Posts: 141
Threads: 16
Joined: 2019 Jun
Thanks: 21
Given 87 thank(s) in 53 post(s)
Country: 
Hey everyone,
I've been capturing bit-perfect PCM using Bronan's guide (Reaper set to 44/24 and then exporting at 44/16). I've got some questions about editing.
Although I capture in Reaper, I've been editing out side breaks/silence between discs using Audacity (a program I know a lot better). I'm wondering if that is messing up my files at all.
I've got two options for editing:
Option 1: Capture in Reaper at 44/24. Edit out side breaks/disc swaps in Reapar. Export edited WAV file in 44/16.
Option 2: Capture in Reaper at 44/24. Export file in 44/16. Put it into Audacity (which seems to upsample file from 44/16 to 44/32). Edit in Audacity. Export from Audacity at 44/16.
Will either option impact the bit-perfect nature of the audio itself? I know upsampling the khz is bad but is there any problem with upsampling the bit depth?
Any help would be great. Thank you!
Posts: 7,153
Threads: 601
Joined: 2015 Jan
Thanks: 1081
Given 1466 thank(s) in 963 post(s)
Country: 
If you capture laserdisc at 44/24, it's already not bit-perfect...
Posts: 141
Threads: 16
Joined: 2019 Jun
Thanks: 21
Given 87 thank(s) in 53 post(s)
Country:
(2020-07-09, 06:55 PM)spoRv Wrote: If you capture laserdisc at 44/24, it's already not bit-perfect...
Ahhh. I was following Bronan's guide: https://fanrestore.com/thread-2742.html
I've used the 44/24 capture method to capture both DTS CDs and LDs (which then are able to be converted from PCM to DTS) which would seem to suggest it doesn't harm the audio in any way. I could be totally wrong about that though.
Posts: 7,153
Threads: 601
Joined: 2015 Jan
Thanks: 1081
Given 1466 thank(s) in 963 post(s)
Country:
(2020-07-09, 09:24 PM)alinskey Wrote: (2020-07-09, 06:55 PM)spoRv Wrote: If you capture laserdisc at 44/24, it's already not bit-perfect...
Ahhh. I was following Bronan's guide: https://fanrestore.com/thread-2742.html
I've used the 44/24 capture method to capture both DTS CDs and LDs (which then are able to be converted from PCM to DTS) which would seem to suggest it doesn't harm the audio in any way. I could be totally wrong about that though.
It "could" not harm PCM (read: even if not bit perfect, you could get the signal more or less the same), but it would definitely harm it for DTS (and AC-3, too)
Posts: 141
Threads: 16
Joined: 2019 Jun
Thanks: 21
Given 87 thank(s) in 53 post(s)
Country:
(2020-07-09, 11:12 PM)spoRv Wrote: (2020-07-09, 09:24 PM)alinskey Wrote: (2020-07-09, 06:55 PM)spoRv Wrote: If you capture laserdisc at 44/24, it's already not bit-perfect...
Ahhh. I was following Bronan's guide: https://fanrestore.com/thread-2742.html
I've used the 44/24 capture method to capture both DTS CDs and LDs (which then are able to be converted from PCM to DTS) which would seem to suggest it doesn't harm the audio in any way. I could be totally wrong about that though.
It "could" not harm PCM (read: even if not bit perfect, you could get the signal more or less the same), but it would definitely harm it for DTS (and AC-3, too)
I see. I've been able to play my edited PCM capture of a DTS soundtrack on Foobar no problem. I did just put that same file into BeSplit and got a weird DTS file. It was all the content but at like 2.2x speed. No idea what is going on there.
Do we have a guide about highest quality editing? Silence before and after, side breaks, and potentially small edits need to be done in order to sync. Is the goal "bit-perfect for as long as posible... until you can't be bit-perfect anymore"?
Posts: 279
Threads: 8
Joined: 2015 Feb
Thanks: 163
Given 124 thank(s) in 83 post(s)
Country:
In my setup at least, it doesn't matter if I set it to a higher bit-depth or sampling rate when capturing PCM, as it's essentially capturing the same data stream regardless.
I once accidentally had Sound Forge set to 24/96 for one LD, but all I had to do was go to "Resample" and then "Set the sampling rate only (do not resample)" to change it back to its proper 44.1kHz, and then used eac3to ("eac3to input.wav output.wav") and it automatically changed it back to a 16-bit file, since Sound Forge simply padded the 16-bit stream with zeroes to make it 24-bit.
