Taperssection.com

Gear / Technical Help => Ask The Tapers => Topic started by: checht on June 02, 2018, 09:55:02 PM

Title: What sample rate are you using?
Post by: checht on June 02, 2018, 09:55:02 PM
I've been recording at 24/96 since I had a deck that could. Now I use Audacity in my post workflow, and it converts to 32/96, leading to large files. Hard drives are cheap, and I use an SSD as a work drive, so it's not a big problem, but I'm wondering what I'd lose going to 24/48 for recording.

Thoughts?

What rate do you use?

Thanks,
Chris
Title: Re: What sample rate are you using?
Post by: Fatah Ruark (aka MIKE B) on June 02, 2018, 09:56:44 PM
24/48.

For concerts I can't tell the difference above that.
Title: Re: What sample rate are you using?
Post by: rocksuitcase on June 02, 2018, 10:13:57 PM
24/48 live. Storage used to be the reason, now its doing 4 or 6 tracks at a time with the space/storage that entails.
Title: Re: What sample rate are you using?
Post by: heathen on June 02, 2018, 10:16:01 PM
24/48. For the stuff I record, there's no reason for anything higher than that.
Title: Re: What sample rate are you using?
Post by: rigpimp on June 03, 2018, 01:04:19 AM
24/96.  I have a 64TB NAS so space isn't an issue.
Title: Re: What sample rate are you using?
Post by: firemt66 on June 03, 2018, 08:24:22 AM
What is the sense of older shows recorded in 16bit re master to 96. The music has been recorded your not adding music..noise is being added,no?
Title: Re: What sample rate are you using?
Post by: rippleish20 on June 03, 2018, 08:42:50 AM
24/48
Title: Re: What sample rate are you using?
Post by: kindms on June 03, 2018, 08:54:40 AM
I've been recording at 24/96 since I had a deck that could. Now I use Audacity in my post workflow, and it converts to 32/96, leading to large files. Hard drives are cheap, and I use an SSD as a work drive, so it's not a big problem, but I'm wondering what I'd lose going to 24/48 for recording.

Thoughts?

What rate do you use?

Thanks,
Chris


audacity defaults to 32bit floating but you can change that in the interface after the import. Also you have your raw master, then your edited exported audacity at ? 24/96 or 16/44.1 why do you need to keep the 32 ?


my workflow is 24/48 original > Audacity edit in 32 floating > export to 24/48 and export to 16/44.1 > CD WAV Editor for tracking > TLH FLAC. I flac the untouched master files, the 24/48 tracked version for sharing and the 16/44.1 for sharing as well. Master files get archived, the tracked versions as well.  No need to keep the project files really just my .02
Title: Re: What sample rate are you using?
Post by: wforwumbo on June 03, 2018, 09:01:28 AM
I personally record at 24/96, thinking about moving up to 192. I don’t inherently hear a difference on the raw tapes, but many digital algorithms for things like EQs and saturation modeling plugins have significantly reduced filter error the higher the same rate; and that I do hear. Space is dirt cheap, I stash my shows on an external and have a 64-gig internal card in my mixpre-6.

What is the sense of older shows recorded in 16bit re master to 96. The music has been recorded your not adding music..noise is being added,no?

Assuming you mean 16 to 24 bit and/or 44.1k/48k/88.2k to 96k, correct. Depending on how the conversion is done there are ways to try and reduce noise but inherently it will add some level of noise. There are reasons why some may upsample, and some of them are valid (e.g. hardware is optimized for a certain bit depth/sample rate, or certain approximations in filter design improve drastically in phase response), but those are extreme special use cases, and far and few between.
Title: Re: What sample rate are you using?
Post by: Gordon on June 03, 2018, 10:01:19 AM
24/48.

For concerts I can't tell the difference above that.
Title: Re: What sample rate are you using?
Post by: ilduclo on June 03, 2018, 10:40:21 AM
24/44.1
Title: Re: What sample rate are you using?
Post by: Sloan Simpson on June 03, 2018, 10:59:54 AM
24/48, and would do 44.1 except to head off complaints from video folks I send files to on occasion. I use Acustica plugins and processing is slow enough as it is at 48k; 96k rendering time is prohibitive.
Title: Re: What sample rate are you using?
Post by: ilduclo on June 03, 2018, 11:03:50 AM
I need 24 bit to do post process amplification without adding hiss. But 44.1 works fine for sound quality, IMO and makes the files smaller and easier to work with.
Title: Re: What sample rate are you using?
Post by: capnhook on June 03, 2018, 11:07:29 AM
24/44.1

yep, but 24/48 if the video folks want some.

Title: Re: What sample rate are you using?
Post by: billydee on June 03, 2018, 11:12:38 AM
I guess I'll be the odd man out and say I only do 16/44 for 95% of what I record these days.
Title: Re: What sample rate are you using?
Post by: u2_fly_2 on June 03, 2018, 11:33:37 AM
96/24 Bit and also 192/24 on a new recorder. It also has the sound option "DSD Format with 2.8 MHz" which might or might not be to any use in the future(?)
Title: Re: What sample rate are you using?
Post by: Scooter123 on June 03, 2018, 12:13:30 PM
24 48

Title: Re: What sample rate are you using?
Post by: jagraham on June 03, 2018, 12:35:38 PM
24/48 for the recording, mastering and archiving, dithered down to 16/44.1 for flac filesets for listening purposes. Of course I always keep the edited 24bit files along with the cue files so I could track them out in 24 later.
Title: Re: What sample rate are you using?
Post by: capnhook on June 03, 2018, 12:57:32 PM
I guess I'll be the odd man out and say I only do 16/44 for 95% of what I record these days.




