Yeah, I also noted that Nero Digital Audio has sometimes evident strong clipping "clicks" when played in foobar2000 without DSP "Advanced limiter" enabled.

You have probably experienced something else than plain decoder clipping.

I have recently tried to ABX the effect of decoder clipping without success.

A track peak number like 2.6 is interesting and may look serious, but in my experience this is not audible. These extreme peaks last only a few samples at max (mostly only one sample), i.e. typically these clipped peaks have a duration of under 0.1 ms.

You're right, just tried to detect the same issue on another machine with the same tracks, but there are no such clicks.

Here is an example:

The first file contains 1 ms of the file with the 2.6 track peak. The "2.6 peak" is at 0.5 ms in the right channel. I decoded the file with foobar's replay gain (about -8.4 dB) to prevent it from clipping.

The second file is the same 1 ms sample after increasing the file gain 8.4 dB with Wavelab, saving the file in 24-bit integer format, reopening the file, decreasing the gain 8.4 dB and saving in 16-bit integer format.

1ms.wv (0.3 kB)
clipped_1ms.wv (0.3 kB)

Do you happen to use ASIO output on foobar? I am unsure, but the driver could cause the clipping there (actually worse, the inversion of the sample).
Note that this is an hipothesis.

Guys, I am very sorry for not being online for so long, but I am facing some PC problems. I re-formatted my Samsung HDD on Friday, spent the whole Saturday setting up Windows, Office and other programs I use just to notice that my disk suddenly started to have bad sectors. After running chkdsk /R, the PC rebooted and my partition table was gone. I managed to restore some data, but the majority of the files could not be read any longer. The most crappy thing is that my OS backup was on the Samsung disk at that time because I needed space on my external WD HDD in order to copy my pictures, documents and WavPack disc images. I was just about to switch the location of the documents and backup when I noticed the bad sectors. In the meanwhile, I downloaded a free program for analyzing disks which also happens to be able to restore bad blocks - no idea how it works, but it worked as it seems. After "regenerating" the HDD, the Gutmann wipe (35 passes) finished successfully and chkdsk doesn't show any space used by bad sectors any longer.
By the way, before running the repair tool, S.M.A.R.T. did not report anything and the Samsung diagnostic tool also didn't show any errors. Its surface test finished at 2% with the message that my unit is in good condition - yeah...

I am now configuring the OS and the other programs and hope to be ready tomorrow or Wed. the latest.

Thanks to everyone who posted samples, by the way. I think I have enough now and will decide which ones to use once the PC works.

Buy Seagate next time

Hey, you were one of the persons who recommended Samsung!

Indeed! But after one of my Seagate SATA HDDs was severely shorted, and afterwards booted normally, I decided I would never buy from another brand (unless they really fuck up and go the way of the Maxtor)

http://www.rarewares.org/rja/IMG_0078_3.jpg

Enough off topic now...

In the end, I decided to use 20 samples because it's actually only 2 more than 18

Alpha: pre-echo, pretty low
aquatisme: weird
bibilolo: bandwidth test
BigYellow: English male voice, easy listening / pop
debussy: low volume
Eleanor_Rigby: stereo separation, violins
Flyin___to_Fly: high tones, sort of Indie music (at least that's what I think of when I listen to it)
Gipsy: stereo separation, electric guitar
kraftwerk: pre-echo, electronic music
LesJoursHeureux: instrumental (harpsichord)
Locomotive_Breath: flute, some piano, cymbals, some electric guitar
macabre: symphonic orchestra
MysteriousTimes: usual pop / trance
Paganini_Allegro_spirituoso: classical, dynamic
Senor: jazz, French female voice
SQAM German male voice
symphony_metal: pretty loud mixture of symphony orchestra with percussion and electric guitar
The_Wizard: acoustic guitar (longer than guitar_potf)
TomsDiner: English female voice
WhiteAmerica: rap, English male voice, "strong"

Any protest?

It sounds good

OK, if nobody complains, I will start preparing the test today, have the samples uploaded by the end of Sunday and start the test early next week (Monday, Tuesday or Wednesday the latest).

Hello Everyone!

I'm so happy I found these forums

I just wanted to share with u guys my own little take on the WMA Pro 10 standard....

I , until recently, have started to encode my cds to WMA PRO 10 and have found the codec to be very good to my ears... I used to encode to wma standard, but have found that the PRO 10 verison of WMA is much better!

I just wanted to share the following files with you... I really hope u guys don't get ticked off or something for posting this post... I just wanted to share my recent experience with it.... Yes, maybe i'm a total neophyte to the audiophile business, but I find the codec is just SUPER! Here are WMA Pro Files i encoded from a CD. They are in spanish, so... (sorry, need to get some english, i know! But I LOVE this song ) Click on the song titles to load the songs.......

