Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[danlavry]

You said:
"The Mike Story Paper is not discussed elsewhere in this thread, and has no other links hereabouts in this thread. It is mentioned by Mr. Lavry responding to Mr Massenburg on page 14 and mentioned in passing by Johnny B without comment. That is all.
With respect, you must have discussed it elsewhere. I certainly haven’t discussed it with anyone hereabouts and only ‘tripped over it’, yesterday."

I say:
I do not view it as a technical paper. It is far from being up to par with what I consider minimum requirement for such consideration. I do not see it as meeting the bar for peer review.

You said:
“although we may not be able to hear energy above 20 kHz, its presence is mathematically necessary to localise the energy in signals below 20 kHz..."

I say:
Yes, and we know mathematically exactly how fast we need to sample to contain 20KHz - ALL THE INFORMATION WITH ALL THE FINEST OF DETAILS, nothing missing! Nyquist theorem is the key, and my paper explains it pretty well. 88.2KHz and 96KHz is more than enough to do so.  
Again, do not get confused into thinking that bandwidth and impulse response (or rise time, or quick reaction time) are anything other than one of the same. If the maximum bandwidth of a mic is so and so, say 20KHz, than it can only react to signals up to 20 cycle a second, or a motion with a rise time of such bandwidth. Putting an AD capable of registering almost 5 times the speed of the fastest signal yields no advantage, and the speakers on the other end would limit it if it were possible. It is not just doing the un needed, there is also a big cost involved, including lost accuracy and more serious problems...

I am very glad you brought up the word math. Lets just use correct math and physics.  

You said:
I thought it was worth a read.

I say:
A lot of non technical people thought so, and indeed I believe that was one of the explanations for why sampling faster is "possibly better". I imagine it was a good marketing tool, but the information is wrong, and I am not pleased to see that such "work" receives as much attention as it does, making my efforts to steer the industry right just so much more difficult.

I guess I should not complain, given that the subject on this site has over 21000 hits (the average is what 1000?) and there was a lot of action on other sites. 192KHz is going away...

I do not read everything, but I try and watch for your comments.

Best wishes to you too
Dan Lavry

[Paul Frindle]

Peter  Oxford wrote on Wed, 21 July 2004 12:42

Doubtless there are internal structures that designers can improve, but I think many people haven’t quite got their gain structures right at the basic level and are using digital meters (especially on computers) with naive over confidence. I haven’t got time to expand this now, and probably many others hereabouts could do so far better than I. But I do think many user’s have quite a bit to answer for (rather than the equipment), and somewhat of a learning curve to negotiate. Best Wishes   Peter Peter Poyser

I would agree that the users are stuck in a never ending learning curve. But who has generated the need for this so fundamental learning and where indeed do the users look to obtain that learning - in a world where people are actively being discouraged from understanding?

I'm sorry that in my opinion the use of such equipment should be a primarily artistic process and it should be up the the designers of equipment to facilitate that artistic process as much as is possible. Thats the key to success IMO - who's side are we on, after all?

As so many of these threads have illustrated over the years, even the most obvious of technical realities are under constant discussion and disagreement. And even the most fundamental of issues are being subjected to continual misappropriation for the purposes of advancing finance rather than art.

On this thread we are indulging in discussion of the possible benefits of reproducing what we can't directly hear at all, whilst many people are actually clipping their programme without knowing how, why or even when. Those same people come back to these threads seeking an answer to their woes, forever hoping that the newest buzz will somehow fix up their worries. "Yeah - maybe that 192K thing everyone's talking about and the stuff we don't hear over 20KHz will finally get rid of that darned harshness I hear every bleedin day thats driving me crazy".

Far too often have I heard the notion that whatever is inconvenient and damaging to the user's environment is due to his own lack of knowledge. Too often have I heard that its the users responsibility to devine the minutae of the all-important truths that are buried beneath veritable mountains of techno-noise which besets our industry in ever-increasing heights.

If it sounds bad - "hell man thats your fault - either you're an uninformed imbecile, or you haven't got the latest kit, or you're the last geek to hear that word on the street says 'math doesn't work' and you've been conned"? No sorry I can't subscribe to that - if I make it and it sounds bad - that's my darned fault and no one else's - and certainly not his!

Yes 'its tough on the streets' these days. And within our protected societies, whilst we become obsessed with the modern day 'sophistication' of turning our 'streets' back into something resembling a primeaval jungle (both physically, intellectually and morally) in order to play out some innate and unsatisfied need for struggle and victory against adversity, that no longer really exists - and the market has signed up to it by making the process of unneccessary 'discovery' of the cons and deceptions part of the newly sophisticated 'consumer involvement plan' - hasn't anyone noticed:

jeez, there just ain't no place left to score reliable and straight up information no more man!!

