Fascinating!

Freebird

Some people are so clever 

Paul_C

I enjoyed that - I enjoyed his voice moving around in stereo too

Freebird

Paul_C said:

I enjoyed that - I enjoyed his voice moving around in stereo too

I learned quite a few things from that, very interesting and useful stuff. I too was a square pixel and stairstep digital waveform believer! 

Here is the original Neil Young article:

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

And the original video in stereo    

https://xiph.org/video/vid2.shtml

Winny_Pooh

That was excellent. This guy would destroy a flat earther

Freebird

Episode 1: (A digital media primer for geeks)



Original source: https://xiph.org/video/vid1.shtml

Freebird

Am I right in thinking that to make digital audio sound "good", you have to degrade the audio quality by using some analogue emulation plugins of old studio favourites, such as tape, console, line in, and pre amp circuitry, and maybe run it through a vinyl emulation too?

Winny_Pooh

Freebird said:

Am I right in thinking that to make digital audio sound "good", you have to degrade the audio quality by using some analogue emulation plugins of old studio favourites, such as tape, console, line in, and pre amp circuitry, and maybe run it through a vinyl emulation too?

Not quite, analog tape worked because it used compression and does 'soft clipping' as the signal peaks go over a certain point. And this is a trick to emulate the human ear's perception of loud noises to some extent. 

Vinyl does not sound good and has always been a poor way to represent the sound of a finished mix on a studio tape machine IMO.

Plugins are polish. Unless you start with a great source and great mic your results will always be substandard regardless of whether whether you're using a console and tape machine or a HD Protools rig from 15 years ago.

Freebird

Winny_Pooh said:

Freebird said:

Am I right in thinking that to make digital audio sound "good", you have to degrade the audio quality by using some analogue emulation plugins of old studio favourites, such as tape, console, line in, and pre amp circuitry, and maybe run it through a vinyl emulation too?

Not quite, analog tape worked because it used compression and does 'soft clipping' as the signal peaks go over a certain point. And this is a trick to emulate the human ear's perception of loud noises to some extent. 

Vinyl does not sound good and has always been a poor way to represent the sound of a finished mix on a studio tape machine IMO.

Plugins are polish. Unless you start with a great source and great mic your results will always be substandard regardless of whether whether you're using a console and tape machine or a HD Protools rig from 15 years ago.

I'm actually experimenting with Nebula sampling technology at the moment, and I have programs that have sampled the circuitry for various kinds of tape, valve distortion, DI input, Line input, Mic Preamp, Buss Groups, and Mix Bus Out, etc.,

My initial audio is pristine, so my experiments involve enhancement (degregation?) of these audio signals, and I can pretty much emulate the entire analogue chain from the Pre Amp, through the console, and onto different kinds of tape.

It's still too early to call, but I will continue to plod along, as it is a fairly new and interesting area of technological development. I would recommend giving some of the new Acustica Audio demos a go, particularly Pink, Gold and Sand (API, Neve & SSL), although I am mostly using the older Nebula 3rd party stuff.

Winny_Pooh

Freebird said:

Winny_Pooh said:

Freebird said:

Am I right in thinking that to make digital audio sound "good", you have to degrade the audio quality by using some analogue emulation plugins of old studio favourites, such as tape, console, line in, and pre amp circuitry, and maybe run it through a vinyl emulation too?

Not quite, analog tape worked because it used compression and does 'soft clipping' as the signal peaks go over a certain point. And this is a trick to emulate the human ear's perception of loud noises to some extent. 

Vinyl does not sound good and has always been a poor way to represent the sound of a finished mix on a studio tape machine IMO.

Plugins are polish. Unless you start with a great source and great mic your results will always be substandard regardless of whether whether you're using a console and tape machine or a HD Protools rig from 15 years ago.

I'm actually experimenting with Nebula sampling technology at the moment, and I have programs that have sampled the circuitry for various kinds of tape, valve distortion, DI input, Line input, Mic Preamp, Buss Groups, and Mix Bus Out, etc.,

My initial audio is pristine, so my experiments involve enhancement (degregation?) of these audio signals, and I can pretty much emulate the entire analogue chain from the Pre Amp, through the console, and onto different kinds of tape.

