Rendering 16 or 24 bit?

[Jmax]

Are you a 16 bit guy or a 24 bit? also 44.1 48?

16 bit will only play through CD players. Yet, 24 bit seems to be the new norm. There is a big difference in file size. I can tell the difference a little I think I need to listen more between 16 and 24 sound. I'm a little unsure what to render to these days so often do both. A lot of publishers prefer and are ok with 16 though it seems to be industry standard but perhaps that is changing.

What do you guys normallly do?

[Rook]

Last time I tried, I couldn't hear the difference. My stuff only ever goes on Soundcloud, if it even makes it that far. So, I just go 16 bit. I guess if I were trying to be super "pro" and sending stuff to someone else for mastering or something, I'd go 24 because...reasons

[gak]

Well, once you upload it to "discover" or "soundcloud" it won't matter (man, both horrendous sound quality!)

[Jmax]

Last time I tried, I couldn't hear the difference. My stuff only ever goes on Soundcloud, if it even makes it that far. So, I just go 16 bit. I guess if I were trying to be super "pro" and sending stuff to someone else for mastering or something, I'd go 24 because...reasons

I hear you on that one. I guess it depends what your using it for. But still I believe there is a difference in the sound quality. I think 24 bit captures a little more. Perhaps a little more richness to the sound. Your right though, most will never know the difference, including me sometimes. It would be interesting doing a blind hearing test.

[Benedict]

16/44 is still the standard for final render as that is what goes on CD. Also virtually no players will play 24 bit files yet. Higher rates my sound a tad nicer but you are only slitting your throat unless you are releasing one of those "audiophile" CDs. Don't let rate snobs suck you into their rabbit hole.

[Jmax]

16/44 is still the standard for final render as that is what goes on CD. Also virtually no players will play 24 bit files yet. Higher rates my sound a tad nicer but you are only slitting your throat unless you are releasing one of those "audiophile" CDs. Don't let rate snobs suck you into their rabbit hole.

Very interesting to hear Benedict. I will continue to render at 16 I believe. But do you yourself notice any difference in the sound? the file size is much larger. Also have you any idea what Discover renders audio to? When you upload to them. They must convert it into a really small mp3? 128 who knows. Perhaps another reason why discover music isn't really ever going to be published and is just a sharing/creating thing.

[Benedict]

In my experience 24 bit seems slightly different but that is often down to the way the synth engines render and not representative of the final output anyway (assuming working at 24/96 and rendering at 16/44). I'd say it was better to work as close to what the CD will sound like is the smart way to go. And yes you will save a heap of data too.

I would assume that Discover is like Soundcloud etc and offers a 128 kps stream and then up/downloads a FLAC or Wav file. I would think that people saying that SC chews or otherwise damages their tracks are either being super-sensitive or making mixes that don't translate to the real world well. I can't say I have ever felt any concern over my pieces on SC or Bandcamp etc.

To me it is the story of the music that matters and not the ability to hear the air of the hats at 15kHz. I recorded my whole music collection (vinyl & CD) to mp3 at 196 VBR and sometimes I hear some issues but mostly I am enjoying the music. Having it all so accessible is worth far more to me than obsessing over a few seconds of crystallized sound.

[Jmax]

Yes I think your right about the synth engines. The audio is perhaps being rendered slightly differently. And sticking to CD quality is a very good idea.

I've never heard anyone saying Soundcloud chews up their sound. Quite the opposite in fact and is a great resource. Perhaps Youtube but that's a different story.

You are completely right about the story of music. You can have a 128 mp3 playing and if it's a beautiful song then that's all that matters.

[Gaja]

I might have gotten it wrong, but afaik the bit depth covers the dynamic range of the signal, (with 2^16 compared to 2^24 different possibilities of level) so it shouldn't have any audible impact on the sound itself, only on signal to noise ratio and overall dynamic range. Now when you're talking sample rate there can be audible differences, for example with synths that don't have integrated oversampling (like thor) thor will sound different on 96kHz/s than when you export it to 44.1 (need to export to 96 first, then downsample the audio when already bounced).

[normen]

As Gaja said, bit depth _solely_ defines the signal to noise ratio of a digital recording, nothing more. Each bit gives you 6dB of S/N ratio, giving you 96dB of S/N ratio for 16bit files and 144dB S/N ratio for 24bit files. 96dB alone is probably more than some of the analog components behind the D/A playing your music can do and given that most music doesn't have more than 24dB of effective dynamic range its well enough for a final output.

However, when you export stems to be imported for further processing or generally export for further processing you want to keep the maximum possible S/N ratio and export at 24bit.

