Different sounding Rip Programs

[ole]

Hi Jochen!

A while ago there was an article by me on this site on the topic, however it is a little outdated and possibly overkill for what you want to do.

It's fairly easy to install one of the common Linux distributions on your notebook and then add MPD to it, assuming your music files are on the notebook. The slightly difficult part is how to configure MPD for bit perfect output. You would have to read about that in the docs, it also depends on the sound subsystem your Linux distribution uses - the latter would of course also have to be configured that way as well, so that the sound device you are using only gets the sample rates it can handle natively, i.e. no up/downsampling on the fly and no mixing. There are many articles on the topic, chances are that somebody did it before. Google is your friend there.

MPD is just the server side, you can then use any of the available front ends you like to control it. They are all on http://www.musicpd.org/ . Many distributions have ready made packages for MPD, too.

Most current Linux distributions (like Ubuntu, for example) can be installed along with other OS. To explain how to do and where to click would be a little difficult here. Just try.

Ole

[ole]

It's been quite a while since I initially put together my Linux music server, and Jochen's question made me think - so I looked around for new stuff.

There is a Debian based Linux distribution called Voyage Linux, http://linux.voyage.hk/ which is also available in a customized music server version featuring MPD. It also has a web frontend for basic configuration. The installation is not too difficult, it can also run off a flash drive. Hardware requirements are very modest.

I already replaced my old system with this, works fine!

Ole

[Jochen]

Hi Ole,

thanks   I'm just checking the website.


Ahhh got news  @ computeraudiophile is a discussion ongoing about sound differences when using wave, Aiff, Flac.

http://www.computeraudiophile.com/content/WAV-vs-new-FLAC-Uncompressed-format

Basically the same arguments, some people here differences - some not -  & some claim that by theory there cannot be a difference.


Jochen

[ole]

It's really a very old issue, and it isn't about just WAV and FLAC but any type of digital audio source or transport. Think of CD transports with belt drives vs. direct driven (this argument was directly taken over from the analog age), and the associated tinkering with it. Point is, if some people hear a difference and can reproduce that in a blind test, then there probably is a difference in the system of those people, but it's not caused by WAV or FLAC or the by CD being spun on a kilogram heavy belt driven plate vs. cheaply direct driven inside a computer drive. Most people don't know how to do anything about the software side of digital playback (other than changing the program) and so they tweak what they can see, which is the hardware or the file format, and if it's any better, the conclusion quickly jumped to is that it was the cause of all the trouble.

There's nothing wrong changing the transport or the file format if it gives a better result. Only the conclusion is wrong.

On the mentioned forum, they can't seem to accept that there are indeed differences for some people, but that there is also direct evidence that it isn't caused by different file formats. Perhaps they are subconsciously afraid that if they accept this fact, they don't know what to do about it?

Ole

[Jochen]

Hi folks,


I want to reply to Oles last suggestion as seen in the above post:


The test was done at my friends apartment, we used his HP notebook, Windows XP, Sonata Music server (iRiver) & his highly modded Monica with a brilliant modded AKG Headphone. I may also add that later I verified this exactly within my high end home system.


1. We used a FLAC file: Patricia Barbera "And beat goes on"

2. We converted the Flac to Wave with DbPowerAmp and later changed to flac and son on...

Goal was to verify if sound changes can be detected during these convertations.
 
My friend and me discussed our listening findings and both shared (suprisingly) the same result. I don't want to go as far to "objectivly" say we detected "improvements" but very definitve differences!!!

Speaking subjectivly its no question we gained enhanced SQ by different format with a winner called WAVE.

Details:

We started by Particia Barberas Flac I had on my HDD since years. Its a 44,1KHz-16bit file from a CD.
Music: In the first seconds the live atmosphere openes up and the crowed can be heard talking, moving chairs... the room acustic can be identified easily = we have a small club were Patrica playes. The bass player openes up ...followed by several hand claps in rythm. The claps have a nice stereo effect. Later Particia sings and she plays hammond organ with sweet harmonics.


We converted the original FLAC to Wave by the program DBPowerAmP and listenen again:
For us two we confirmed about the new produces WAVE file that the stage, the air gained a lot, In our ears the noise floor was improved. Especially the first seconds of club enviroment before the "beat goes on" was much more detectable!!!! Felt like the Flac did cut this away = the "air" inbetween wasn't important. With Wave the swinging claps were more lively - the whole scene was playing more authentically.



OKAY we converted the WAVE back to FLAC:
This FLAC was now slighty better than the original FLAC but not as good as the wave.
That is a very strange expirience no one had thought about But definetly this flac sounded different than its original. For us its the better flac now


And once more. The 2nd FLAC back to WAVE again.
Again the improvements are back.BUT!!! This 2nd WAVE isn't as good as the first converted one. Somehow logical - but still these effects were shocking us


We made this until the 5th conversion and noticed a minor degradation growing each process.






