Technical: bit rates, sampling frequencies, and digital audio.

[APK]

This is a mighty fine technical essay on digital audio that shows why higher and higher bit and sampling rates is a waste of time ... and space. Well worth the read if you don't already know this stuff. Or if you thought you did, but actually dont 

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

[El culto]

Old story, but just let people believe in this myth…there is clearly a market with huge money for it.

Congrats to the industry 

[Julio Di Benedetto]

Thats a great essay APK.

I have been running my studio through a Lavry AD/DA setup for quite a while.  The man will not create anything above 96k.  Below is an introduction to a paper he wrote.


The Optimal Sample Rate for Quality Audio By Dan Lavry, Lavry Engineering Inc. May 3, 2012

Imagine that you and a friend were standing at the base of a lone hill in an otherwise flat plain, and you start walking up that hill towards the top. Now imagine that when you reach the top, your friend tries to convince you that you can go higher up the hill if you simply walked twice the distance forward from the base. You would probably pause for a moment, look around, and quickly realize that continuing forward would just take you back down the other side of the hill! This concept of an optimum is not hard to understand, but in digital audio circles more and more people are taking the word of their friend the salesman, who tells them to keep on walking.
In this paper, I will cover some of the myths of higher sampling rate and illustrate how higher sampling rates can actually reduce accuracy in audio conversion. Moreover, I will attempt to elucidate the existence of an optimal sample rate and how conversion at higher or lower rates compromises the accuracy of the audio signal......

Heres the link to the continued white paper highlighted above. Have to say a fair bit of this goes way over my head   http://www.lavryengineering.com/pdfs/lavry-white-paper-the_optimal_sample_rate_for_quality_audio.pdf

[petekelly]

'Music for Bats' ?

Of course 24/192 is the way forward - but more for our canine friends than us

[ffcal]

The jury still out for me on this one.  I've seen similar articles.  Robert mastered my recent recording at a 24-bit level and it seems to have more definition to me than the 16-bit version.  It don't think it's a placebo effect because I experienced the same thing within my own software.  We'll see.

Forrest

[Ekstasis]

This is a pretty well known article, on studio forums, there is usually two parties, one for and one against bitrates.
If this article was made so that the opposite side could respond to each argument in the article I am sure the opposite side would
sound just  reasonable.  we do choose what HiFi philosophies  to believe of course as I said there will always be different parties.

I think it is a mistake to generalize this, for me it depends on more WHAT you use this high sample rate for,  and also what DAC design is used, to say that high sample rates is useless for all designs ,  the main argument with higher sample rate is to reduce processing in the DAC,  to reduce the need for upsampling in the interpolator component in the DAC, the DACs today are working in 5 bit audio, which needs  audio that comes into the DAC needs to be upsampled up to 30 times of the audio you feed the DAC with.  In a perfect world you would not need any upsampling within the DAC, but all DACs today need it (besides DSD format DACs) since the DAC chips today is working in 5 bit audio. So the argument of high sample rate seem to be mostly about the internal processing and not about the actual human/physical abilities to hear the difference of the high sample rate, it is more about internal signal processing, and how the DAC is responding to a higher sample rate and having a less need for upsampling,

In the real world however,  problems such as CPU power is a bigger problem, I am still using 44khz myself because of limited CPU power, and this problem will most likely remain for a long time too, because intel has a slow progression and within the small steps of cpu power progression software developers are also fast to respond with more cpu hungry plugins, so I do not seem myself going over to 96k in the next 5 years to be honest. The other problem is that all music today is downsampled to 44khz anyway,  even though lossless services do exist they are still using the CD format often, 16bit/44khz. To record something in high sample rate  and down sample it later just seem like a big hassle for nothing.

