Checking Dither and CD error detection

[dean1964]

I was wondering what ways there are to check and confirm if a dither plugin was working as it should be. I'm using waveburner loaded with 24/44.1 files and the waves IDR dither on the master plugins plane.

I've got a demo pf spectrafoo for the next few days but unfortunately there is not an OSX plugin version, so I cant stick the bit scope on it until after it is encoded. This being the case I have no real point of reference.

Once the disc is burnt, what methods are there for error checking the disc?

Thanks folks

[Ethan Winer]

Originally Posted by dean1964

I was wondering what ways there are to check and confirm if a dither plugin was working as it should be.

LOL, I'd love to know that too. Gee, if only dither actually made an audible difference, it'd be so much easier!

--Ethan

[Chili]

To check your disks for errors, you can use Nero DiscSpeed.

http://www.cdfreaks.com/software/Dia...DVD-Speed.html

You're looking at your C1/C2 errors. Whether you can actually read the codes or determine error rates depends on your cd-drive, it has to be able to return error codes. There's a plethora of info on www.cdfreaks.com website.

As for dithering: I've read Ethan's comments in the past and honestly, I don't understand it all. I believe dithering is more valuable for very dynamic material where the quiet parts are really quiet. Don't know how true that is. I dither when I convert to a lower bit depth simply because it's cheap and easy to do.

And I don't know of anyway to 'check' it.

Hope this helps.

(and let the dither debate begin again )

[dean1964]

Thank you sir. Ill check DiscSpeed out.

I really cant tell an audible difference which is why I wish spectrafoo had a plugin version for OSX. You could plugin the bit scope, and turn the dither on and off during silence to see the difference, I think. I like to measure exactly whats happening with re-quantization to 16 bit.

Details Details..

When it comes down to it....

"...what does it sound like baby!" - Ray Charles

[Massive Master]

If you can't hear the dither - most of the time, most people won't - Especially on music that has a crest of 10 or 15dB.

AFA error checking - I can tell you without little doubt that DiscSpeed isn't worth its weight in worthless. I had high hopes for it (as Plextor stopped making "real" drives) as the NKOTB in BLER checking, but it failed me miserably on several drives. "Perfect" discs over and over - Inconsistent results from one scan to the next - I'd even scratch discs on purpose (which of course, showed up like fire using PTPro) and DiscSpeed showed nothing.

I can't imagine how bad a disc has to be before DiscSpeed would actually register a few BLER bursts. I wouldn't trust it as far as I can throw my brother.

[mshilarious]

I'm in Ethan's camp; after long argument and experimentation, I don't believe it is possible for quantization distortion to exist on program material, which includes any solo acoustic instrument. You could experience it with softsynths though, on certain patches. So no chance of quantization distortion leads to the conclusion that dither is not necessary.

On the other hand, it should be quite easy to measure what the dither is doing: dither a blank file and note the dither spectrum. Compare with a very quiet part of your track, and see if you can note the same spectrum. If your track never gets as quiet as the dither at some frequency, then you didn't need it in the first place.

[dean1964]

I ran DiscSpeed and it hit me with 100% no errors, my eyebrow was raised. It just does not seem likely that the disc would be perfect considering the margin of error allowed by the manufacturing plants.

So are there any alternatives?

Just FYI. I burned the disk with me MacBook superdrive and tested on the NEC 3500 in the PC.

Thanks

[Massive Master]

That's what I mean - I've burned and checked a few discs before... I've never had a "perfect" disc. I've had some amazing discs to be sure - But never one that had a BLER of "0" --

Plex Tools Professional or a Clover system. The only things I trust at the moment.

[dean1964]

Massive thanks Massive Master. Im with you 100%

Looks like I'll picking up one of those plextor premium drives sometime soon.

[Ethan Winer]

Originally Posted by Chili

I don't know of anyway to 'check' it.

It's easy to check. Do what I did here:

Dither Report

Then you'll know.

