Skip to main content ↑ ↓

Notice

Please note that most of the software linked on this forum is likely to be safe to use. If you are unsure, feel free to ask in the relevant topics, or send a private message to an administrator or moderator. To help curb the problems of false positives, or in the event that you do find actual malware, you can contribute through the article linked here.
Topic: Multiformat Listening Test @ 128 kbps - OPEN (Read 164729 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

Multiformat Listening Test @ 128 kbps - OPEN

Reply #150 –
Quote
Quote
Quote
Once the results come in, post that, you'd most definately not want the /. masses praticipating since 9/10 they'd just purposely screw up the stats.


Interesting. How would they purposely screw up the stats?
[a href="index.php?act=findpost&pid=349130"][{POST_SNAPBACK}][/a]


Maybe he thinks they would just rate the samples at random?
[a href="index.php?act=findpost&pid=349263"][{POST_SNAPBACK}][/a]


Chances are good that they would rate the reference sample then, too. Such results are easy to sort out.

Multiformat Listening Test @ 128 kbps - OPEN

Reply #151 –
Quote
hey guys.

just finished setting up all the files needed for the test including installation of java, and i have a problem. the wav's dont play. what am i doing wrong?
reference file doesnt play, ABX option playback doesnt work either. checked settings and my sound card is selected.....
need help if you want some results

thanks all
[a href="index.php?act=findpost&pid=348010"][{POST_SNAPBACK}][/a]


What sound card is selected?

Multiformat Listening Test @ 128 kbps - OPEN

Reply #152 –
Quote
Do you have MacOS X 10.4 Tiger?
You should have a "lib" folder in bin_osx, otherwise wavpack will not decode files.
[a href="index.php?act=findpost&pid=349133"][{POST_SNAPBACK}][/a]

Yes 10.4.3 - Can't remember if there was a lib folder. (Tried to run test at work, now weekend )
If it's not there, should I download it seperately somewhere or just create an empty lib folder?

Multiformat Listening Test @ 128 kbps - OPEN

Reply #153 –
Quote
Quote

Do you have MacOS X 10.4 Tiger?
You should have a "lib" folder in bin_osx, otherwise wavpack will not decode files.
[a href="index.php?act=findpost&pid=349133"][{POST_SNAPBACK}][/a]

Yes 10.4.3 - Can't remember if there was a lib folder. (Tried to run test at work, now weekend )
If it's not there, should I download it seperately somewhere or just create an empty lib folder?
[a href="index.php?act=findpost&pid=349390"][{POST_SNAPBACK}][/a]


A "lib" folder was included inside "bin_osx". It contained the following files:

libwavpack.0.0.0.dylib
libwavpack.0.dylib
libwavpack.a
libwavpack.dylib
libwavpack.la

Multiformat Listening Test @ 128 kbps - OPEN

Reply #154 –
I've downloaded the torrent with all samples but can't find the "bin" directory referenced in the readme. I already had ABC/HR Java from before so I didn't download that again. Also there's no .bat files, just the audio samples. Is the stuff I'm missing in the ABC package you're hosting?

edit: Nevermind, I DL'd your ABC zip and everything is in it. You may want to clarifiy that on the readme. As it is now, it looks like you don't need to download it if you already have it.

Multiformat Listening Test @ 128 kbps - OPEN

Reply #155 –
Quote
Also, @All - I did Advanced-PEAQ analysis (with Opticom Opera) of the listening test samples, and did statistical analysis afterwards (ANOVA).

I will publish results when the test is over - it will be very interesting to measure correlation between Advanced PEAQ and real-world listening test results.
[{POST_SNAPBACK}][/a]


I’ve also finished loading the same contenders but different sound samples to SoundExpert testing engine. Starting ratings based on my own listening tests are ready but will be published at the same time with the current test results. Corresponding web page [a href="http://www.soundexpert.info/coders128.jsp]“Special 128 kbit/ test”[/url] is ready now.
keeping audio clear together - soundexpert.org

Multiformat Listening Test @ 128 kbps - OPEN

Reply #156 –
Quote
Quote
Quote
Do you have MacOS X 10.4 Tiger?
You should have a "lib" folder in bin_osx, otherwise wavpack will not decode files.
[a href=\"index.php?act=findpost&pid=349133\"][{POST_SNAPBACK}][/a]
Yes 10.4.3 - Can't remember if there was a lib folder. (Tried to run test at work, now weekend )
If it's not there, should I download it seperately somewhere or just create an empty lib folder?
[a href=\"index.php?act=findpost&pid=349390\"][{POST_SNAPBACK}][/a]

A "lib" folder was included inside "bin_osx". It contained the following files:

libwavpack.0.0.0.dylib
libwavpack.0.dylib
libwavpack.a
libwavpack.dylib
libwavpack.la
[a href=\"index.php?act=findpost&pid=349393\"][{POST_SNAPBACK}][/a]

OK, issue solved, following is dump of output when running the .command.

