How do you master a DSD recording?

[71 dB]

Because you can see the pulse height and width change as well as latency, and the introduction of pre and post ringing on the waveform.

Seeing is not hearing. As digital PCM audio is bandlimited, it is not optimal for very short impulses that ask for very large bandwidth, but luckily it doesn't have to be, because music is pretty much NEVER just separate impulses. The pre and post ringing happens, because of the limited band, but you only "see" it in very short pulses/impulses. The longer your signal is, the less ringing can be seen. In theory infinitely long signals have zero ringing, but even short musically relevant pulses have so little ringing it can't be seen.

Where does the ringing go in longer signals? Well, each sample point "rings" mathematically and in longer signals those "ringings" cancel each other mathematically out. In very short impulses there is just one sample point ringing and not previous or later ringing to cancel it away.

Pre and post ringing is a popular talking point in pro analog audiophool community, because it makes digital audio look bad, but fortunately only look bad, not sound bad, because seeing is not hearing.

 

[71 dB]

Define yourself by your achievements and what you can do. Don’t require validation from others about who you are, and focus on your strengths to overcome your weaknesses. Don’t surrender to them.

Well, what things count as achievements? I can do many things, but it doesn't help if other people don't value those skills and pay me for them so I can pay my bills.

So, I am not even sure if I have achieved anything in life. Depends on what things count as achievements. Nobel prize? Olympic medal? What?

 

[bigshot]

The longer your signal is, the less ringing can be seen. In theory infinitely long signals have zero ringing, but even short musically relevant pulses have so little ringing it can't be seen.

That’s true. When it comes to time, it’s easy to just look at numbers in the abstract and not take into account their scale. Audiophiles regularly point to things like jitter, which has never been an audible problem, because it’s so hard to wrap your head around the difference between milliseconds and picoseconds. But the difference is immense. It helps to have a basic idea of the thresholds of human hearing. If something is an order of magnitude or more below the threshold, it probably isn’t worth discussing.

 

[bigshot]

Well, what things count as achievements? I can do many things, but it doesn't help if other people don't value those skills and pay me for them so I can pay my bills.

So, I am not even sure if I have achieved anything in life. Depends on what things count as achievements. Nobel prize? Olympic medal? What?

This forum isn’t the place to address that. You do that in real life with people who know you, not hope that strangers on the internet will value you and pay you money.

Cultivate friends. Be as generous and supportive of them as you want them to be with you. Don’t demand anything from them or use them as a dumping ground and you’ll start getting the acceptance you want. You won’t find that on the internet.

 

[71 dB]

This forum isn’t the place to address that. You do that in real life with people who know you, not hope that strangers on the internet will value you and pay you money.

Cultivate friends. Be as generous and supportive of them as you want them to be with you. Don’t demand anything from them or use them as a dumping ground and you’ll start getting the acceptance you want. You won’t find that on the internet.

Nice to hear if you have social skills to cultivates tons of friendships, but you are probably not an INTJ with aspergers! Our brains are wired very differently on fundamental level and it means social interactions are easy to some people and difficult for others. Also, introverts (and INTJs are extreme in that sense) enjoy being alone. That's their way to "charge batteries" while extroverts need to spend time with other people. The "need" to be alone a lot makes it even more difficult to be among other people socialising. People have problems for valid reasons and there are not easy solutions! For years I blamed myself for all my problems before I studied personality types and understood what kind of struggle different people have. I live in a society that favours people of different personality types than my own, good looking people (I am not good looking), tall people (I am short), athletic people (I am not athletic at all) etc. NO WONDER I have struggled so much, when the cards I got for life are so damn crappy! So, I have stopped blaming myself for everything and I feel much better.

Again, I am writing all of this so other people here can understand WHY I don't have 30 year long career in sound engineering. Instead I struggle to have any work at all! I have done things such as calculated heat losses of buildings, but that has nothing to do with sound engineering.

