24bit and CD sector boundries?

[homer420]

Howdy folks.  I did a couple quick searches but didn't turn anything up, so apologies of this has been asked/answered already .....

I just moved into the 24bit world and wasn't sure if that meant I no longer need to concern myself w/ CD sector boundaries?  Since it won't land on CD I figured this was no longer necessary, but wanted to check to make sure I'm not missing anything.

Thanks in advance!

Peace,
~Dave~

[fozzy]

It is not an issue but I still split according to CD markers in wavelab in case files are dithered/resampled down to 16/44.1

gapless 24bit playback is up to the DVD-A/DVD-V authoring and/or software/hardware used to playback the files.

[homer420]

It is not an issue but I still split according to CD markers in wavelab in case files are dithered/resampled down to 16/44.1

But wouldn't the sector boundary alignment go out the window once dithered/resampled?  The filesizes would all change and thus any alignment/padding would be gone .... wouldn't it?

[fozzy]

sector boundary has to do with sector length not actual file size.

[JasonSobel]

It is not an issue but I still split according to CD markers in wavelab in case files are dithered/resampled down to 16/44.1

But wouldn't the sector boundary alignment go out the window once dithered/resampled?  The filesizes would all change and thus any alignment/padding would be gone .... wouldn't it?

no, I routinely track out my 24/96 masters in CD Wave, and then use WaveLab or r8brain to batch process the files to resample/dither to 16/44.1.

I always check my 16/44.1 files with SHNtool, and I never have SBE's.

the key here is that the files are cut on even increments of 1/75th of a second.  at 24/96, 1/75th of a second is obviously more sample and bits.  but when it gets resampled to 16/44.1, the total file size is an exact increment 2352 byte/588 samples at 16/44.1.

[wbrisette]

I wonder if we shouldn't make this a sticky. This topic does come up several times a year.

Wayne

[morst]

sector boundary has to do with sector length not actual file size.

Also it has to do with sample rate and length, and not bit depth. If we were to record at 88.2 KHz, we could split those files on 1/75 second points, and that would sample-rate-convert neatly to SBE-free 44.1Khz. 24 bit, 16 bit, 8 bit, still has the same number of samples per second, the 24bit ones just have more resolution!

Hey, so if I record at 48KHz and split files into even multiples of 640 samples, my 44.1KHz output will be in 588-sample increments? This gives me an idea!! 

[fozzy]

If i set "CD markers" in wavelab 5 on a 24bit file, when i resample/dither the sector boundaries are accurate(no SBEs) on the 16/44.1 files.

Does other software behave the same way?