Irony: The "boring" parts interest ME!

M

Moseph

New member
Hey all,

I think I've got a pretty good concept of the types of things that most of the posts here are about. I'm not saying I'm particularly good at any of it, but I do think I at least have a basic understanding of Mastering at the artistic level, but less so at the more technical level: preparing the actual audio for the final medium.

In particular, this started off when I saw this post a few months ago:

SouthSIDE Glen said:
Fala, do you mind if I ask you a few questions about your mastering technique and process? (Everybody else, shhhhh! ;) )

1. What do you use to do your PQ code editing?
2. How do you generate your disc error list?
3. How do you generate your track list?
4. What do you feel is the proper ratio of C1 to C2 to CU errors on the premaster?
5. How do you determine track order?
6. How do you determine the reference or guide amplitude to adjust the songs to?
and finally
7. What is mastering?

G.
Click to expand...


These are all great questions to me, since I'm kind of in the dark about the nitty-gritty aspects of a lot of them. So I'm hoping maybe we can talk about this kind of stuff?

For the record, I did pick up a copy of Bob Katz's book already. I read through (some parts twice or more) and it had some good information, but a lot of it was also directed to the artistic side of things, and I don't know that the more technical aspects really "stuck" in my brain. If anybody can point me to some good links about the medium preparation (not necessarily just for CD either), that'd be swell. Free/online resources are always preferred, but if you know of a "must-have" book then don't be shy about it.

Just for some context of this post, I recently worked a session for a band and they also asked me to mix/master (I realize this isn't the best policy, but this was more for fun/learning than any top-notch commercial endeavor), and had to do a little bit of research into this sort of stuff. What I found was that I know f-all in terms of the technical aspects of mastering.

Take PQ Codes. All I really know about PQ codes is that they determine how the CD is structured in terms of track number, track length, and transitions.

I didn't have a "stock" answer to how to handle this. In terms of generating the final disc, I ended up importing all the finished songs into a session in Sonar and burned out one long continuous file. Then I used Sonar to find the points between tracks where I wanted the crossover to happen, and took note of the exact time code info for those points. Then I made up a CUE sheet and used CDBurnerXP to put it together. I have used Nero in the past (minus CUE sheet part), but found that some players had issues with truly "seamless" transitions between tracks when I did. So far I think this new strategy worked, but I haven't gotten to test it on some players I knew were a problem before (either they stopped working or I no longer have access to them).

As for C1/C2/CU errors (another part of SSG's post that made me realize how little I knew), all I can really say about them is that they're not normally detectable by listening and ideally they each should be on the order of 0-10 errors/minute (I think?). I don't really know what the different errors indicate though, and my detection/insurance against them was basically burning at a slow speed and crossing my fingers.

I don't do a lot of mastering, but I'm definitely still up to learn more about the process, and I think this is the area where most people really don't know how much they don't know. So any advice/tips/pointers/sagely wisdom is appreciated.

Thanks!
 
I've arranged CD projects in Sony Sound Forge 10 before. However, it doesn't produce a DDP fileset, which many disc manufacturing plants like to see. So, I'm currently checking out a program called DDP Creator. You might want to give that a look, for coming up with the final arrangement of tracks etc.

As for C1/C2 errors, what you want to find is a tool to tell you the BLER or block error rate of your disc. If you're submitting a CDR copy to the plant - not recommended - they usually have a maximum average BLER rate they will tolerate.

I don't have it in front of me, but the Katz book does go into all the technical details of how a CD is laid out, what a DDP fileset is, etc.

Another thing you might want to do is, when you have burned a CDR of the master, check it for errors by extracting the audio from it reliably (EAC or Exact Audio Copy is a good program for this), load the resulting wave files into your DAW of choice, along with the pre-CDR versions of the same audio. Then, play them back simultaneously, but with the phaze of one copy flipped.

this is called a "null test" because, ideally, the pre-cdr and the CDR coppies of the audio should be identical, and since you have the phaze reversed for one of them, they should cancel or NULL each other out. You should hear only silence through your monitors.

Anything you hear is a difference between the two, and is quite possibly an error.
]
 
jmz93 said:
As for C1/C2 errors, what you want to find is a tool to tell you the BLER or block error rate of your disc. If you're submitting a CDR copy to the plant - not recommended - they usually have a maximum average BLER rate they will tolerate.
Click to expand...

The tool in question would be used to actually burn the CD-R, right? Or am I looking for an analysis tool for after-the-burn?


jmz93 said:
I don't have it in front of me, but the Katz book does go into all the technical details of how a CD is laid out, what a DDP fileset is, etc.
Click to expand...

I'll have to go reading through it again. I has been awhile, maybe I'll retain more of it on a re-read.


jmz93 said:
Another thing you might want to do is, when you have burned a CDR of the master, check it for errors by extracting the audio from it reliably (EAC or Exact Audio Copy is a good program for this), load the resulting wave files into your DAW of choice, along with the pre-CDR versions of the same audio. Then, play them back simultaneously, but with the phaze of one copy flipped.

this is called a "null test" because, ideally, the pre-cdr and the CDR coppies of the audio should be identical, and since you have the phaze reversed for one of them, they should cancel or NULL each other out. You should hear only silence through your monitors.

Anything you hear is a difference between the two, and is quite possibly an error.
]
Click to expand...

Null testing never occurred to me! I'll probably sit down and do this in the near future, just to see what's up.

For the most part, I do this sort of thing to kind of "help out" small-budget bands/musicians, so I don't know exactly where they'll go with the "finished" disc when they get it. So far, I've only done a couple of "mastering" type projects, but they just resulted in digital audio "releases" (e.g., MySpace streams).