As for sample rate, just keep the sample rate consistent during the whole process. If you recorded / composed at 44.1, keep it that way. As gaja also indicated the outcome of sound generation and processing is somewhat dependent on sample rate (in some cases) although the difference in a recording itself isn't really distinguishable between 44.1k and 192k (don't let Neil Young tell you otherwise ). Reducing (or upping) sample rate is a relatively complicated process and Reason does admittedly not have the best sample rate conversion on the market.

As soon as you're going to compressed formats like mp3 the concept of sample rate vanishes anyway as these types of compression store the data in the frequency domain and not in the time domain. "Inside" an mp3 there is no such thing as sample rate, only when its decompressed by the player this concept comes back in and the quality depends on the decoder (and encoder of course).

[Kenni]

http://people.xiph.org/~xiphmont/demo/neil-young.html

There's no audible difference between 16 and 24 bit.

[submonsterz]

16/44 is still the standard for final render as that is what goes on CD. Also virtually no players will play 24 bit files yet. Higher rates my sound a tad nicer but you are only slitting your throat unless you are releasing one of those "audiophile" CDs. Don't let rate snobs suck you into their rabbit hole.

Very interesting to hear Benedict. I will continue to render at 16 I believe. But do you yourself notice any difference in the sound? the file size is much larger. Also have you any idea what Discover renders audio to? When you upload to them. They must convert it into a really small mp3? 128 who knows. Perhaps another reason why discover music isn't really ever going to be published and is just a sharing/creating thing.

discover seems to use M4A format .

[Goriila Texas]

The question should be who really still uses cd's? Most car radio's have a usb connection where you can hook your phone up and play music. At home I play whats on my computer. So my question is should we really stick to the standard for cd's?

[4filegate]

Bit Depth
A recording or selected RE/Refill on a 24 bit DAW such acoustic guitar as highly dynamic instrument - the difference will be there. It's not that 24 bits of data makes the sound better. It actually does not. Just build on your audio more room to breathe in the numeric realm of digital audio.
I talk about numbers, calculations, not analog waveforms. With 24 bits of data on your DAW, it's possible mixing extremely dynamic music, with very quiet soft passages and extraordinary loud passages. Quiet passages will be less likely struggling to stay above the noise floor on your system. One can record with no compression at lower levels, with more headroom. This ensures that the occasional peak is not truncated at the top and it give converters some room to breathe. Not pushed by the limits of the 16 bit bandwidth - instruments find good placement, the song will mix in a better way and there will be less noise. In fact they may sound just as bad or worse than 16 bit. But 24 bit give a noise floor and better headroom. It's a tool that can blow you away.

why I focus at HD audio RE/Refill and playground SC for dwnld in the upload format

[normen]

The question should be who really still uses cd's? Most car radio's have a usb connection where you can hook your phone up and play music. At home I play whats on my computer. So my question is should we really stick to the standard for cd's?

I think the question would be why do you need additional S/N ratio in the digital domain that isn't avilable in the analog domain. Theres almost no converters that actually reach 144dB of S/N ratio. The fact that 16/44.1 is CD standard is just a sidenote in this case.

[Exowildebeest]

Music in my "scene" i.e. progressive house, dubtechno, electronica etc., is supplied to the labels as 44.1 16bit masters.

Whatever happens before delivery of the final master, in production and processing, is up to the producer and mastering engineer. The benefits and downsides of higher sample rates in Reason are well known by now. Higher resolution CV, better/different sounding synths due to "forced oversampling", possibly audio recording/timestretch differences - all at the cost of CPU and memory.

[mcatalao]

(...)96dB alone is probably more than some of the analog components behind the D/A playing your music can do and given that most music doesn't have more than 24dB of effective dynamic range its well enough for a final output.

I have to digress on this. Maybe true for most music done nowadays in the Pop genre, but if you hear styles like jazz and its derivatives, New age, and a bunch of world music styles (for example Portuguese fado, or for example norwegian folk inspired jazz - Jan Garbarek anyone?), you will have very dynamic stuff from 50 db up! Not to mention that many classical music concerts can have a wide dynamic range of 60 to 80 db's.

Most good monitors, and standard gear, made up to the 2000's (not hedious mp3 players and yucky soundbars) and still today's good hi-fi stuff will have sensitivities of -80 to -90 dbv from unity.
Of course for each component you put on the chain you will take out some sensitivity, but at max, most complete systems have a -80 db sensitivity, and that's A LOT!

[normen]

I have to digress on this. Maybe true for most music done nowadays in the Pop genre, but if you hear styles like jazz and its derivatives, New age, and a bunch of world music styles (for example Portuguese fado, or for example norwegian folk inspired jazz - Jan Garbarek anyone?), you will have very dynamic stuff from 50 db up! Not to mention that many classical music concerts can have a wide dynamic range of 60 to 80 db's.

