[sensing]

Originally Posted by audioconsult

I belittle those only in cases of advice resistant behaviour to give them a hint to overthink their position.. There are things that are a matter of taste and there are things that are just facts.. and its almost insulting to be forced to provide proove for such standards everybody should know by now..

Like in this thread..

As Acid Mitch has put very well.. a test about the possibiluty to mask 5 ms dont tells a thing about the possebilty to hear such timing differnces in positions where they are not masked or even emphazised.

As i said earlier..timing trouble only hurts at positions where it hurts.. And they come along in production praxis quite regulary ...dependend on the setups..state of computer os´s..bad wordclock conections..

wordclock .. good example.. a jittery comection there is causing timing derivations in the micro second range.. but still effects the sound in nasty audible ways.. 5 ms are miles against that.

I have to assume that people that dont know about timing problems where 5ms hurt dont have much experience with computer based systems.

The problem with the test proposed here is, that the question it trys to investigate, is just absurd.. And i expected it to be a wiener test that trys to mask a problem instead one that trys to show it.. so a propaganda test.

And voila..it is... People participating in it make themself to the dolls of an uneducated kid that toys around..

because 5 ms is not much.. you need to double that to go into the obvious department.. 5 ms is mayby 4 mm on tape.. when you cut tape and do it 4mm wrong you are in the area wher you hear a little glitch already.

The real border here is around 2 ms.. thats the scientivic border of perception.. somtimes in discussion to be just 1 ms..

but.. also a question.. 1 or 5 ms of what? derivation from the real position? or jitter?..

a 5 ms jitter window equals a +/- 2,5 ms derivation from the real or indented position

So when the OP moves files 5,5 ms in one drection from the grid.. we are actually in the 11 ms jitter department

So his test trys to proove that we cant hear 11 ms jitter..

11 ms jitter ...

that is the number events can be max apart in relation from each other in a multitrack recording than..

still not much.. its like moving the speakers 4 meter to the back.. or to the front or in case of jitter having them jumping 4 meters forth and back..

But thats actually enough for the human brain to get it easily..when something does 4 meter jumps into your direction you get better alarmed..

Or musicans on headphones will start to refuse to play or at least complain.


more than 10 ms between the expected sound and its real arival is too much.. even on a plug in synth 10 ms latency is the area where it starts to hurt the playing..


So moving events along the timeline by 5,5 ms equals a jitter or max error between tracks of 11 ms..

do you really want to prove to yourself that this dont matters with this wiener test here?

Thanks god he has choosen 5,5 ms..because thats an absurd number..with 2 ms we would have a bigger problem to show that this matters..

But as victims of jitred computer clocks on otherwise straight drummachines know all to well.. 2ms is allready well enough to be experienced as a problem under such conditions..
And it never was the question wether you can tell which of the many events the drummachine produces under the wobbely clock is off or 2 ms early or late.. the feel of the thing is different..

You dont even need to evaluate better or worse.. just the plain fact of being different can hurt a production when the musicans are used to a certain feel for hours and suddenly its gone because the extra load of a certain plug makes your computers midi performance go ballistic.

And you measure whats going on..and ...suddenly you have 5 ms clockjitter what was before just 0,2 ms clock jitter..

so a +/- 2,5 ms derivation from the supposed position.. so half of that what the OP trys to proove to be irrelevant

Such situations are the prove that 2,5 ms matter !
And i had many of them in the last 17 years.. since i deal with ITB DAW´s its a permanenet topic..some years mores than in others.. depends on the projekt load now.. was better on my black macbook and the titanium than on the new ones.. the new cpu´s are stronger but the midi system goes balsitic boarder is earlier reached,.. not verified yet..but as it looks i can use less plugs on my i7 than on my old core duo in regards of the midi clock output... that sucks big time :-/

YOu get a littlle agressive when people try to tell timing is no issue and seeing the industry reacting to that by lowering the performnce in that regard again... because " nobody needs/hears it"
So we cut in the timing department"

that pisses me off..

Some very well explained stuff here, but i feel there are some issues with it.
Wordclock; thats a mechanism for synchronising digital systems, bits if you will.
It will have an effect on the audio, but the deviation in the wordclock is not the same as the deviation in audio. So, that 0.5 ms wordclock jitters is audible does not mean that 0.5 ms audio jitter is noticeable. I am sure you know that, but wordclock is not a good example here.

Also, isnt jitter a deviation that is not constant?
So if the audio is shifted a constant 5.5 ms to one direction, there is no jitter.
It would only be 11 ms jitter if the deviation fluctuated 5.5 ms to both sides.

For the OP; there is nothing in this test which makes it universal enough to state that such a timing difference can not be heard. In that way its probably not scientific. What you might prove is that it cant be heard under the specific circustances in your recording.