This sounds like good stuff to start looking into though, thanks.
 
I really think you'd be much better off just using software designed for the job -- Wavelab, CD Architect, Samplitude (if you go Samp Pro, you'll probably never need another program for anything except BLER check and DDP output if you need it).

** CD-R to the plant is absolutely fine -- Some plants don't even accept DDP filesets anymore. Why? No idea.

** BLER checks can be performed only by very specific hardware/software -- Generally with drives that aren't even made anymore (Plextor PWP, 712, 716, 755, 760) and the acceptable BLER is determined by orange/redbook specifications (although basically every mastering engineer I know finds those numbers FAR too high and have their own specs which are much, much more stringent).

** Burning slow is usually worse than burning fast when it comes to block error rates. Working in the "comfort zone" of around 25-33% of a drive's rated speed is almost universally where to be.

** Null tests work well with audio but have nothing to do with the BLER -- and the other way around -- An extremely low OR high BLER is NOT indicative of an accurate data transfer.

I'd make more of these bullet points, but it's early and I need more coffee...
 
It's all pretty much been covered very well by John (who ALWAYS needs more coffee, I think ;)) and jmz (of whom I know nothing of his coffee needs), but just to reinforce a point:

Moseph himself correctly said :
Moseph said:
As for C1/C2/CU errors...all I can really say about them is that they're not normally detectable by listening
Click to expand...
Which means it is correct that "null test" differential listening tests won't necessarily help in that regard. In fact, I'd argue that if you could hear a difference at all, that you probably have a bigger issue on your hands than just a high Cx error rate

You have to have software that can perform a differential on the actual bit-by-bit data itself, along with a drive that will actually provide that unadulterated data stream. Personally, I use a Plextor 716 burner/drive along with either/both the Plextools software that came with it and Nero CD-DVD Speed utility to check the disc.*

And, personally, I grew up on the Sonic Foundry/Sony stuff, so I'm a CDArchitect guy myself when it comes to the actual CD pre-mastering software, and use it's built in CD Subcode/Track List editor for my PQ editing.*

*CAVEAT: I am not a mastering engineer as my main course, do not claim to be, and do not solicit my services as one on either a pro or amateur level. Honestly therefore don't often have the opportunity to officially take the process this far myself, so I defer to the others on the final word on much of this stuff.

G.
 
Thanks for the input, gentlemen. Definitely some stuff to think about.

A couple of quick questions/comments though:

Massive Master said:
** BLER checks can be performed only by very specific hardware/software -- Generally with drives that aren't even made anymore (Plextor PWP, 712, 716, 755, 760) and the acceptable BLER is determined by orange/redbook specifications (although basically every mastering engineer I know finds those numbers FAR too high and have their own specs which are much, much more stringent).
Click to expand...

If I'm reading this right, then one of the standard rubrics for duplication is no longer supported by most modern hardware? Is the assumption that things are staying in the digital realm until mass production nowadays? Is there a way to find out if the drive I already have can report this sort of thing?


Massive Master said:
** Burning slow is usually worse than burning fast when it comes to block error rates. Working in the "comfort zone" of around 25-33% of a drive's rated speed is almost universally where to be.
Click to expand...


Noted, will look more into this.


Massive Master said:
** Null tests work well with audio but have nothing to do with the BLER -- and the other way around -- An extremely low OR high BLER is NOT indicative of an accurate data transfer.
Click to expand...

If it's not used to determine accuracy in data transfer, what is it used for? I thought the idea was to get a general idea about how well the data was preserved in changing to the physical medium?
 
Moseph said:
If I'm reading this right, then one of the standard rubrics for duplication is no longer supported by most modern hardware? Is the assumption that things are staying in the digital realm until mass production nowadays? Is there a way to find out if the drive I already have can report this sort of thing?
Click to expand...
This is an unfortunate trend that I don't completely understand, but the quality of consumer equipment in this regard is horrendous, with the best stuff generally disappearing. I think it probably has much to do with a lack of a market for it. Generally speaking, consumers are just not interested in true (pre)mastering of audio discs, and in fact don't even KNOW what's involved in this regard. They just want to make their CD-Rs, with "mastering" meaning to them only trying to make up for a lack of mixing skills, and then send them off for documentary duplication (if they even want CDs at all any more), or let the mastering engineer take care of that "boring" stuff. Kind of hard to sell more expensive burners in that kind of uninformed and disinterested market.

G.
 
So, in other words, the reason manufacturers aren't adding the analysis features might be because they know the product can only output sub-par results? In other words, they're trying to hide their shame, and get away with it because nobody knows any better?
 
Moseph said:
So, in other words, the reason manufacturers aren't adding the analysis features might be because they know the product can only output sub-par results? In other words, they're trying to hide their shame, and get away with it because nobody knows any better?
Click to expand...
No, that's not really what I'm saying. What I meant was - and this is just speculation on a partial answer, don't take it as gospel - there's not a big enough consumer market to bother including those features.

Probably a good 99.9% of purchasers of CD-R burners are using them either for non-audio data purposes or simply cutting audio CD-Rs to give to themselves or their friends and family, where things like Redboox Cx error rates, PQ code editing, DDP files, etc. just aren't really needed or even known about.

Kind of hard to justify to stockholders why one should spend the extra cost on selling one out of a thousand specialty drives and software to folks like you and me, when they can sell lower cost packages to the other nine-hundred-ninety-nine home consumers who just want to play a few MP3s in their car, archive their office data, save their holiday pictures for posterity, or give a few demo discs of their bar band to a few friends, bar managers or college radio stations.

G.
 
You must log in or register to reply here.
Back
Top