Apologies for rant - but you have to agree it's totally absurd

[PP]

[ted nightshade]

Paul Frindle wrote on Wed, 21 July 2004 16:22

jeez, there just ain't no place left to score reliable and straight up information no more man!! Apologies for rant - but you have to agree it's totally absurd

Well Paul I don't know why I trust you so much, maybe 'cause you're kinda sad and dreamy and empathetic with me, and always have your eye on the interaction between kit and artist and art (actually, kit comes between artist and art...), but your posts seem pretty straight up to me.

I imagine there are quite a few promising careers, from an artistic point of view if not a financial one, that have crashed and burned from this insane hype and lies environment... could be mine is one of them. Or maybe I get wiser and survive, but it's hard to trust that gear does what it says it will, even at big price tags...

I think of all the folks (how many?!?) out there with their creativity dying on the vine as they invest all they have and some they don't have into baffling and frustrating systems and replacing pieces again and again at more expense in a search for something that makes it possible for artist to create healthy art- and their heads spinning from the hype and the $ signs and the zeroes and formats and the unfathomable math of it all...

I'll tell you, it's just about destroyed my marriage, I've become such a frustrated obsessive geek. If I'd known then what I know now... but actually if I knew now what I really needed to know in order to get a simple set-up (any such thing?!) together and get some work done.

[steve parker]

Nika Aldrich wrote on Wed, 21 July 2004 17:17

Steve, It is here:  http://www.tllabs.com/index.php?option=content&task=view &id=37 It looks like you need to register to get it now?  Hmm. Nika.

thanks!

steve parker

[Johnny B]

I started another thread on the headroom issues, maybe the moderator could cut some of the posts off and move them over there. Otherwise, we will have sort of hijacked Dan's thread.
What do we have to do to get some of this moved?

And just maybe, if people understood some of this gain structure and signal flow stuff better, the sound quality of CD's would improve. Maybe they'd even stop asking for sample rates going at the speed of light and bit depths as big as the Atlantic Ocean! LOL.

[Level]

Calibration has not been documented properly for most folks to do it properly. Sine waves, pluses, music and individual instruments, a great O-scope and understanding impedance matching, not to mention designing and building your own line amps and pres...a full understanding of what happens between YOUR particular chain in your room with all kinds of situations..some unorthodox is required to find individual calibration for your studio to insure purity throughout.

This is why my calibration has no bearing to what others should do "spec" wise. I simply assure that what comes in, goes out, as untouched as possible. That is all we can and should do with calibration.

Then the art has a chance to happen graciously.

[Johnny B]

Moderator,

Could you please divide this thread beginning on the top of page 49 and append it to the bottom of the Zero Headroom Thread. Thanks.

[bblackwood]

[message from admin] This topic was split from the 192kHz discussion so both may continue without interaction...

[Johnny B]

Thanks Brad, you're one of the good guys.

[Paul Frindle]

Peter  Oxford wrote on Thu, 22 July 2004 01:49

Hi Paul, How are you? I’m sorry I’m so tied up. I’m posting at crazy times for me at the moment. It’s ‘night and day’ as the old standard goes! Furthermore, all the better engineers I have encountered have a tendency to have with them (wherever they go and especially in new situations) a very accurate, properly calibrated meter so that they can be sure of what is really going on as regards to levels, and for setting gain structure between critical bits of gear properly. They don't take anything for granted and certainly don't assume the meters are spot on (You can’t help but notice such things). So again, I think that there is something here to be learnt. An important lesson. <snip> Perhaps I’m completely wrong about this and will get flamed! At least I am big enough to apologise if I am. But it’s my ‘take’ that people have too much confidence in the accuracy of their meters, and a tiny degree of scepticism would serve them very well indeed. I MUST get some sleep now! My best wishes to all, and especially to Mr Lavry whose life work is all about recording hit’s! Bye Now! Peter Peter Poyser

I agree with your sentiments and sorry for the rant. It's just that it gets so frustrating to try presenting truth and a balanced argument in a world so messed up with marketing that even my efforts are becoming indisinguishable from the very hype I deplore so greatly. Increasingly there is no 'untainted' language I can draw upon

Anyway the problem goes much deeper than the 3dB issue you speak of - although that is one valid issue among others.
This is a prime issue where all the hype does so much harm that effectively cannot be undone (and its much the same right across our society on many levels). I'm not going to rant on about it - I'll just give one example of one word - 'resolution'.

That rotten and naive concept of 'resolution' thats responsible for as much harm as almost anything else I can think of.