I record in 24bit to save my bacon when the levels get too hot, and I'm not watching..


I output 16/44.1 .flac when I'm finished post-processing, 100% of the time.




Title: Re: What sample rate are you using?
Post by: checht on June 03, 2018, 01:04:01 PM
Seems like most use 24/48. I'll likely go to 48 khz instead of 96khz, as I haven't heard any argument against.

I think I'll stay with Audacity's 32 bit depth for post. From the manual:

"Audacity uses "float" format for 32-bit recording instead of fixed integer format as normalized floating point values are quicker and easier to process on computers than fixed integer values and allow greater dynamic range to be retained even after editing. This is because intermediate signals during audio processing can have very variable values. If they all get truncated to a fixed integer format, you can't boost them back up to full scale without losing resolution (i.e. without the data becoming less representative of the original than it was before). With floating point, rounding errors during intermediate processing are negligible.
The (theoretically audible) advantage of this is that 32-bit floating point format retains the original noise floor, and does not add noise. For example, with fixed integer data, applying a compressor effect to lower the peaks by 9 dB and separately amplifying back up would cost 9 dB (or more than 2 bits) of signal to noise ratio (SNR). If done with floating point data, the SNR of the peaks remains as good as before (except that the quiet passages are 9 dB louder and so 9 dB noisier due to the noise they had in the first place).
In many cases you will be exporting to a 16-bit format (for example if you are burning to a standard audio CD, that format is by definition 16-bit 44100 Hz). The advantage of using 32-bit float to work with holds even if you have to export to a 16-bit format. Using Dither on the Quality tab of Audacity Preferences will improve the sound quality of the exported file so there are only minimal (probably non-audible) effects of downsampling from 32-bit to 16-bit."

Note the use of 'theoretically' and 'probably'.
Title: Re: What sample rate are you using?
Post by: capnhook on June 03, 2018, 01:18:03 PM
Seems like most use 24/48. I'll likely go to 48 khz instead of 96khz, as I haven't heard any argument against.

I think I'll stay with Audacity's 32 bit depth for post. From the manual:

"Audacity uses "float" format for 32-bit recording instead of fixed integer format as normalized floating point values are quicker and easier to process on computers than fixed integer values and allow greater dynamic range to be retained even after editing. This is because intermediate signals during audio processing can have very variable values. If they all get truncated to a fixed integer format, you can't boost them back up to full scale without losing resolution (i.e. without the data becoming less representative of the original than it was before). With floating point, rounding errors during intermediate processing are negligible.
The (theoretically audible) advantage of this is that 32-bit floating point format retains the original noise floor, and does not add noise. For example, with fixed integer data, applying a compressor effect to lower the peaks by 9 dB and separately amplifying back up would cost 9 dB (or more than 2 bits) of signal to noise ratio (SNR). If done with floating point data, the SNR of the peaks remains as good as before (except that the quiet passages are 9 dB louder and so 9 dB noisier due to the noise they had in the first place).
In many cases you will be exporting to a 16-bit format (for example if you are burning to a standard audio CD, that format is by definition 16-bit 44100 Hz). The advantage of using 32-bit float to work with holds even if you have to export to a 16-bit format. Using Dither on the Quality tab of Audacity Preferences will improve the sound quality of the exported file so there are only minimal (probably non-audible) effects of downsampling from 32-bit to 16-bit."

Note the use of 'theoretically' and 'probably'.



Stick with Audacity 2.1.0 if you have to use the "Change Speed" tool.  Newer versions are broken, truncating error.  I have not examined V2.2.2, maybe someone can report on it.  V2.1.0 works perfectly, and is OK enough for me.






Title: Re: What sample rate are you using?
Post by: checht on June 03, 2018, 05:12:03 PM
^ Thanks!
Title: Re: What sample rate are you using?
Post by: daspyknows on June 03, 2018, 08:15:39 PM
24/48 for the recording, mastering and archiving, dithered down to 16/44.1 for flac filesets for listening purposes. Of course I always keep the edited 24bit files along with the cue files so I could track them out in 24 later.

same here
Title: Re: What sample rate are you using?
Post by: noahbickart on June 03, 2018, 09:47:04 PM
I use 24/48, and share those and 16/44.1 “mastered” versions.
Title: Re: What sample rate are you using?
Post by: nulldogmas on June 03, 2018, 10:16:41 PM
24/48 for the recording, mastering and archiving, dithered down to 16/44.1 for flac filesets for listening purposes. Of course I always keep the edited 24bit files along with the cue files so I could track them out in 24 later.

same here

Yup, that.
Title: Re: What sample rate are you using?
Post by: opsopcopolis on June 03, 2018, 11:53:57 PM
16/44.1 usually. Occasionally 24/48. I almost never use 96 or higher even in my studio work (only really orchestral stuff/entirely acoustic music) so I sure as hell ain’t using it to record two track tapes of PA mixes from the back of a room.
Title: Re: What sample rate are you using?
Post by: ycoop on June 04, 2018, 03:21:53 AM
I guess I'll be the odd man out and say I only do 16/44 for 95% of what I record these days.