High sample rate is simply something for the future,  from a sonic quality perspective there is only benefits,  but the drawbacks with higher sample rate are bigger,   cpu power is one of them, but also file size and also in the age of audio streaming the network bandwidth, which drains much battery from people with mobile devices etc,   we need more headroom in both processing power in computers, network speeds and also power in mobile devices until we are mature to change to a higher sample rate standard.  I do hope however that more music is released in higher bit resolution/sample rate.  I myself find always that when I downsample my projects to 16 bit there is a big difference in sound quality,  and from wav to mp3 is even omre noticeable, I would not have heard the difference before, but with the more sensitive equipment i hear now it is more obvious,  with that said.. 99% of the music I listen to is MP3,  but the reason why is because of the problems I mentioned earlier,  file size is a one of them, and availability is another, to have a big archive of flac music takes just a lot of space.. but it is only when you compare the mp3 and flac version side by side you know what you are missing.. Spotify shold be illegal, they are the biggest threat to music, they are like a weapon of mass destruction against music, their OGG music sounds just terrible..destroys all what makes music good.

[ffcal]

One thing that bugged about the article at the top was a focus on whether the listeners thought they could hear a difference.  That really isn't the question for me at all.  If it were, I would have released everything on MP3 at a 256 kbps bitrate.  The test for me has always been whether I can hear the difference and whether that difference more accurately reflects what I intended.  On that score, I would lean in the direction of 24-bit masters for myself, though that doesn't mean I would want or expect the same as a consumer of other musicians' work.

Forrest

[Ekstasis]

24 bit is definitely no placebo effect.. It is always frustrating to have to down sample to 16 bit.. When you  hear such drastic difference.. But it does also matter what music..  But especially with ambient music I hear a difference..  If I buy an digital flac album I would easily choose the 24 bit if there was a choice.

[El culto]

24 bit is definitely no placebo effect.. It is always frustrating to have to down sample to 16 bit.. When you  hear such drastic difference.. But it does also matter what music..  But especially with ambient music I hear a difference..  If I buy an digital flac album I would easily choose the 24 bit if there was a choice.

Many blind tests regarding 16 vs 24 bit have proved that peopled can´t hear a difference. In any case, when it comes to CD anything has to go down to 16 bit anyway. To hear a difference especially (!) in ambient music is a huge myth for me. When it come to classical music with a lot (!) of dynamics there is a reason for this higher bit/sample rate rate, but when when it comes to Ambient I don´t know any music in this genre using the full dynamic making it worth at all to for the additional headroom nor a higher sampling rate.

Are people really be able to hear a difference between 192kbit and 320kbit mp3?   

But OK, it´s usual that those believers feel better if they can say they feel have the tools or special gear (the price for the additional investment in those cases needs an explanation too) or rendering options. I´m pretty sure, if I would offer 2 unnamed files for a blind testing concerning this issue, those people would avoid to be part of that test.

 

[Julio Di Benedetto]

Recording and mastering @ 24bit makes sense to me.  Why not capture more detail if its available and when I switched from 16 bit to 24 bit recordings there was a difference.  The thing is I now have to use words like "more presence" to express what I mean and this is not science but emotion.  I cant prove it but I hear it. 

Now how high you go in sample rate......music for bats says it all.  I have always recorded at the target cd rate, 44.1khz. because I dont feel comfortable with up and down sampling.  For what you gain you will loose or add what was not original there so for better or worse I keep it simple and the only conversion is from 24 bit to 16 bit. 

in other media such as photography there is absolutely no argument over image quality.  An 8"X10" negative captures so much more information than a 35mm, or I should say a 50 megapixel digital back on a medium format camera and a full frame 35mm sensor.  Same size print...the difference is astonishing.  Perhaps our eyes are more sensitive than our ears....i dont think so.  Our brain is not as well trained to understand the information the ears delivers compared to our eyes...perhaps, but the key word is training.  With sight you get what you see...it is how you perceive it that changes what you see, not the actual information. Is the same true for the ears?

[petekelly]

I really don't know what to make of this, on one hand 24 bit makes more sense to me than higher sampling rates. While there are a lot of myths around, some of these are also on the technical side. The 'boffins' can tend to follow a lot of accepted wisdom regarding how we hear, without challenging this.

