CCETargetSectors Question

[btspm]

A few questions about sizing:

1. Does the version of CCE used matter? I used to be spot on with 2260000 for 2.67, but the various versions of 2.70 seem to be all over the map. Is there a similar sweet spot for 2.70, or is 2.70 just less consistent?

2. Would using more passes give a more consistent result? If I understand correctly, part of the sizing 'error' is caused by being unable to perfectly predict bitrate allocation. Would feeding more passes (information) into the equation help either consistency (same size for same CCETargetSectors setting) or accuracy (output matches up with target according to definition of bytes per sector)?

I know there is no appreciable quality increase by filling the last ~300 MB, but sometimes I have discs with upwards of 250MB in DVD-ROM extras- sizing becomes very tricky with those- even if I leave CCETargetSectors at default.

[JFerguson]

I've been using CCE 2.70 for quite a while.

Assuming that 2236400 = 4.32GB, the following formula is pretty consistent for me:


tgtsize (MB) 1 2236400
----------- x ---------- x
1024 4.32

This hits my target size within 10MB 90% of the time, if using the old editor. Note: I have not had consistent results with the new editor.

If I were targeting 4440MB, but needed an extra 250MB to merge DVD-ROM content, then I would use:

((4440-250)/1024)/4.32)*2236400 = 2118263

for TargetSectors= setting.

BTW, I never save DVD-ROM content. I'm usually oversizing discs for post-processing, so I don't have any undersizing projects to reference, but my oversizing ones fall within that 90% figure I gave above.

p.s. - I rarely use more that 2 passes for CCE. If a movie's bitrate falls to around 3500kbps, I might go 3 passes. I nuke most bonus content on my backups, so most movies hit 4000+kbps easily...

[rahzel]

a DVD+R has 4482mb and a DVD-R has 4489 or something, so about 4.37gb.

CCE's output does vary, but i use 2254000 with CCE 2.70.2.10, and i've never gone over 4482mb. most encodes are usually around 4465-4480mb.

[jonny_ftm]

I personally undersize all my DVDs
I use TargetSectors=2122509, which gives me stable output at 4.11Gb, occasionally 4.10Gb, and really quiet never outside these ranges. One or two on hundreds DVDs, I got a 4.05Gb

I prefere undersized DVDs. I never saw quality issues in the movies (I mostly use movie only mode, occasionally full DVD mode). Undersizing should make your DVDs live longer, as they usually alter on the outside edges first.

[rahzel]

I personally undersize all my DVDs
I use TargetSectors=2122509, which gives me stable output at 4.11Gb, occasionally 4.10Gb, and really quiet never outside these ranges. One or two on hundreds DVDs, I got a 4.05Gb

I prefere undersized DVDs. I never saw quality issues in the movies (I mostly use movie only mode, occasionally full DVD mode). Undersizing should make your DVDs live longer, as they usually alter on the outside edges first.
actually, if anything, the inner edge would degrade quicker than the outer edge. the bonding on the outer edge is stronger than it is in the inner edge. quite a few DVD-/+R manufacturers currently are having issues with bonding, and it's the middle thats the most suspect to seperation, therefore, degrading quicker.

also, a lot people like to use DVD cases and/or CD cases, so through normal wear and tear of taking your DVD's in and out, that puts more pressure on the middle.

ALSO, if there is ANY seperation at the outer edge, even if you only burn around 4GB, air will still eventually seep in and the disc will still be unreadable. of course, seperation at the outer edge should really never happen through normal wear and tear.

theres nothing wrong with burning a DVD to its max, and it won't degrade any quicker IMO.

[jdobbs]

I can tell you quite confidently that problems occur much more frequently on the outer edges of the discs.

[Rippraff]

@rahzel

I completely disagree with your above post.
I've seen hundred of scans and problems occur mostly on the outer side.

Cu Rippraff

[jdobbs]

This hits my target size within 10MB 90% of the time, if using the old editor. Note: I have not had consistent results with the new editorWhat kind of variance are you getting with the new editor? I'm looking through the code and can't find any miscalculations. Is it coming out too small, too large?

[rahzel]

even though it was a pretty long post, it was kind of rushed and it was at like 4am. my post doesn't have much to do with jonny's post at all. i'm talking about bonding, and hes talking about burn quality.

anyway, id agree that the outer edge is usually the most problematic, but the outer edge IS bonded better than the inner edge. as long as the burn was ok at the end (most media should be ok burning full) the inner edge would degrade quicker if anything, not the outer.

[jdobbs]

There are lots of theories why, and I'm not sure which is right... my best guess is that the outer edge would experience more up/down "wobble" than the part closer to the hub -- so the laser focus might not be as precise.

