I'm no expert, but I have concluded that EAC in some instances alters the original CD sound when it converts to a wav file. I'm no expert so I can't fully explain why, but I have ears and am sure this is happening. I noticed this when ripping Santana's Beyond Appearances. EAC transforms it by greatly exagerating the high end --- this is very real!!!!!! I have experimented with this CD and noticed that when iTunes converted this CD to a wav file it was much more accurate sounding as compared to the original CD (believe me I am no fan of iTunes which is why I use LAME). So I compared the two different wave forms using a wave editor (e.g. Nero Wave Editor or Cool Edit Pro) and what I found was that the iTunes wav file displayed more narrowly than the EAC wav file. I know EAC is supposed to be an exact copy blah blah blah but can someone explain what's going on here. It may only be an issue with the songs on this album (a sampling of other CDs yielded what appeared to be identical wav forms from EAC and iTunes) but the songs on this Santana album create noticably different wav forms and it has me wondering if this differential exists elsewhere. Anyone?

I HAVE RESOLUTION TO THIS PROBLEM!!....HERE IT IS.
Thanks to all those that posted. This problem only relates to older CD's which were recorded using a now obsolete method called pre-emphasis. If the "Pre-emphasis" column in EAC says YES when you have a CD inserted, then this situation exists. (I didn't even know that the "Pre-Emphasis" column existed in EAC as it was hidden on the far right and I had to shorten the title column in order to see it). I now know to check this column.

If a CD has pre-emphasis, I simply will use iTunes to convert it to a wav and then convert that wav with LAME. EAC was not performing the de-emphasis needed to bring the equalization back to normal but iTunes makes this adjustment.

Thanks again for everyone's help. I once again feel comfortable using the EAC/LAME as my ripper of choice.

Ouch.

If you would please, would you go into EAC's settings and check to see if it is set up to normalize the audio files as it rips them? Also, you might consider using Accuraterip in conjunction with EAC to give a high degree of confidence that the tracks you have ripped are perfect / flawed.

http://www.accuraterip.com

I do not have Normalize checked as an EAC option nor is it checked in iTunes. I should have mentioned that in the original post.

You might ant to submit samples and proof. "Sounds better" just does not work on this board and will get you heaps of scorn.

Screenshots?

I understand what your saying but this is not a subtle difference in sound and besides as I said I can see the difference in the wave forms. I will try to submit some screenshots later when I get more time.

Is it possible that you are listening to CDs using an analog connection for playback, and then once they are ripped you are listening to the WAV files directly as digital.

When you open EAC with the Santana CD inserted, what does the Pre-Emphasis column say?

The way I first noticed this was because my LAME encoded mp3 sounded like it had way too much high end. So I conducted a direct comparison between the LAME mp3 and the CD, both played through a very good analog system. Upon further investigation, I concluded that it was not the LAME encoder that caused the drastic difference in the sound. Instead, I noticed the differences in the wav files when ripping with EAC and iTunes. When I rip the iTunes wav file using LAME and play it through my analog system, it sounds identical to the CD played through that same system.

Again, once I noticed this difference I compared other CDs wav files as ripped by these two programs and they looked identical. So I am still baffled by the clear difference noticed in this one CD and wonder if it is somehow impacting other CDs in a less noticable fashion.

I'm almost certain the issue is pre-emphasis. I see the album was originally released in 1985. Is your CD an original? Many early CDs were pressed with pre-emphasis.

The question is whether the pre-emphasis tag should be there or not. If pre-emphasis is supposed to be "yes", then the treble-boost is the intended and "correct" sound. iTunes and probably most other hardware these days might not detect it, as it's an antiquated feature of the redbook specification.

However, it's also possible the pre-emphasis flag is not supposed to be there.

If preemphasis is there, shouldn't it be attenuating the high and and boosting the bass? So it almost sounds like iTunes and your CD player are doing the right thing, and EAC isn't.

I was afraid I had that backwards

Axon's right. I was thinking of the boost on the recording side. Either way, I still think it's the issue.

