mitigates this considerably in many cases is the acoustical noise in the recording venue--but that isn't constant at all times or all frequencies. Our ears and brains are most sensitive in the 2 - 5 kHz region and there, every dB and every bit really counts sometimes, at least in the classical recording that I mainly do.

--best regards

[yug du nord]

^^^Once again man, you've schooled me in a fashion that this simpleton can understand!  Thank you!!!  Makes complete sense.  It's always a pleasure to catch onto a thread that you're involved with.  +T every time I see ya!

[Lil Kim Jong-Il]

Even if today's target is CD, someday you will likely upgrade your playback system to 24-bit playback.  I record at 24/96 and use DVD.  Although I have not heard a difference between 96k and 48k/41k, the difference between 24 and 16 bit is clear on my system and storage is cheap.  Processing to 16-bit is not that big a deal and even with a low powered machine you need to do it only once for a show you intend to archive and circulate.  

If you are mastering down to CD for the car, you aren't likely to perceive any real difference between a native 16-bit version and a dithered 24-bit version.  The ability to capture a high SNR recording with conservative levels at 24-bit is priceless when you are in a situation where monitoring the signal level is not possible and you need to be concerned about large transient signals.  

I would even tend to question if you can get such a thing as a true native-16 bit copy from a 24-bit recorder.  The ADC or firmware may be downsampling to 16-bit a 24-bit native sample stream anyway between the sampler and the wav file.

[sml42]

But if I'm starting from an 18- or 20- or 24-bit live recording, then the noise floor of the recording doesn't increase at all when I boost the gain by 3.5 dB.

To be pedantic, the noise floor does increase, it's just still way way below the 16-bits offered by CD

Slightly more seriously, there was another discussed on these forums some time ago. Think it was originally started by TRB, and damned if I can find it, but the gist of it is as follows: even if you were running at 24-bit, and could set your levels perfectly (ie. coming close to 0dB but no clipping) you shouldn't. Instead you ought to aim to peak at -6dB, possibly even -12dB, and add gain in post-processing. The rationale given was, traditional class A amplifiers have more distortion when you crank them hard, so back it off a little - you have extra bits for the dynamic range. The assumption being, the preamp exhibits such behaviour. Now, that's how the thread went, and for the record I'm not sure I buy into it totally; in practice the real-world means you won't come close to 24-bit, I'd be surprised if my MT2496 gave anything better than an effective 18-bits, so no way would I want to be peaking at -12. Maybe if I had some higher-end gear... maybe not.

[sml42]

However, 3 - 4 dB is easily audible as a difference in loudness--and even a 1 dB difference or less in playback levels can have a real impact on people's opinions about the quality of a recording

On the basis that 'louder is better' I will also point out that a small amount of compression can help achieve this... that is of course a whole differnt discussion, so I will shut up now

[Lil Kim Jong-Il]

that is of course a whole differnt discussion, so I will shut up now

light fuse, step away? 

[digifish_music]

However, 3 - 4 dB is easily audible as a difference in loudness--and even a 1 dB difference or less in playback levels can have a real impact on people's opinions about the quality of a recording

On the basis that 'louder is better' I will also point out that a small amount of compression can help achieve this... that is of course a whole differnt discussion, so I will shut up now

This YouTube video on the 'Loudness War' presents one side of the story...

http://www.youtube.com/watch?v=3Gmex_4hreQ

http://www.youtube.com/watch?v=xkkqsN69Jac

Being in the music production business I am interested to look at how many commercial tracks are mastered, and it's not pretty. That said, I regularly see tracks that look like they have been destroyed by over-compression (judging from the audio-editor peak-view), and yet they sound good at all listening volumes. It does however depend on the style of music.

digifish

[boyacrobat]

loadness war is by grand design.
portals that are dimensional found in sound
must be closed,squashed .

this task has been given to the mastering engineers.
blame them more than the producer of the album, but blame them as well.
most still sleep the big sleep.


g

[DSatz]

just sml, thank you for your postings. Compression is certainly an option to get higher long-term average levels onto the transfer of a recording. When to use it, if ever, is a whole other discussion. But from a technical standpoint, if you're going to use it, it's unquestionably preferable to start out with more than 16 bits in the live recording--otherwise the noise will "breathe" or "pump" from moment to moment as the compressor changes the overall gain.