File 1
------------------------

Song: Te Pertenezco by Fey
Codec: WMA Professional 10
Encoder program: dBpower AMP 12 (Alpha 7)
Bitrate: 48Kbps Constant Bitrate
Sample Rate: 44Khz
Depth: 16BIT
Encoding Mode: 2-pass encoding

*************************************************************

File 2
------------------------
Song: Te Pertenezco by Fey
Codec: WMA Professional 10
Encoder program: dBpower AMP 12 (Alpha 7)
Bitrate: 256Kbps Variable Bit-rate
Sample Rate: 96Khz
Depth: 24BIT
Encoding Mode: 2-pass encoding

*************************************************************

Now, i know the 2nd song is encoded at a higher frequency than a normal cd (44 or 48khz?)... Like i say, i'm a total neophyte to this....But I am very happy with this new audio standard.. Just wanted to share my experience!
Eddi

P.S. Please contact me if u have any questions or problems with this post! Thanx!

a) gotta say i love fey, too! do you have her DVD?
b) they don't like samples longer than 30-60s around here...
c) did you get the audio from a DVD-A source or something? i wasn't aware of a 24-bit version... if it's just from a CD, just use "whatever (bitrate) sounds best to you" ... i failed at ABXing WMA9Pro @ 128kbps

also read the FAQ >_>

also, why not just use MP3?

Sebastian Mares: why not use eig for pre-echo? I found it's much easier to hear pre-echo in it than in Alpha (with vorbis, LAME, and CT-AAC, don't have Nero so didn't test that). Then again, Vorbis handles Alpha pretty poorly.

Is eig the same as ae_gg?

no it isn't, here is a link to the eig.wv sample.

OK... Well, personally I don't mind which one is chosen. Anyone else thinks eig is better than Alpha?

eig was chosen as sample.

By the way, I just noticed that iTunes which I use for encoding anchors does not support VBR encoding at 48 kbps. I will use CBR instead - it's low anchor anyways.

I'm afraid that iTunes AAC LC at 48 kbps is too good as a low anchor. With certain samples it may be better than for example WMA standard. The low anchor should generally have clearly lower quality than the contenders. Perhaps you could lower the bitrate to 32 kbps.

As a high anchor I would be interested of seeing aoTuV beta 5 at 96 kbps instead of iTunes AAC. iTunes AAC quality at 96 kbps is pretty well-known (at least for me). It would be interesting to see how close to transparency the newest aoTuV reaches at 96 kbps (-q 2), even though that is not the purpose of this test.

Sorry, but this comes too late. The choice of settings for contenders and achors is known for quite some time. Samples were prepared already. I don't think 48 kbps AAC is too good to be honest. We'll see in the test.

I just checked again and I think LC-AAC is good as low anchor with its current setting. The sound is also somewhat lowpassed because iTunes resampled to 32 KHz.

iTunes AAC-LC should be fine as a low anchor, in my opinion. In the AAC 48kbps test, it was a good low anchor.

Perhaps WMA 9.2 standard is then better than the older WMA standard versions have previously been. At least at 64-80 kbps range WMA standard has not been better than iTunes AAC. Though, at 48 kbps the situation may be completely different than e.g. at 64 kbps, but has anyone actually tested and reported this?

rjamorim's public 64 kbps test: http://www.rjamorim.com/test/64test/results.html
guruboolez's personal 80 kbps test http://www.hydrogenaudio.org/forums/index....showtopic=35438


Edit:

It was a good anchor against AAC-HE, since the quality difference was big enough, but this test is different. aoTuV b5 may be better than the current release version, but only marginally. I doubt that WMA 9.2 standard is radically different from 9.1. Quality of WMA 10 Pro is unknown before the test.

On a second thought, perhaps iTunes could be promoted to be a test contender if the results are on a par with one or more of the other contenders' results. However, in this case the test would be more difficult for the testers since there would be an additional encoder that is not clearly distinguishable.

Edit

I listened through 3 x 10 samples (iTunes & WMA 9.1 Standard & the first version of aoTuV b5 from last summer).

It appears that iTunes AAC really uses a lowpass that makes it distinguishable from the others. The lowpass frequency is a bit over 9 kHz. This may help it to produce more pleasent quality with some samples.

For example, WMA 9.1 standard is much brighter since it uses an about 12 kHz lowpass. On the other hand WMA 9.1 standard produces some very nasty artifacts. (I didn't have WMA 9.2 samples readily available).

I used the same samples that I used when making these tests:
http://www.hydrogenaudio.org/forums/index....st&p=402289
http://www.hydrogenaudio.org/forums/index....st&p=401514

I am living the anchors as-is. As I said, I was clearly able to tell apart WMA Standard and Low Anchor and found Low Anchor to be worse than WMA Standard which you are right, has a quite metallic sound.

Actually, it will be interesting to see which one the testers prefer more, the lowpassed iTunes or the brighter WMA Standard with its "metallic" artifacts. iTunes has bad artifacts too, but at least with certain samples these are not as pronounced.

Did I screw up:

Forget the last post. Does anyone have an elegant solution for decoding samples under Linux and OS X (PPC AND Intel)? Edit: Except providing samples as WAV. For Linux I could imagine using "wine wcmd /c SampleXX.bat" if user has wine installed. In that way, the same binary would be used for Linux and Windows (no trouble with possible different offsets, whatsoever).

I would like to ask OS X people to use a Linux live CD for this test (Ubuntu for example). This makes it easy to decode samples (by using Wine) and I can also use the latest version of ABC-HR which works with JRE 1.5 only.

Edit: News: I have a ABC-HR version that should work with JRE 1.4, too. OS X users are asked to handle decoding of samples themselves, though. I think the script I posted above should still be available. You can also use the bin_osx stuff from my last test (in the ABC-HR ZIP).

Seems like Nero has made a press announcement?
CDRinfo.com (11-1-2007)

Interesting. By the way, my listening tests have a new home: http://www.listening-tests.info.