It's still too early to call, but I will continue to plod along, as it is a fairly new and interesting area of technological development. I would recommend giving some of the new Acustica Audio demos a go, particularly Pink, Gold and Sand (API, Neve & SSL), although I am mostly using the older Nebula 3rd party stuff.

No harm in that, especially if you focus on mixing. However, if you want professional resuts pre and post processing can only do so much vs just switching out a mic. I've previously tried to match two tracks recorded with different mics and post processing still can't get close enough to compensate for failings in the source signal and capsule of the cheaper (by £700) mic. 

Freebird

Winny_Pooh said:
No harm in that, especially if you focus on mixing. However, if you want professional resuts pre and post processing can only do so much vs just switching out a mic. I've previously tried to match two tracks recorded with different mics and post processing still can't get close enough to compensate for failings in the source signal and capsule of the cheaper (by £700) mic. 

My recorded sounds are generally ok, as I track through a Kemper Profiler via SPDIF, and I use Roland Vdrums with Superior Drummer 3. My vocal microphone is a tube equipped SE Electronics z5600a II, which isn't too bad, but I'd probably be looking at tracking the vocals in a proper studio if the project required it, and maybe the drums too.

The Nebula stuff is just experimentation, and my main aim is being able to compose and record demo tracks of sufficient and improving quality. If anybody ever gets to a commercial level, it's probably best to farm out the recording, mixing and mastering to those who know what they are doing.

sev112

I really really enjoyed that and I think I learned some interesting stuff.
however, I really didn’t have a clue what most of it was about !

so why does sound follow a sine wave - can someone please explain, or point me to a primer or idiots guide that explains that please

Freebird

sev112 said:

I really really enjoyed that and I think I learned some interesting stuff.
however, I really didn’t have a clue what most of it was about !

so why does sound follow a sine wave - can someone please explain, or point me to a primer or idiots guide that explains that please

The main take away for me, is that 16 bit (or 8 bit) is as good as 24 bit for capturing source information when converting an audio signal from analogue to digital, and amazingly you can reproduce the same audio signal with a subsequent digital to analogue conversion. The only differences between each bit-depth are the noise levels, which can't be heard once you get to 16 bit.

Regarding sine waves, I know that sound travels through the air in waves until it hits the membranes in our ears, and then our brain turns it into sound. Maybe somebody else can chime in with a more technical explanation 

wave100

Freebird said:

sev112 said:

I really really enjoyed that and I think I learned some interesting stuff.
however, I really didn’t have a clue what most of it was about !

so why does sound follow a sine wave - can someone please explain, or point me to a primer or idiots guide that explains that please

The main take away for me, is that 16bit (or 8bit) is as good as 24bit for capturing source information when converting an audio signal from analogue to digital, and amazingly you can reproduce the same audio signal with a subsequent digital to analogue conversion. The only differences between each bit-depth are the noise levels, which can't be heard once you get to 16bit.

Regarding sine waves, I know that sound travels through the air in waves until it hits the membranes in our ears, and then our brain turns it into sound. Maybe somebody else can chime in with a more technical explanation 

I'll play - I haven't watched the videos though so can't comment on them. However if the take away is that 8 or 16 bit is as good as 24 bit, that's just wrong. The bit depth defines the available dynamic range of the converted signal ie the difference in volume between the quietest and loudest things you can record. Using 24 bit means you can leave tonnes of headroom while recording thereby avoiding overloading the analogue electronics of your system and also peaks in the signal going above 0dB after which you "run out" of bits which causes nasty distortion which you want to avoid - as opposed to nice sounding analogue distortion that all guitarists like. So it is good practice to record at 24 bit with signals peaking about -12 on the meters. Plug ins are also designed to work with plenty of headroom so sound better this way. 

The frequencies you can record are defined by the sample rate, and according to the Nyquist theorem, the highest frequency you can record is 1/2 the sample rate so if you record at 44.1 KHz the highest frequency you can record is 22.05 KHz, just above the range of human hearing.

As for the sine wave thing, I suspect they are just used as an example of a simple sound wave - an organ or flute can make sounds which are <close> to sine waves. An old French dude called Fourier decided that any sound timbre can be recreated by combining sine waves of different frequencies in the right proportions - this is how Additive Synthesis works. However from the context I suspect this is not what is being referred to.

