FLAC signal levels

[deenice]

Hi,

I've been encoding a few CDs in FLAC...I want to make sure my settings are as optimal as possible before I go on an encoding 'binge'! I'm using EAC/FLAC/AccurateRip and Foobar to add Replay Gain...I'm playing these files through Slimserver/Squeezebox. The rips I'm doing are 'approved' by the software.

One thing in common in the five musically varied CDs I've encoded so far is that the FLAC files (when listened with my ears and eyes through the VU meter on the Squeezebox interface) have low levels compared to the equivalent MP3 files I've encoded myself with Lame...as well, the sound is not as dynamic as when I A/B with the CDs (both SB & CD going through the digital outs in the same amp).

Is there something potentially wrong with my EAC/FLAC settings to explain such results?

Thank you.

[SuperQ]

Hi,

I've been encoding a few CDs in FLAC...I want to make sure my settings are as optimal as possible before I go on an encoding 'binge'! I'm using EAC/FLAC/AccurateRip and Foobar to add Replay Gain...I'm playing these files through Slimserver/Squeezebox. The rips I'm doing are 'approved' by the software.

One thing in common in the five musically varied CDs I've encoded so far is that the FLAC files (when listened with my ears and eyes through the VU meter on the Squeezebox interface) have low levels compared to the equivalent MP3 files I've encoded myself with Lame...as well, the sound is not as dynamic as when I A/B with the CDs (both SB & CD going through the digital outs in the same amp).

Is there something potentially wrong with my EAC/FLAC settings to explain such results?

Thank you.

I thought there was a bug posted a while ago about VU meters and replay gain not making things look right. (no audio problem, just VU meter problem)

Turn off replay gain, and then do the A/B comparison.. "not dynamic" is probably just mis-matched gain.

[tyler_durden]

FLAC can add the tags and replay gain data if you tell it to, so you can simplify your process...

The VU meters on the SB3 are not accurate. They are eye candy only.

TD

[MeridianMan]

The VU meters on the SB3 are not accurate. They are eye candy only.
TD

Slightly OT, but does this remain the case on the TP?

[jeffluckett]

Slightly OT, but does this remain the case on the TP?

Yes ... they both use the same code to generate the display.

[jeffmeh]

FLAC can add the tags and replay gain data if you tell it to, so you can simplify your process...

TD

Can FLAC add both "track gain" and "album gain" in this simplified process? Since all the tracks must be available to calculate "album gain," I assume not, but I may be missing something....

[tyler_durden]

Yes, flac will add both track and album gain. It converts all the files of a CD first then adds the replaygain tags all at once, adding about 10 seconds to the process.

If you set up your library structure ahead of time, flac will also populate the title, artist, etc. tags for you automatically. You have to play with the settings to get it working but you only have to do that once, then you can rip, store, and compress with everything automated except swapping discs and correcting errors in the FreeDB data when EAC reads it.

I am told MusicBrainz has a better data base with fewer errors. Maybe EAC can be pointed at it instead of FreeDB.

TD

[Robin Bowes]

tyler_durden wrote:
> Yes, flac will add both track and album gain. It converts all the files
> of a CD first then adds the replaygain tags all at once, adding about 10
> seconds to the process.

flac only adds track and album gain if you convert several files at
once, i.e. "flac *.wav" or similar. When you use it with EAC it is only
converting one file at a time so only adds track gain.

I use a script to add both track and album replaygain tags to my flac
files post-ripping:

http://robinbowes.com/projects/apply_replaygain

R.

[tyler_durden]

Which begs the question, why convert 1 track at a time in EAC when you can simply extract the entire CD in EAC them start flac and compress all the tracks?

I actually ripped 20-50 CDs at a time, placing them in the appropriate directories, then started flac and compressed/tagged/replaygained all of them. Very simple.

TD

[slimkid]

Which begs the question, why convert 1 track at a time in EAC ...
TD

And also begs an audiophile type question: Unless it's done for DJ purposes, why mess with track/album gain at all. It kills dynamics. If I was bitter, I'd add it's also OK for rap/techno/hip-hop :)

[jeffmeh]

And also begs an audiophile type question: Unless it's done for DJ purposes, why mess with track/album gain at all. It kills dynamics. If I was bitter, I'd add it's also OK for rap/techno/hip-hop :)

Can you provide an explanation as to why the attenuation done by replaygain is going to kill dynamics at one's preferred volume level any more than turning down the volume to reach that level? I would like to understand the factual basis on this one. Thanks.