Noise that rises and falls rapidly in level--even if it remains at fairly low levels overall--is far more noticeable than constant noise of an otherwise similar character and amount. This is why we record with linear PCM rather than logarithmic quantization, which would undoubtedly save bits. The problem with logarithmic quantization is that the noise floor rises and fall with the instantaneous levels of the recorded material, causing a kind of modulation noise. Modulation noise was one of the main shortcomings of analog tape, particularly at slower recording speeds (and to some old-timers I've known, anything slower than 30 ips was considered a slow recording speed).

--When I said that "if I'm starting from an 18- or 20- or 24-bit live recording, then the noise floor of the recording doesn't increase at all when I boost the gain by 3.5 dB" I think I was closer to the truth than you seem to think. I was referring to the noise floor of the resulting 16-bit CD, and I think you were referring to the noise floor of the 24-bit recording that's been boosted 3.5 dB--that difference in assumptions would explain your reply, and make us "both right."

But to be totally honest I must admit that there would be a very, very slight increase even in the noise floor of the 16-bit recording--a few thousandths of a dB. It wouldn't even be measurable, let alone audible, but it could be calculated if the hypothetical recording situation were to be specified in enough detail.

--best regards

P.S.: The only point of leaving some headroom in any recording is to avoid overload. It isn't desirable for its own sake. Headroom of (say) 3 - 4 dB or more that is never, ever used is simply wasted dynamic range. It only leads to increased noise, because listeners have to crank up the volume on their playback systems. Personally I'm happiest if my recordings peak at around -1 or -2 dB.

[ScotK]

Of course the 1-bit folk might have a slightly different answer to this question ...
Thanks for the nice discussion!
scot

[DSatz]

ScotK, what do you mean?

A 1-bit A/D converter can be considered as a roundabout way to get to a multi-bit result. Maybe that's not the whole picture but for most practical purposes other than DSD recording (now a thoroughly dead technology), that's how it is used.

It's an interesting technology, very nice in several respects, while in others, not so much. But once you convert its results into the multi-bit realm (16, 24, whatever), the same limits and concepts apply as with conventional multi-bit A/D converters.

--best regards

[ScotK]

Agreed, DSatz.  Just thinking that in the DSD realm, it's low # of bits and high sampling freq - sort of the opposite
priority we're stating here for "traditional" recording methods.

And yes, once you convert the 1-bit stream to something more genreally listenable, you're right that the same arguments apply.

And out of curiosity, why do you call DSD a thoroughly dead technology? Just that it hasn't garnered much
industry acceptance?  Aren't SACD (not that they are particularly popular) 1-bit?

s

[Gutbucket]

Noise that rises and falls rapidly in level--even if it remains at fairly low levels overall--is far more noticeable than constant noise of an otherwise similar character and amount. This is why we record with linear PCM rather than logarithmic quantization, which would undoubtedly save bits. The problem with logarithmic quantization is that the noise floor rises and fall with the instantaneous levels of the recorded material, causing a kind of modulation noise. Modulation noise was one of the main shortcomings of analog tape, particularly at slower recording speeds (and to some old-timers I've known, anything slower than 30 ips was considered a slow recording speed).

Interesting nugget of knowledge there.
Thanks & +T.

[DSatz]

ScotK, now I see what you were referring to, and of course you're completely right. 1-bit recording is a little like real estate in Tokyo--there's no more room for lateral expansion, so the only direction left to build in is upward. (Or maybe downward, but the mole people already have that market cornered.)