[El culto]

I really don't know what to make of this, on one hand 24 bit makes more sense to me than higher sampling rates. While there are a lot of myths around, some of these are also on the technical side. The 'boffins' can tend to follow a lot of accepted wisdom regarding how we hear, without challenging this.

Reminds me of something:

Once I´ve spend hours on a track to find the right EQ settings in den lower mid range and with each new listening i´ve found new bothering frequencies/resonances. When coming back from a short coffee break, to finally confirm my last editing, I realized that the EQ was all the time bypassed…...

[Ekstasis]

24 bit is definitely no placebo effect.. It is always frustrating to have to down sample to 16 bit.. When you  hear such drastic difference.. But it does also matter what music..  But especially with ambient music I hear a difference..  If I buy an digital flac album I would easily choose the 24 bit if there was a choice.

Many blind tests regarding 16 vs 24 bit have proved that peopled can´t hear a difference. In any case, when it comes to CD anything has to go down to 16 bit anyway. To hear a difference especially (!) in ambient music is a huge myth for me. When it come to classical music with a lot (!) of dynamics there is a reason for this higher bit/sample rate rate, but when when it comes to Ambient I don´t know any music in this genre using the full dynamic making it worth at all to for the additional headroom nor a higher sampling rate.

Are people really be able to hear a difference between 192kbit and 320kbit mp3?   

But OK, it´s usual that those believers feel better if they can say they feel have the tools or special gear (the price for the additional investment in those cases needs an explanation too) or rendering options. I´m pretty sure, if I would offer 2 unnamed files for a blind testing concerning this issue, those people would avoid to be part of that test.

Two unnamed files created to prove your point or two unnamed files to show there is a difference.  ?
Tests can be designed with different goals. 
I myself I hear a difference and I know a lot of other people to do,  you need sensitive equipment and ears and know what you are looking for.  In heavy high energy distortion  guitar I think the 16 bit have harsher edges to me. 

I think it is funny say  classical music has more complex audio information then in ambient music.
Ambient can be as complex as audibly possible. Classical music in general consider very lofi in comparison to modern music.  Adding max amounts of transients is usually the best way to find differences in 24 bit or 16 bit  or different sample rates.  On all my recordings in my signal chain I always use audio processors to maximize transients,  especially with ambient there is nothing worse then ambient where all transients are drown into reverb tails,  the transients bring out all details and making complex audio harmonics.  Yeah so the amount  of transient information is probably one thing that matter,  without the complex audio information it might be possible there is no difference in sonic experience in resolution or sample rates

[El culto]

I think it is funny say  classical music has more complex audio information then in ambient music.
Ambient can be as complex as audibly possible.

I think you completely misunderstood something…I´m not talking about "complex" at all!

"When it come to classical music with a lot (!) of dynamics there is a reason for this higher bit/sample rate rate, but when when it comes to Ambient I don´t know any music in this genre using the full dynamic making it worth at all to for the additional headroom nor a higher sampling rate."

[petekelly]

Here we go again...?

[Altus]

[...] I realized that the EQ was all the time bypassed...

Been there, done that. 

[Seren]

[...] I realized that the EQ was all the time bypassed...

Been there, done that. 

me too.

I prefer the 24 bit myself and I think my VS2480 is set up for 96K....

I worked in British Telecoms as a youngster and we had to send test tones down the lines - we used to play around and I could always hear higher frequencies than anyone else (whether this would still be true after time spent in Motorhead bass bins is another matter).

I have the korg MR1 with its  1-bit/2.8 MHz set up - the sound quality is amazing - I seem to remember it being said that this is SACD technology and quality....SACD is considered better quality than CD - is that reality?

I prefer to record in as good quality as possible and work down. I find the higher sampling rates allow more subtlety as well as dynamics in my sounds and music - I know that the difference is much more obvious at lower rates, where things like stereo image can be lessened and sounds become more muddy..