Thanks Cyaneyes. Yes this CD is an old one.
I had never heard of pre-emphasis till your post. I am not home right now, so I can't check the CD yet, but I just did a little research on that topic and agree that my problem appears to involve pre-emphasis. As you highlighted from my original post, it has two of the symptoms: 1)only noticed on this CD and, 2) greatly exagerates the high end.

I still am a newbie with this topic so I don't fully understand your present reply. All I know is that EAC and iTunes are presently ripping this CD differently. So, I'm assuming that one of these programs is not properly adjusting the wav file for this pre-emphasis setting. I am still unsure which program is handling it correctly. It is possible that the manufacturer intended that the CD be very heavy on the high end in which case EAC is converting it correctly and my CD player is not recognizing the flag and playing it without the high-end emphasis. On the other hand, the iTunes may be handling it correctly and my CD player is properly recognizing the flag. I don't know enough yet to be sure which it is. (Off hand my guess is that iTunes is handling it correctly because the EAC rip just sounds way to high...but this is a personal preference judgement).

My present crude understanding of this topic leads me to believe that iTunes is properly carrying out the deemphasis process when converting to a wav file. This deemphasis process is taking the place of what would be performed by the CD player when playing a CD with the flag. Any thoughts

If I were you, I'd get myself some flame-resistant clothing ...

If EAC performs an error-free rip of an audio CD (that isn't HDCD-coded although EAC keeps the HDCD info - but software players don't get along with it), the resulting wav file will be 1:1 with the original track on CD.

Pre-emphasis could be the reason (especially if it happens only on one CD) but I always thought that EAC would carry out de-emphasis properly ...

Screenshots and short samples of the different extractions would be useful for the purpose of investigating your claims.

I wouldn't be so sure about EAC (and most rippers) carrying out deemphasis.

Edit: deemphasis at the digital domain is a lossy process, in the sense that some of the low-level signal of the pre-emphasized ripped signal will be lost in the process of deemphasicing. Not that anything audible will be lost, if done properly. It could be lossless only in case of the deemphasicing process resulting in a final 24-bit deemphasized rip that kept all of the original 16-bit data.

Here's my latest research on this topic.
EAC may indeed be performing an error-free rip, but apparently in the case of these rare CDs manufactured with pre-emphasis, the resulting wave files contain the pre-emphasized wave forms along with a flag which is supposed to inform decoders that deemphasis is needed during playback. But, because this emphasis/de-emphasis process is obsolete, most decoders don't apply de-emphasis. Therefore, listening to the file without deemphasis will not be a true reproduction of the intended sound (which is why it is very heavy on the high end). So, it seems the best way to rip these CDs is to apply the deemphasis in the ripping process in order to ensure that the playback signal is returned to a linear (non-emphasized) frequency response.

So, it seems that iTunes is taking care of the de-emphasis process in the ripping to a wave file. I have not heard whether or not there is an EAC option which can also perform this function.

Go ahead and flame me just because EAC may technically be generating a perfect rip. The fact of the matter is that EAC gave me a file that sounded like crap. Thanks to everyone in these forums that help resolve problems like this because learning and having good music is what it's all about!

Wow.. very interesting to me that iTunes would know to apply the deemphasis curve and EAC wouldn't. I can't find any option in EAC relating to preemphasis other than the column displaying the yes or no flag. I just wish I had a disc with it so I could test myself.

Did a little googling and found a post on another forum by our esteemed Pio2001 that includes a very helpful link. http://club.cdfreaks.com/showpost.php?p=400909&postcount=23

If you have access to a good parametric EQ, you could try applying the deemphasis manually yourself.

I had completely forgotten that I posted this !

Anyway, the best way to copy such a CD is to use EAC's wave files, and set the emphasis flag to "yes" on the copy. This way, the copy will be faithful to the original.

I checked four programs
EAC ( http://www.exactaudiocopy.de ) and Feurio ( http://www.feurio.de ) allow to set the emphasis flag to "yes" on a copy.
Nero and Plextools Pro don't.

