The pros and cons of mixing at 44.1 KHz 16-bit

[Manfred Philips]

This is a subject that was suggested in the presentation topic because I was expressing my opinion as an amateur producer to work on so-called CD quality sampling which was for me quite sufficient in quality to release listenable models which returned my first cassettes of 4 tracks at the museum.

 

And what quality are you working on and above all why?

 

[john]

Hi

 

A great question!

 

The shortest answer is to record, mix and master at the highest quality your system can cope with for future proofing your audio. You can lose detail from your audio at any time by reducing quality, but you cannot add back lost detail by increasing quality. Instead, all you do is interpolate values… ie adding pseudo or false detail that merely preserves the same audio signal at that better quality. 44.1kHz 16 bit is a 40 year old standard, but still being used in some 8k video, albeit in 22:2 audio channels. Apart from that you are looking at 48kHz for HD video and 48kHz or 96kHz for 4K (which is also 5:1 surround sound) and 44.1kHz or 48 kHz in 22:2.

 

Video is vital for modern musicians. If you ever plan to have your music used in video formats, then master at 48kHz minimum.

 

Just now 96kHz 24 bit is used by sound designers for movies, audiophiles but as processing continues to speed up audio quality will only improve. Why have your audio languish?

 

Disadvantages of recording, mixing and mastering at higher quality? You need higher spec computer, software, audio interface. That’s it.

 

Cheers

 

John

[dogstar]

I compose for a few different production music libraries and most want 16 bit-48k, so this has become my default setting for mixing and mastering.

[HoboSage]

The 32-Bit Float audio recording revolution and its HUGE headroom before clipping (making clipping and gain staging for it a thing of the past) is underway.  DAWs and VSTs have been using it a long time for internal processing.  Zoom may have have been first out of the gate with a consumer audio interface - this one:.  

 

https://www.amazon.com/Zoom-UAC-232-Converter-Interface-Headphone/dp/B0BW1VDHNS/ref=asc_df_B0BW1VDHNS/?tag=hyprod-20&linkCode=df0&hvadid=647293221758&hvpos=&hvnetw=g&hvrand=9200862543381157809&hvpone=&hvptwo=&hvqmt=&hvdev=c&hvdvcmdl=&hvlocint=&hvlocphy=9009415&hvtargid=pla-1968708094295&psc=1&mcid=df63d616fb4f3e5bbf09b08b69b770d6

 

 

Edited December 25, 2023 by HoboSage

[Mahesh]

I wanna ask because I'm not sure of this myself. If most audio and video platforms online through which audiences consume media, uses upto 16-bit 44.1Khz for relay, would it still make sense to record at higher rates like 24 bit 48Khz and THEN let that high quality audio be rendered down? 

 

Or does it make more sense to record at relay quality from the get go? 

[john]

2 hours ago, Mahesh said:

I wanna ask because I'm not sure of this myself. If most audio and video platforms online through which audiences consume media, uses upto 16-bit 44.1Khz for relay, would it still make sense to record at higher rates like 24 bit 48Khz and THEN let that high quality audio be rendered down? 

 

Or does it make more sense to record at relay quality from the get go? 

 

You can lose detail on going high to low bit depths and sample rates, because the detail is there to be lost.

 

You cannot add detail going low to high bit depths and sample rates, because the detail does not exist. You might be able to approximate pseudo detail, but not the actual signal detail.

 

At a minimum if it overly affects track and effect numbers, record at 48kHz, because of it’s prevalence in video and film.

[Manfred Philips]

That is why I handle all my sound lineage in 44,1 KHz 16 bits called CD Quality so I have not to downsample on mastering.

Higher sample rate is just needed if you handle video for DVD's or other video support.

 

For amateur music imho no need to use very high samplerates.

[Steve Mueske]

I create in 48 32, and tailor rendered files based on medium. It's pretty stupid that most aggregators want 44.1 16 bit, which is, literally, 38 years old in terms of quality. In each case, I dither for the release audience based on format requirements. The raw mixed file at 48 32 makes this pretty easy. I used to record and mix at 96 32, but the audiophile market is pretty small, and I doubt anyone can tell the difference between 96 24 and 48 24. I wish the market supported 32 bit floating point because that would be ideal. In terms of flexibility, this basic approach covers every market I am interested in. Mixing in 44.1 16 means you'll have to resample for any video applications. Downsampling is much better than resampling at a higher rate. FLAC supports 48 24, so that is what I shoot for.