[jeffmeh]

Which begs the question, why convert 1 track at a time in EAC when you can simply extract the entire CD in EAC them start flac and compress all the tracks?

I actually ripped 20-50 CDs at a time, placing them in the appropriate directories, then started flac and compressed/tagged/replaygained all of them. Very simple.

TD

It does sound like a more efficient approach. How is the tag data available if you use EAC to rip to WAV? Are you limited to whatever intelligence is embedded in the file/directory names, or is there something else going on here that I do not understand? Also, will FLAC do the album gain for files in the same directory?

After playing around with some ripping schemes (starting with EAC > MAREO > (FLAC and LAME), then Foobar for Replaygain), it's time to get more efficient. Thanks.

[Robin Bowes]

tyler_durden wrote:
> Which begs the question, why convert 1 track at a time in EAC when you
> can simply extract the entire CD in EAC them start flac and compress
> all the tracks?

I'm not sure if you're suggesting ripping a disc to a single file, or
simple ripping tracks to wav files then encoding afterwards.

With that in mind, here are the reasons I rip direct to flac with EAC:

1. (single file) Not everyone is happy with/wants to rip CDs to single
flac files - I personally prefer to have one track per file.

2. On a reasonably modern processor, the flac encoding takes less time
than the ripping so EAC effectively encodes and rips at the same time
(apart from the last track, of course) so it takes less time than
ripping followed by a separate encoding step.

> I actually ripped 20-50 CDs at a time, placing them in the appropriate
> directories, then started flac and compressed/tagged/replaygained all
> of them. Very simple.

If you rip all tracks to wav first then encode separately, where do the
tags come from?

R.

[slimkid]

Can you provide an explanation as to why the attenuation done by replaygain is going to kill dynamics at one's preferred volume level any more than turning down the volume to reach that level? I would like to understand the factual basis on this one. Thanks.

Oh, gladly.

Take example of Beethoven's 5th.

First movement with its infamous 'pa pa pa paaa, pa pa pa paaaa' goes rather high in Dbs or shall we say, it's rather loud. Then right after that there's a second movement with very silent, quiet string parts. Together, they form a huge dynamic range. Now, Mr. Karajan and his crew have been chosen for their talent, and have worked for years to perfect how they do that, so I'd say they are quite an authority in that field.

What various replaygains do is dynamycally equalize those tracks to what is considered (by the programmer) reasonable dynamic range. So it would silence pieces of the the first mevement and enhance some pieces on the second one, just like you would do taking the wolume down when it feels too loud and then back up again when you can't hear it.

What you are effectively doing is saying: ' S...w you, Mr. Karajan, I know better how that Beethoven is supposed to sound. In my system, to my ears this sounds better' :)

So, while from time to time volume really needs to be taken down or up a notch while listening (WAF reasons, mostly :) ), that is better done using the volume knob, as, once the track is processed, the information is lost and there's no going back.

Well, this was a theory, however, I was hit hard in practice with it too. Not so long ago, I was in the market for a speakers, so what I did was to prepare a compilation of maybe 50 differen fragments of various types of music I listen and burn it to the CD. Didn't notice that 'normalize' box was checked on my burning software. As a result, since 50 pieces of recording can have a huge dynamic range as they come from different albums, some of them were compressed so much, that the next day they would just create loud buzzing from tweeters.

K

[ceejay]

What various replaygains do is dynamycally equalize those tracks to what is considered (by the programmer) reasonable dynamic range. So it would silence pieces of the the first mevement and enhance some pieces on the second one, just like you would do taking the wolume down when it feels too loud and then back up again when you can't hear it.



I think you are being a bit harsh on "all" replaygain techniques here. First, adding a replaygain tag to a file doesn't lose any information at all, you can choose whether to apply it or not at play time. Second, they don't twiddle the volume at all within a track (which would be most annoying) or (I'm not sure whether you think this or not) apply any compression. And if you apply "albumgain" then the same volume adjustment will be made equally to all tracks/movements, so there is no question of overriding Mr Karajan or his colleagues.

I'm not saying, BTW, that all gain adjustment techniques are without any of these problems - merely that it is quite possible to get volume normalisation without messing with the music...