I record in 24bit to save my bacon when the levels get too hot, and I'm not watching..


I output 16/44.1 .flac when I'm finished post-processing, 100% of the time.

This is something I've wondered about. What about the extra 8 bits allows for more work in post without losing as much quality?
Title: Re: What sample rate are you using?
Post by: aaronji on June 04, 2018, 06:48:43 AM
This is something I've wondered about. What about the extra 8 bits allows for more work in post without losing as much quality?

A major advantage, as I understand it, is that the extra bits greatly reduce the noise introduced from quantization error during analog to digital conversion (each extra bit approximately halves the noise).  So even if the recording peaks considerably below full-scale, you can amplify it without making the quantization noise audible...
Title: Re: What sample rate are you using?
Post by: yates7592 on June 04, 2018, 08:48:33 AM
24/192 - storage space isn't an issue.
Title: Re: What sample rate are you using?
Post by: capnhook on June 04, 2018, 12:52:08 PM
This is something I've wondered about. What about the extra 8 bits allows for more work in post without losing as much quality?

A major advantage, as I understand it, is that the extra bits greatly reduce the noise introduced from quantization error during analog to digital conversion (each extra bit approximately halves the noise).  So even if the recording peaks considerably below full-scale, you can amplify it without making the quantization noise audible...

Precisely, thanks aaronji
Title: Re: What sample rate are you using?
Post by: nak700s on June 04, 2018, 05:56:20 PM
The only thing that's important is that you're happy with your recordings.  I use 24/48, and I'm happy with that.  Depending on how you record, and your editing process, would make a difference as to what you should be recording at.  My suggestion, try out different things and determine what works best for you.
Title: Re: What sample rate are you using?
Post by: Cheesecadet on June 05, 2018, 01:02:01 AM
24/44.1
Title: Re: What sample rate are you using?
Post by: morst on June 05, 2018, 03:07:03 AM
I record and work at 24 bits / 48 kHz from start to finish. Audacity is 32 bit internally, but I export as 24 bits.

Since I don't resample, I don't dither, but I would pick triangle dither if I had to do so.

I tried recording something at 96kHz once, and converted it down to 48kHz, then inverted and subtracted, and the difference file sounded blank to me. ¯\_(ツ)_/¯
Title: Re: What sample rate are you using?
Post by: Gutbucket on June 05, 2018, 09:59:55 AM
I record at 24/48, edit in floating point (some plugins doing their own upsampling), then use triangular dither to truncate back to 24bit files for output.. or SRC to 44.1 + triangular dither to truncate to 16 bit.
Title: Re: What sample rate are you using?
Post by: goodcooker on June 05, 2018, 10:48:11 AM
24/44.1

Same here. It's fine for 95% of what I do in the field.
Title: Re: What sample rate are you using?
Post by: IMPigpen on June 05, 2018, 11:02:53 AM
I use 24/48 for everything.  Most of my stuff I upload to the LMA.  I used to put up a 24/48 version and a 16/44 version but stopped doing that a while ago.  I've found that most people just listen to the streaming mp3 on there anyways.  And if archive my 24/48 raw files and the 24/48 FLAC.  So if someone wants something different, I can convert the FLACs to whatever they need.  I just had an artist ask me for mp3s of his solo show after I sent him the FLACs, so it took about 5 minutes to convert and zip them and send them to him.
Title: Re: What sample rate are you using?
Post by: DSatz on June 11, 2018, 09:09:07 PM
I record at 24/44.1 if the eventual requirement is audio-only, and 24/48 if my sound track will be matched to video. Then I deliver whatever is specifically requested, whether it be mp3 files, CD-Rs, etc.

--aaronji wrote:

> A major advantage, as I understand it, is that the extra bits greatly reduce the noise introduced from quantization error during analog to digital conversion (each extra bit approximately halves the noise).  So even if the recording peaks considerably below full-scale, you can amplify it without making the quantization noise audible...

Aaronji, that just isn't true as a rule. It all depends on the noise in the signal that's being quantized versus the converter's own noise floor. The quantization noise of the converter only needs to be a few dB lower than the noise in the incoming analog signal, however many bits that takes. Any further bits beyond that will (or should!) be random; therefore they don't contribute to the precision or accuracy of the recording in any way. The notion that computation (post-processing) will somehow work out more accurately with more bits is a total fallacy when those bits contain only noise.

Of course if they contain meaningful signal information, then by all means keep them--but from what I've read on this board over the years, I think that a lot of people here are kind of fooling themselves on that point. For live classical recording, which is what I mostly do, the signal going into the a/d converter might have a noise floor maybe 60 or (if I'm very lucky) 65 dB below the peak levels. 14-bit quantization can handle that (keeping in mind that the converter noise needs to be below the incoming noise at all frequencies). 16-bit recording allows for more conservative level settings and fewer accidental overloads--plus it's been decades since 14-bit recording was even an option on available recorders. But for example the BBC used 10-bit (analog companded) studio-to-transmitter links for their classical broadcasts for years, and they were justly renowned for their audio quality.

