Different sounding Rip Programs

[ToniM]

Hello all,

after my experience with the famous QA-550 player, Black Crow and Mojo 2  I´ve done an evaluation about the cd - wav convertion process. I used the rip programs "Exact Audio Copy", "Nero 8" and the "Windows Media Player"I ripped the same CD`s with this programs on the same SD Card and what a surprice they sound all different. And the difference is not subtle. Its like SPDIF and I2s.To make a long story short, Nero 8 wins by a large margin. The sound is much more analog! Exact Audio Copy sonds like a crappy Cd player and WMP somewhat in between.
Have someone the same experience?

Tony

[vt4c]

Tony,
  Thanks for the heaps-up....You're always ahead in trying out things. 

[yeo]

hi tony

hmm... now i'm worried!

since i'm using mac, i use itunes only. eac is supposed to be really good. unless yours isn't configured properly?

i like to use itunes as the handle the library in a nice way.

yeo

[ToniM]

Hello Yeo

EAC is configured to optimal quality and not speed. Test it yourself.

[jon L]

Hi TonyM

Hmmm. very Interesting. I have a more basic rig than yours usb monica 3 and charliize, both modded quite heavily. I use a PC and Cplay memory player playing thru 1gb usb drive. I regard my system as Hi rez.

I have found that ripping EAC in burst mode (See under drive options tab, It allows no error correction at all !!!) to my complete surprise easily betters secure mode by a quite a margin. I found this out only recently.

Why? This should not be ? Many believe error correction interlacing i.e. Reed Solomon code as a bad thing in terms of CD playback, hence mem players sound good. Check out Nova physics. Of course your CDs have to be in pretty good nic to use this mode, but i'm surprised how many rip very well even with a few errors here and there but usually they're inaudible or one or two a faint tics per CD at worst. Music sounds much more present and correct. The rips are real fast too.

I will check out Nero 8 and see how it compares.

Regards

Jon L






  

[ToniM]

Hi all,

installed I.tunes yesterday and riped a cd with it.Also EAC in burst mode. Conclusion is Nero 8 sounds best the flow the sense is better than the other programs.

[ole]

This never fails to amaze me.

The CIRC error correction on the CD is really correction, i.e. it provides redundancy to completely repair read errors on the fly. The resulting data is absolutely correct. Disabling this function is a real bad idea. Errors that are too severe to be fixed by CIRC are interpolated, which are the only errors that can possibly have an audible effect. Some ripping programs actually display a statistic about which types of error occurred and how they were handled. The resulting WAV files should not be subject to interpretation - they should be exactly the same every time.

Some computer CD drives get sloppy when reading out audio data, though, especially at higher speeds. Some programs like EAC try to compensate for that by reading multiple times and comparing and then taking a guess. I think varying results with different software or even different runs with the same program points to a problem with the drive. Some drives are better for ripping than others due to their firmware supporting audio ripping better.

Try to rip several times and compare the resulting WAV byte wise. Under Linux, you can just use the "diff" program. It should be the same every time. If it's not, try a different drive.

Ole

[jon L]

Hi Ole, Tony M, and all

Thanks Ole for your input on error correction.

EAC should do the best job by thoroughly checking by each sector. You suggest poss problems with individual drives, maybe ? Mine is a TSST DVD-ROM TS-H352C only 3 yrs old with light use and works fine. It is puzzling how different ripping software produces significant variations in SQ on the same drive. i've just tried express rip free and this sounds fuller than EAC in secure mode.

I thought EAC was king, and maybe it is, but my drive could be letting the side down. I've also found simple things like Laser scatter in the CD itself affects SQ. I have lined the CD tray with turquoise sheet. The rip shows 100% sucess in quality no errors and bytes exactly the same, but still noticable SQ differences. It seems the optical interface, and issues associated is definitely a weak area in digital.

Roll on high rez downloads straight from servers and master files. In the meantime getting CD to HD,SD or stick drive is a precarious business!