YMMV

Ceejay

[slimkid]

I think you are being a bit harsh on "all" replaygain techniques here. First, adding a replaygain tag to a file doesn't lose any information at all, you can choose whether to apply it or not at play time. Second, they don't twiddle the volume at all within a track (which would be most annoying) or (I'm not sure whether you think this or not) apply any compression. And if you apply "albumgain" then the same volume adjustment will be made equally to all tracks/movements, so there is no question of overriding Mr Karajan or his colleagues.

I'm not saying, BTW, that all gain adjustment techniques are without any of these problems - merely that it is quite possible to get volume normalisation without messing with the music...

YMMV

Ceejay

Ok, just to clarify a couple of things. I'm not trying to debate various technical solutions here. Surelly, some are better than others. And of course, when I said that information is lost and compression introduced, I meant if dynamic equalisation or normalisation is done during the format conversion/burning proces.

My main point is, in the context of an audiophile world, where (some) people would pay $x.000 for just cables that don't change the nature of their signal, where tone controls are long gone, where the best pre-amp is no preamp, doing something like autogain would be considered rather radical.

And, BTW, replaygain will change the volume of the track (even digitally, but that is being debated in different topic), so in my prior example, dynamic difference between first and second movement will be changed (decreased), which is messing with Karajan's intentions :)

[Phil Leigh]

There is a problem with ANY kind of digital normalisation, in that unless it is a pretty dumb "auto-gain control" it may introduce compression - otherwise a quiet track might be boosted so much that the loudest parts clip. So, if it doesn't boost or cut the track by a simple fixed amount of dB, it is compressing and that is a BAD thing since it is sucking the original dynamics out.

...and if it is a dumb auto-gain shifter...how does it avoid clipping?

I think replaygain does have a use for casual listening, parties etc...but I'd never use it when I really want to hear what the artist intended.

[snarlydwarf]

And, BTW, replaygain will change the volume of the track (even digitally, but that is being debated in different topic), so in my prior example, dynamic difference between first and second movement will be changed (decreased), which is messing with Karajan's intentions :)

Thats why there is trackgain and albumgain (and "smart gain" to recognize when to use each).

Take an album we all know: Pink Floyd's Dark Side of the Moon. You wouldnt want the clocks in Time to be turned down as "too loud" and Us and Them turned up to be just as loud... that would, well, suck. But you would want to have it about as loud as other albums so you didnt have to mess with volume control all the time. So AlbumGain (or SmartGain) would decide if the peak of the album was too high or low and adjust all tracks equally.

TrackGain is for when you are playing a playlist and some songs were mastered by deaf people who believe that everything should be set to 11, even quiet songs... It makes all the tracks roughly the same range. Yes, that can be evil in some cases, but it is okay for "i want a party mix and am getting annoyed at futzing with the volume all the time"

To do gain right, you really need track and album gain and ideally SmartGain so you dont have to keep futzing with which is right (though it can guess wrong..)

In theory it should work.

That said: I am lazy and use the volume control, so as usual, I really don't care what you use on your system. :P

[snarlydwarf]

...and if it is a dumb auto-gain shifter...how does it avoid clipping?

I think replaygain does have a use for casual listening, parties etc...but I'd never use it when I really want to hear what the artist intended.

I'm not sure I get that first part: it decodes the whole track, finds the peaks and adjusts the gain so that the peaks are just hitting the tippy-top of the range. It isnt hard to do that without clipping in software.

Think of a series of numbers (which is all digital music is): 39, 29, 23, 47, 48, 32... if the range is 0-100 without clipping, then multiply all the numbers by a constant so that the largest is 100. Ie: 100/max. There will be rounding errors, for sure, but no clipping.

And, again, I dont use replay gain. Maybe to keep the purity, but really just so I have a use for the remote other than 'skip'.

[Phil Leigh]

OK - if that is how replaygain works it's just messing with the gain, not the dynamics...that's OK I suppose...

[snarlydwarf]

OK - if that is how replaygain works it's just messing with the gain, not the dynamics...that's OK I suppose...

Yeah, it is done on a track level. I am not sure what the iTunes gain does: it shoves a whole bunch of numbers in comments, so maybe it is doing something really evil that adjusts parts of a track. I would hope not.

(Especially since, harping back to Pink Floyd... so much of the later Pink Floyd output was loud-soft-loud-soft even on one track, that would be downright evil...)