Code: [Select]
Last login: Mon Dec 12 07:13:00 on ttyp1
Welcome to Darwin!
dhcp1s608:~ az$ /Users/az/Desktop/ABC-HR/bin_osx/Sample01.command; exit
/Users/az/Desktop/ABC-HR/Sample01.zip
Archive:  Sample01.zip
   creating: Sample01/
  inflating: Sample01/BigYellow.wv   
  inflating: Sample01/BigYellow_1.m4a  
  inflating: Sample01/BigYellow_2.mp3  
  inflating: Sample01/BigYellow_3.mp4  
  inflating: Sample01/BigYellow_4.mp3  
  inflating: Sample01/BigYellow_5.ogg  
  inflating: Sample01/BigYellow_6.wv  
dyld: Library not loaded: /usr/local/lib/libwavpack.0.dylib
  Referenced from: /Users/az/Desktop/ABC-HR/bin_osx/wvunpack
  Reason: image not found
/Users/az/Desktop/ABC-HR/bin_osx/Sample: line 38: 10196 Trace/BPT trap          $wvunpack ${basename}.wv
 ****** FAAD2 (Freeware AAC Decoder) V2.0 RC1  ******

        Build: Sep 11 2003
        Copyright: M. Bakker
                   Ahead Software AG
        http://www.audiocoding.com
        Floating point version

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

BigYellow_1.m4a file info:
Track   Type    Info
1       audio   MPEG-4, 24.543 secs, 128 kbps, 44100 Hz

  ---------------------
 | Config:  2 Ch       |
  ---------------------
 | Ch |    Position    |
  ---------------------
 | 00 | Left front     |
 | 01 | Right front    |
  ---------------------

Decoding BigYellow_1.m4a took:  0.58 sec.
input:  BigYellow_2.mp3  (44.1 kHz, 2 channels, MPEG-1 Layer III)
output: BigYellow_2.wav  (16 bit, Microsoft WAVE)
skipping initial 1105 samples (encoder+decoder delay)
Frame#   940/940     32 kbps   MS            
 ****** FAAD2 (Freeware AAC Decoder) V2.0 RC1  ******

        Build: Sep 11 2003
        Copyright: M. Bakker
                   Ahead Software AG
        http://www.audiocoding.com
        Floating point version

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

BigYellow_3.mp4 file info:
Track   Type    Info
1       audio   MPEG-4, 24.517 secs, 136 kbps, 44100 Hz
2       od      Object Descriptors
3       scene   BIFS

  ---------------------
 | Config:  2 Ch       |
  ---------------------
 | Ch |    Position    |
  ---------------------
 | 00 | Left front     |
 | 01 | Right front    |
  ---------------------

Decoding BigYellow_3.mp4 took:  0.58 sec.
input:  BigYellow_4.mp3  (44.1 kHz, 2 channels, MPEG-1 Layer III)
output: BigYellow_4.wav  (16 bit, Microsoft WAVE)
skipping initial 1105 samples (encoder+decoder delay)
Frame#   930/930    128 kbps                 
OggDec 1.0.1
Decoding "BigYellow_5.ogg" to "BigYellow_5.wav"
        [100.0%]
dyld: Library not loaded: /usr/local/lib/libwavpack.0.dylib
  Referenced from: /Users/az/Desktop/ABC-HR/bin_osx/wvunpack
  Reason: image not found
/Users/az/Desktop/ABC-HR/bin_osx/Sample: line 38: 10203 Trace/BPT trap          $wvunpack ${basename}_6.wv
logout
[Process completed]
By copying the content of bin_osx/lib to /usr/local/lib it decodes perfectly fine.
Someone might either fix the script or add fixnotes to the readme under a OSX section.