Because people have been wrongly advised that recording at anything less than flat out will result in 'less bits' and therefore 'less resolution' and therefore 'worse sound' (obviously), they are mistakenly and understandably obsessed with recording stuff that is actually illegally high - which sample value metering cannot even detect - even if they are perfectly accurate to value.

Again misappropriation of language is hiding truths that affect everyone - and sadly the hyped words are MORE compelling to the user than the truth - because the marketed concepts are necessarily (and deliberately) simplistic and apparently obvious - whereas the truth requires thought and effort that no one has time to embark on. The effect of such misappropriation of language is akin to poisoning a system and rendering it useless - language will no longer be effective to convey truth - once all words have been 're-purposed'.

The single expresssion 'resolution' is of course responsible for massively greater misunderstandings than the example of headroom and limiting - and of course there are many other misappropriated words that are almost as dangerous - it is not the only one.

It is not the user's responsibility to seek a college degree in sampling theory before he can cut the multiple layers of crap and search for an understanding of how he's being conned and thus avoid the pitfalls. And if he actually gets there - he isn't even guaranteed a 'balanced' and unbiassed view - in a world where increasingly research is funded by commerce. And the human language of communication has been cleverly subdivided into virus-like soundbites, that by their very simplicity and self-obviousness, spread unaided thoughout all society, poisoning the process of understanding - implanting manufactured notions and concepts, such that we can only demand that which is already intended for our 'consumption'.

Sorry - another rant I'll stick to technical matters next time - honest

[Johnny B]

Paul,

I think the problem with the word "resolution" as applied to audio got mixed up with the word as used when applied to "digital." Then the marketing and inept sales guys really butchered the word so the poor end user became completely confused.

I've often been told that you can lose bits if you record what they term as "too low" a level, any truth to this? Where are the places that the bits disappear? Bad plug-ins? How should one approach recording and mixing in digital to avoid any of these pitfalls. In short, how do you avoid royally screwing things up?

[Paul Frindle]

Quote:

Paul, I think the problem with the word "resolution" as applied to audio got mixed up with the word as used when applied to "digital." Then the marketing and inept sales guys really butchered the word so the poor end user became completely confused.

Yep - now we helpfully also have a new marketing term called - you guessed it - 'High Resolution Audio', which basically means anything that has more of 'anything' than the original system. Explicit and highly useful isn't it

Quote:

I've often been told that you can lose bits if you record what they term as "too low" a level, any truth to this?

Of course not! - The data is not physically scaled by programme level. ALL the signal is always represented in ALL the data.

Quote:

Where are the places that the bits disappear?

Obviously they don't go anywhere. If you reduce the number of bits artificially (say 24 to 16) you increase the noise floor - thats all.

Quote:

Bad plug-ins? How should one approach recording and mixing in digital to avoid any of these pitfalls. In short, how do you avoid royally screwing things up?

Firstly and most importantly - FORGET the word 'resolution', remove it and it's irretrievably tainted connotations from your brain as far as is humanly possible!

It would help immeasurably if mixing applications included considerable headroom internally between processes (i.e. 20dBs or so) and the industry adopted a notional operating level somewhere sensible and had meters calibrated to reflect this.

But in the absence of this (apparently lost cause) - your only option is to modulate at least -3dB down on record (-6dB being even better). On mixing attenuate all the returns by around -14dB to -20dB to avoid clippling when you add processing like EQ and Dynamics. Do all your balancing at these levels and bump the final output up to the required level (i.e. around -3dB maximum peak value) - as the very last thing you do before your mix is being printed to file.

In this way you will be released from permanently ruining your balances as you adjust levels to avoid overs each time you try to get your 'sounds' and crucially you will produce a result that is legal under all conditions, which will avoid clipping in the user's playback systems. Absolutely everything will become vastly easier for you as a sound engineer and everything will likely sound immediately better.

Simple isn't it? A direct reversal of what everyone has been encouraged to believe

[Bob Olhsson]

Johnny B wrote on Fri, 23 July 2004 02:00

I've often been told that you can lose bits if you record what they term as "too low" a level, any truth to this?

Considering that the best available converters manage around 21 bits, "too low" probably means peaking below -18 assuming there aren't analog issues with the converter being used. One of the biggest problems with peaking to full level is that there are serious analog overload issues in all but the very finest digital converters.

[Nika Aldrich]

Bob Olhsson wrote on Fri, 23 July 2004 16:45

One of the biggest problems with peaking to full level is that there are serious analog overload issues in all but the very finest digital converters.

...Digital overload issues as well.  Most signals that get near full scale in a converter end up clipping the converter's internal upsampling algorithms in addition to, as Bob mentioned, the analog backend.

Nika.