--Ethan

[Massive Master]

Originally Posted by dean1964

Massive thanks Massive Master. Im with you 100%

Looks like I'll picking up one of those plextor premium drives sometime soon.

They stopped making them - All of them. The Plextor PlexWriter Premium, the 712, 716, 755 and 760. There are some NOS units available here and there (for more than they listed for, of course) and I think the Japan market still has a number of PlexWriter2's. But few and far between finding them anywhere you can get your hands on them.

I'm glad I bought up as many as I did, but I'm down to 5. I might actually have to start sending them out for repair when they fail.

**RUMOR HAS IT** that a few new Plextor drives might be using a new Plextor utility that reads BLER. I'll believe it when I see it.

[dean1964]

Damn. I felt screwed for a second there. I guess Ill have to find another way.

Ethan, very cool write up about Dither.

Thanks Guys.

[Chili]

Originally Posted by dean1964

I ran DiscSpeed and it hit me with 100% no errors, my eyebrow was raised. It just does not seem likely that the disc would be perfect considering the margin of error allowed by the manufacturing plants.

So are there any alternatives?

Just FYI. I burned the disk with me MacBook superdrive and tested on the NEC 3500 in the PC.

Thanks

Originally Posted by Massive Master

That's what I mean - I've burned and checked a few discs before... I've never had a "perfect" disc. I've had some amazing discs to be sure - But never one that had a BLER of "0" --

Plex Tools Professional or a Clover system. The only things I trust at the moment.

Not that I'm disagreeing, but I believe your cd drive has to be able to return error codes. Not all do. Maybe you got no errors for that reason. I only tried Discspeed once when I downloaded it and got errors on a cd-r I burned in the same drive. I was going to use it to check my cd's before shipping to CDBaby. (Haven't gotten to that point, yet)

If it's not reliable, I've got to find something else. What's the Clover system?? Sounds expensive already.

Thanks for the info...

Edit: I just found Clover Systems.... it is expensive.

[SouthSIDE Glen]

Actually dither is one of those Zen meditation koans: If you can't hear it, it's working.

G.

[Massive Master]

Not that I'm disagreeing, but I believe your cd drive has to be able to return error codes. Not all do. Maybe you got no errors for that reason. I only tried Discspeed once when I downloaded it and got errors on a cd-r I burned in the same drive.

I had zero-error returns on discs burned in the drives the program came with. Then of course, I'd run it on one of the Plextor drives to find out what the real story was - Let's just say they weren't even close to similar...

[Chili]

Originally Posted by Massive Master

I had zero-error returns on discs burned in the drives the program came with. Then of course, I'd run it on one of the Plextor drives to find out what the real story was - Let's just say they weren't even close to similar...

Well, unfortunately, that's good to know. Any other suggestions for checking cd-r's before shipping them out? Other than the Clover System. Might just have to do the ol' Old CD Player check.

[SouthSIDE Glen]

As long as we're on the subject, can I throw another question in here for you John? It's been a while since I looked into this and I'm wondering if maybe this has changed lately: What is considered an acceptable error rate these days?

G.

[Massive Master]

220 C1 average errors per second over any given 10 second period. Which personally, seems ridiculously high (IMO) and I'd dump a disc for having 1/10th of that. A "good" disc (again, IMO) has an average of under 2 (ideally under 1) per second. "Fantastic" would be around 0.5 per second with a *peak* under 20.

No C2, no CU errors.

I think a lot of the replication houses think the "acceptable" numbers need to be updated also - I still get calls from plants if my numbers (I print the error rates on the disc before shipping them out) differ much from theirs. Even if they're still very good (maybe a peak of 50 but the disc says that the peak was 25).

Although at that point, I usually ask them if there's a fingerprint or something on it... And there usually is. One time I had them send me one back that was *way* off and it had a foot (shoe) print on it. Reality really is stranger than fiction sometimes.

[SouthSIDE Glen]

Gratzi!

G.