I don't think you understand what was going on here. I should have made it clearer in my updates/replies that my Santana CD was flagged as having pre-emphasis. Also, I conclude that EAC will make a copy that is faithful to the original (without me changing any settings because EAC did recognize the flag and I believe would carry that forward to a copy).

Again, the Santane CD had already had the Pre-emphasis flag set as "yes." As such, my CD has wav files that are "Boosted" and need to be de-emphasised for playback. I believe that EAC copy the wave files with the boost and will carry that pre-emphasis flag over to a copied CD. And even though this flag was only used on CDs manufactured a long time ago, it seems that most CD players will recognize the flag and perform a de-emphasis of the EQ on playback.
My problem related to the conversion of such a wav to an mp3. Even though the wav and mp3 may still carry the flag, most decoders won't recognize it and thus the playback is not de-emphasized, therefore it's high end is exagerated. For example, I think that computer programs such as win-amp will not recognize the flag, even on a wav file, and therefore if you play any pre-emphasized files on your computer it will exagerate the high end. Similarly, I came to notice this problem on my iPod.

To get around this playback problem when ripping these now rare CDs, it seems the best solution is to use software to convert the file to one that is no longer pre-emphasised. I have found that iTunes will do this.

So in response to your suggestion, EAC appears to properly recognize the pre-empasis flag, and will properly carry it forward to a copied CD. There is no need to change it unless the original CD was not correctly flagged...which is probably an even more rare case. I came across another post that described how to change an incorrect flag with EAC by adding "FLAGS PRE" to the cue sheet. See:

http://www.stevehoffman.tv/forums/archive/...-10981-p-3.html

IIRC, the fhg encoder allows for encoding of files with pre-emphasis. I would image that this would "de-emphasize" the mp3 files.

Hey Cyaneyes, thanks again for your help. It goes without saying that you were right that my Santana Beyond Appearances CD had the Pre-emphasis flag. Your graph clearly shows that if de-emphasis is not applied, the high frequencies would be heavily exaggerated.

I have since found another of my old CDs, Pink Floyd The Wall, to contain pre-emphasis. I have a newer remastered copy of the same album that does not have pre-emphasis. So, I made a comparison of 3 wav files: one from the old CD ripped with EAC, one from the old CD de-emphasized by a rip with iTunes, and one from the newer CD. I looked at the waveform and spectral views and found the old one ripped with EAC was clearly heavier on the high end while the other two were virtually identical. This is conclusive proof that iTunes is applying de-emphasis in the rip process thus reshaping the wav file to the form in which it was intended for playback.

Now it would be nice to have a free tool that can properly apply de-emphasis on WAVs securely ripped by EAC.
BTW, I noticed that LAME has -e switch (marked as obsolete) which only flags MP3 as pre-emphasised (it doesn't apply de-emphasis to WAV - expects decoder to do the job). More details in the docs.

What version of EAC are you using? Wonder if this issue is fixed with the latest version of EAC

I have been using V0.95 beta 2 from June 1, 2005. I just downloaded V0.95 beta 3 dated 8/26/05 and do not see any new extraction feature that would perform the de-emphasis.

Has anyone tried asking Andre about de-emphasis application in EAC?

Is the emphasis in MP3 exactly the same thing as in compact disc

I always assumed so.

That would be the ideal solution. I definitely prefer EAC because of its precision and error detection abilities.

This command line sound file processor, Sox can do de-emphasis. Its standalone usage to do this is just "sox infile.wav outfile.wav deemph" If you pop sox.exe in Eac's directory it should be possible to pipe Eac's output through Sox on the way to the encoder, something like "commandline encoder: sox.exe - - deemph | lame --preset standard - outfile.mp3" -but it will take some fiddling to find the exact syntax.

But... does anybody knows if this emphasis (rarely) used on CD-AUDIO correspond to the RIAA curve used on vinyl LPs? Doesn't seems so.

Sergio

Is it easy to tell if a CD has pre-emphasis or not? Should I be concerned about having to go back and listen all my older CDs to see if they need to be de-emphasised now?