OK but we didin't really started a science here. I'm sure there are things to discover - buts not our task.

what can my friend and me conlude after this night? Well with a windows system you'll expirience differents in uncompressed formats. For us the WAVE is the winner. It has less noise floor. Its produces a different staging image
(Really). WAVE is at the musicians side. You'll get a natural impression that humans play instruments in front of you. With FLAC we had a rather "overall" sound presentation. Details or the "close up" view were not as present- at least influenced by the feeling of less darker background.


We're just catching the first thing of an investigations ==> the effects. What factors may be the root cause? I don't know??

                                                             

Ole probably you can interpret this findings better than me... I'm interested in what's your opinion.

For all the windows user it may be helpful to try this conversion as well and probably gain some sound quality without investing money. If you're able to confirm it - I would be interested to read it here.

Thanks Jochen



                   

[ole]

FLAC compression is lossless. In fact, there is a test function (use option -V) in the original FLAC binary that ensures the encoder works correctly by doing exactly what you did, that is by decoding again and comparing against the original. The FLAC developers also have a comment on this, see http://flac.sourceforge.net/faq.html#tools__hardware_prob.

Please note that the decoded WAV files can actually differ from the original, but only in their header part, not in the audio data - so if you just do a binary file comparison you might draw the wrong conclusions.

Bottom line of this is: if your encoder/decoder chain works properly, you can repeat encoding/decoding infinitely without changing the audio data. In other words, if your audio data degrades over repeated encoding/decoding then there is something wrong with software or hardware.  Since you did not use the original FLAC encoder but some bundled version included with DbPowerAmp, I would assume that the latter has bugs. Or, perhaps DbPowerAmp does something to the decoded data before saving it or playing it back?

If you really want to know, you might want to repeat the experiment with the original available at http://flac.sourceforge.net/download.html, using the command line encoder/decoder. For direct playback of FLAC however I don't know which Windows program works correctly, I'm a Linux user. There is however a Winamp plugin included with the original FLAC package on Windows, so perhaps try that.

Ole

[Jochen]

Hi Toni,

tried to use Nero 8 as trial but it doesn't work. Under XP and Vista Nero 8 tells me that the trial is expired??? Did you have the same troubles??? Or did you check any newer Nero Version?
 

Jochen

[ToniM]

Hi Jochen

both Nero 8 and 10 works on my Vista Notebook. Is it the full version?

regards Tony

[Jochen]

Hi Toni,

got the failure: the Nero 8 version is not permitted to try as "trial anymore". Nero 11 is up to date now. Its stated on a shareware website.

Jochen

[Jochen]

Sooooooooo now I fooled my PC : I set up a date of the past and could install the "old" nero as trial.

Nero understood fast and blocked it again   but I could rip a song of my wifes EMI FUJITA - best audio CD.

As alternative I used Foobar, it has a rip function for those who don't know.



Same Notebook, same power supply, same OS (WXP), same CPU settings, same processes running, only rip speed was slightly different with both. Anyway they stayed between 1x-4x. I used WAVE as format and saved to my external HDD.


Result: Well very easy I can confirm a different sound - Nero sounds cleaner, less background noise, more focus to the instruments. Staging is rock solid. A note played with Nero is stronger and intense. Foobar isn't bad but Nero is one stage ahead. The only critisim I have: For me Emi Fujita sounded a little too analytic.


So what's coming next? I again fool Nero (forgive me)  and check out more music and compare other converters and other system settings like CPU speed and accu usage of my notebook.

Toni very interesting indeed  ....Grueße Jochen

[Jochen]

.....The slightly difficult part is how to configure MPD for bit perfect output. You would have to read about that in the docs, it also depends on the sound subsystem your Linux distribution uses - the latter would of course also have to be configured that way as well, so that the sound device you are using only gets the sample rates it can handle natively, i.e. no up/downsampling on the fly and no mixing. There are many articles on the topic, chances are that somebody did it before. Google is your friend there.

MPD is just the server side, you can then use any of the available front ends you like to control it. They are all on http://www.musicpd.org/ . Many distributions have ready made packages for MPD, too.

Hi Ole,

a german magazine "Eins und Null" helped me understanding now what to do for Linux.

I'll start with Ubuntu + MPD and user interface. I use again my notebook just with a third partion.


I've got two question, maybe you can help me:
You spoke about bit perfect playback. I understood that at Linux it means to "bridge" the ALSA mixer. You have to set the samplingfreq and bit depth to a fixed value for the DAC in MPD config.

Is that right? and is that the only thing to optimize?    or is there anything else   


Secondly: I definetly enjoy the function of my JRiver that it automatically recognize when playing high res files that they get resampled to 48Khz 16bit and so I'm able to also play these @ Monica without manual settings.

Is there anything comparable at Linux / MPD ?



Regards Jochen