External Clocking - is it all that?

krd

kfoltman wrote: »

Ok, another test:

- test environment: one PC, two decent quality soundcards (not clock-synced), output of first soundcard fed into input of second soundcard; by "decent quality" I mean anything that has dedicated ASIO drivers (not really high bar to be honest)

- some software tone generator playing a sine wave through first soundcard (the level must be high enough to ensure good S/N ratio and not high enough to cause clipping)

- some software narrow-band notch filter cutting out everything at the frequency of the original sinewave (+/- clock drift effects), with output sent to speakers or some kind of spectrum analyser

If the DAC in the first soundcard or the ADC in the second one is really causing any frequency modulation to occur, it should be visible as peaks in the spectrum, or audible in the speakers (the original tone has already been cut out, so any residue other than noise should be caused by jitter, intermodulation distortion of soundcard's inputs or some other unwanted effect).

I don't want to do the full calculations, but in theory. If you have two identical sine waves coming from a different devices ariving at the same sound card. And the sampling rate of the recieving sound card is 192Khz - for both sound waves to be in phase, without any modulation, the delay between each sound wave being process must not be more then 1/96,000th of a second - 0.0104 miliseconds - for the waves to be in phase (accounting for the nyquist theorem - I think my calc may be wrong - but the lower the sample rate of the card the greater the time you would have - this is down to Nyquist sampling)

But why would you want to do that in the first place?

krd

kfoltman wrote: »

MIDI has very little to do with this particular topic.

I mean, jitter exists both with audio and MIDI data, but the scale of the problem is much worse with MIDI. And it's not limited to soundcards - many hardware synthesizers use slow microcontrollers that can't keep up with processing events. Some MIDI events are processed later than they're supposed to, and the delay isn't constant - almost a textbook example of jitter. And those delays may be in millisecond or above range, comparing to audio converter jitter, which is in nanosecond/microsecond range.

On the other hand, some software synthesizers brag about sample-accurate event processing, because events are timestamped using sample position within a buffer. But that's a completely different story...

I would put money on it the soft synth companies are talking out of their arses. Every sound card has an implementation of Direct X for windows - each manufacturer does their own implementation. The Soft synth companies are writing to card.

But even if they generate each sample exclusive of the card - the maths they use will always be slightly off each time it's iterated - impossible to hear - but I can't imagine it not causeing the tones generated to be ever so slightly out of phase if your compared them.

they only way they could make these claims is if the software was writing directly to buffers on the sound card - don't how or why they'd really want to make these claims I don't see how the could gaurantee it.

kfoltman

krd wrote: »

I don't want to do the full calculations, but in theory. If you have two identical sine waves coming from a different devices ariving at the same sound card.

No, it's not two different signals. The general idea is:

- one source of a sine wave (tone generator, either software or hardware)

- one measurement device capable of distinguishing between a pure sine wave and a sine wave corrupted by (alleged) extra harmonics due to (alleged) FM caused by jitter - this pretty much has to be software-based, for practical reasons

The "measurement device" may be a spectral analyser or a notch filter - the point is to show or make audible any "crappiness" added by the process of D/A and subsequent A/D conversion with jittery clocks of "soundcards for ordinary people".

I mean, if before the D-to-A-to-D we have a pure sine wave -12dBFS loud, and after we have a -12dBFS sine wave plus -80dbFS of noise, maybe it's not such a big deal, frequency masking and all.

However, if we end up with -12dBFS sine wave plus -40dBFS second harmonic, or plus a bunch of -40dBFS sines n*50Hz apart from the sine wave (as in "classic" FM), it probably IS important.

kfoltman

krd wrote: »

The Soft synth companies are writing to card.

Or, say, reading from a sample-accurate event buffer and writing to an audio track in a DAW (say, a WAV file). Given the same source MIDI track and no added randomness from sources like noise generators, not-time-synced LFOs, oscillators not being reset on first note etc. you should get the identical track each time. Or at least the notes should start exactly at the same position within a track.