If you constrain an audio data stream by definition to just a single bit of precision, then oversampling becomes necessary for anything more complex than Morse Code messages--thus the sampling frequencies in the (low) MHz range. How the sampling rate of a 1-bit system relates to the audio bandwidth and dynamic range of the system isn't quite as straightforward as it is with conventional multi-bit linear PCM.

--best regards

[ScotK]

DSatz,

I was re-reading some of the 1-bit stuff today and it struck that what you really want is not binary logic, but
tri-state logic, which people have worked on for some time: up, down, or same. I guess with such fast sampling
rates, it doesn't matter that same is represented as up down up down ..., but it's got to be better if you really
did have a single tri-bit instead, though!
s

[DSatz]

ScotK, yes--it's like in marching band, where the saying used to be, "if you forget your part, trill." When a one-bit converter wants to indicate a steady state, it has no inherent way to do so, and has to (literally) dither rapidly and randomly between "Go up!" and "Go down!".

The requirement for randomness is one of the Achilles' heels of this approach, since in reality the idling noise of any particular converter will have a response pattern which makes the dithering effect less than perfect (creating low-level birdies, whistles and/or modulation noise). It's a real flaw, though outweighed in some people's opinion by other advantages that the method undoubtedly can have.

I think it's fair to say that DSD's biggest "advantage," however, was that the patents had run out on most proprietary technology for multi-bit linear PCM and red-book CD audio. The manufacturers needed something new purely for their own commercial reasons, whether consumers wanted it (mostly they didn't) or whether it was sonically and technically superior or not. Linear, multi-bit PCM and the CD audio medium had become democratized; the genie was out of the bottle.

Having rubbed that bottle for years, and turned it every which way and shaken it and even swung some hammers trying to open it up, I'm not sorry about which side my sympathies are on.

--best regards

[ScotK]

DSatz,

Ah, hence the "dead" comment.  Makes sense now!

thanks,

s

[illconditioned]

javertim, if your "ultimate delivery medium" is a CD, then one thing is for certain: If you make your live recording at a sampling rate higher than 44.1 kHz, you will eventually need to convert it down. The processes available for this conversion range somewhat in quality, and can never improve the sound quality of recordings except by some kind of wacky accident; at best they can be qualitatively neutral while adding a little bit of noise.

Valid generalizations about which approach is best are therefore hard to make. However, I've noticed that this doesn't stop people from making broad statements; it doesn't even seem to slow them down. I view such advice with skepticism. It may be valid for the particular equipment or software that a given person is using, or it may just be what they'd like to believe they're hearing (or can hear), or it may be some of each. But if someone says "It sounds better if you start at 96 kHz" they can't possibly be speaking for all 96 kHz equipment, or all ways of converting 96 kHz recordings down to 44.1; that's simply not possible. And yet there is so much of that kind of careless talk. People only discredit themselves by stating that type of opinion.

There's one definite practical and (potentially) audible advantage to 24-bit live recording even when your eventual "delivery format" will be 16-bit. The advantage is that, with suitable equipment used correctly, you can set your recording levels conservatively. You can avoid accidents if an unexpectedly loud sound occurs, without giving up the full dynamic range that 16-bit recording offers.

What I mean is, if I'm recording live at 16/44.1 and I'm going to deliver a CD to the client, then I really will be trying to get the peak recorded levels to land somewhere between -1  dBFS and maybe -3 or -4 dBFS. If they're lower, people can't help imagining that they might have guessed more accurately themselves, while if they're higher, then there's distortion (overload). And that makes for a somewhat tense situation because, what if I'm wrong? I'll have to change the levels during the recording, keep careful notes, and compensate afterwards during the transfer to the final product. That's a fair amount of extra hassle and the dynamic range of the product won't be ideal.

With 24-bit live recording and 16-bit delivery, though, I can allow (say) 7 or 8 dB headroom. If the performers want to surprise me, they can go right ahead. In the end, either way I'll reset the gain and dither down to 16-bit, and come out smelling like a rose every time. I like that.

