Digital myths - a must read

  • Thread starter Thread starter sjoko2
  • Start date Start date
EXcellent article - takes the wind out of the sails of both analog bigots and digital bigots.

I guess the short summary is that digital has advantages and disadvantages - and so does analog.
 
I think it clearly outlines some basic facts and principles everyone should know
 
Most of this was fairly elementary, but I really liked #8.

Clear answer to a question I asked here a week or two back with no real luck: What's the qualitative difference between hardware and software digital effects?

None, unless they're programmed in.
 
Exactly bong - so it comes down to the following points:
- Most hardware units cannot be upgraded, their soft versions can.
- When recording digitally, the latency of software is much lower than hardware. If the hardware has analogue I/O its made worse in both latency and quality through additional DA and AD conversion requirements.
- Use of hardware is restricted by number of I/O - soft is restricted only by DSP power - which in general is cheaper.

As an example - I use a lot of TC plug-ins. On a recent mix I used EQsat, VoiceTools, and MegaReverb, on a whole bunch of channels. Total plug-in costs about $2.500 Someone counted them, calculated the number of units required to do the same with TC hardware, and then costed that ...... Over $30.000 You can buy a pretty decent Pro Tools rig for that, with plug-ins and plenty of prosessing power.
 
Gidge, that article contains some very definate misconceptions
Waht was SP's opinion?
 
I really don't like the marketing smile on that John Vestmans face. Gives me the chills... :eek:
And while his points might make sense, he doesn't really know what he's talking about judging from his explanation. Computers crash. Yep, that's a proof. Certainly when you're running windows. The man's not smart enough to simplify those explanations. Something like that would be it. Well... I don't care.

Next article? :D
 
Jezus Gidge! Glad I wasn't around then :)
Neither am I going to rake up any of that shit.
Just 3 comments, then everyone can form their own opinions:

- There are different quality CD's available for writing audio data to. For good reason. I believe Dan is talking about a system where he is using the right tools to burn a correct type of CD.

- Data = data, no matter which way you look at it. Burn it at 1, 2, 50 times. If the tools are right, if its copied right, it works or it doesn't work.

- Is there a difference between 'normal' and audio data? No. Is there any other thing that differentiates audio data? YES! In that audio data has to be delivered IN TIME. This is the one OVERRIDING factor. If audio data has to rely on any kind of correction, thats to late already. This is not just the case with (even though accentuated by) CD's, this is the case throughout an entire digital chain.
 
To point out another difference between audio and data on CD... With data CD's, there is extra error-correction. CD's only have a simple error-correction(With interleaving etc), because, for consumer-audio, it doesn't matter all that much. Since there is an analog part behind it that filters the high freq out of the signal, the errors that cannot be corrected by the errorcorrection will be less noticable because of the integrating behaviour of the filter!

The error correction that is foreseen for audio CD is on most players very bad implemented, because it simply isn't all that important. (The standard error correction is implemented, if the error is noticed by the system, but cannot be corrected, you can integrate the signal, regenerate the value some way, but since the analog circuitery allready does that partly...)

Anyway, the audio data on a cd has got standarized error-correction and interleaving. This is nowadays implemented in hardware, and is handled in realtime! (Actually, the encoding/decoding part of the CD/DVD chips is what I test at work... :cool: ) What you copy when you copy a cd, is the DECODED data! So first you decode the data, get most errors out by the errorcorrection, and then you decode it again and write it. So only VERY bad errors will result in noticable degradation, when there are to much of them. (I doubt that you will hear one wrong sample, taking into count that it's stereo and integrated by the filter...)

Not that I'm an expert on the field. ;)

(By the way, the cd was developped for audio. A data CD is just a VERY BAD sounding audio CD. The standard is totally conform with the audio cd, but defines a higher level encoding in bigger blocks of data. Most audio CD players will mute the data because they know there's something strange about the cd.)
 
Roel, what you've said leads me to believe there shouldn't be any difference in the sound quality between cdr's unless there is significant error in the inferior cdr. How can there be that much difference between a data and audio cdr especially since the non-audio data, as you said, requires more error correction data?
 
Sjoko,

I didnt get involved in it then and wont get involved in it now...just thought it was interesting comparing the two articles....but what do I know, I thought The Klits was a good name for a band;)
 
But it is! There was a band in the Netherlands called the Klits.
And my son's band was called MonkeySpank
 
Oh yeah, I also meant to warn everyone as stupidly curious as myself from reading any of the three threads gidge listed a few posts back. They are all useless and a complete waste of time unless you're into the bizarre extremes these discussions can go to when someone takes it personal. I wasted my whole morning(my only time to play) reading nearly all of every post, hating myself the whole time for not doing something more constructive.

Curious now, aren't 'cha? Hah! Now you can't resist... DON'T DO IT!!
 
Yep, Sjoko, but didnt that band change their name....well, you wouldnt know anything about that, would ya;)
 
ap said:
Roel, what you've said leads me to believe there shouldn't be any difference in the sound quality between cdr's unless there is significant error in the inferior cdr. How can there be that much difference between a data and audio cdr especially since the non-audio data, as you said, requires more error correction data?
Click to expand...

Just some last words on it... Hehe. Data cd's have extra errorcorrection because data can be any totally not-correlated order of zero's and one's. So extra errorcorrection is needed because here the data will be useless if just a bit is wrong, and the data cannot be estimated using previous and following data.

The data however doesn't have to be read right the first time, you can go back and try it again if it fails. Cannot do this with audio...

The error's that occur can be of all kinds. Scratches, fingerprints, black dots, anything wrong with the disc. But it could also be circomstances, trying to read to fast, a shock, ... Or data that wasn't correctly written. Lot's of reasons. Not really what I'm doing. I have no idea how many errors would occur. But I think recent technology is pretty stable. (Reading at 50x is possible, so 1x won't be a problem no more...)

Another serious misconception is the point he starts with. The data on a cd will not degrade by playing it. THAT's the real statement. And that's true. It will degrade slightly when copied (I think it'll take more a few copies before you'll notice...) and over time. (Guaranteed over 40 years, for REALLY bad conditions, think sahara...) Just don't worry about it. It's good enough for us.
 
You must log in or register to reply here.
Back
Top