I will stick with my remote tho.

[jeffmeh]

I think SnarlyDwarf has this one right. Replaygain (specifically Replaygain; different algorithms certainly approach it differently) does not change the data, only inserts tags with the gain adjustments to hit the reference level for each track (track gain), and for an entire album (album gain).

I'm still curious as to whether there is any real difference between this and making the adjustment myself with the volume knob. It seems to me that if implemented correctly, there should be no difference, but I remain open to enlightenment.

[snarlydwarf]

I'm still curious as to whether there is any real difference between this and making the adjustment myself with the volume knob. It seems to me that if implemented correctly, there should be no difference, but I remain open to enlightenment.

It should be exactly the same, at least for negative adjustments. Due to the trend-to-clip in most cds these days, virtually everything is a negative adjustment.

[tyler_durden]

And also begs an audiophile type question: Unless it's done for DJ purposes, why mess with track/album gain at all. It kills dynamics. If I was bitter, I'd add it's also OK for rap/techno/hip-hop :)

Replay gain does not kill dynamics. All it does is set attenuation so that selections from multiple albums come out at about the same volume level. It is the same thing you would do using a volume control if you didn't have replaygain enabled. It is very useful if you make playlists or play random song selections from your collection and don't want to have to keep fiddling with the volume control with each new title.

There is no dynamic range compression involved, except for the loss of theoretical dynamic range because the the maximum volume has been lowered while the noise floor remains unchanged. There are few, if any CDs that actually make use of the full dynamic range the disc is capable of recording, so there is plenty to "throw away" for replay gain. IIRC, in my collection, replaygain typically lowers the track gain by 3-5 dB.

Replay gain is your friend.

I recall reading somewhere that when you use the smart gain setting in slimserver, it uses the album gain value if all the songs in the playlist come from the same album, and if not, it uses the track gain values. The album replaygain tag is set by finding the loudest track on the disc and setting its level to the "standard" volume value. That same attenuation/gain value is applied to each tracks album gain tag, so with album gain, all tracks are adjusted by the same amount (look at the tags for one CD- you'll see that all the album gain values are identical). That means if you play a disc with some loud and some quiet tracks, that is the way they will be heard, as the producer of the CD intended.

When setting the track replaygain tags each track is set to the "standard" volume level so that when you play random songs from different discs they will all come out at about the same level. Quiet tracks will be boosted and loud tracks will be attenuated. Replaygain tags are merely tags. They do nothing to the audio data in the file. Even if the replaygain tags are there, you can still decide whether slimserver should use them or not, so it is a good idea to populate the replaygain tags- it only takes a few minutes to scan the library and fill the tags.

TD

[tyler_durden]

If you rip all tracks to wav first then encode separately, where do the
tags come from?

R.

The tags are populated from the directory structure of the library. I rip each CD to individual .wav files named with the track number and song title only.

The library structure looks like this:
D:\MUSIC\genre\artist\year - disc title\track no. - song title

When the ripped CDs are placed into their appropriately named locations (this too, can be automated in EAC), you run flac and have it populate the tags based on the directory structure. You can rip 100 CDs, then, before you go to bed, start flac and run it on all the ripped files. It will tag them properly and add replay gain as it completes each CD. In the morning you'll have a working library with everything neatly tagged and placed in directories that make it easy for you to find your stuff. Simple and it works. I did it for my 600+ CD collection.

Of course, you may want to edit some of the tags, especially for classical or multiple artist type discs, but those are a problem with ANY tagging/storage system.

TD

[Robin Bowes]

tyler_durden wrote:
> Robin Bowes;165299 Wrote:
>>
>> If you rip all tracks to wav first then encode separately, where do
>> the
>> tags come from?
>>
>> R.
>
> The tags are populated from the directory structure of the library. I
> rip each CD to individual .wav files named with the track number and
> song title only.
>
> The library structure looks like this:
> D:\MUSIC\genre\artist\year - disc title\track no. - song title

[snip]

Gotcha.

Thanks,

R.

[Phil Leigh]

This might be "stupid question of the week" - but does replaygain work with the digital output or just the analogue?

[Robin Bowes]

Phil Leigh wrote:
> This might be "stupid question of the week" - but does replaygain work
> with the digital output or just the analogue?

Both.

If ReplayGain processing is turned on, the signal level is adjusted in
the digital domain according to the ReplayGain tags in the file then
processed as normal.