--best regards

I measured the converter noise on the Edirol R09 (the original, not the HR) and it was about 15bits accurate.  That is, if I short the line in terminals and set the gain to the lowest it can go (#1/30), and record "silence" at 24bits/44.1k, I get noise (ie., random variation) in the last nine bits!  (Measured using Wavelab 5.0 "bit meter".)  So this thing is only getting an SNR of about 92dB, not even 16 bits, let alone 24.  This is consistent with the specs of the chip inside, which reports 92dB SNR.

So, what should I be doing?  Should I record in 24 bit anyway or 16 bit?

Thanks for all the help...
  Richard

[Gutbucket]

^^^^
I've been wondering about this since I started using R-09's a couple years back.  Since I've never measured it or reached an airtight conclusion, I've run them at 24bit 'just in case' since the storage cost of the extra data to cover my ignorance seems reasonable.  Obviously the performance of a $350 flash recorder isn't going to actually approach anything like the full theoretical range of a 24bit sample when even well engineered, high-dollar equipment usually tops out around 20+ bits or so of actual usable range [so I'm told]. At the same time, I know that signal can be discerned below the noise floor which is part of the reason for dither. At the risk of exposing my mental lack of bit depth farther, I ask how deep into that noise floor is it reasonable to record?  1 bit? 3 bits? 6 bits? a dollar? 

If I'm getting a usable extra bit or two, I feel the extra storage needed is well worth the additional 6 or 12db. If the noise measurement Richard made at the 15th bit level means that there is no advantage to recording a 24bit file over a 16 bit one...

[edit'd fer spellin']

[stantheman1976]

Has anyone measured the Sony D50?

[Will_S]

Re: whether there is any benefit to recording 24 bit on the R09:

Quoting from this thread (John Siau as quoted by jerryfreak): http://taperssection.com/index.php/topic,107099.msg1431081.html#msg1431081

Here are some easy rules-of-thumb for adding noise sources:

1) Two equal amplitude noise sources added will degrade the SNR by 3 dB.
2) If a new noise source is added that is 6 dB lower than the existing
noise, the resulting noise will increase by 1 dB.
3) If the noise sources differ by more than 6 dB, the quieter source can
be ignored.
4) Use RMS calculations for summing noise sources when you need exact
numbers.

The above assume random noise and no correlation between the two noise
sources.  This is usually the case with microphones and electronics, but
be careful about room noise because it is often not random when the room
is empty.  On the other hand, the random rustle of a crowd can approach
white noise under certain conditions.

Another interesting and useful rule of thumb:
The human ear can detect a tone that is 30 dB lower than the ambient
noise level.  I frequently demonstrate this is the lab.  What this means
is that a TPDF dithered 16-bit signal with an SNR of 93 dB can audibly
reproduce a signal that is recorded at -123 dB FS.

So...if the signal to noise ratio of the analog component of the R09 ADC is 92 dB, it would seem to follow that digitizing this at 16 bit resolution with a theoretical SNR of ~ 96 dB (actually a bit worse...maybe even as low as 93 dB based on the above quote) would be intermediate between combining two sources with equal SNR (which would add ~3 dB of noise) and combining two signals that differ by 6 dB (which would add ~1 dB) of noise.  So it would seem like, if the quote above is true and I'm not applying those rules inappropriately, you can save yourself about 2 dB of noise on the R09 by recording at 24 bit rather than 16bit.  Of course your mics need to be providing a very clean signal indeed for that low level noise to be relevant.

[Gutbucket]

So...if the signal to noise ratio of the analog component of the R09 ADC is 92 dB, it would seem to follow that digitizing this at 16 bit resolution with a theoretical SNR of ~ 96 dB (actually a bit worse...maybe even as low as 93 dB based on the above quote) would be intermediate between combining two sources with equal SNR (which would add ~3 dB of noise) and combining two signals that differ by 6 dB (which would add ~1 dB) of noise.  So it would seem like, if the quote above is true and I'm not applying those rules inappropriately, you can save yourself about 2 dB of noise on the R09 by recording at 24 bit rather than 16bit.