Multiformat Listening Test @ 128 kbps - OPEN

Reply #158 –
I've done all the samples now, but I'm not so sure if I should submit my results. The thing is, I'm quite sure I've only managed to identify the low anchor in the tests, so all the other sample have got 5 from me. And I think I rated the low sample a little too high in the start. Do you want my results?

Multiformat Listening Test @ 128 kbps - OPEN

Reply #160 –
I've got a problem with playback in the java/abchr.

My setup: Win2000. Installed latest 1.5 Java today.
I have some ESS Maestro on a laptop. Pretty crappy sound setup, I know.

The problem occurs when I select "Primary sound dirver" or "ESS Maestro" in the abchr settings.

1) if I simply click to play something it plays ok, but the sound is muffled.

I get the same effect if I select the "Direct sound" and "Allow hardware mixing" in Foobar. (Foobar says: "Device info: hardware mixing available" but apparently this option doesn't work on my setup). In Foobar I'm using "Direct sound" without hardware mixing, and it sounds fine. Foobar "WaveOut" works too but it skips sometimes.


2) so java/abchr plays muffled, but then if I click to switch A or B or Ref - it also changes the sound - from muffled to what I think is normal sound. Then if I switch (EDIT: I mean I switch during playback) A,B,Ref again it switches back to muffled and so on:
it always starts with muffled, then normal,muffled,normal,muffled,.....


3) There are two more choices available in the abchr settings:
the default "Port ESS Maestro" doesn't produce any sound.
"Java sound engine" sounds normal but much quieter and it skips when switching A-B.


Note: this is not very urgent problem. However I am planning to try the next 48 kbps test. (Edit2: deleted unnecessary remark about my abx-ability in this test)

Multiformat Listening Test @ 128 kbps - OPEN

Reply #161 –
Quote
Quote
hey guys.

just finished setting up all the files needed for the test including installation of java, and i have a problem. the wav's dont play. what am i doing wrong?
reference file doesnt play, ABX option playback doesnt work either. checked settings and my sound card is selected.....
need help if you want some results

thanks all
[a href="index.php?act=findpost&pid=348010"][{POST_SNAPBACK}][/a]


What sound card is selected?
[a href="index.php?act=findpost&pid=349356"][{POST_SNAPBACK}][/a]


my ABCHR selected Bluetooth Audio by default (and my Skype and Gizmo like to choose the soundcard as i/o device  as of late)

Multiformat Listening Test @ 128 kbps - OPEN

Reply #162 –
Mmmm.... This remembers me of an old problem with Soundblasters, with directsound.

If it used primary buffers, it played at 22.5khz. I doubt this is the same problem, but it would be a good start to investigate.

Multiformat Listening Test @ 128 kbps - OPEN

Reply #163 –
Quote
Quote
Quote
Once the results come in, post that, you'd most definately not want the /. masses praticipating since 9/10 they'd just purposely screw up the stats.


Interesting. How would they purposely screw up the stats?
[a href="index.php?act=findpost&pid=349130"][{POST_SNAPBACK}][/a]


Maybe he thinks they would just rate the samples at random?
[a href="index.php?act=findpost&pid=349263"][{POST_SNAPBACK}][/a]


Hmmm what's that quote, never underestimate the power of idiots in large numbers?

Multiformat Listening Test @ 128 kbps - OPEN

Reply #164 –
Quote
Quote
Quote
Quote
Once the results come in, post that, you'd most definately not want the /. masses praticipating since 9/10 they'd just purposely screw up the stats.


Interesting. How would they purposely screw up the stats?
[a href="index.php?act=findpost&pid=349130"][{POST_SNAPBACK}][/a]


Maybe he thinks they would just rate the samples at random?
[a href="index.php?act=findpost&pid=349263"][{POST_SNAPBACK}][/a]


Hmmm what's that quote, never underestimate the power of idiots in large numbers?
[a href="index.php?act=findpost&pid=351397"][{POST_SNAPBACK}][/a]


Yeah, but as I said, chances are good that they are going to rate the reference sample then, too.

Multiformat Listening Test @ 128 kbps - OPEN

Reply #165 –
We are slowly approaching the deadline and I still don't have enough results for several samples.
Considering 10 results per sample are minimum, this is what I ask people to focus on :

Moved to first post: [a href="index.php?act=findpost&pid=347839"][{POST_SNAPBACK}][/a].