Assuming you still have your original CDs, just open Exact Audio Copy, insert them in the drive and give a look at the rightmost column in EAC. It tells you if Pre-Emphais is used for each track.

Sergio

Also, I wonder if there's only one standarized curve for CDDA, by looking at LAME -e switch, there seem to be at least two kinds of emphases (emphasises? emphasizes? ).

No, it has nothing to do with it. The de-emphasis curve for CD ranges from 0 to -10 dB according to the chart above, and the de-emphasis curve for vinyl ranges from +20 to -20 dB.



For CDDA, yes, but I don't think that CD emphasis and MP3 emphasis are the same.

I have previously used correction values from a 15/50us de-emphasis table for making rough adjustments with a software EQ. My table seems to be a bit different from the curve in the earlier post. I have no idea which one is correct. I guess the difference is not audible.


As a test I tried if iTunes could be used as a de-emphasis tool. First I ripped a CD with EAC in single image wave file & cue sheet format. Then I used Alcohol 120% and mounted the wave file as a single track audio CD by using this custom cue file:


"FLAGS PRE" is the pre-emphasis flag.

After that I ripped this virtual disc with iTunes in wave format. Apparently iTunes used a de-emphasis correction and made a new corrected wave file. This corrected file can be used instead of the original wave file. It works fine with the original cue sheet. Though, the FLAGS PRE lines should now be removed.

I don't know how good the DSP equalizer used in iTunes is, but I can't hear any obvious problems. ABX testing is impossible because the original file is uncorrected.

Edit: typo

An interesting test (that I cannot do as I don't have, and don't want to have iTunes installed on my system), would be to burn a CD (or make a virtual image of it) with white noise or a frequency sweep, insert the PRE flag, rip with iTunes and compare the original.

From that one can see what equalization curve iTunes is using and eventually adopt it within Audition or any other software EQ.

Any voulnteer?

Sergio

Here's a frequency sweep before and after the iTunes treatment:



Edit: fixed the image link

WOW! it is totally different and a lot "stronger" equalization that the one published in this previous post!

Edit: sorry, I missed that the base line was at -20 dB, so... maybe not that much stronger. Anyway I must now leave. Will give a deeper look ASAP...

Sergio

No it's not. Read the graph values more carefully.

The file I used was recorded at -20 db. I didn't bother to fix the db values in the graph.

I checked the source of the table I posted earlier. It was this thread at EAC forum:

http://www.digital-inn.de/showthread.php?t=19034

As I'm re-ripping my entire collection I came across something that might be of interest. The CD "The Digital Domain - A Demonstration" (1983) contains 4 tracks of pink noise:

"22 - Pink Noise @ -26 dB with emphasis"
"23 - Pink Noise @ -26 dB no emphasis"
"24 - Pink Noise @ -46 dB with emphasis"
"25 - Pink Noise @ -46 dB no emphasis"

Unfortunately EAC lists Pre-Emphasis as No for all 4 tracks.

Anybody interested in performing an analysis of these tracks?

- Bill

After reading this thread I went and tried sox.exe to demphasize the few CDs that I have that were mastered with preemphasis. Sox does the job very fast and with excellent results. I've had sox for a long time but until now I'd only used it to convert some aiff files to wav. Thanks to the poster who pointed out a new use to me.

If you have the old version of EAC with the "detect TOC manually" feature, try it. It may set the pre flags properly by reading them from the subchannels instead of the TOC.

Just rip with a CUE and add the PRE flags to the CUE file. When I tested the frequency sweep I used a custom CUE sheet. The sweep file itself was an old wave file from my archive without flags or tags whatsoever.

Alas, I'm using the latest version of EAC, V0.95 beta 3.

I get the impression that these tracks don't use the flag, instead they are examples of what the difference in sound is.

Hi, guys!

I'm collecting classical CDs since the beginning of the CD era.

In my collection of ~ 5000 classical CDs I have discovered some 150 CDs with pre-emphasis so far (1 among ~33 - is this rare???).

Attention: It's often said that only recordings mastered before 1986 use pre-emphasis --> definitely wrong !!!
Some labels do use it just up to date (but not regularily) e.g. Erato and the Swedish label BIS (on almost every second CD) !!!