 

[bigshot]

I have friends with Autism and it's a challenge for them, but they don't surrender to it and they resist feeling sorry for themselves. They take a proactive stance by seeking out people who relate to them and they make a concentrated effort to get out and do things in the real world. The internet isn't a solution for autistic people. It's a trap.

 

[71 dB]

I have friends with Autism and it's a challenge for them, but they don't surrender to it and they resist feeling sorry for themselves. They take a proactive stance by seeking out people who relate to them and they make a concentrated effort to get out and do things in the real world. The internet isn't a solution for autistic people. It's a trap.

Autism is a spectrum. People with severe autism can hardly function in society, while someone with mild autism may only have a challenge with social interaction.

I don't really have a friend problem, since we INTJs enjoy being alone a lot. The few friends I do have feel enough. My problem is that my skills, strengths, and work history are so weird combination of things, that 1) The world as it is doesn't give much value on it and 2) It is very difficult for me to even promote myself to anywhere. The result is I don't really know where I belong in this World. What kind of career should I have? On what field? Architecture? Acoustic engineering? Sound Engineering? Interrior Designer? Developper of green energy solutions? I don't know... ...the World has never given a clear sign about this? All signs are confusing and contradictory.

Naturally I don't seek solutions to these things on a board about DSD mastering! All of this was "insprired" by the bragging contest about how many years people here have been working as sound engineers etc. I raised my hand and said "I can't brag about that at all and here is why."

 

[IanB52]

Actually, 12bit is close to transparency but then we don’t use 12bit, we use at least 16bit which is audibly transparent.

As far as I'm concerned 12 bit is a special effect. Now if you do know anything about vocal recording chains you know that most commercial releases are still using Neve preamps, tube mics etc, tube comps like a CL1b or LA2A, and by equivalency should be far lower fidelity than 12 or 14 bit digital. And yet...convert them to 12 bit and you've got the basis for lofi hip hop, not a listenable lead vocal. The same thing happens all the way up to 24 bit, save that the offensiveness decreases.

Sure, analogue audio/processing doesn’t have noise/distortion and magically defies the laws of physics. While digital audio has all these huge problems even though it is not constrained by the laws of physics. Sure, one can screw things up with plugins, just as one can screw things up with analogue processing.

A concept I am trying to get across is that digital systems are very complex, and thus the problems they deal with are complex and diffuse. It is naive and inappropriate to assume that if you simply apply the same knowledge used to evaluate more simple, and fundamentally different, analog systems . This kind of mathematical or THD+N reductionism sure makes it seem like digital systems are perfect, and yet displays no awareness of how they actually work at either end in the real world. Bring up jitter, phase noise, noise modulation, power supply interference, filter artifacts/tradeoffs, latency, impulse distortion/Gibbs phenomena, DSP artifacts, challenges in the converter design, alias challenges with the DSP, and all I hear is "Oh, those used to be problems, but now they are all perfect" which is BS. Of course you can't address the flaws in a digital system if you don't have understand how they are implemented in real life, and don't listen critically.

Depends how it was mixed ITB and how analogue mixes were made. But a well mixed album absolutely should not sound grainy or bleh in the box! How can you possibly not know this?

If you mix with lots of plugins, it's audible. Period. Maybe in the future there will be some kind of equivalent with analog processing, but not yet.

1. So now you’re contradicting yourself. Previously you supposedly represented every engineer, then you admitted there were some outliers who didn’t agree with you and now you’re saying you are the outlier. At least get your story straight!

I said no such thing, but I can't always account for other people's reading comprehension.

2. When the recording industry moves from converter to converter or plug-in to plug-in it’s because the technology improves in terms of functionality NOT because of transparency. There are some specific exceptions but still it’s not really about transparency. Again, as a professional engineer you should know what the imperfections are and how it relates to transparency, instead of just spouting this incessant audiophile marketing BS!