Most good monitors, and standard gear, made up to the 2000's (not hedious mp3 players and yucky soundbars) and still today's good hi-fi stuff will have sensitivities of -80 to -90 dbv from unity.
Of course for each component you put on the chain you will take out some sensitivity, but at max, most complete systems have a -80 db sensitivity, and that's A LOT!

Well basically you're agreeing The tech around the A/D D/A doesn't support much more than 90-100dB S/N ratio and theres only few music that has very high dynamic range.

Sure an actual orchestra in a concert can have more than 80dB dynamic range but few recordings of orchestras do. Theres a big difference between hearing music in a live context and hearing it on speakers. If you'd actually use 120dB of dynamic range in a recording you'd basically go from inaudible to damaging your listeners hearing, so...

[Biolumin3sc3nt]

Chalk me up for 24bit! I've been using it for the last 10+ years and I'm not going back! I think the real question/argument is whether to use 44.1 / 48 / 88.2 or 96k! For plugins (VI), there's a noticeable difference between using lets say 44.1 vs 96k - at least for me.

[ScuzzyEye]

I think the real question/argument is whether to use 44.1 / 48 / 88.2 or 96k! For plugins (VI), there's a noticeable difference between using lets say 44.1 vs 96k - at least for me.

I have an anti-aliasing filter that I'm working on. There's a noticeable difference between 1x, 2x, 4x, and 8x, but regardless of the internal oversampling, it's always outputting at 44.1 kHz in the end. So the difference isn't in the sampling rate, but how well aliasing artifacts are minimized.

I tried to argue a (long) while back on the PUF that I could easily hear the difference with Thor running at 44.1 vs. 96. What I didn't do was export both, and then down sample the 96 to 44.1 using Audacity (Reason is OK now, as it's down sampling has improved). After I did that, and could still hear the difference between the 44.1 and 96k files, I knew it wasn't the sampling rate.

[Biolumin3sc3nt]

Thanx for replying Scuzzy. I regard your technical know how in high regard! I was just looking at the Graphs of down sampling from a Gearslutz link the other day. Fart, I forget what site it was, but yeah Audacity (the earlier versions) and Reason didn't look too good. Anyway, I'm not an Engineer in the classical sense, however I understand what "Aliasing" is, and what my ears tell me. Do you remember that Lavry article that mentioned 60k to be an ideal rate? I'm interested in what comes of this filter. Still learning new things everyday : )

[ScuzzyEye]

The site with the graphs was probably this one: http://src.infinitewave.ca/

I don't think I saw that article you mentioned. With digital oscillators, which haven't been specifically designed to minimize aliasing, 60 kHz would likely prevent most of the reflected energy from reaching back down into the audible range. But after something is rendered at a higher frequency, and brick wall filtered to remove everything at or above the Nyquist limit of 44.1 kHz, it can be down sampled and should sound just as good.

[Biolumin3sc3nt]

Yes indeed Sir, that was the link. Brick wall filtering is a " Doozy". In my understanding, that's what separates the good from the bad n ugly?
I'm sorry I'm having a few drinks at the moment - it might be best if I put the puter to bed

[mcatalao]

Well basically you're agreeing The tech around the A/D D/A doesn't support much more than 90-100dB S/N ratio and theres only few music that has very high dynamic range.

Sure an actual orchestra in a concert can have more than 80dB dynamic range but few recordings of orchestras do.

I was disagreeing about the 20 dB's you stated. And i still disagree about that for some normal rock music, reverb tails on stops, benefit from more headroom except if you squeeze the hell of you tracks at mastering.

I have the full Rach recordings, as matter of fact everything from the romantic period will have 70 dB+ for the whole concert, they span from a simple piano line at ppp, to full fortissimo with the whole orchestra playing.

[normen]

I was disagreeing about the 20 dB's you stated. And i still disagree about that for some normal rock music, reverb tails on stops, benefit from more headroom except if you squeeze the hell of you tracks at mastering.

I have the full Rach recordings, as matter of fact everything from the romantic period will have 70 dB+ for the whole concert, they span from a simple piano line at ppp, to full fortissimo with the whole orchestra playing.

So everything is well within the 96dB of a 16bit file - all good.

I don't know what "the" Bach recordings are supposed to be but as recording technology only had about 60dB S/N ratio at best just a few decades ago I take it you're talking about some recent recordings?

Edit: Sorry, misread that - you said "Rach", which I suppose means "Rachmaninow"?