IPB

Welcome Guest ( Log In | Register )

Audio from vinyl (24 / 96-192 khz). Converting/downsampling?
Taishou
post Aug 27 2012, 03:01
Post #1





Group: Members
Posts: 12
Joined: 27-August 12
Member No.: 102679



Hi again

I have several digitized vinyls in FLAC, at 24 bit, 96-192 khz. Is it worth all that sample rate? I was thinking of making CD quality (16/44.1) FLACs or even mp3 320. I doubt I'll ever in my life spend more than, idk, 200 usd in headphones, so the over-quality looks a waste. What do you say? I'm far from an audiophile and I feel pretty happy with mp3 320 (I can't even tell from FLAC).

Anyway, I need to re-encode that at least for my mobile devices. Should I have any special care when converting? I use fre:ac (LAME 3.99.5). Should I just set 44.1 khz when encoding there and done? Or should I downsample with another tool and then encode? I just tried encoding something setting 44.1 in LAME and the result sounds the same as the original 24/192 (with my crappy headphones at least).

The same goes for FLAC or ogg (I tend to use the latter for my DAP). Just setting the sample rate when encoding is enough or is it better to do it with another program? (I doubt, but I wouldn't like regretting later because I made a crappy conversion).

Thanks!


(maybe this should go in the "vinyl" forum, but my concern is more with those high sample rates and downsampling, regardless the source)
Go to the top of the page
+Quote Post
 
Start new topic
Replies
jensend
post Aug 27 2012, 16:45
Post #2





Group: Members
Posts: 143
Joined: 21-May 05
Member No.: 22191



Try the FAQ for good info and advice. Also this page, which is one of the first google results for "LP noise floor": Transferring LPs to CDR.

Any frequency content in audio above 20kHz is completely inaudible to human listeners. On your LPs, though you're picking up content up to 40kHz, practically all of the content beyond 16kHz is going to be noise, not signal. (Anything above 20kHz is definitely going to be noise). Keeping 96kHz versions is a waste.

The LP noise floor is rather high- maybe -70dB under very good conditions. 12-bit sampling (RMS noise floor of -72dB) would be sufficient for LP use as long as your levels are right (peak signal above -6dB). (12-bit sampling was used for DV but hasn't seen any other widespread use).

You would most likely find that 32kHz 12-bit would sound just as good in regular listening as anything else.

Any extra detail you capture beyond that is basically going to be useful for editing, not listening. (People try a lot of different things to help clean up LP transfers, and leaving yourself a little room in case you want to try that in the future could be nice.) Extra bit depth is usually more helpful for editing than a higher sampling rate. 16-bit is plenty of extra room. There may be a little bit of real signal on your LPs in the 16-20 kHz range, even if it's not really audible. Downsampling to 48 or even 44.1 kHz retains all that and more, and gives you room for a reasonable transition band on top of that.

I'd really recommend you go with 48kHz 16-bit FLAC.

BTW, even if you had high-fidelity digital originals rather than an LP, here's a paper on why- at least for playback purposes- >16 bit sampling is just a waste of storage while extra-high sampling rates are actually *harmful*.

This post has been edited by jensend: Aug 27 2012, 17:17
Go to the top of the page
+Quote Post
benski
post Aug 28 2012, 18:37
Post #3


Winamp Developer


Group: Developer
Posts: 670
Joined: 17-July 05
From: Brooklyn, NY
Member No.: 23375



QUOTE (jensend @ Aug 27 2012, 11:45) *
The LP noise floor is rather high- maybe -70dB under very good conditions. 12-bit sampling (RMS noise floor of -72dB) would be sufficient for LP use as long as your levels are right (peak signal above -6dB). (12-bit sampling was used for DV but hasn't seen any other widespread use).


The SNR of an MP3 is around 25-30dB. Does this mean that I only need to decode them to 6-bit PCM to capture all the details?

I am curious if anyone has ever done a detailed analysis of an LP's SNR in different frequency bands.

This post has been edited by benski: Aug 28 2012, 19:07
Go to the top of the page
+Quote Post
Woodinville
post Aug 29 2012, 03:27
Post #4





Group: Members
Posts: 1402
Joined: 9-January 05
From: JJ's office.
Member No.: 18957



QUOTE (benski @ Aug 28 2012, 10:37) *
QUOTE (jensend @ Aug 27 2012, 11:45) *
The LP noise floor is rather high- maybe -70dB under very good conditions. 12-bit sampling (RMS noise floor of -72dB) would be sufficient for LP use as long as your levels are right (peak signal above -6dB). (12-bit sampling was used for DV but hasn't seen any other widespread use).


The SNR of an MP3 is around 25-30dB. Does this mean that I only need to decode them to 6-bit PCM to capture all the details?