24-bit recording (which might have 18 or 19 actual bits to offer, when the noise level of the converters and the rest of the signal chain is taken into account) is even nicer than 16, but it's clearly a luxury for our benefit as live engineers. I use it because it doesn't hurt anything*, and once in a blue moon it helps a little. But I've never delivered a 24-bit recording to a client; people not directly involved in recording have no use for the extra data, which only causes them trouble and expense.

(Notes: If the extra, unneeded converter bits beyond the incoming signal's noise floor aren't random, then they're adding distortion to the recording, and any such converter should be removed from service immediately. And please keep in mind that the noise floor of a converter is whatever its ACTUAL value is, with the number of bits setting a theoretical limit that is never reached in practice; no physically realizable 24-bit converter has an analog input noise floor anywhere near 144 dB below peak level. So please don't make a mental leap from "24 bits" to "8 bits quieter than a 16-bit converter" because it's much more likely in reality to be only 2 or 3 bits quieter, if that.)

--best regards

_________
* Edited later to add an asterisk above, and an important qualifying remark as a footnote: 24-bit recording "doesn't hurt anything" ONLY if you set your recording levels appropriately. At the risk of seeming to be unkind, let me say this as clearly as I can: If you're one of those people who completely misunderstand the concept of headroom, who actually AIM to achieve peak levels of -12 dB or even lower (instead of aiming to get your peak levels as close as possible to 0 dB without actually hitting it), then the use of 24-bit recording is hurting you by "enabling" your basic failure to understand the concept of recording levels, and your "24-bit" recordings are quite possibly noisier than they would be if you learned how digital recording actually works, and used 16-bit recording appropriately.
Title: Re: What sample rate are you using?
Post by: morst on June 12, 2018, 06:21:44 AM
24-bit recording (which might have 18 or 19 actual bits to offer, when the noise level of the converters and the rest of the signal chain is taken into account) is even nicer than 16, but it's clearly a luxury for our benefit as live engineers. I use it because it doesn't hurt anything*, and once in a blue moon it helps a little.
I'll take it!
Quote
But I've never delivered a 24-bit recording to a client; people not directly involved in recording have no use for the extra data, which only causes them trouble and expense.
D'oh!
Quote
...So please don't make a mental leap from "24 bits" to "8 bits quieter than a 16-bit converter" because it's much more likely in reality to be only 2 or 3 bits quieter, if that.)

--best regards

_________
* Edited later to add an asterisk above, and an important qualifying remark as a footnote: 24-bit recording "doesn't hurt anything" ONLY if you set your recording levels appropriately. At the risk of seeming to be unkind, let me say this as clearly as I can: If you're one of those people who completely misunderstand the concept of headroom, who actually AIM to achieve peak levels of -12 dB or even lower (instead of aiming to get your peak levels as close as possible to 0 dB without actually hitting it), then the use of 24-bit recording is hurting you by "enabling" your basic failure to understand the concept of recording levels, and your "24-bit" recordings are quite possibly noisier than they would be if you learned how digital recording actually works, and used 16-bit recording appropriately.

Thanks for another informative post DSatz! Just for max clarity, could you please elaborate on the * part?
What would be the best practice for making use of 24-bit recording? Get as close as possible to 0 dBFS without going over, and don't just aim for a conservative -12 dB peak?
I must admit, I'm not scared of the occasional red light flashing on a transient peak, and I'm happy to find out if I can hear my limiters in action. (mostly never!)
Title: Re: What sample rate are you using?
Post by: heathen on June 12, 2018, 09:26:59 AM
Just for max clarity, could you please elaborate on the * part?

Seconded.
Title: Re: What sample rate are you using?
Post by: Gutbucket on June 12, 2018, 10:59:21 AM
What would be the best practice for making use of 24-bit recording? Get as close as possible to 0 dBFS without going over, and don't just aim for a conservative -12 dB peak?
I must admit, I'm not scared of the occasional red light flashing on a transient peak, and I'm happy to find out if I can hear my limiters in action. (mostly never!)

Get as close as possible to 0 dBFS without going over
^This is the optimal strategy for all digital recording.. if only it were so simple in the real world, where we don't know in advance how loud it will get.

The common misconception DSatz is referring to is the notion that peaking no higher than -12dB or whatever is somehow beneficial in and of itself.  Its not, its only good because it's not over. The goal isn't peaking specifically at -12dBfs, -6, -18, -3 or whatever number, the goal is simply not going over 0dBfs.  The second (unstated) goal which this achieves at the same time is to try and keep levels high enough that the quietest parts do not drop beneath the recording system's noise-floor.  That second part happens automatically if we push levels to get at close as possible to 0 dBfs, leaving almost all the "extra space" at the bottom as randomized noise.  A portion of that extra stuff at the bottom gets tossed out when truncating the 24 bit file down to 16 bits, but not all of it.  It's like trimming excess fat off that side, but still leaving a sufficeint fatty edge without cutting into the meaty portion. The full dynamics of the recording fits entirely within the 16bit file container.  It just has less excess fat than before.

Find a comfortable balance between not going over at the top of the dynamic range nor dropping down beneath the noise floor of the recording system at the bottom of the range.
^This strategy is effectively the same with respect to the end result, as long as the second half of the statement is true.  And it's a lot easier to manage in the real world.

