Lynx, FLACs, and emphasis bit?

[MR]

Anybody understand how to use the emphasis bit? A search of the forum for "emphasis bit" yielded no posts. The Lynx mixer allows the bit to be turned on or off on each output channel.

 

As I understand it, emphasis represents processing applied during recording; frequencies from 1 kHz to 20 kHz are emphasized. When the bit is found in the channel status data, the DAC applies de-emphasis to bring the frequency response back to flat.

 

With the bit turned off, most of the FLAC files I am playing sound zipped-up (upper frequencies emphasized), a problem I have been fighting with clocks, cables, and some purely mad ideas. When I check "Emphasis", the DAC's flattening darkens the sound appreciably.

 

What I don't understand is whether/how this information survives (or doesn't) the process of ripping FLAC files.

 

At first blush, what I am hearing sounds better with the option checked.

 

[vortecjr]

try converting the file to wav or aiff and repeat your tests. I'm curious about flac with emphasis to wav without emphasis, but try other combinations as well. Keeping notes would be best because it could get messy:) Then report your findings!

 

Jesus R

www.sonore.us

 

[MR]

The presence or absence of pre-emphasis doesn't appear to survive the rip; for those who might be interested, EAC has a writeup:

 

http://blowfish.be/eac/Rip/rip12-preemph.html

 

This is apparently an issue for some of the CDs I ripped; however, it appears to be mostly an issue on older CDs and with certain classical labels.

 

Applying pre-emphasis during the rip would result in modified copy, so I am not immediately drawn to that idea; the procedure outlined by EAC first creates a bit-accurate wav and then produces a modified file without changing the original rip.

 

I haven't found a way to automatically set the AES16/AES16e 'Emphasis' flag or simply pass the channel status data along in the rip.

 

[Synfreak]

Hi MR!

 

AFAIK, there is (at least as of today ...) no way to (re-)play files with an "automated" de-emphasis flag.

 

The flag is stored in the TOC of the CD, and if you copy the disc, this setting gets lost.

The only way to reproduce the files correctly - de-emphasized - is to convert the files with some (third party) software.

 

If you rip your discs with ITunes, the software (usually) keeps an eye on this and performs the filtering.

 

If you are ripping your discs with EAC (or dbPoweramp ...), you have to do the conversion manually.

 

Yes, this will "alter" your "bit perfect" files. But the DAC does the same before the data gets "D-Aed" (and there are a lot of disc-players and DACs out there which aren't performing this at all!).

 

This link might be of interest:

http://wiki.hydrogenaudio.org/index.php?title=Pre-emphasis

 

I use "WaveEmph" (Windows) software for the task, and it seem to do a great job, so I don't care anymore

(It is way more tricky to catch all Pre-emphasized discs when ripping ...)

 

Cheers

Harald

 

[vortecjr]

Thanks for the link. I thought this was some kind of base boost:) How does WaveEmph know if a cd is pre-emphasized if the data is lost at ripping?

 

Jesus R

www.sonore.us

 

[Synfreak]

Hi Jesus!

 

WaveEmph doesn't "recognize it" - it is a software to calculate (perform the filtering) on files on which you know that they are emphasized.

 

Here a link to the homepage (in english):

http://www.picosound.de/E_index.htm

 

Worth a try, IMO ...

 

Cheers from Germany

Harald

 

[ziggyzack]

'Applying pre-emphasis during the rip would result in modified copy, so I am not immediately drawn to that idea; the procedure outlined by EAC first creates a bit-accurate wav and then produces a modified file without changing the original rip.'

 

Pre-emph is designed to be removed before playback. The unfiltered rip will always need to be filtered, and you'd only want to play back the filtered rip, so create a filtered version and discard the original rip, which would never be used again once you have created a filtered version. The only reason you'd hang on to the original is if you weren't confident in the de-emph process.