Hope that helps.
;-)

Freebird

I'll add this video which further explains what I wrote above

wave100

OK, a couple of points - that vid is talking about replaying digital audio, not recording it, so if you are recording multiple tracks at less than 24 bit, the noise which he demonstrates is multiplied with each track you record. I just thought that you were implying in your post  that there was no point in using higher bit depths for recording, which could give a misleading impression, especially to beginners.

Winny_Pooh

wave100 said:

OK, a couple of points - that vid is talking about replaying digital audio, not recording it, so if you are recording multiple tracks at less than 24 bit, the noise which he demonstrates is multiplied with each track you record. 

Fallacy

Musicwolf

The guy in the video is a very good communicator and what he presented was correct, however, I completely agree with wave100 that he failed to get the point across that the benefit of 24bit vs 16bit is the dynamic range which means that you do not have to record 'hot' as we did back in the days of tape (in order to achieve the best possible signal to noise ratio).  Yes, 24 bit eats up more storage space but that really isn't an issue these days.

The video, whilst making reference to Nyquist, failed to make the point about highest frequency equaling half the sample rate which would have been informative.  Understanding about Fourier transforms would have helped but that would have been a whole new video.

Freebird

wave100 said:

OK, a couple of points - that vid is talking about replaying digital audio, not recording it, so if you are recording multiple tracks at less than 24 bit, the noise which he demonstrates is multiplied with each track you record. I just thought that you were implying in your post  that there was no point in using higher bit depths for recording, which could give a misleading impression, especially to beginners.

Yes, we may be discussing two different topics here. The guy in the original video was responding to the Neil Young/Steve Jobs proposal of 24 bit 192kHz music downloads, whereas us lot on here are also interested in the recording side of things.

I always record using 24bit for the headroom, but I still found the original video very informative, especially the bit about the stairstep waveforms.

I also think people prefer listening to music recorded on analogue equipment, as digital sounds too pure and it needs to be dirtied up a bit, hence the trend for analogue emulation plugins.

wave100

Winny_Pooh said:

wave100 said:

OK, a couple of points - that vid is talking about replaying digital audio, not recording it, so if you are recording multiple tracks at less than 24 bit, the noise which he demonstrates is multiplied with each track you record. 

Fallacy

How so?

Nerine

This is one of the problems with digital audio. It makes testing very easy for the people that like arguing about it on Internet forums. 
Some of the best songs ever written/recorded are probably mixed very averagely by today's standards. Does it detract from the beauty? No. Does it enhance it? Probably, yeah. 
There is literally no point in arguing about this kind of thing. 
The fact that anyone even cares enough about it to argue about it is basically lunacy. They're tools. Use them. Don't use them. They all sound good. We have it a lot better nowadays than ever before. Has that improved music? No. Probably not. 
Concentrate on writing a good song rather than worrying that your oh so precious fidelity, sample rate or bit depth has got to go through another conversion which might degrade the audio to the point where no one will notice nor give a fuck. 

If it sounds good, it is good. The minutiae that internet engineers pore over is ridiculous. I've been doing it professionally for about 10 years or so now, and yes, of course I have my preferred sample rates and bit depths etc. But fucking hell, am I bothered if I have to sample rate convert some audio a couple of times?? NOPE. Has anyone EVER NOTICED?? NOPE. Have I noticed?! NOPE. Not enough to concern myself with. 

This isnt aimed at anyone, by the way. Pro audio and mixing and engineering is just a pointless thing to argue about. No one's ears are the same, we perceive things differently, we are prone to confirmation bias, we make changes to an EQ whilst it's accidentally in bypass and think we've improved shit... (We've all done it.) 

NOPE. 

Winny_Pooh

wave100 said:

Winny_Pooh said:

wave100 said:

OK, a couple of points - that vid is talking about replaying digital audio, not recording it, so if you are recording multiple tracks at less than 24 bit, the noise which he demonstrates is multiplied with each track you record. 

Fallacy

How so?

@wave100 easier to demonstrate than explain. Create a track in your DAW and record a tone at say 800hz into it or a synth drone. Now copy and duplicate that track 20 times and play all at once. Is the tone louder?