It leaves more space at the top and somewhat less at the bottom as the recording is being made.  We can then trim some of that extra stuff from both the top and the bottom.  We cut excess fat from both sides this time, and end up with the same full dynamic recording in the 16 bit file.

If you never have a problem with noise-floor in the quiet parts of your recordings, then there is no problem with peaking somewhat lower and amplifying later.  It makes things easier in a practical sense by diminishing the possibility of overs.. or of limiters audibly cutting in.  Clipping doesn't sound good, and neither do limiters if you can audibly hear them working.  But otherwise there is no advantage to peaking lower other than not having to worry about overs.

No need to bump up too close to either end of the range. 17 or 18 bits worth of real-world recorded dynamic range leaves plenty of room in which to fit the full range of recorded sound with "headroom" and "footroom" to spare at both the top and bottom.

Sufficient headroom and footroom are not ends in themselves.  It's fitting comfortably within the limits which is the real goal.

Title: Re: What sample rate are you using?
Post by: DSatz on June 12, 2018, 11:07:04 PM
Hmm. I'm not sure what needs to be explained, because I don't know what you currently know and/or believe. How about this as an outline:
Thus as long as the channel has sufficient dynamic range and the levels are set optimally--where the peaks of the program material come as close as possible to the channel's maximum level without actually reaching it--any additional extension of the channel's dynamic range can't audibly improve things, because the signal itself is the limiting factor at that point, not the channel.

This is actually nothing peculiar to digital audio; analog works exactly the same way. In fact I would strongly recommend "thinking in analog" in order to get the correct mental picture.
Title: Re: What sample rate are you using?
Post by: heathen on June 12, 2018, 11:27:15 PM
Maybe I misunderstood the initial point you were making.  After reading the more recent posts, I'm assuming the initial point was that we shouldn't be using -12 (or -6 or -16, etc) dogmatically as our goal for where levels peak, but instead use 0.  But, based on Gutbucket's comment and your last post, if we don't get the levels right at 0 and fall a little under, it's okay as long as the low parts of the audio signal still ride above the noise floor of the recorder.  Is that (somewhat) accurate?

Sorry for my ignorance here...and thanks for everyone's patience and explanations.
Title: Re: What sample rate are you using?
Post by: checht on June 13, 2018, 12:26:45 AM
Super appreciate where this thread has gone.

How does this relate to post-processing bit depth? Audacity defaults to 32 bit, any harm in going with 24?

TIA
Title: Re: What sample rate are you using?
Post by: edtyre on June 13, 2018, 02:01:31 AM
24/44.1 or 24/48 no need for anything else
Tryed DSD, 24/192 doesn’t work for me.
Title: Re: What sample rate are you using?
Post by: DSatz on June 13, 2018, 02:06:57 AM
Heathen, that's exactly right as far as the general relationship between a signal and a channel is concerned.

However, a recorder's circuit design also matters. When the record level control on a recorder is set way too low, there's a real risk that unnecessary, extra noise is being added to the recording that doesn't come only from the noise floor of the a/d converter stage, but also from the noise floor of one or more earlier stages of the analog electronics. The user is basically inviting extra, unnecessary noise into the recording at each analog stage after the record level control.

One well-known recorder that I'm almost certain has this characteristic, for example, is the Sony PCM-M10. According to what I saw when I did some testing and measuring about five years ago, a 16-bit recording that reaches -2 on peaks would be considerably quieter than an exactly equivalent 24-bit recording that reaches only -12 but is then boosted 10 dB in post-processing. That's the opposite of the result that I think some people would expect.

I should probably repeat those measurements to make sure, but this isn't so much about that particular recorder as it is about the fact that there are always multiple analog stages before the a/d converters in any recorder--and if you set the record levels too low (or the sensitivity switch to its less sensitive setting when the more sensitive setting could be used safely without overload), it is rather likely that there will be unnecessary, audible noise contributions from one or more stages of the electronics prior to the converter.

Ideally, for the best signal-to-noise ratio. you want to nearly "saturate" each successive analog stage before the converters--keep them pumped full of signal, so that their noise doesn't make an audible contribution. Again, that's a basic way of thinking that an analog engineer would have from experience, that people have lost sight of in the digital era. I think some people might think about individual sample values too much, and/or think about them as if they only contain energy from the desired signal. But any one quantized sample reflects the instantaneous sum of the desired signal plus noise--and for any given sample, you can never know its "value due to signal" or its "value due to noise," just as you can't "unstir" a cup of coffee that already has cream and sugar in it.

--best regards
Title: Re: What sample rate are you using?
Post by: noahbickart on June 13, 2018, 09:16:55 AM
Dsatz is one of the main reasons that Ts.com remains the best place on the internet.

Thanks for your continued presence here.
Title: Re: What sample rate are you using?
Post by: aaronji on June 13, 2018, 10:21:06 AM
--aaronji wrote:

> A major advantage, as I understand it, is that the extra bits greatly reduce the noise introduced from quantization error during analog to digital conversion (each extra bit approximately halves the noise).  So even if the recording peaks considerably below full-scale, you can amplify it without making the quantization noise audible...