You mean because of the addition of the dither noise of the ADC at 16bit to the analog noise?

Here's the part that addresses what I was getting at above-

Another interesting and useful rule of thumb:
The human ear can detect a tone that is 30 dB lower than the ambient
noise level.  I frequently demonstrate this is the lab.  What this means
is that a TPDF dithered 16-bit signal with an SNR of 93 dB can audibly
reproduce a signal that is recorded at -123 dB FS.

I love the nice black, 'out of sight - out of mind' noise floor in my better digital recordings, but I grew up listening to reel to reel dubs of LP's, AM radio and (gasp!) cassettes where it seemed that most of the delicious details of the music were buried in the tape hiss.  I would have preferred the noise not to be there, but the music was still alive deep into it and to me that is more important than the presence of the noise itself.  Not sure if that applies the same way in this case.

If the noise floor is above the 16th bit, does recording 8 more of noise help capture those tails? or would the analog stage noise in the R-09 best case scenario Richard describes, act as dither feeding the last bit of 16 and allow me to encode that information below the noise floor just as effectively with a 16 bit file as a 24 bit one?

[Will_S]

So...if the signal to noise ratio of the analog component of the R09 ADC is 92 dB, it would seem to follow that digitizing this at 16 bit resolution with a theoretical SNR of ~ 96 dB (actually a bit worse...maybe even as low as 93 dB based on the above quote) would be intermediate between combining two sources with equal SNR (which would add ~3 dB of noise) and combining two signals that differ by 6 dB (which would add ~1 dB) of noise.  So it would seem like, if the quote above is true and I'm not applying those rules inappropriately, you can save yourself about 2 dB of noise on the R09 by recording at 24 bit rather than 16bit.

You mean because of the addition of the dither noise of the ADC at 16bit to the analog noise?

Yes, dither noise or quantization noise at ~-96 dB or a little higher.  Added onto what I interpret from Richard's tests as analog noise in the pre-quantization circuitry at -92dB.  Resulting in a combined noise floor around -91 or -90dB.  Does that sound right?

[DSatz]

illconditioned, by the time you boil circuit noise down to any one number for amplitude, you can be very near the point of having a meaningless result. Noise is dynamic by nature (its amplitude varies from moment to moment), and only rarely is it evenly distributed throughout the audio frequency spectrum. Those alleged "easy rules of thumb for adding noise sources" depend on preconditions which, in real life, are never met except under contrived circumstances.

Furthermore, shorting the inputs to a preamp and dialing the gain to an extreme are definitely going to change the noise performance of the circuit from what it would be when driven by the specific source impedance of your microphones, and set to gain levels that you typically use for music recording.

Further-urthermore, some kind of weighting curve should always be used when evaluating the effect of noise, but the proper curve depends very much on the amplitude of the noise in playback--a chicken and egg problem if ever there was one.

I don't mean to discourage serious attempts at reaching an answer to your question, but I do want to hint that there is a very poor likelihood of getting a trustworthy answer simply by looking for consensus on an Internet forum.

I know from careful measurement that, at the recording levels I typically use for classical material, the noise levels of my microphones and preamps are low enough that 16 bits is not enough to capture their full range at all frequencies. (The inherent noise of a condenser microphone is heavily weighted toward the lowest frequencies. Below a few hundred Hz or so, 16 bits can be enough, or even more than enough--but not at, say, 3 - 4 - 5 kHz where the ear is most sensitive to noise.)

--best regards

[Gutbucket]

I do know from careful measurement that at the recording levels I typically use for classical material, the noise levels of my microphones and preamps are low enough that 16 bits is not enough to capture their full range at all frequencies. (The inherent noise of a condenser microphone is heavily weighted toward the lowest frequencies. Below a few hundred Hz or so, 16 bits can be enough, or even more than enough--but not at, say, 3 - 4 - 5 kHz where the ear is most sensitive to noise. For example, preamp noise testing has to be done 24-bit or the results will be meaningless, except for very noisy preamps.)