R.

[adamslim]

What could be interesting is seeing if ReplayGain can actually *improve* the sound. A (poorly-mastered?) CD might not get up to 100% of level. Applying (positive) ReplayGain in the digital domain might improve DAC SNR, even if there is no more information. Of course, you would need to be using an analogue volume control.

Anyone tried this? I would have thought it would make little change on a 24-bit DAC, but might do on 16-20 bit ones.

Adam

[tyler_durden]

What could be interesting is seeing if ReplayGain can actually *improve* the sound. A (poorly-mastered?) CD might not get up to 100% of level. Applying (positive) ReplayGain in the digital domain might improve DAC SNR, even if there is no more information. Of course, you would need to be using an analogue volume control.

Anyone tried this? I would have thought it would make little change on a 24-bit DAC, but might do on 16-20 bit ones.

Adam

You can't improve S/N by applying gain because the gain raises the noise floor along with the signal.

TD

[Phil Leigh]

OK I'm going to try Robins apply_replaygain thingy to see what happens.
Do I need change the "digital output is fixed" setting?
Thanks
Phil

[slimkid]

How would positive gain work in SS/SB3 combination where the attenuation is at 0 (no attenuation) and the volume is 100 (max)?

[pfarrell]

Phil Leigh wrote:
> There is a problem with ANY kind of digital normalisation, in that
> unless it is a pretty dumb "auto-gain control" it may introduce
> compression - otherwise a quiet track might be boosted so much that the
> loudest parts clip.


I agree that normalization is generally dumb, but it never introduces
compression.

Normalization means taking the existing signal and making it be as loud
as possible in a 16 bit PCM signal. in 16 bits, you can hold 0 to 64K,
but in PCM, it is +32K to -32K

So if you have a signal that has a max of only 4K, to normalize it, you
multiply all signals by 8. This raises the max signal to 32K, and raises
any noise signal by 8 as well. It does nothing to the signal to noise
ration.

Compression takes a loud signal and makes it smaller.
It does _not_ take a quiet signal and make it louder. But
most people use compression with make up gain. Thus when you take the
loud signals and make them smaller without changing the small signals,
and then make all signals as loud as the loudest used to be, you get the
impression that loud sounds stayed the same, and quiet parts are louder

>So, if it doesn't boost or cut the track by a simple
> fixed amount of dB, it is compressing and that is a BAD thing since it
> is sucking the original dynamics out.

No, see above

> ...and if it is a dumb auto-gain shifter...how does it avoid clipping?

By scanning the file, finding the highest value, and adjusting to suit.


> I think replaygain does have a use for casual listening, parties
> etc...but I'd never use it when I really want to hear what the artist
> intended.

Or more precisely, what the producer and recording engineer and
mastering engineer wanted you to think that the artist intended.


--
Pat Farrell PRC recording studio
http://www.pfarrell.com/PRC

[Phil Leigh]

Thanks Pat - that's clear. I realised it would have to read ahead to find the max level value and then boost the track by whatever the available headroom was (along with the noise floor!). I've used that kind of normalisation before in Sonar - on individual tracks in a mix. What I often found was there were high-level transients somewhere in the track that were a lot higher than the average level so actually the normalisation didn't have much headroom to work with...until I compressed (or zapped) the transients.

This shoudn't be the case on a full mix coming off a CD (almost certainly brick-wall limited during astering these days!).
Cheers
Phil

[pfarrell]

Phil Leigh wrote:
> This shoudn't be the case on a full mix coming off a CD (almost
> certainly brick-wall limited during astering these days!).

Depends on the genre. All pop/rock/etc CDs are cursed with the loudness
wars, which all professional mastering engineers hate. They do it
only because the client insists (artist, producer, record label, etc).

Jazz, chamber, classical, etc. typically have some dynamic life left,
probably because they don't get much radio play.

The really sad thing about loudness wars (other than that it doesn't
work and destroys music) is that the claimed aim is to get it
sound loud on radio. All, and I mean _all_ radio stations have
compressors between the studio and the transmitter. They compress
the hell out of music anyway. There is no reason to compress music for
retail sale.

Since the world listens to music on iPod, and has totally rejected SACD
and DVD-a, the market for quality sound is growing more limited by the day.