Aaronji, that just isn't true as a rule. It all depends on the noise in the signal that's being quantized versus the converter's own noise floor. The quantization noise of the converter only needs to be a few dB lower than the noise in the incoming analog signal, however many bits that takes. Any further bits beyond that will (or should!) be random; therefore they don't contribute to the precision or accuracy of the recording in any way. The notion that computation (post-processing) will somehow work out more accurately with more bits is a total fallacy when those bits contain only noise.

I have to admit (and with all due respect) that I am not quite sure what you mean.  My comment had nothing to do with post-processing accuracy.  What I was saying is that, as I understand it, the major (only?) advantage of the extra bits is to increase the channel's dynamic range by reducing quantization noise.  That makes it easier to fit the dynamic range of the signal, even with very low peak levels, into that space with (to use Gutbucket's term) sufficient footroom.  So when amplified, the signal's noise will dominate the converter's noise (rendering it inaudible).  Am I missing something here?  There are even some examples available (such as on the Sound Device's website) that appear to show this and it also seems to be in line with both your post and Gutbucket's.

Granted, the real world, live music recordings scenarios in which this will actually help are likely rare, but stuff happens due to accident or circumstance...
Title: Re: What sample rate are you using?
Post by: H₂O on June 13, 2018, 10:50:19 AM
1bit 2.8224Mhz for 2 channel
24bit 96Khz for 4-8 channel
Title: Re: What sample rate are you using?
Post by: rigpimp on June 13, 2018, 11:26:08 AM
Dsatz is one of the main reasons that Ts.com remains the best place on the internet.

Thanks for your continued presence here.

This^^.  If I am perusing and I see a post of his I take off my coat and read his post thoroughly.  I do the same with a few others like Morst and Gutbucket, for example.  The historical and technical contributions these folks make is so valuable. 

With that said, I am still going to run at 24/96.  :-p
Title: Re: What sample rate are you using?
Post by: wforwumbo on June 13, 2018, 02:53:17 PM
I agree with much of what’s said here, particularly coming from DSatz (whose praises I want to echo throughout). BUT... I DO want to iron out something here, and I’ve brought it up before. The “1 bit == 6 dB of dynamic range” statement.

That is a theoretical rule that mostly follows how bits work at lower bit depths. A 24-bit word length will always contain 24 bits of information. It’s not that the extra bits are wasted; they are still there, and they are accessible in memory and computation. Where we run into limitations is in the design of ICs (mostly in amplifier design and topology) and getting THOSE low-noise enough. That is a pretty complex and technical challenge.

Something else worth noting... the average dynamic range of a young, healthy human hearing system is about 110 dB of dynamic range. So as long as you are peaking below 0 dB-FS, the purpose of having more bits and a wider dynamic range is to make it such that your ears - not the technical limitations of your system - is the bottleneck for getting things to sound “as good as you can” so to speak.

And again, another advantage of larger word lengths, is reduced filter error. The more data you are pumping through a system doing truncated (discrete) calculations, the more accurate the end product will be. It’s why 32-bit floating point operations are used internally in DAWs, and we aren’t stuck to using just 16- or 24-bit systems. Post work does improve significantly in my (and many other studio pros’) experience and to my/our ears.

Again, this isn’t to say that practically most of what’s said here is right, and it’s not to say that you can get away with 16/44.1 without issue; in many cases, it will do the job just fine. Just want to clarify, from a technical perspective.
Title: Re: What sample rate are you using?
Post by: goodcooker on June 13, 2018, 03:30:41 PM
the average dynamic range of a young, healthy human hearing system

I hope you realize that a lot of us here are old, worn out, half deaf and barely human... :guitarist: :zombie02:
Title: Re: What sample rate are you using?
Post by: Gutbucket on June 13, 2018, 07:35:36 PM
I'm happy wforwumbo has joined TS.  As far as I can tell by way of a few personal conversations he is sharp, extremely well versed in matthmatics, not old, only partially deafened thus far (yet retains a very keen ear), and is fully human.  Keep an eye open for his future contributions on the board.

I do want to bring up something which came to mind while reading his post above, because it is something I find commonly misunderstood at TS and could be easily misconstrued from what he wrote.  I regularly see statements in these kinds of threads which say something to this effect- "I record 24 bit files because I plan to do post work on the recordings, and 24 bit files are better for that", but that doesn't actually correlate with what he wrote above. 

Performing mathematics with sufficient precision such that rounding/truncation errors are avoided makes for more accurate calculations, and that can translate to better sounding audio.  But the bit-length of the data being manipulated doesn't dictate the bit-length of the calculation space.  The calculation space within the DAW is larger (32-bit floating point) so calculation precision is preserved.. until one outputs it again.  As DSatz describes above-

Quote
..as long as the channel has sufficient dynamic range and the levels are set optimally--where the peaks of the program material come as close as possible to the channel's maximum level without actually reaching it--any additional extension of the channel's dynamic range can't audibly improve things, because the signal itself is the limiting factor at that point, not the channel.