The only reason I don't implement it in my own plugins is that certain optimizations are easier to do when processing data in larger blocks (say, 32 samples), and I don't really think sample accuracy is such a big deal. Especially in open source plugins made for a very narrow group of people.

But even if they generate each sample exclusive of the card - the maths they use will always be slightly off each time it's iterated

I don't agree with "always". Though, the more complex the synth (or a patch) is, the more chance for getting different audio data.

What then happens when that audio data is sent to the soundcard - that's another story. But that's outside of the scope of softsynth creators' claims.

madtheory

Bluebirdstudios wrote: »

I use big ben to master clock PT 192s, Lynx auroras ,bricasti , Pcm 96 , eventide eclipse and other outboard.

Yes, that's exactly what a Big Ben is really for- clock distribution. It's not a superior clock, that's marketing bull.

Each of those devices you have are all extremely low jitter, with PLLs on the clock input.

When distributing word clock, you can:
1. Chain using BNC T junctions. Get out the oscilloscope and check for reflections caused by incorrect impedance matching and termination in the wrong places, then check each device and correct the terminations. Many devices are terminated by default. Either put the terminated device at the end of the chain, or if there's more than one, you need the schematic for each device to be certain that you're removing the right resistors.

2. Ignore electronic engineering, and fork out a load of cash for a Big Ben or WHY.

But the fact is, a Big Ben will not improve the sound of your HD192 on its own.

PaulBrewer

madtheory wrote: »

But the fact is, a Big Ben will not improve the sound of your HD192 on its own.

There are many who disagree !

Bluebirdstudios

madtheory wrote: »

Yes, that's exactly what a Big Ben is really for- clock distribution. It's not a superior clock, that's marketing bull.

Each of those devices you have are all extremely low jitter, with PLLs on the clock input.

When distributing word clock, you can:
1. Chain using BNC T junctions. Get out the oscilloscope and check for reflections caused by incorrect impedance matching and termination in the wrong places, then check each device and correct the terminations. Many devices are terminated by default. Either put the terminated device at the end of the chain, or if there's more than one, you need the schematic for each device to be certain that you're removing the right resistors.

2. Ignore electronic engineering, and fork out a load of cash for a Big Ben or WHY.

But the fact is, a Big Ben will not improve the sound of your HD192 on its own.

Actually have tried looping word clocks came across 2 problems -

1- the unit last on loop suffered sound quality wise and also had clock drop outs - ( even when changing what was last on line . As you have covered I was convienced there should be no problem but running seven lenghts of 75 ohm cable no matter how short causes audible jitter even mutible bnc t junctions will add to that jitter. Also note MR Lavry himself on some gearslut arguments stated master clocks are neccessary in situations as describe above.

2 - needed to be able to sync units sometimes working at different sample rates ie Master CD machine, Eventide eclipse - which the master clock allowed by having the ability to multiply or divide rates on their clock outputs.

I've already stated that the difference I've heard between master clock and none was very slight and money could be spend on other often overlooked products first in searching for audio bliss.
That said the big ben has proved a problem solver for me and with the very slight improvement in sound quality made the purchase easier.

madtheory

Agreed- it's for clock distribution where you cannot be certain what each device is doing, and for video work. It's not for improving an existing clock, despite the marketing.

PaulBrewer wrote: »

There are many who disagree !

And that is a pity. Hopefully, a bit of scientific thinking will have people buying it for the right reasons, as outlined above, and not hifi snakeoil notions about sound quality.

PaulBrewer

I don't really know where to go with this now - if you can't hear a difference you can't and I accept that.

PaulBrewer

http://www.antelopeaudio.com/en/shops1.html

Now that will sort your problems !

madtheory

Sell me a Prism Orpheus for the same price as a big Ben. That would certainly be an appreciable improvement in sound.

Just to clarify- two of us heard no difference. Blind A/B test. Acoustically treated room.

Maybe we should set up a blind A/B test with a larger group. Would be fun to get together too.