In short (I can only tell about classical, though) - which labels are suspicious to use it more frequently than others?

I never found any CD by Universal (Deutsche Grammophon, Philips, Decca...), RCA (BMG) or CBS/Sony (but that's no guarantee)
EMI didn't use it after 1987 (but Maazel's Zeffirelli-Otello soundtrack has it - besides this one I only got some CDs from EMI Japan with pre-emph).

Pre-Emph is (still !) used frequently on Supraphon, BIS, Erato, New World Records and other smaller labels - especially from Northern & Eastern Europe.
( I believe you'll find it often, too on smaller jazz labels...)

So pre-emph still remains a serious problem for ripping (because most people simply don't know about it)!
Best puzzles are CDs with mixed pre-emph status (some tracks on the CD have pre-emph, others not!). For a long time I was quite sure such CDs wouldn't exist anyway - but meanwhile I have discovered at least 2 examples of New World Records CDs with mixed status

There's a simple (but precise) free tool available developed by an acoustics engineer (Mr. Ahlersmeyer) to de-emphasize pre-emphasized audio data (so we could do compressing, converting, releasing - all we need afterwards).

It's just great, of course better than the often recommended WAVE 10 band equalizer Q10 - an equalizer can only give an approximation to the de-emphasis curve. Mr. Ahlersmeyer uses exactly the mathematics needed. You need the VB5 runtime libraries installed to your system. The tool is available as freeware here:

http://www.picosound.de/Waveemph100a.zip

Mr. Ahlersmeyer says that the tool wouldn't run under XP --> it does (you have to choose Win98 compatibilty modus though, else you would indeed get an overflow error).

Unfortunately the tool produces an overflow error, too, if the original CD data contains exactly 100% peaks. The only workaroung at the moment would be to normalize to 97% before using the tool...

So - if you want to produce compressed audio from a pre-emphed CD, try the following procedure:

1). Open the CD with EAC and figure out if the PRE-Flag is set
2). Extract the digital audio with EAC - > Copy Image & Create CUE Sheet
3). Apply Waveemph100a to the extracted .wav
4). Split the de-empasized WAV to separate tracks with EACs Tool -> Split WAV by CUE Sheet
5). Produce your mp3s (oggs... whatsoever) from the single tracks...

If you want to archive a CD as a whole (for burning later as CD Audio - not for listening on the PC!) it's sufficient to use EACs
- > Copy Image & Create CUE Sheet -> compressed
because EAC produces correct cue-sheets with the Flag PRE lines (many other rippers don't).

Attention: NERO (even V.7) simply ignores the PRE-Flag in an audio cue sheet (even when burning an image) and doesn't set the bit on the burnt audio CD.
So never burn an audio CD image of a CD containing pre-emph with Nero!
(This is a real shame - only if you will be using Nero's plain Copy CD function the PRE-Flags are preserved!).
Burning with EAC itself of course will produce perfect copies (I would like to know which alternative burning proggies do evaluate the PRE-Flag of an audio CD cue correctly - please post your experiences).

Out of curiosity: Is there any place on the web showing the magnitude response of a proper de-emphasis filter ?

All I ever saw was some kind of analogue circuit which is supposed to do the de-emphasis filtering, but hey ... I've no clue how I should generate a digital filter from something like that -- Also I can't find it right now, nor any good description about such filters via google / wikipedia.

BTW: I once saw an older proposal for storing PCM audio on a DVD with optional (digital) pre/de emphasis. Did this make it into the official standard or did they ditch that ? Does somebody know ?

Sebi

Wow! Resurrecting a very old thread, but I'm just discovering the joys of pre-emphasis...
Now. I was preparing my big move to a music server and I already copies 400 of my CDs on my HD, using EAC, with cue sheets. The I did The Wall, my 1979 copy. Pre-emphasis. 1st time.
What I would like to know is:
is there any utility program that can perform a earch on all my CUE sheets to check if the FLAGS PRE appear?
I've tried the Search function of XP, with no success.