--
Pat
http://www.pfarrell.com/music/slimserver/slimsoftware.html

[karma mechanic]

Since the world listens to music on iPod, and has totally rejected SACD
and DVD-a, the market for quality sound is growing more limited by the day.


My naive solution to that is very simple. The iPod or other music player should be doing the compression. Then the users who prefer that can have as much compression as they want. The non-iPod users get high quality music.

I know, too simple.

[adamslim]

All, and I mean _all_ radio stations have
compressors between the studio and the transmitter. They compress
the hell out of music anyway. There is no reason to compress music for retail sale.http://www.pfarrell.com/music/slimserver/slimsoftware.html

Radio 3. Worth moving to the UK for :)

[tyler_durden]

The really sad thing about loudness wars (other than that it doesn't
work and destroys music) is that the claimed aim is to get it
sound loud on radio. All, and I mean _all_ radio stations have
compressors between the studio and the transmitter. They compress
the hell out of music anyway. There is no reason to compress music for retail sale.
Pat
http://www.pfarrell.com/music/slimserver/slimsoftware.html

Maybe the goal in compressing the CD is to make CD or mp3 playback sound the same as it does on the radio.

TD

[Phil Leigh]

Radio 3. Worth moving to the UK for :)

The Optimod is only off in the evenings.. :o)

[pi.r]

When the ripped CDs are placed into their appropriately named locations (this too, can be automated in EAC), you run flac and have it populate the tags based on the directory structure. You can rip 100 CDs, then, before you go to bed, start flac and run it on all the ripped files. It will tag them properly and add replay gain as it completes each CD. In the morning you'll have a working library with everything neatly tagged and placed in directories that make it easy for you to find your stuff. Simple and it works. I did it for my 600+ CD collection.

TD

This sounds exactly like what I would like to do with my wav library. - I just can´t figure out how to make it happen... Is it possible from the kommand line with flac, or with the frontend for windws, or do you need to use some kind of skript?

[tyler_durden]

I always used the windows front-end. There are just a few check boxes to set the options. You can set the same options from the command line, but it would be a PITA to specify a bunch of different files to compress, unless maybe a *.wav would work.

In the windows front-end, you click "Tag Conf." button. In the box that pops up, check "custom", then enter this string (or a suitable one to match your library structure)"G\A\Y - L\N - T". This works for my library set up this way: D:\MUSIC\genre\artist\year - disc title\track no. - track title. There is a similar string to set up EAC to store the .wav files in that structure. In falc, be sure to check the "Add Tags" and "Replaygain" boxes, too. flac will figure out which tracks come from the same album and apply both album gain and track gain to them.

The problem I live with right now is that the library disc was formatted NTFS. I can read NTFS in linux, which I have recently switched to, but I can't write to NTFS, so when I want to rip another disc to add to the library, I have to drop back to windows (I feel like I need to take a shower every time I have to use windows).

TD

[pi.r]

I don't know if I'm just being stupid? But it dont work for me. Encoding and tagging based on filename works fine. But I have 2 issues:

1. Adding multiple albums. With 250 albums to encode it would be nice to drag and drop the entire file tree to flac. I can only drag album folders one by one. It doesn't work with subfolders. The "Add" dialog can only add individual files, not folders. - Or is thre a way to do it?

2. Album gain don't work when I add multiple albums (in separate folders). If I check both checkboxes for Replaygain and "Treat input files as one album", flac returns one error for each file saying "ERROR: can't open input file <filename.wav>: no such file or directory". (there is no path in the filename). By reading "go.bat" I can see that flac assumes that all files are in the same folder (from the last added file). If I don't check "Treat input files as one album" I don't get album gain (album = track gain).

I just can't see any way to get it work like you describe. Do I have a different program version? Mine are - Flac: 1.1.3, Frontend: 1.7.1

[tyler_durden]

I apologize- it has been a while since I did it but your post reminds me of how I made it work.

Put your .wav files in all the appropriate directories -let EAC do it as you rip the files. When it is time to flac them, do a windows search for .wav files. Select all of them (notice that they have paths!) and drag them to the flac front-end window. Watch out for windows sound files (like beep, etc.)- you don't want to .flac them!

This will solve both problems. Now when you tell it to encode it WILL work. If you tell flac they are multiple albums, it will figure out which tracks come from the same album and apply album gain properly.

TD

[pi.r]

Indeed a nice trick to add all the files in one go. But still no cigar... - I just get error from flac because lack of file path.