You are totally wrong here, and it is you pretending to speak for a profession that you are ignorant of. Maybe you are still using an HD 192 I/0 system in a cave and haven't checked in on converters in a long while.

Firstly, of course everything is quantified/measured, because that’s the definition of digital audio! And, how can “people weigh various metrics differently” when the metrics are below the threshold of audibility to humans? There’s only two answers to that question and obviously neither of them have anything to do with audible differences.

As I said, tons of people love vocals through a distortion laden U47+Neve 1073 combo, and nobody will go out of their way to bit crush or record a vocal to 12 bit, because one is inoffensive/sonically pleasant, and the other sounds bad. Just so, many amazing recordings done on tape with a -60db noise floor. And many done with better measuring digital systems that sound far inferior.

Again, it’s just wall to wall audiophile marketing BS with no indication of any professional engineering experience or formal education in the subject. You wouldn’t even have passed a short course or diploma with your assertions, let alone a degree!

But indeed I did, graduated CRAS in 2005, have certifications in Pro Tools and Logic, interned at Conway, Pulse/CastleSound, and Scream Studios, worked as an engineer at Cider Mountain Recorders in Idaho before I went freelance and built my own studio. Have worked on projects mastered by Ted Jensen, David Glasser, Adam Gonsalves at Telegraph Audio, and Bruce Brown. I have worked on or owned around 15 different different professional interfaces, ADCs and DACs, I have mixed plenty of tracks exclusively with plugins and I have mixed on API Legacy, SSL G Series, Neve VR (ugh), and a few Soundcraft Ghosts, have also recorded on Otari MTR 90-MkIII, Studer A820, A827 Gold Edition (mainly), Ampex ATR-102, and Tascam-38. I have recorded DSD directly from the mic input, from the mix buss on analog and digital multi track projects, and have literally sat in on DSD to DSD mastering sessions, using Tascam, Korg, Merging, and Mytek converters. I have recorded, compared, and converted every format of digital audio at 16/44.1 and above into every other format multiple times, and done the downconversion to 16/44.1 countless times, on every single project.

What you aren't hearing is that I'm not disputing the THEORY which I have also read, I'm disputing what actually happens in practice, and the notion that there are perfect (or audibly perfect) audio systems anywhere on earth, and that a one-size fits all reading of a couple different metrics is relevant to high quality recordings, or a competent way of making them.

 

[old tech]

As I said, tons of people love vocals through a distortion laden U47+Neve 1073 combo, and nobody will go out of their way to bit crush or record a vocal to 12 bit, because one is inoffensive/sonically pleasant, and the other sounds bad. Just so, many amazing recordings done on tape with a -60db noise floor. And many done with better measuring digital systems that sound far inferior.

You haven't bothered listening to that 8bit v 16bit video I posted have you? Put your pride to one side and listen to it. Then come back an tell us how 8bits (let alone 12bits) has crushed that recording.

 

[71 dB]

About 20 years ago I transferred maybe 50 hours worth of vinyls on CD-Rs. I used a pretty crappy PC sound card to digitize the music and I cleaned the music on a wave editor removing the loudest pops etc. Then I burned the music on CD-Rs. The CD-Rs sounded exactly the same as the vinyls, just with less pops. The "analog" feel was 100 % there. My crappy sound card was transparent enough for vinyl. So, when I compare my own experiences with digitising analog music to what IanB52 writes about working with PROFESSIONAL audio gear and what PCM does to (analog) sound, things don't add up at all!

 

[gregorio]

A concept I am trying to get across is that digital systems are very complex, and thus the problems they deal with are complex and diffuse. It is naive and inappropriate to assume that if you simply apply the same knowledge used to evaluate more simple, and fundamentally different, analog systems .

It’s trying to get that concept across which is so wrong (because it’s false), which any knowledgeable engineer should know!