The raw 16 or 24 bit recorded file is a channel containing the signal.  Mathematics are performed on that signal in the DAW in a 32-bit floating point space regardless of the signal's bit-length.  We then export the data from the DAW via another file (channel) of sufficient bit-length to contain the manipulated signal.  Because we have full control over signal levels when manipulating and exporting that signal from the DAW (unlike when we were originally making the raw recording) we can make sure the signal fits comfortably within an output channel of an optimum size, so we can actually output a 16 bit file for almost everything without compromising the signal. We can make sure that "the signal itself is the limiting factor at that point, not the channel".  Or we can output 24 bit files with some extra unrelated noise at the bottom, which is easy, and space is cheap.  But in most cases a 16 bit output file can be of equivalent quality because the full dynamic range of almost all music can be fully represented by less than 16 bits.

The portions quoted below should be understood as being relevant within the DAW (or possibly within the recorder prior to a file being written in some cases), and not as specifying what file length is necessary in file storage formats-

A 24-bit word length will always contain 24 bits of information. It’s not that the extra bits are wasted; they are still there, and they are accessible in memory and computation.

Quote
The more data you are pumping through a system doing truncated (discrete) calculations, the more accurate the end product will be. It’s why 32-bit floating point operations are used internally in DAWs, and we aren’t stuck to using just 16- or 24-bit systems. Post work does improve significantly in my (and many other studio pros’) experience and to my/our ears.



tldr- As long as the full dynamics of the raw recording fits within 16 bits, recording at 24 bits will not make "post processing more accurate or sound better", but doing the post processing in a larger bit-length calculation space can do so. 

Sample rate conversion is a different animal.
Title: Re: What sample rate are you using?
Post by: checht on June 13, 2018, 08:35:11 PM
....
tldr- As long as the full dynamics of the raw recording fits within 16 bits, recording at 24 bits will not make "post processing more accurate or sound better", but doing the post processing in a larger bit-length calculation space can do so. 

For math-challenged folks, what's the tldr of the tldr?

I record trying to get close to but not exceed 0db. Will my recordings be better in 16 or 24 bit?

Thanks for simplifying way way down.
Title: Re: What sample rate are you using?
Post by: wforwumbo on June 13, 2018, 11:46:50 PM
....
tldr- As long as the full dynamics of the raw recording fits within 16 bits, recording at 24 bits will not make "post processing more accurate or sound better", but doing the post processing in a larger bit-length calculation space can do so. 

For math-challenged folks, what's the tldr of the tldr?

I record trying to get close to but not exceed 0db. Will my recordings be better in 16 or 24 bit?

Thanks for simplifying way way down.

Try this post I wrote:

http://taperssection.com/index.php?topic=184569.msg2251598#msg2251598

Still on the technical side, but a simple explanation of the overall concept.
Title: Re: What sample rate are you using?
Post by: checht on June 14, 2018, 12:52:41 AM
^ That clarifies a lot.

Thanks!
Title: Re: What sample rate are you using?
Post by: Gutbucket on June 14, 2018, 10:50:43 AM
Quote
I record trying to get close to but not exceed 0db. Will my recordings be better in 16 or 24 bit?

In practical terms it likely depends on what recorder you are using, the ADC chip and specific implementation of it in the machine.  If you don't have measurements, don't care to do critical listening tests, don't want to worry about it too much, and don't care about storage size, my take is that it makes sense to default to 24 bit recording in this day and age, if simply as a belt and suspenders approach more than a definitely better kind of thing. This is a practical take rather than an engineering answer.


Practicality rules.  In the end, I suspect the two largest practical determinants on this is what era one started recording in, and how many TB of raw recordings one has amassed and needs to deal with! Storage space may be growing ever cheaper, but managing and backing up large catalogs is a PITA.
Title: Re: What sample rate are you using?
Post by: DSatz on June 14, 2018, 11:40:34 AM
aaronji, howdy. As for why I mentioned post-processing in my reply to you, please look again at the earlier message of yours that I replied to. It quotes a message from ycoop as follows: "What about the extra 8 bits allows for more work in post without losing as much quality?"--to which you replied that essentially, more bits in a/d conversion = lower quantization error. To which, in turn, I replied that no, that's not a valid general statement; it all depends on the noise in the signal being converted.

If that noise isn't correlated with the desired signal, and is entirely above the noise floor of the converter, then more bits don't (can't!) improve the accuracy of the conversion. Hypothetically and simplistically, if the noise floor of a signal is at a 13-bit level, you could convert that signal with a 14-bit, 16-bit, 20-bit or 24-bit converter all of which would give equal audible results (all other things being equal), because the noisy signal was the limiting factor, not the converter.

However, if the noise IS correlated with the desired signal, then it is a form of non-linear distortion and you've got yourself a defective converter. Most discussion assumes tacitly that the residual noise of a converter will be uncorrelated with the signal, because for decades now, that has normally been the case. That is another way of saying that the non-linear distortion is very low.

Still, the older ones among us, or anyone who's ever used an inadequately dithered A/D converter, will know what that kind of distortion sounds like, because some of the earliest available digital audio recording equipment didn't dither properly (ahem Sony, including their professional PCM-1600 which a lot of early CDs were recorded with). Some people describe it as "noise breathing" or "noise pumping" because it rises and falls with the signal levels, particularly on very soft sounds such as reverberation decay (or imagine taking a 400 Hz tone and fading it down slowly to nothingness with a smooth, analog fader). An accompanying phenomenon is (or was) "digital deafness"--signals below the scope of the lowest-order bit are simply blanked out.

