PHASE

[tgstgs]

expected brutal math here?
well maybe later_
resonance depends on feedback you know;
so let me start with a poem . .
------
phase and delay
they are 2 brother
belive it or not
they belong 2 each other
change the delay
and the result will be
you changed the phase
2 a speziffic degree
-------

good vibes from vienna

[dante]

A Delay is a Delay
Relative to itself
Its only brother Phase
Relative to another of itself

[tgstgs]

for(bool interest=0;!interest;) interest=waitforvibes();

[katano]

$waitforvibes=hugevibefromSUI

[wayne]

vibes everywhere i turn my head

[winger]

I seem to missing something here. Is there a point or question?

[dante]

No question, I think just a reference to previous discussions about phase, turned into some poems and funny programming code.

[tgstgs]

no previous discussion but maybe there should be one to rise quality of discussions ?_

just some good vibes

[tgstgs]

the point
even 1 sample delay change the phase (equal the samplerate btw.);
if you mix the changed phase signal with the original you get a nice LP;
this is the most simple LP filter man may think of;

-------
could be you have more than 1 sample delay between dsps who knows . .
now the question
what would happen if you mix it up?

----
sorry thought there is more interest in some serious discussion resulting in some funny code for a nice free device;

good vibes from vienna

[David]

I'm interested in the topic of phase. It's been revealing to read previous discussions as they have led to me questioning my ears and altering my mixing practices. Xite's possible use of multiple dsp chips make understanding what's going on with phase a must.
Would like to read more on this topic.

[astroman]

well... as I'm not the brightest soul under the math sun, I take the opportunity...
Given a complex signal mixed up from (say) 5 diffent frequencies, a change in 'phase' will result in 5 different 'delays', with a specific amount for each of the frequencies.
A 'delay' in processing by one or more samples will 'shift' each component by an identical amount of time.
Which is obviously a quite different process.

While we're usually talking just about 'phase', we really mean the 'phase integrity' of a signal mix (as it's not about single sines).

Imho it's important to know what can happen, but otherwise ignore it as long as it doesn't disturb.
For my (personal) taste there is way too much focus on mathematically precise processing, instead of just recording the sh*t.

To be honest a post of Mr Arkadin totally opened my ears recently - he mentioned a record by some dude called The Man with the Stereo Hands which was done entirely on a Tascam 8 track cassette recorder

I didn't believe that and mailed him (yes, I bought the record...)

...it was recorded on a Tascam 488 Portastudio with no
bouncing of tracks at all, in other words only 8 tracks of sound were
recorded for each song

I think I can quote his answer in public 'for the sake of the song' (an old Townes Van Zandt album)

cheers, Tom

[tgstgs]

thank you for opportunity its the bread in each discussion
as well as the phase/delay is the bread for filters;

i would answer
if you just record that sh*t
noone will ever listen to that sh*t;
you had to have standards if you want to play with the big boys_
if not play with the kids for personal satisfaction only_

the smallest math number is 1 sample delay right;
but the probl. is that already 1 sample delay has a hugh effect to the frequency spectrum;
you will hear it for sure even in a low level mp3 on a 15euro headset;


btw. the cassetrecorders of back then had much better quality than any mp3 today;

as rule i would say:

check your lines (for lenght / delay)
be it the cables in your studio the reflextions of wall in a concerthall;
the reflextions of the speakers desk compared to the directsource or
the blue wires in your scopexiteproject window;
as soon as there are different delays you are creating primitive filters;

combfilter vibes from vienna

[astroman]

... if you just record that sh*t
noone will ever listen to that sh*t;
you had to have standards if you want to play with the big boys_
if not play with the kids for personal satisfaction only_

I don't mind them... at least not because of their name
I will honour their work when they deserve it and I will call it sh*t when it is

Bob Clearmountain epically failed on Costello/Van Otter 'For the Stars'
whoever mixed the Funkbrothers in 'Standing in the Shadows of Motown', obviously never heard on original track...
or he/she would still be blushing...
the latest release by Phil Collins is an acoustic desaster
how about JK's vocals on the recent Jamiroquai album ?

The producers of these tracks were all certainly real BIG BOYS, with the greatest tools on earth and budgets up to the sky
yet they all spoiled the job, more or less
and (imho) some of them did because they worked over-precise, over-balanced, over-transparent, over-brittle

those guys would never succeed on a 2nd 'Daddy was a Rolling Stone' (Temptations)
1 chord, the same 6 bass notes - repeated for 13 minutes of groove with increasing tension.
they could never record another 'Gloria' as played by the Patty Smith Group
you simply can't sequence THAT timing... but THAT's what makes the track go through the roof.

imo those 'Big Boys' lack any soul (I'm afraid) and I don't wanna be associated with them, not a single moment
(if you spoil a Funkbrothers Tour and the record you have to take the credits where due...)
ok, this just had to be said
possibly a bit harsh, but then it's me suffering whenever the wife switches on the radio in the car...

ranting vibes from the other side of the mountains, Tom

and now we can return to the subtle details of digital signal processing
of course I know about the comb filter effect (which is fairly easy to avoid)
I have a much harder time monitoring VST processing, which seems driver related
(sounds like there's data leaking in the process, adding the same sample a 2 nd time with a tiny displacement or so...)

cheers, Tom

ps: of course we should have good tools (and know them well), but if thinking about tool takes more time than the job ... something goes awefully wrong...