Regards

Jon L

PS: Can anyone suggest a good ext drive that can be tweaked easily
(ie not slim line and easy to get inside)

[ole]

I think you should really try to compare the digital data, rather than guessing which program sounds better. If they do not sound the same, the at least one of them makes mistakes. Under Linux it's "diff", under Windows the "comp" command. Take two rips from the same track with the same program and compare the files. As there is no timing involved, there cannot be a "better" rip - only correct rips and faulty rips. If you get two different files, try a different drive.

All ripping programs analyze the drive first and determine how to obtain audio data from this particular drive model. Some drives have higher level functions that read out the actual audio data and correcting it, possibly hiding errors from the software. Others can read only raw data and the software has to process the read timing and error correction. Some may have firmware bugs, too. I used to have a drive that was so stupid that audio data could be read only at 1x!

Have a look at http://forum.dbpoweramp.com/showthread.php?t=16442 , there is a discussion about which drives perform best. Over the years, some Plextor drives seem to have stood out pretty well.

Ole

[CoughinHK]

Hi,

I have noticed differences in rips using the same drive and different software, but for me WMP was worst and dmc (dbpoweramp software) and EAC were best.

I noticed much greater differences with drives - the best so far is a Pioneer BDR205 (bluray writer) - the worst was a TSST notebook CDwrite/dvd reader.

Disk condition is exceptionally important also. 'As new' disks are a must. Love them !

I completely agree with Ole; when all these factors are right, the rip is as good as it can be and then it is down to the storage media, etc, to get the playback right, and for that, the number of variables is so huge that it is probably the source of most 'problems'.

Just my 2 cents. 

[Jochen]

Hi my two cents:

I believe in what Tony says! I'm also in my experimental rip phase

Current conclusion:

WAVE sound better than FLAC (both soundwise uncompressed).

TEST IT  and your amazed, its also better than WMA!

I use dbPowerAmp and an IBM notebook R51 - 1830

I've just dismantled Monica again but in my christmas holidays I'll do tests with A)Ripping software.
B) Processsor speed.
C)Mechanical connection of my notebook to the floor. meaning spikes, puks....
D)CD threatment like L'art du son
E)Supply issues of accu vs PSU, WLAN noise, TFT noise


Greeeeets Jochen

[ole]

If you have a track which you ripped once as 1.flac FLAC and once as 1.wav WAV and you think that your 1.wav sounds better - just convert 1.flac to 2.wav, and 1.wav to 2.flac. Use nothing fancy, just the free FLAC encoder/decoder.

Now try to find out which one sounds best, and which worst, by letting a friend choose the tracks at random.

If you find that 1.flac and 2.wav are equally worse than 2.flac and 1.wav, then there is something wrong with the ripping program or drive that you used to rip the original FLAC.

If however you find that both 1.flac and 2.flac sound equally worse than 1.wav and 2.wav, then you should be trying a different playback software because it obviously does something to the uncompressed FLAC stream that it doesn't do to the WAV stream.

My point is that both formats preserve the integrity of the original data 100%, which means you can convert them back and forth as much as you want without changing anything. If it still sounds different, then it cannot be related to the format. And timing is not part of an audio file, no matter which format. Jitter should also not vary significantly with different formats, provided that the output interface driver is written properly.

Of course one could say so what, if it sounds better with WAV then just use it. Nothing wrong with that, but - there is obviously another issue that causes the problem, and if you find it, it could improve performance altogether, not just with FLAC. The effect you see is just a symptom of something else.

Ole

[Jochen]

Hi Ole,

thanks for your suggestions.

I've also wondered when my friends came over the first time with this discovery. They even judge the difference much higher than me (Wave - Flac). For me its more a small step - but anyway naturalness and soudstage is better - you will not step back.
As my friends aren't technically interested they thought obvisiously that FLAC is soundwise compressed.  But I clearified it that its the 100% identical content just written in a much economical way.