PaulBrewer

http://mixonline.com/products/review/audio_apogee_electronics_big/index.html

madtheory

You've linked to that article in this thread already Paul. That review was not a blind A/B test.

jtsuited

this debate will go on and on and on and eventually the conclusion will be:

1. external clocking CAN make a difference
2. whether the difference is actually worth the couple of k's you'll spend on it is justifiable is completely dependent on your point of view (and approach to the world in general).

Personally I think clocking is a complete waste of time and money for the difference it makes (and for the record, I've clocked 192's with a big ben and it HAS made a difference).

woodsdenis

jtsuited wrote: »

this debate will go on and on and on and eventually the conclusion will be:

1. external clocking CAN make a difference
2. whether the difference is actually worth the couple of k's you'll spend on it is justifiable is completely dependent on your point of view (and approach to the world in general).

Personally I think clocking is a complete waste of time and money for the difference it makes (and for the record, I've clocked 192's with a big ben and it HAS made a difference).

If I remember correctly I started this one off in a different thread by suggesting that rather than buying a summing box you would be better off
buying an external clock, I didn't specify which brand though. I can certainly hear and have heard a difference, when externally clocking PT HD systems.
I also would tend to listen to very experienced engineers and mixers who also concur. Be that as it may, is it an absolute must have if you are just using a stand alone system not syncing/synced to anything else, of course not. Probably for most people your 2k could be spent on a good pair of monitors before any clocking or summing comes into play.

PaulBrewer

woodsdenis wrote: »

If I remember correctly I started this one off in a different thread by suggesting that rather than buying a summing box you would be better off
buying an external clock, I didn't specify which brand though. I can certainly hear and have heard a difference, when externally clocking PT HD systems.
I also would tend to listen to very experienced engineers and mixers who also concur. Be that as it may, is it an absolute must have if you are just using a stand alone system not syncing/synced to anything else, of course not. Probably for most people your 2k could be spent on a good pair of monitors before any clocking or summing comes into play.

Just for clarity a Big Ben sells for Euro 1027 ex Vat. So from a Business buyer's perspective it's a 1k box not 2k.

However that's only a detail and I take Denis and JT's points.

PaulBrewer

I was chatting to Leizer at Antelope Audio today regarding this subject.
Here's an article in Tape Op.

http://www.antelopeaudio.com/en/images/TapeOpAntelopeReview.pdf

We have had no dealings with Antelope.




Yet !

jtsuited

so we're all agreed then........er, are we?

interesting issue, but very very deep, and tbh my head gets a little sore with some of the in-depth arguments some of the engineers involved in clocking get into.

Personally, would I buy a big ben? No.
Would I put a Big Ben in my system if it was given to me? Yup.
Will it make a difference? Depends.

PaulBrewer

Well Antelope Audio are sending over some units for us and anyone else who might like a go, to try.

Their Rubidium 10M Atomic Clock looks cool.

Ultra-precise Atomic oscillator with stability of 1 second in 1,000 years!


No idea what Rubidium is but sure, if it sounds good ........

kfoltman

PaulBrewer wrote: »

Ultra-precise Atomic oscillator with stability of 1 second in 1,000 years!

What's the typical oscillator drift in Minimoog?

Does Minimoog sound bad?

PaulBrewer

kfoltman wrote: »

What's the typical oscillator drift in Minimoog?

Does Minimoog sound bad?

One doesn't run the whole studio off a Moog though do you !?

kfoltman

PaulBrewer wrote: »

One doesn't run the whole studio off a Moog though do you !?

Hah, true!

Okay, so what do you think is more noticable, at least for people without a gift of absolute pitch perception: subtle drift of tuning of the whole mix, or tuning drift of one instrument against everything else in the mix?

Next question, how many cents of drift does "1 second in 1000 years" (or even 1 second in 1 year) correspond to?