[Video Dude]

I have read on various sites that errors are prevalent because the dye around the outer edge is not evenly distributed during the manufacturing process. Errors are more common in cheaper discs.

[rahzel]

all DVD+/-R's get errors, but there are different kind of errors. and honestly, the amount of errors a disc has after a burn, has very little to do with the longevity of the media. the quality of the dye, how well the disc is bonded/manufactured, how its stored, and again, the burn quality has little to do with it. there is media that has amazing initial burn quality (ie Taiyo Yuden) however, there are DVD-/+R's out there that have inferior write quality, but would last even longer.

but yeah, i agree, it's usually the cheaper discs that are problematic near the end, and most DVD-/+R's are ok burning to the max.

[~bT~]

I can tell you quite confidently that problems occur much more frequently on the outer edges of the discs.
I can back that up coz it always seems that after a long period of not using a certain disc I get problems playing near to the end. The rest is fine.
PS> These are not cheap discs Im talking about. TDK, Verbs, TY & Panasonic to name a few.

[JFerguson]

What kind of variance are you getting with the new editor? I'm looking through the code and can't find any miscalculations. Is it coming out too small, too large?

Too small -- I'll run another test with 1.10.5...

Actually, can I not predict this somewhat by comparing the "vbr_brate_avg=" attribute in the REBUILDER.ECL files?

That is, if I run a project using the old editor and it produces correctly sized output, then I should be able to blank, adjust, etc. in the new editor, save, and then compare these values to the old editor?

[jdobbs]

Yes. As long as you know the reduction percentage that was used in the first pass (which is in the INI file).

[SAPSTAR]

I can tell you quite confidently that problems occur much more frequently on the outer edges of the discs.
Absolutely, especially if you're using "low" quality DVDs...a big difference between the different brands is the quality of the outer ring, some are well finished, others are lacking of dye medium.
But with DVDs like Tayo Yuden, TDK, Verbatim, ... all the quality tests were made to allow a full usage of your DVDs.
About the degradation with the time, all depends once again, of the initial quality of the media...whatever size you burnt on it.

[jdobbs]

The better discs are more stable and the likelihood is lower. But my experience has been that even the best ones have a tendency to occasionally fail on the outer edges. That is the main reason DVD-RB defaults to a 4.32GB image instead of a 4.37GB image -- early tests showed that even the top notch discs were weak there.

I don't think the problems I found in testing the "outer edges" were age related -- you'd see the problem on your first playback. I've seen it on virtually every kind of disc and with several different burners.

[jdobbs]

@JFerguson

Just FYI. Here's the calculations associated with target sectors that will accurately predict the output size (assuming encoder perfection).

(TargetSectors x 2048) x 1.015 = Output_Size

DVD sectors are always 2048 bytes long. The .015 adjustment represents the overhead (NAVPACKS, Packet headers, etc.) associated with packetizing the video/audio for DVD.

So the default value of 2236400 would result in:

(2236400 x 2048) x 1.015 = 4,648,849,408 = 4.329GB

The result usually falls between 4.32GB - 4.33GB.

[therat]

So the default value of 2236400 would result in:

(2236400 x 2048) x 1.015 = 4,648,849,408 = 4.329GB

The result usually falls between 4.32GB - 4.33GB.

But isn't 4,648,849,408 bytes 4.433 GB?

[rahzel]

But isn't 4,648,849,408 bytes 4.433 GB?
no, its 4433mb, not GB. 4433mb is 4.33GB

[Sharc]

1 kilobyte = 1'024 Bytes
1 Megabyte = 1'024 kilobytes = 1'048'576 Bytes
1 Gigabyte = 1'024 Megabytes = 1'048'576 kilobytes = 1'073'741'824 Bytes

1 byte = 0.0009765625 kilobytes = 9.536743164E-7 Megabyte = 9.313225746E-10 Gigabyte

Hence:
4'648'849'408 bytes = 4'539'696.688 kilobytes = 4'433.297546 Megabyte = 4.329392135 Gigabyte

The conversion factor is always 1024 or 1/1024, respectively.

[jdobbs]

Exactly. In calculating storage you use binary numbers. So one Gigabyte is 2^30. That's 1,073,741,824.

1 Kilobyte = 2^10
1 Megabyte = 2^20
1 Gigabyte = 2^30
1 Terabyte = 2^40

This fact is sometimes obfuscated by hard drive manufacturers who like to call 1,000,000,000 bytes a GB on the box. But that's just a way to make their discs look bigger. Interestingly, though, when you are talking about frequencies binary numbers aren't used. So 1Mhz is 1 million cycles per second, 1GHz is 1billion cycles per second etc... no wonder everyone gets so confused.

[therat]

quite correct, I forget to divide the 4433mb by 1024. my mistake

cheers