Take for example EQ: A digital EQ is simply applying mathematical formulae, which by computational standards is not complex or demanding. The problems are not complex or diffuse, they are relatively straight forward and well defined. Now let’s take an equivalent analogue EQ, what is it you think it’s doing and how do you think it’s doing it? You think it’s just some sort of magic and therefore it’s simple? A knowledgeable engineer knows that an analogue EQ is doing fundamentally exactly the same mathematical procedure except it’s applying that mathematical procedure not in the domain of just numbers but through circuit design. Maths applied to an analogue signal through circuit design is obviously more complex than just the maths on it’s own. And, with circuit design you obviously have a lot more problems to deal with, a power supply, signal interference, component distortion and the unavoidable thermal noise. How could an experienced, knowledgeable engineer not know this? It’s inconceivable!

If you mix with lots of plugins, it's audible. Period. Maybe in the future there will be some kind of equivalent with analog processing, but not yet.

That’s nonsense, there has ALWAYS been practical problems/difficulties with analogue generational losses, noise and distortion build up with sizeable analogue chains! With “lot’s of plugins” it certainly can be audible depending on what plugins we’re talking about and how they’re used but it can also be transparent, which is not the case with lots of generations and/or large analogue processing chains. Again, this is basic knowledge for an engineer.

As I said, tons of people love vocals through a distortion laden U47+Neve 1073 combo, and nobody will go out of their way to bit crush or record a vocal to 12 bit, because one is inoffensive/sonically pleasant, and the other sounds bad.

Is this really just ignorance or is it deliberate misinformation? You passed the Pro-tools entry level certification right? So how do you not know that pro-tools includes a dedicated plug-in specifically for bit crushing and there are numerous bit crusher plugins available? It’s not commonly used on vocals but to say “nobody will go out of their way to bit crush” vocals is nonsense! There are numerous examples, for instance Sail by Awolnation went triple platinum, has 379m views on YT and the lead vox is horrendously bit crushed, lower than 12bit! Any knowledgeable engineer knows there really isn’t such a thing as “sounds bad”, there’s just appropriate and inappropriate depending on the context. It would be inappropriate to bit crush a vocal in a Puccini opera but not so for some other pieces/genres. Jimi Hendrix sounds great with massive distortion but that exact same distortion would sound horrendously bad on a guitar concerto. How can you not know this?

Just so, many amazing recordings done on tape with a -60db noise floor. And many done with better measuring digital systems that sound far inferior.

True but also completely false, by way of a “lie of omission”, there are ALSO many terrible recordings done on tape and many great recordings done “with better measuring digital systems that sound far” superior! Again, you can screw up a recording no matter how good the recording system or format but all else being equal, a good digital recording system will have higher fidelity than even the very best analogue system. How do you not know this?

But indeed I did, graduated CRAS in 2005, have certifications in Pro Tools and Logic, interned at Conway …

Maybe you didn’t just quote audiophile marketing BS but instead actually gave factually accurate answers in your coursework/exams? An outside possibility is that the short course you completed simply covered how to use some studio equipment rather than the basics of how analogue and digital work. You could get away with not knowing that for the lowest/introductory level of Pro-tools certification but not so easily for the higher actual “Operator” or “Expert” certifications.

Bring up jitter, phase noise, noise modulation, power supply interference, filter artifacts/tradeoffs, latency, impulse distortion/Gibbs phenomena, DSP artifacts, challenges in the converter design, alias challenges with the DSP, and all I hear is "Oh, those used to be problems, but now they are all perfect" which is BS.

Ah, you must have a serious problem if that’s “all you hear”, absolutely no one has stated that, so “all you hear” is something that doesn’t exist! That explains a lot, though not how you could be a professional engineer!