Multiformat Listening Test @ 128 kbps - OPEN

Reply #166 –
Sorry, can't contribute at all with the Christmas coming and due to personal commitments elsewhere. I hope others can do better. I'll have to do the test for myself after the Christmas holidays. The results won't be useful for the test at large, but for personal learning it can be interesting.

Multiformat Listening Test @ 128 kbps - OPEN

Reply #167 –
Wel, Gambit already warned you about this earlier, didn't he? Most people's holidays (including mine) start with Christmas, so most people (including me) won't have time to perform this test until after Christmas.

Sorry.
Over thinking, over analyzing separates the body from the mind.

Multiformat Listening Test @ 128 kbps - OPEN

Reply #168 –
Well, I've done all samples 1-11. I'll send them in on the 25th if you don't extend the time. If you do I'll try to do the rest as well.
davidnaylor.org

Multiformat Listening Test @ 128 kbps - OPEN

Reply #169 –
Quote
We are slowly approaching the deadline and I still don't have enough results for several samples.
[a href="index.php?act=findpost&pid=351772"][{POST_SNAPBACK}][/a]

After spending hours listening to the samples, I have to declare my defeat: besides some encodings of sample nr. 3, all other encodings are transparent to me (except for the low anchor of course, even I can distinguish that). No matter how intensely I listen, I seem to get fooled by all modern codecs at 128k - and I can't say I'm unhappy about that at all, to be honest 
However, to make more than one sample contribution, I've recruited my wife to take part in the test. She's the opposite of me WRT listening experience: She's a trained musician who plays several instruments but has no lossy artefacts training at all  She also started with sample 3 and was more successful than me right from the start. I'll try to convince her in the name of public good to try some more samples 
Proverb for Paranoids: "If they can get you asking the wrong questions, they don't have to worry about answers."
-T. Pynchon (Gravity's Rainbow)

Multiformat Listening Test @ 128 kbps - OPEN

Reply #170 –
Quote
We are slowly approaching the deadline and I still don't have enough results for several samples.
Considering 10 results per sample are minimum, this is what I ask people to focus on

i would test all of them. but i think you should extend it till 5 of january.

Multiformat Listening Test @ 128 kbps - OPEN

Reply #171 –
Quote
Well, I've done all samples 1-11. I'll send them in on the 25th if you don't extend the time. If you do I'll try to do the rest as well.
[a href="index.php?act=findpost&pid=351858"][{POST_SNAPBACK}][/a]


It would be cool to have the results ASAP if you don't plan on testing additional samples. That saves me from sorting and evaluating a bunch of results at once - it's easier for me to sort four or five result sets per day.

The problem with extending the test is that I start school on January 6th and we're writing a couple of exams in the first two weeks, so I don't know how much time I am going to have. I will think about it until tomorrow evening.

Multiformat Listening Test @ 128 kbps - OPEN

Reply #172 –
I'm not sure how much a week's extension will help overall, but if your own schedules allow it, please consider it. I can promise to test half of the samples if there's an extension.

One thing to consider.

If you have to sit on the results without analysing for a couple of weeks, because of your own schedules, I think people will understand.

It's probably better to have more data (and higher reliability) with delayed reporting than on-schedule reporting, but lacking in data (and reliability). What do you think?

Multiformat Listening Test @ 128 kbps - OPEN

Reply #173 –
Quote
I'm not sure how much a week's extension will help overall, but if your own schedules allow it, please consider it. I can promise to test half of the samples if there's an extension.

One thing to consider.

If you have to sit on the results without analysing for a couple of weeks, because of your own schedules, I think people will understand.

It's probably better to have more data (and higher reliability) with delayed reporting than on-schedule reporting, but lacking in data (and reliability). What do you think?
[a href="index.php?act=findpost&pid=351949"][{POST_SNAPBACK}][/a]


I am not sure if the lack of results is caused by people having to work or prepare themselves for Christmas / New Year, or simply because the test is hard at this bitrate. If the former applies, extending the test is a good idea. However, if the later applies, I see no point in extending the test just to obtain two or three additional results.

Multiformat Listening Test @ 128 kbps - OPEN

Reply #174 –
OK, let's be more precise...

Who could submit results if the test is extended by one or two weeks? So far, I have Halcyon, kwanbis and sTisTi (or his wife ).

BTW, if you don't plan testing all samples, it would be cool to test the samples for which I have too few results.