Edited December 27, 2023 by Steve Mueske

[Mahesh]

17 hours ago, Steve Mueske said:

I create in 48 32, and tailor rendered files based on medium. It's pretty stupid that most aggregators want 44.1 16 bit, which is, literally, 38 years old in terms of quality. In each case, I dither for the release audience based on format requirements. The raw mixed file at 48 32 makes this pretty easy. I used to record and mix at 96 32, but the audiophile market is pretty small, and I doubt anyone can tell the difference between 96 24 and 48 24. I wish the market supported 32 bit floating point because that would be ideal. In terms of flexibility, this basic approach covers every market I am interested in. Mixing in 44.1 16 means you'll have to resample for any video applications. Downsampling is much better than resampling at a higher rate. FLAC supports 48 24, so that is what I shoot for.

 

This is super helpful @Steve Mueske. Thank you!

[Glammerocity]

 I found the audio codec in my video editor was dependent on the imported file. If I imported 96KHz I got a popup warning that most codecs only handle up to 48KHz, except PCM and AAC. Youtube recommends AAC anyway, so that was ok but there was a configuration I was using where it was resampling down from 320kbps to 288kpbs. Most of my imported audio is from Youtube at 44.1 anyway, so I work with that now.

 

It's all tech stuff most listeners wouldn't notice, especially in a YT video, and techs do sometimes get caught up in the trees rather than the wood, but it makes me wonder why standard video formats have increased - many more YT music videos are at 4K now and even I'm using it - but audio codecs are slugging along.

[john]

12 hours ago, Glammerocity said:

 I found the audio codec in my video editor was dependent on the imported file. If I imported 96KHz I got a popup warning that most codecs only handle up to 48KHz, except PCM and AAC. Youtube recommends AAC anyway, so that was ok but there was a configuration I was using where it was resampling down from 320kbps to 288kpbs. Most of my imported audio is from Youtube at 44.1 anyway, so I work with that now.

 

It's all tech stuff most listeners wouldn't notice, especially in a YT video, and techs do sometimes get caught up in the trees rather than the wood, but it makes me wonder why standard video formats have increased - many more YT music videos are at 4K now and even I'm using it - but audio codecs are slugging along.

 

I think because home tv/cinema set ups run at higher rates and resolutions, although the trend for ever-higher quality audio has slightly reversed for 8k, largely due to the increase in processing bandwidth for visuals and a perceived lack of need for higher quality audio. There’s an ongoing need for higher visual bit-depth to accommodate perceived visual contrast. The amount of film shot in low light conditions in particular is only appreciated in high contrast televisions. The similar fashion for “realistic” low level audio, especially around speech, was largely already served by the existing higher quality audio, at least to the point that general viewers had got to a point where they could not really appreciate the difference. To a lesser degree, on computers, I think few can afford the expensive processing power required for 8k at least to render in a reasonable time scale especially with the lack of drive for ever-higher quality audio in audiences. Video creators want to spend their processing power at render time on quality visual effects and surround sound audio. Audio processing has dropped back to 48kHz albeit in multichannel surround sound formats… and that is where audio processing has gone. 7 channels @48kHz as opposed to 2 channels at 96kHz or higher.

[MisterB]

On 11/30/2023 at 10:04 AM, Manfred Philips said:

This is a subject that was suggested in the presentation topic because I was expressing my opinion as an amateur producer to work on so-called CD quality sampling which was for me quite sufficient in quality to release listenable models which returned my first cassettes of 4 tracks at the museum.

 

And what quality are you working on and above all why?

 

 

I always end up with 44.1/16 or worse (MP3), but what I work in depends on context. Some plugin effects and some VSTi's work better at higher sample rates, and I always work in 24 bit floating inside of Cubase. I could change over to 32 bit floating now that its available, but I can't really see the point based on my content, which is always sample or algorithm based material from VSTi's. There's no need for anything to be pushed to the limit during the writing and production phase, and I don't have huge templates with masses of instruments, so 32 bit floating is kind of redundant to me. Since I'm 100% in the box, if someone asked me for a file at a really high sample rate, I could just convert my project.  Whether or not it would make a difference would likely depend on the VSTi's I'm using, but I sure as hell can't hear a difference on final output.

 

I've used all kinds of audio sources on videos in the past with all manner of conversions to meet an output spec. No one cares. No one ever comments on the "low sample rate" of a source.