However, the noise floor of nearly all available recording venues swamps the noise of the microphones and preamps most of the time when the gains are set as I would normally set them for live recording. So for live concert recording, I am quite happy with 16-bit resolution as the delivery medium. I record live concerts at 24-bit resolution nowadays, but that is just so that I can set conservative levels and avoid stress, while still not giving up the eventual dynamic range of the finished 16-bit product.

I thoroughly understand the last paragraph above and completely understand what you saying in the previous one too, it's the implication of the two together that gets very close to addressing a question that has rattled around the back of my head for a couple years now.

In light of the above and concerning using noisier mics in those not-perfectly quiet venues like many of us do here..
Take for example the not uncommon scenario of using a pair of DPA 4060 miniature microphones into a relatively quiet external preamplifier feeding a 16 or 24 bit flash recorder-

Relevant DPA 4060 specs:
Equivalent noise level A-weighted: Typ. 23 dB(A) re. 20 µPa (max. 26 dB(A))
     [edit to add - Equiv. noise level ITU-R BS.468-4: Typ. 35 dB (max. 38 dB)]
S/N ratio, re. 1 kHz at 1 Pa (94 dB SPL): 71 dB(A)
Total harmonic distortion (THD): <1% THD up to 123 dB SPL peak
Dynamic Range: Typ. 100 dB
Max. SPL, peak before clipping: 134 dB

If the ADC in my recorder could capture 17 bits, shouldn't I be able to set the gain permanently for those microphones and never need to adjust it, capturing the full range between the 23dB noise floor of the mics (or more likely the higher noise floor of the room) all the way up to the 123dB SPL 1% THD rating? That's high enough for headroom in most cases and I should be able to add whatever make-up gain was needed for quiet sources digitally in post without any additional noise penalty over changing gain before the recorder. With 16 bits I should be able to do the same and record up to 116dB SPL before clipping.

Here's my thinking-
I can never get lower than the noise floor of the mics (23 dB(A) re. 20 µPa in this case).  If I adjust gain so that the noise floor of the analog stage is slightly below the mic self-noise so as to keep the sum of those noise sources low, and then add whatever 'real world dynamic range the ADC in the recorder is capable of, I should be able to calculate the highest SPL level I can record at that particular gain setting before clipping.

   23 dB (mic self-noise (A) re. 20 µPa)
  - 3 dB (set analog noise floor 3dB beneath mic self-noise - is this reasonable?)
-------  equals
   20 dB (bottom of the available recordable bit depth)
 +96 dB (dynamic range of 16 bit PCM)
-------- equals
  116 dB SPL @ 0 dBfs

Limited at the bottom by the noise of the mics, if I can capture the full 96dB range of 16 bits I should be able to record up to 116 dB SPL before clipping. With a better 24 bit ADC that could realistically capture at least 17 bits or a 102 dB range I could take advantage of the full dynamic range of the microphones (given their >1% THD at 123dB SPL peak spec) and reach 0dBfs around 122 dB SPL, which should leave headroom enough.

Is this correct or am I way off base and exposing my weak dB understanding, bit depth knowledge and mic spec comprehension?

[edit to add more realistic mic noise specs]

[DSatz]

Gutbucket, what you're trying to do is ambitious but from a quick look it seems to me that you're using the dB correctly. (I'll have more time later to look carefully at what you posted.)

But my point is that to reach a meaningful, practical result with this type of calculation--one with real predictive value as to what human ears are likely to hear in playback--you would need more detailed information than what you've got. You'd need to consider different noise spectra (frequency distributions) as well as the time-domain aspects of noise (e.g. the relative contributions of impulse or "shot" noise to a signal as compared with steady "background" noise). Those aspects of noise aren't just incidental; they can make truly major differences in the audible result.