[tgstgs]

what did you expect from a man who 'cant dance'?
-----
but if these examples pass your taste test or not;
im sure there is no phasedistortion in;
and to be honest these 'boys' could even sell stuff with phasedistortion and earn millions_

U2? or anyone in here?
me not!
---

but what i realy dont get is_

you tell the people dont care about a displaced audiostream causing all sort of filtereffects in xite
and searching for a tiny displaced sample yourself???
didnt you tell us forget about that sh*t just record it. . . .
twotonguetom?

good vibes from old danube

[astroman]

sorry for my expression, 'tiny displacement' was intended as additional description - and 'tiny' is more than 1-3 samples.
Attacks seem to be almost exponential and are kind of distorted while the meter doesn't reflect this.
It happens under certain conditions related to VSTs when monitoring on a RME card, which also CAN deliver a clean monitor signal in another mode.

I'm certainly not going to dig it out, but will exchange the card (or setup a different monitor path).
It's not about a nuance one person may notice and another not - it's a severe decrease of the signal.

Actually I wouldn't mention it at all...
if I hadn't had the very same audible effect with a Scope card (or Pulsar?) once running a demo of IK's Ampeg SVX plugin.
At least one other person here heared it, too - because we both commented the plugin sounds like sh*t.
Which it really did - but not because of it's processing... embarassing.
The funny thing is that the 'effect' seems to be implementation dependent.
It was NOT with all VSTIs on that installation.
But since I don't do that much native processing it didn't reach a high priority on my list.

cheers, Tom

[astroman]

back to the 'real' sample displacements in DSP processing...
iIrc the 'problem' was completely overlooked in early Scope plugins, most prominent example the 'classic line' of Sonic Timeworks devices BEFORE Warp69's reverbs.
Since then developers have paid attention to that part of the processing.

It's certainly very important to have everything aligned perfectly in traditional orchestral music with dozens of mic channels.
The setup itself is sensitive in this domain already by it's physical dimensions and raw signal paths.
On the other hand one wouldn't expect heavily post processing there... (at least I wouldn't)

How about the rock/jazz/country band ?
The usual listening environement (of a live experience) is anything but ideal
... and has an influence on our listening experience.
Of course the 'problem' increases with degree of post processing in a studio record - a natural matter of fact.
But in my estimation it still remains relatively small compared to other things you can do wrong, starting with the mic and pre-amplification.

Then there's the contemporary solo artist: an entirely synthetic process of making a record.
As a producer you can run into trouble indeed, as heavy post-processing seems state of the art
.... really art ?

that's why my point above was: just record that sh*t...
focus on the moment and that great take - it contributes way more than those minor tweaks (possibly) failing
Imho you can't really argument with an unintentionally introduced filter (by sample delay) when they eq the sh*t out of a voice - see (hear) the Jamiroquai singer as an example.

Some of that stuff certainly is up to personal taste and preferences.
But there ARE recording references that probably everyone 'tuned' his or her senses on.
JK's squeezed voice and Babbit's bottom only Precision bass ARE fails
(you really wouldn't spend a 5 figure amount on that instrument (if it were on sales) to record THAT tone)

Many posters here seem more afraid about a (potential) tiny error somewhere in the processing chain, than committed to do their record.
Even WITH some misaligned stuff it might still have scored... if one only had trusted in the content and just DID IT.

cheers, Tom

[tgstgs]

i started this topic to tell the people some basics;
i think they should know now in time of dspselfassignment;
in time of ..
i know becorse i read that one hears about a man who hears that one had red
something _ sometime _ somewhere YO;
but there is not much interest it seams;
maybe becourse most red about one . . . .
----

i know you for a long time now;
i think i already read your posts at old tdatforum at CW later dspsystems;
i like your posts;
so i guess you have the same intention and take opportunity
just to go on discussion and maybe help some people improof their setups who knows
here we go


-----
lets take some white noise dublicate delay one stream mix together
and see what happen:


1 SAMPLE delay!!!

even a lowquality mp3 has a better frequency response no?
maybe some people aboth 80 cant hear a difference;

[tgstgs]

2 SAMPLE delay

here even people with a heavy tinitus sensed somthing wrong;

[tgstgs]

3 SAMPLE delay

do with any music of any style
nuances?

[tgstgs]

4 SAMPLE delay

starting to get whats going on?