-> I will try your suggestions.
-> I'm also interested what happens with my old FLACs when I simply convert them to Wave. I currently have some ideas that the higher computing rate when playing FLAC is a Jitter causer - buts just an idea. The proposed test may hopefully clarify.


But Ole: Why not test it in a short trial as well?  (I'm very curious about others expiriences here). Just use DbPowerAmp or something else to make this small test. Take a good audiophile record for this - I initially chose the Koeln concert of Keith Jarret. The piano and the whole atmosphere was simply more breathing and the feeling of "3D" staging was improved when using Wave.

I use Vista with J.River (memory play & small buffer/short latency settings)

Bye

[ole]

Jitter would be the only problem that's possibly left if the rip is identical. In that case, the culprit is the driver of the sound output device, which ideally should not depend on CPU load at all, but if it does, then things go south.

It is by the way not per se a good idea to use a low latency sound driver for this purpose. Low latency means low (or no?) buffering, and that could mean more dependency on CPU load. Of course, that also depends then on what comes behind the output device, i.e. what jitter countermeasures are there on the receiving end.

I don't have any Windows system here, I use MPD on Linux for playback, and it's configured for bit perfect output. There is no mixer involved, no resampling or other things like that. There are also no other applications running, all system activity is reduced to a minimum. It doesn't' even have a harddisk.

Just to be sure I don't miss anything, I connected my Kumisa III headphone amp and a closed studio monitor headphone (Superlux HD662F, gotta love it...but that's another story).  This combo is like a magnifying glass.

My favorite test track is Jazz at the Pawnshop, Lady Be Good. I alternated between a FLAC and a WAV, at random, for several minutes. Didn't even look at which one it was at the time. And it was dead silent in the listening place, too.

Lovely sound, lots of detail and ambience and all the goodies, but - absolutely no difference at all. I really couldn't tell which is which. Case closed 

Now one could of course argue that my hardware is perhaps not good enough to tell subtle differences like that, and I'd probably say that my hardware just doesn't generate these differences in the first place, and looking at just the facts it should be clear that there shouldn't be any if everything works as it's supposed to.

FLAC and all the other lossless codecs are essentially nothing else but glorified Zip programs. The only difference is that they have knowledge about the nature of audio data and can make certain assumptions when compressing that Zip could not (being generic), resulting in a better compression ratio. One example would be the assumption that left and right channels have usually a lot in common, so what's present in both channels at the same time doesn't have to be stored twice. Again, this is bit perfect and of course varies case by case. If both channels have nothing in common, the ratio just gets worse, the resulting file gets bigger, and it's still bit perfect.

So, if you really hear a difference between WAV and FLAC - then the culprit is somewhere else in your system.

Ole

[Jochen]

Hi Ole!

Woow very intersting!  thank you that you've done a short test!

Very interesting indeed: No difference at all.  
Well I do agree with your arguements that the culprit can be at my hardware or even a S/W issue in the chain of how my music is replayed. FLAC and Wave by theoretical approach should sound identically, yes I think you're right.

Never the less it makes a different in my system and its therefore (for me) quite interesting to come after this effect.  
I also strongly believe that my PC system set up is a "typical" one seen everywhere and so other Windows PC users expirience the same. As proof I have my two friends already. Meaning that discovering the effect is helpful for other Windows users.

BTW: Very funny - I started the wave/flac test initially with my headphone system as you. But I have only a "spontainious" and not really a representative set up: Its a battery driven T-Amp and a Yamaha orthodynamic headphone which is slightly modded.
Result: Flac was better than Wave ...anyway the difference was there     Anyway I trust my speaker system much more   and here Wave was clear winner.


One last question Ole: Could you help to build up a Linux music system with my IBM Notebook? I tried but I'm not skilled enough
What software are you using and can I run it as a third OS toghter with WXP and Vista? I only need my USB Monica and my external harddisk to setup running in good performance- the rest is not important - I would like it for music only.