Also, and I don't like to say something that could be taken as disparaging a serious microphone manufacturer, but--DPA is currently alone among the "majors" in the way they specify the noise levels of their products. The only numbers which they generally give in public follow a standard method (A-weighted "rms") which gives the least practical information of the available standard methods for specifying microphone noise. It is the one which gives the lowest number in dB, however, so naturally it's the one used by marketing people for "specification battles."

Compare this with any spec sheet from AKG, Beyer, Neumann, Sennheiser or Schoeps and you'll also see a second figure which is normally some 8 - 12 dB higher (CCIR weighted "quasi-peak"). That figure has far more predictive value in terms of the audible result of a microphone's noise, and is currently the one to pay attention to if you simply must have one single number.

When I get time I'll post some frequency graphs to illustrate what I mean. When you plug a condenser microphone into a preamp, you get an overlay of the noise from both of them. With most high-quality equipment the microphone's noise will dominate in the low frequencies while the preamp's noise will dominate in the upper midrange and above. From the dB figures published in spec sheets, or from a "shorted input, maximum gain" measurement of a preamp, you'd never know this but it's what almost always happens.

The "crossover" point depends greatly on the preamp gain you're using, and that in turn depends on what you're recording and how sensitive your microphones are. But a preamp's noise performance at one gain level simply can't be extrapolated from knowing what it is at some other level--different circuits react differently to changes in gain.

So this has to be measured under more realistic conditions, and can't be predicted from spec sheet values alone. If you do that, you'll get a number and your arithmetic may check out but the result may be very, very wrong. You might well choose preamp "A" over preamp "B" on this basis when preamp "B" would actually be several dB quieter under real-world conditions.

I can tell you for example that in my own tests, one preamp which has a reputation for not being super-quiet (the FMR "Real Nice Preamp"--it's small and not very expensive, and its manufacturer almost apologizes for its A-weighted full gain EIN spec) turned out to be among the very quietest when it was driven by the source impedance of an actual microphone, and when its gain was set to what I actually use for music recording.

--best regards

[Gutbucket]

DSatz, thank you for taking the time to explain the non-theoretical realities that only experience and the knowledge of what to measure and how to interpret that information can uncover.  I've suspected that there must be complexities and complications that would compromise the simple approach I propose above even if the numbers add up, but examining the theoretical side and then uncovering the 'real world' issues really helps for developing a thorough understanding.

It was my oversight on just posting the A-weighted mic self-noise.  DPA also includes a ITU-R BS.468-4 specification for all their mics which is listed as 35 dB typical (max. 38 dB) for this microphone.  I suspect that may correspond to the (CCIR weighted "quasi-peak") measurement that you mention as the more meaningful noise measurement. They include that specification for all their mics, including these relatively less expensive miniature versions which lack the individual measured frequency response and polar plots documentation that comes provided with their standard mics.  I just posted a few of the mic specs as raw data for the question I was posing and since the more realistic 35-38dB noise figure decreases the total dynamic range that the mic is capable of, using that higher noise floor as the low level limit would (in theory at least) make setting a single gain setting for all sound levels recorded by this mic less demanding.

I think you've probably set gain too high in your example.  Aren't you much more likely to run into 123dBSPL than 23dBSPL?..

Depends on what you record of course. With my gear I usually set my levels to peak somewhere between -12 and -6dB into a 24bit recorder, depending on the situation the type of music and what I expect from the dynamics of the musicians and audience.  For various reasons there have been times where I've had levels set more conservatively than was warranted and have had recordings that peak -24 dB or less below full scale.  Those recordings need much more amplification to playback at the same level than recordings that peaked higher, but I've been pleasantly surprised that they do not seem to suffer much.  That experience, along with a general curiosity about the implications of the numbers vs reality and the opportunity to exercise my weakling audio mathematics muscles was part of my motivation for asking about this.

Thanks for the great discussion!