Posts: 109
Threads: 17
Joined: 2019 Jan
Thanks: 75
Given 50 thank(s) in 21 post(s)
Country: 
2023-08-31, 01:19 PM
(This post was last modified: 2023-08-31, 01:30 PM by Falcon.)
(2020-07-09, 06:55 PM)spoRv Wrote: If you capture laserdisc at 44/24, it's already not bit-perfect...
I know this post is old, but I'm going to correct this sentence so new people don't have to worry:
24-bit capture can be perfect (and usually is), when you capture 16-bit digital audio in 24-bit you will simply have 8 bit of superfluous zero bytes added. This does not degrade the 16 bits inside in any way. You can delete superfluous zero bytes in many ways. If your track is originally AC3 or DTS, you will need to delete superfluous bit to be able to decode the files.
You can delete it like this:
Put the 24-bit audio into EAC3to, and simply convert to WAV (input WAV output WAV) or Flac (DO NOT DO -down16) EAC3to will tell you that the track actually contains 16 bits + superfluous zero bytes, and will automatically delete it.
However, sometimes it doesn't, for one simple reason: if a single sample was not captured correctly (which became noise), this sample will be in real 24 bit. If EAC3to detects a single sample at 24 bit, it will leave the entire track at 24 bit. This case is rare but it can happen.
For PCM audio it is not important to have one missed sample. For the AC3 and DTS I don't know if the decoding works with a missed sample but I suppose that there is a safety margin and redundancies to avoid this kind of problem (like AC3 on 35mm which had redundancies in case a QR code is damaged).
You can also export from Audacity or Adobe Audition in 16 bit by deactivating the dither, it will cut you're 24 bit track at 16 bit, and from the tests I carried out, the perfect bit was maintained.
Another thing regarding bit-perfect capture, for me, the easiest way is to capture in 24 bit with Audacity (Wasapi mode only).
Indeed, if you capture in 16 bit on Audacity, it adds a dither to the capture (not bit-perfect). By capturing in 24 bit, it does not add dither, and keeps the 16 bits intact ( + 8 bit of superfluous zero bytes ). You can then export in 16, although for beginners, I recommend letting EAC3to delete the superfluous bits (you never know if you left the dither active when exporting from Audacity)
Posts: 759
Threads: 35
Joined: 2018 Feb
Thanks: 651
Given 1071 thank(s) in 405 post(s)
Country:
2023-08-31, 05:00 PM
(This post was last modified: 2023-08-31, 05:04 PM by bronan.)
(2023-08-31, 01:19 PM)Falcon Wrote: Another thing regarding bit-perfect capture, for me, the easiest way is to capture in 24 bit with Audacity (Wasapi mode only).
Indeed, if you capture in 16 bit on Audacity, it adds a dither to the capture (not bit-perfect). By capturing in 24 bit, it does not add dither, and keeps the 16 bits intact ( + 8 bit of superfluous zero bytes ). You can then export in 16, although for beginners, I recommend letting EAC3to delete the superfluous bits (you never know if you left the dither active when exporting from Audacity)
That makes sense, good to know. Even in Reaper you are recording 16-bit audio into a padded 24-bit file. That's why I added support to BitPerfect to detect if its padded 16-bit data. I'm lazy and didn't want to export from Reaper for a BitPerfect test when the recorded file is already on your drive 
My check works the same as eac3to I believe. I just scan the 3 bytes for each sample and if the 3rd is always 0 I consider it padded.
Posts: 2,051
Threads: 56
Joined: 2016 Dec
Thanks: 162
Given 1011 thank(s) in 614 post(s)
Irregardless, if you can capture 16bit bit-perfect then surely that's the easiest overall, with fewer processing steps and a smaller file size
Posts: 109
Threads: 17
Joined: 2019 Jan
Thanks: 75
Given 50 thank(s) in 21 post(s)
Country:
(2023-08-31, 07:36 PM)zoidberg Wrote: Irregardless, if you can capture 16bit bit-perfect then surely that's the easiest overall, with fewer processing steps and a smaller file size
I would have agreed with you, but Audacity being free and accessible, easily adjustable to make bit-perfect (even if you have to delete the superfluous bits afterwards) more than Reaper, this is also to be taken into consideration.
|