...If you tell flac they are multiple albums...

How/where do I do this?

I have checkboxes for "Verify", "Add tags", "Replay gain" and "Treat input files as one album" checked, no others. The last one contradicts the fact that it is multiple albums. But if I uncheck it, I don't get error but all files gets encoded individualy with album gain = track gain. Tagging is no probblem but just fore refereence: My file structure is "genre\artist\album\track no.-title" and my tag scheme is set to "G\A\L\N-T".

I begn to wonder if we are using the same Flac Frontend? I found and tested an older version: 1.7 ( from jan. 2003) but the result was the same.

[tyler_durden]

Indeed a nice trick to add all the files in one go. But still no cigar... - I just get error from flac because lack of file path.


I just tried it and it works fine for me. I do a windows search for *.wav. Select the files I want to flac from the list (all of them in this case), then drag them to the flac frontend file list box.

I am now using the latest version of flac but my library was set up using the previous version. Maybe there is some issue with your flac installation. Also check the flac "tag conf." string - make sure it uses "\". not "/"

I am also seeing that the album and track replaygain tags are being set the same in each track. The "treat files as one album" box is unchecked. Hmmmmm. I would swear that it worked that way, but maybe I am remembering a single CD operation. I'm having a lot of "senior moments" these days. I must have used foobar to set replaygain for batch operations.



OK, here's the whole process (and I just verified that it works):

1) RIP in EAC with EAC set to store .wav files in desired library structure. For example, D:\MUSIC\genre\artist\year - ablum title\track no. - track title
This can be set up by using the EAC Options menu (F9). Select the "directories" tab. Click "use this directory" and enter "D:\MUSIC". Now select the "Filename" tab. Fill the "naming scheme" box with this string: "%B\%A\%Y - %C\%N - %T". Now you are ready to rip.

This is the painful part: each time you insert a fresh disc, correct errors in the FreeDB/CDDB data before you start the actual rip. Go ahead and rip a big pile of CDs.

2) Compress and tag: do a windows search for *.wav on the HDD where you have stored all the .wav files. Start the windows flac front-end and drag the .wav files found in the search to the file box in flac. Click the "verify", "add tags", and "delete input files" buttons.

If you've ripped a single CD you can have flac add replaygain tags by clicking "replaygain" and "Treat input files as one album". If you are ripping a bunch of CDs, DON'T add replay gain tags. Use foobar for that after all the files have been flac'd.

Tell flac to populate the tags from the file structure: click "Tag Conf." button, then select "custom" and fill it with the appropriate string- from the example above, use "G\A\Y - L\N - T".

3) Add replaygain tags: do a windows search for *.flac files that you just made. Start foobar and drag the .flac files from the search onto the foobar window. Select all (ctrl-A), right click, select "replaygain", then "scan selection as albums (by tags)".

Now point Slimserver at D:\MUSIC and it will find everything and set up its database so you can browse/search by artist, album, song, year, genre, etc.

That will do it. Each operation- ripping, flacing, and replaygain tagging can be done as a batch operation on as many CDs as you care to rip. Ripping takes the longest and is the biggest PITA because you have to sit there and change discs every few minutes and correct all the freedb errors (if you care to). I had two computers with 2 CDROM drives in each doing the ripping and it took me about a week to rip 600 CDs.

Multiple artist and classical discs are a PITA, and you'll have to work out your own preferred scheme for handling those.

If you DL music you'll often have to add or correct the tags. A freeware program called EasyTag (http://easytag.sourceforge.net/) works for that. The user interface takes some getting used to, but once you do it is easy to use. Another program called "FileRenamer Basic" (http://www.sherrodcomputers.com/products_filerenamer.cfm) can be very useful. Of course, I never DL music... ;)

TD

[pi.r]

OK. That finaly sets it. I thought it was just me that coulden't get it... I vas just about to give up and use Foobar for replay gain anyway. I guess one could use foobar for encoding and tagging as well. It still needs to be done in several steps though.

The errors i talked about is only because the frontend ignores the individual file path when "Treat input files as one album" is selected. It simply assumes all files to be in the same directory. When not selected, it works just like you says.

Thanks A LOT for your time...

Ripping is already done, and I thought I would be satisfied with wav files. But then I heard about replay gain and realized the benefits of tagging my classical albums. So now my pc will have some work to do...