1 day is 86400 seconds.
1000 years is almost 365250 days.
Windows calculator says 0.00000238 of a cent. Or 0.0000000238 of a semitone. I might have got the equations wrong though, so you may check for yourself.

(-:

PaulBrewer

kfoltman wrote: »

Hah, true!

Okay, so what do you think is more noticable, at least for people without a gift of absolute pitch perception: subtle drift of tuning of the whole mix, or tuning drift of one instrument against everything else in the mix?

Next question, how many cents of drift does "1 second in 1000 years" (or even 1 second in 1 year) correspond to?

One instrument shifting agin the mix would be more noticeable I reckon.
However we're not talking tempo here, regarding a whole mix.

I have no idea what these specs mean in real life, in fact there could possibly be a point above which an the increase clocks 'accuracy' has no impact on perceived audio quality.

However I can say I've either spoken to, emailed or IMed -

Pat McCarthy ( REM, Madonna)
Mike Masters (Flight of the Conchords)
Mick Glossop (Van Morrison/PIL)
Eddie Ciletti ( Tech writer for Mix Magazine )

who all agree converters CAN sound better Externally clocked.

madtheory

Just thinking:
1. A blind A/B test with a group of us from here would be interesting, and fun.

I get the impression that no one has done a blind A/B test on clocking. It removes bias. And bull****.

2. I don't know if there's more than one version/ revision of the 192. Maybe they don't all have a PLL?

3. Ethan Winer has an interesting article on the source of this type of disagreement- it's the room. "Comb filtering is the root of all evil"
http://www.ethanwiner.com/believe.html
Just moving your head an inch gives a different sound, if the room is not treated.

Anyway, I think I've outlined the science of this issue as clearly as I can. If people want to take the easy option and "believe" that's their problem!

madtheory

PaulBrewer wrote: »

I was chatting to Leizer at Antelope Audio today regarding this subject.
Here's an article in Tape Op.

http://www.antelopeaudio.com/en/images/TapeOpAntelopeReview.pdf

Ok, so he says they printed mixes for four different clocks and did a blind A/B test using laptop spekaers. It doesn't say how they printed the mixes. Were the bouncing down on Pro Tools? If so, then that's thick, because the clocking makes no difference to an internal bounce. Which means they were imagining things on the laptop.

But I'm only assuming that. I'll email them and confirm...

madtheory

The silence is deafening!

kfoltman

madtheory wrote: »

The silence is deafening!

No, it's just jitter interfering with the ability to hear things properly!

madtheory

Nah, can't hear jitter on this forum, the noise floor is too high...

PaulBrewer

madtheory wrote: »

Just thinking:
1. A blind A/B test with a group of us from here would be interesting, and fun.

All been done - that's how we came to our conclusions - which is why I sent you the Big Ben to try. (Not that you 'tried' too hard by the sounds of it either, where's the science there ?)


Your argument was that there couldn't be any change - there clearly is.


I and many World Class experienced Recording Engineers both here and abroad who I've contacted are in agreement with that.

In fact I can't find anyone to disagree that there's a change - except you.

Doesn't the fact JT and I agree on something not tell you something?

Just let it lie will ya ? It's over.

kfoltman

PaulBrewer wrote: »

Doesn't the fact JT and I agree on something not tell you something

Actually, I don't see it as agreement at all.

Also, "appeal to authority" is a weak argument. It's like citing doctors who trust homeopathy.

Not saying that clock distribution systems are exactly like homeopathy - they *do* have their uses, when multiple devices in the studio must be clock-synced, or when device's internal clock is utter ****e. Hell, I externally-clock my primary soundcard, because I record some inputs from external A/D converter (not using Big Ben or anything like that, though).

On the other hand... You can obsess about jitter, which produces amounts of audio distortion that is even hard to detect using objective methods, and very hard to A/B test. But then the money spent on super-accurate clock to sync a single device is the money not spent somewhere else, where one could get some obvious gains: room acoustics, better instruments or outboard gear, furniture (yes!), music lessons (in case of amateur home musician like me), books, music CDs...