Unfortunately, the way a lot of people think about digital audio is still based on the incredible flood of bullshit that was unleashed when digital audio was introduced around 1980, and which (when it was about anything real) harped greatly on shortcomings of conversion that were really due to not having proper dither. When an a/d converter is properly dithered, there's no increase in distortion at lower levels, no "stair-step levels" to be minimized by striving for more and more bits, nor is there "digital deafness" at the bottom of the range; it acts just like an analog channel as far as dynamic range is concerned. Since actual distortion in modern a/d converters is so low, I tend to shift the discussion away from the concept of "precision" in conversion (which would be a more appropriate concept if a signal were utterly noiseless, such as a computationally generated test signal) to dynamic range, which is much more pertinent to live, recorded audio.

--best regards
Title: Re: What sample rate are you using?
Post by: wforwumbo on June 14, 2018, 02:30:36 PM
aaronji, howdy. As for why I mentioned post-processing in my reply to you, please look again at the earlier message of yours that I replied to. It quotes a message from ycoop as follows: "What about the extra 8 bits allows for more work in post without losing as much quality?"--to which you replied that essentially, more bits in a/d conversion = lower quantization error. To which, in turn, I replied that no, that's not a valid general statement; it all depends on the noise in the signal being converted.

If that noise isn't correlated with the desired signal, and is entirely above the noise floor of the converter, then more bits don't (can't!) improve the accuracy of the conversion. Hypothetically and simplistically, if the noise floor of a signal is at a 13-bit level, you could convert that signal with a 14-bit, 16-bit, 20-bit or 24-bit converter all of which would give equal audible results (all other things being equal), because the noisy signal was the limiting factor, not the converter.

However, if the noise IS correlated with the desired signal, then it is a form of non-linear distortion and you've got yourself a defective converter. Most discussion assumes tacitly that the residual noise of a converter will be uncorrelated with the signal, because for decades now, that has normally been the case. That is another way of saying that the non-linear distortion is very low.

Still, the older ones among us, or anyone who's ever used an inadequately dithered A/D converter, will know what that kind of distortion sounds like, because some of the earliest available digital audio recording equipment didn't dither properly (ahem Sony, including their professional PCM-1600 which a lot of early CDs were recorded with). Some people describe it as "noise breathing" or "noise pumping" because it rises and falls with the signal levels, particularly on very soft sounds such as reverberation decay (or imagine taking a 400 Hz tone and fading it down slowly to nothingness with a smooth, analog fader). An accompanying phenomenon is (or was) "digital deafness"--signals below the scope of the lowest-order bit are simply blanked out.

Unfortunately, the way a lot of people think about digital audio is still based on the incredible flood of bullshit that was unleashed when digital audio was introduced around 1980, and which (when it was about anything real) harped greatly on shortcomings of conversion that were really due to not having proper dither. When an a/d converter is properly dithered, there's no increase in distortion at lower levels, no "stair-step levels" to be minimized by striving for more and more bits, nor is there "digital deafness" at the bottom of the range; it acts just like an analog channel as far as dynamic range is concerned. Since actual distortion in modern a/d converters is so low, I tend to shift the discussion away from the concept of "precision" in conversion (which would be a more appropriate concept if a signal were utterly noiseless, such as a computationally generated test signal) to dynamic range, which is much more pertinent to live, recorded audio.

--best regards

Well-stated, eloquently and accurately. I 100% agree with every aspect of this post.

And while the digital “crunch” you mention from poor converters is considered charming or nostalgic by some, it definitely isn’t pleasant to listen to. I’ve heard my fair share of bad converters and understand why digital gets such a bad reputation. 

DSatz I hope I’m blessed enough to tape with you at some point and learn a thing or two from your experience.
Title: Re: What sample rate are you using?
Post by: nak700s on June 14, 2018, 02:44:47 PM
Wow, a big thank you to DSatz, and as always Gutbucket as well as others which made this stuff so much easier to understand.  Luckily, I've been doing the right thing, more or less, possibly, as pointed out above, because I'm an old timer that started out with analogue. Reading these specs put everything in perspective and helped me to understand why I do what I do, and that I may even be able to improve my craft a little to boot.  Can't wait to experiment...Thank you!
Title: Re: What sample rate are you using?
Post by: robeti on June 14, 2018, 04:01:17 PM
I tape, archive and listen at 24/48

Title: Re: What sample rate are you using?
Post by: aaronji on June 14, 2018, 04:49:36 PM
aaronji, howdy.

Howdy yourself, DSatz!  Thank you for the detailed response!

Since actual distortion in modern a/d converters is so low, I tend to shift the discussion away from the concept of "precision" in conversion (which would be a more appropriate concept if a signal were utterly noiseless, such as a computationally generated test signal) to dynamic range, which is much more pertinent to live, recorded audio.

To be honest, I have viewed bit depth solely in terms of dynamic range, based on a lot of articles I have read.  I guess I am one of those folks that just assumed decent conversion.  I always figured that 24-bit would give an extra cushion if I left way too much headroom (as has happened by accident or circumstance a few times), and I will still use it when recording, but the likelihood that the noise floor of anything I record even approaches 16 bits is virtually nil...