 

From the Sox documentation:

 

'Pre-emphasis was applied in the mastering of some CDs issued in the early 1980s. These included many classical music albums, as well as now sought-after issues of albums by The Beatles, Pink Floyd and others. Pre-emphasis should be removed at playback time by a de-emphasis filter in the playback device. However, not all modern CD players have this filter, and very few PC CD drives have it; playing pre-emphasised audio without the correct de-emphasis filter results in audio that sounds harsh and is far from what its creators intended.... [The Sox de-emph filter's] maximum deviation from the ideal response is only 0.06dB (up to 20kHz)'

 

Ripping off-line, then playing at leisure, is different from synchronously playing from a CD. Any manipulations that a CD player would correctly perform before the stream hits the dac should usually be applied and committed to your master rip.

 

[MR]

Certainly, one would most often play a copy to which pre-emphasis has been applied. I suggest hanging on to the original bits in keeping with Chris' principles of digital preservation for CD audio data. I think he makes an excellent case for hanging on to the CRC-correct bitstring as a read-only master for any subsequent processing or modifications, including playing them through a DAC that has pre-emphasis processing of its own. The bit can be turned on manually in the Lynx mixer, arguably too much of a hassle for everyday use, but possibly useful as a point of comparison.

 

[ziggyzack]

Well there will be so few of these, I guess it's not too much hassle to keep the original rips safe somewhere, but I still don't think it's right in principle. The only reason I would do that is if I weren't perfectly confident about the de-emph I'd applied. There's no reason to think that the de-emph built in to a DAC would be any better than in software - on the contrary, the DAC is restricted to doing it in realtime. De-emph is not a hard thing to implement, it has no subjective element, and SOX is apparently up to the job.

 

I don't agree with Chris' approach at all. Suppose you lost your working copy - you'd have to go back to the original rips, then re-apply de-emph (in a few cases), and then re-tag everything. And what about music you've downloaded, which you've also tagged the way you want? Embedded album art? For this reason, CDs are the worst, last resort backup you have - a tremendous hassle to re-rip and re-organise a large collection. Some CDs will have taken a lot of work to rip accurately. Keeping the untagged rips is pointless when you can keep the tagged rips. And there's no point keeping a separate archive copy from a working copy, when lossless codecs are in play. Make backups of your live library, at regular intervals or after significant library changes (e.g. you may go and fill in all the composer tags), and keep them off-site. Two backups are better than one. rsync can automate this.

 

Of course, there are itunes/flac issues, and where this happens, you may need to make workarounds - but I'm convinced that the live, tagged, lossless library should normally be considered the master, and that backups should target that.

 

[The Computer Audiophile]

Hi ziggyzack - I think you are misunderstanding my approach. If you read my article on ripping etc... you see what I mean. There is no need to re-tag, re-rip, re-download, re-embedd art, or even touch a physical disc.

 

I'm not sure where in the world you got the idea that I would even recommend an untagged rip. My article is about using FLAC because it its superior tagging capability.

 

When you lose the working copy the archival copy is as good as it gets to recreate anything you need. It already has all the tags and has been ripped securely. It allows you to convert into any format you want with all bit and sample rates supported.

 

 

 

[ziggyzack]

'I'm not sure where in the world you got the idea that I would even recommend an untagged rip.'

 

From this bit: '(2) A ripped bit for bit "Archival Copy” of the CD only used to produce other copies, not used for playback.'

 

And from MR, who had said you made 'an excellent case for hanging on to the CRC-correct bitstring as a read-only master for any subsequent processing or modifications'.

 

I've read on further into the ripping post now, and apologies, you do indeed talk about tagging the archival copy, but I don't think the emphasis is right.

 

The idea of creating a static archive copy is contingent upon getting the tagging complete, and satisfying all future needs, at rip time. This simply never happens, and furthermore it isn't possible. As collections grow, as people discover new software, there is motivation to add and correct tags which isn't necessarily there at rip time. Policies evolve over time, like how to treat multi-disc albums, classical works and so on, which are down to personal preference and may change. The need for these refinements sometimes only become apparent later. Clean CDs can rip accurately in 3 minutes, while the time invested in tagging can be much longer - that investment needs to be protected with a backup. Should you then lose your data, would you revert to the significantly out of date archive, or a recent backup? You'd use the recent backup. Any older copies are superfluous.

 

This slightly broadens the topic in MR's post, of course, but he says he's adhering to the principles you set out, and I don't think they've given him clarity on what's worth keeping and what's worth discarding.