Shannon proved that ALL the information can be captured with digital audio provided certain conditions are met, this CANNOT be the case with analogue. Practical implementations of Shannon/Nyquist do not allow for absolute perfection, primarily because ADCs and DACs by definition must have analogue sections, which must follow the laws of physics and introduce significantly higher levels of noise than does digital audio. Absolutely no one has said all the items you listed are perfect, there will always be some jitter, there will always be some DSP and filter artefacts but even in some very cheap converter examples, these are all outside or way, way outside audibility. Some of the problems you mentioned are also applicable to analogue equipment and typically much more pronounced. But again, we have a “lie of omission”, of all the problems specific to analogue audio, some of which cannot be solved without breaking the laws of physics. Again, every knowledgeable engineer should absolutely know all this!

I'm disputing what actually happens in practice, and the notion that there are perfect (or audibly perfect) audio systems anywhere on earth, and that a one-size fits all reading of a couple different metrics is relevant to high quality recordings, or a competent way of making them.

And again, you are disputing a notion that no one other than you has even suggested! There are no transducers that are even close to perfect, so there cannot be any perfect audio systems and that’s without even mentioning room acoustics or HRTFs! Also, absolutely no one has suggested that reading “a couple of different metrics” is a competent way of making recordings. In fact I’ve already clearly refuted that assertion previously!

Your accusation of others’ reading comprehension is therefore hypocritical!

G

 

[71 dB]

I admire the patience gregorio has in meticulously going through every claim made by IanB52.

 

[bigshot]

When it’s clear the other person isn’t listening and is continuing in the face of evidence to the contrary, I don’t see much reason. But he seems to enjoy it, so it’s fine.

 

[gregorio]

I admire the patience gregorio has in meticulously going through every claim made by IanB52.

To be fair, I didn’t go through every claim but I did cover many of them.

It’s bad enough when we have false audiophile marketing presented as fact but even worse (as far as I’m concerned) when it’s presented by a (supposedly) professional music engineer. I can’t claim it’s impossible that a professional engineer could be as deluded by the nonsense marketing as most audiophiles but in nearly 30 years of visiting/working in studios and concert venues all over the world, I’ve never come across or even heard of one. You definitely don’t find them in top class commercial studios, you don’t find them in sound studios. If there are any, the most likely place to find one would probably be in a home bedroom/basement studio and be an “engineer” with very little or no formal education/training in the subject.

G

 

[71 dB]

To be fair, I didn’t go through every claim but I did cover many of them.

My bad! It looked like you covered it all! Good job anyway...

It’s bad enough when we have false audiophile marketing presented as fact...

Capitalism has its price...

...but even worse (as far as I’m concerned) when it’s presented by a (supposedly) professional music engineer.

Yes, it can be very confusing. His credentials (as far as he is telling the truth) look very convincing, but many of his claims are total BS even to MY eyes and I am not a professional sound engineer myself! I can totally understand if someone favours subjectively analog sound for the distortions and colourization it creates, but his claim is that analog sound is distorted and noisy in a "good transparent way" while digital (PCM) audio causes some horrible and even unknown things to the sound that makes it less transparent.

I can’t claim it’s impossible that a professional engineer could be as deluded by the nonsense marketing as most audiophiles but in nearly 30 years of visiting/working in studios and concert venues all over the world, I’ve never come across or even heard of one.

Well, if you remember, I was very deluded about cross-feed a few years back thinking it must work well for everyone because it works for me. It really took me years to comprehend and believe that people hear spatiality differently. Even today I don't understand it, but at least I accept it and I label it as a thing I just do not understand. I'm just saying that even if you can talk sense to him, it may take years for him to really "accept" it. Rabbit holes are deep and it takes time and effort to get out of them.

You definitely don’t find them in top class commercial studios, you don’t find them in sound studios. If there are any, the most likely place to find one would probably be in a home bedroom/basement studio and be an “engineer” with very little or no formal education/training in the subject.

G

Looks like there are different kind of communities among music engineers and different kind of styles of doing things. Those who advocate analog sound "get away" their misconseptions by doing the analog sound well and giving the lovers of analog sound what they want... ...if your client is happy who cares how right or wrong you are about the facts...