I am curious if anyone has ever done a detailed analysis of an LP's SNR in different frequency bands.


This just goes to make a point I've made in talks every since about 1990: "SNR is Mostly Harmless"


--------------------
-----
J. D. (jj) Johnston
Go to the top of the page
+Quote Post
krabapple
post Aug 31 2012, 14:08
Post #5





Group: Members
Posts: 2182
Joined: 18-December 03
Member No.: 10538



QUOTE (Woodinville @ Aug 28 2012, 22:27) *
This just goes to make a point I've made in talks every since about 1990: "SNR is Mostly Harmless"



But when it gets too low in online threads, I tend to bail.
Go to the top of the page
+Quote Post

Posts in this topic
- Taishou   Audio from vinyl (24 / 96-192 khz). Converting/downsampling?   Aug 27 2012, 03:01
- - Porcus   QUOTE (Taishou @ Aug 27 2012, 04:01) I ha...   Aug 27 2012, 06:16
- - Taishou   I see... Thinking about it I'll stay with FLA...   Aug 27 2012, 06:57
- - Taishou   Sorry for the double post, I can't see an ...   Aug 27 2012, 08:25
|- - Porcus   Go 44.1 or 48, as those are the well-supported sam...   Aug 27 2012, 10:14
||- - db1989   QUOTE (Porcus @ Aug 27 2012, 10:14) [Oh ....   Aug 27 2012, 14:38
|- - splice   QUOTE (Taishou @ Aug 26 2012, 23:25) ... ...   Aug 27 2012, 21:37
- - jensend   Try the FAQ for good info and advice. Also this pa...   Aug 27 2012, 16:45
|- - benski   QUOTE (jensend @ Aug 27 2012, 11:45) The ...   Aug 28 2012, 18:37
|- - Woodinville   QUOTE (benski @ Aug 28 2012, 10:37) QUOTE...   Aug 29 2012, 03:27
|- - krabapple   QUOTE (Woodinville @ Aug 28 2012, 22:27) ...   Aug 31 2012, 14:08
- - jensend   Oh, BTW: resampling in multiple stages is unlikely...   Aug 27 2012, 17:06
- - Taishou   Great answers, thanks. I just read the links you p...   Aug 27 2012, 20:56
- - jensend   If you're already using SoX, no reason to not ...   Aug 28 2012, 00:09
- - Taishou   QUOTE (splice @ Aug 27 2012, 17:37) QUOTE...   Aug 28 2012, 05:40
|- - Porcus   QUOTE (Taishou @ Aug 28 2012, 06:40) Afte...   Aug 28 2012, 07:42
|- - jensend   QUOTE (Taishou @ Aug 27 2012, 22:40) Afte...   Aug 28 2012, 17:34
- - db1989   Itís a result of one of audioís many compound myth...   Aug 28 2012, 07:09
- - Woodinville   There is some use to that over-20kHz stuff, though...   Aug 28 2012, 08:16
- - Arnold B. Krueger   QUOTE (Taishou @ Aug 26 2012, 22:01) I ha...   Aug 28 2012, 12:40
- - greynol   We had the same inconvenient point raised about th...   Aug 28 2012, 20:32
- - Taishou   Thanks for all your input, it's really great ...   Aug 29 2012, 02:48
|- - Porcus   QUOTE (Taishou @ Aug 29 2012, 03:48) Shou...   Aug 29 2012, 15:05
|- - jensend   QUOTE (Taishou @ Aug 28 2012, 23:47) If i...   Aug 29 2012, 15:13
- - Taishou   If it matters of something, the files with clippin...   Aug 29 2012, 06:47
- - Porcus   But ... if the peak measures to 0.999, and the res...   Aug 29 2012, 20:43
- - pdq   The resampling produced samples that were beyond +...   Aug 29 2012, 21:02
- - Taishou   Well, I decided to leave it as it is. I tried the...   Aug 29 2012, 23:30
- - greynol   Re. compression: I've seen worse. At the end ...   Aug 29 2012, 23:41
- - Taishou   Indeed. And it's the same on the source, so it...   Aug 29 2012, 23:51
|- - Porcus   QUOTE (Taishou @ Aug 30 2012, 00:51) it w...   Aug 30 2012, 07:35
- - greynol   Could it be due to possible inter-sample overs occ...   Aug 30 2012, 02:51
- - greynol   Off-topic discussion on the SNR or MP3 has been sp...   Aug 31 2012, 19:48


Reply to this topicStart new topic
1 User(s) are reading this topic (1 Guests and 0 Anonymous Users)
0 Members:

 



RSS Lo-Fi Version Time is now: 29th July 2014 - 07:59