Recently I purchased some intensely expensive (to me) Grado cans. I've archived all my discs via EAC directly to FLAC, and have been shopping around for a portable format that suites my (supposedly) superior ears.
I received an iPod for Christmas, and had recently bought an iAudio X5 for myself so I could take Vorbis on the road with me.
So my tests were supposed to help me decide between using Lame 3.97b2 as my format for both platforms, or iTunes AAC for the iPod, or alternatively ditching the pod for the X5 and using AoTuv Vorbis.
I've been majorly torn between platforms and lossy formats.
So today I set up an ABX test using my Grado headphones and my laptop.
I was going to transcode my FLAC source file to each of the above formats at various bitrates and let the best (to my ears) win.
But my test stopped short when I COULD NOT EVEN ABX BETWEEN THE FLAC FILE AND AOTUV VORBIS AT 64kbps!!!!
You have no idea what that did to me. I am so embarrassed to post it here, but on the other hand, I think I should because there are possibly many folks out there who, like me, ASSUME that they have golden ears, when really the truth is less than flattering.
Of course, AoTuv is fabulous. I already knew I loved it, but it still hurts when you're the type that assumes you need the latest LAME at high variable bitrates and then tell yourself that you still prefer the FLAC files. I'm the guy who wouldn't touch a 128kbps AAC file. Wake up call for me I guess.
ROTFL! =D
That's classic dude.
it still hurts when you're the type that assumes you need the latest LAME at high variable bitrates and then tell yourself that you still prefer the FLAC files. I'm the guy who wouldn't touch a 128kbps AAC file.
Hmmm. That's me. I've been meaning to do some ABX testing for ages and ages... I
never have. You've inspired me though. I'm gonna set aside some time this week.
Perhaps we should start a group. Non-ABXers Anonymous or something... hah.
Perhaps we should start a group. Non-ABXers Anonymous or something... hah.
Totally. I always just assumed that the tests were for 'those other people' who weren't yet sure of their awsomely perfect ears.
Well, I just realised yesterday that I have a hard time hearing things above 14 Khz. The only thing I did was listening to some samples of 3 seconds with diferent lowpass filters applied.
Well, I think that means that my ears aren't golden, flac would be be senseless for me in terms of quality but it still offers the possibility to do whatever I want to without any loss acoustical and data.
So that's my reason for using flac at home.
Yeah, I had a similar experience after getting my KSC75s. I'd always told myself that I couldn't ABX low-bitrate mp3s because of my cheap stock mp3 player earbuds. After getting my new headphones, I was surprised to discover that I couldn't ABX lame v7!
I'd like to think it's because of my poor audio equipment/listening environment, but that only counts for so much. I honestly doubt I'd be able to ABX most of my music at v7 under optimal conditions. Maybe it takes practice (which I don't have) to be able to identify audio artifacts.
Or maybe my ears just suck.
dpaint4, what's your laptop's soundcard? A better external soundcard may help, unless the existing one is from m-audio or something.
I think the reason is in large due to the common misconception that audio compression heavily alters the sound. Less dynamics, weaker bass and all those other descriptions "audiophiles" like to throw around, and that in fact are nothing more than just placebo. But in reality, the artifacts are much more subtle, and often require actual training for an inexperienced user to be able to hear them. So when somebody tries an ABX test for the first time, without previous training the results are most of the time surprising.
But I would say that sometimes it's not really a good idea to train for compression artifacts. I guess in this case you could really say that ignorance is a bliss. Enjoy your music and forget about the golden ears.
So my tests were supposed to help me decide between using Lame 3.97b2 as my format for both platforms, or iTunes AAC for the iPod, or alternatively ditching the pod for the X5 and using AoTuv Vorbis.
.......................
But my test stopped short when I COULD NOT EVEN ABX BETWEEN THE FLAC FILE AND AOTUV VORBIS AT 64kbps!!!!
Looks like we have a winner!!
You might consider getting rockbox for that ipod and use vorbis on it too
Don't forget also that lossy codec have greatly improved during the last years.
So today I set up an ABX test using my Grado headphones and my laptop.
Hey man! What if the sound card sucks??? You still have a chance of having golden ears!
Don't feel too bad. I failed miserably in the last 128 kbps test-all were 5.0's to me. This led me to conclude that we probably do not need bother to run that test again-I can only imagine the codecs continuing to improve from this point onwards.
With that said, I still store all my music as flacs, and then lossy encode from those on the fly for whatever purpose. Much easier than digging out cds-and I can encode to any format as needed.
But my test stopped short when I COULD NOT EVEN ABX BETWEEN THE FLAC FILE AND AOTUV VORBIS AT 64kbps!!!!
In Sebastian's 128 kbps test Vorbis was near transparency at -q 4.25. I personally had great difficulties to make any difference with the reference files. With many types of music Vorbis has excellent quality already at -q 1.5 (check this thread out: http://www.hydrogenaudio.org/forums/index....showtopic=39233 (http://www.hydrogenaudio.org/forums/index.php?showtopic=39233)).
However, it is easy to find music samples that make Vorbis suffer at -q 0. For example:
http://rarewares.soniccompression.com/test...les/chanchan.wv (http://rarewares.soniccompression.com/test_samples/chanchan.wv)
http://rarewares.soniccompression.com/test...es/kraftwerk.wv (http://rarewares.soniccompression.com/test_samples/kraftwerk.wv)
http://www.mp3-tech.org/tests/aac_48/samples/sample12.zip (http://www.mp3-tech.org/tests/aac_48/samples/sample12.zip) (Liszt_in_B.flac)
http://www.mp3-tech.org/tests/aac_48/samples/sample13.zip (http://www.mp3-tech.org/tests/aac_48/samples/sample13.zip) (orion_ii.flac)
Encode the files at -q 0, ABX and hopefully you'll get some of your ego back.
Here are my results (aoTuVb4.51, -q0)
foo_abx v1.2 report
foobar2000 v0.8.3
2006/06/13 23:16:41
File A: file://E:\test\Vorb_Q0\chanchan.wv
File B: file://E:\test\Vorb_Q0\chanchan.ogg
23:16:42 : Test started.
23:17:22 : 01/01 50.0%
23:17:39 : 02/02 25.0%
23:17:49 : 03/03 12.5%
23:18:04 : 04/04 6.3%
23:18:12 : 05/05 3.1%
23:18:19 : 06/06 1.6%
23:18:27 : 07/07 0.8%
23:18:45 : 08/08 0.4%
23:18:53 : 09/09 0.2%
23:19:05 : 10/10 0.1%
23:19:08 : Test finished.
----------
Total: 10/10 (0.1%)
foo_abx v1.2 report
foobar2000 v0.8.3
2006/06/13 23:22:46
File A: file://E:\test\Vorb_Q0\kraftwerk.wv
File B: file://E:\test\Vorb_Q0\kraftwerk.ogg
23:22:48 : Test started.
23:23:16 : 01/01 50.0%
23:23:23 : 02/02 25.0%
23:23:31 : 03/03 12.5%
23:23:38 : 04/04 6.3%
23:23:46 : 05/05 3.1%
23:23:56 : 06/06 1.6%
23:24:03 : 07/07 0.8%
23:24:26 : 08/08 0.4%
23:24:35 : 09/09 0.2%
23:24:43 : 10/10 0.1%
23:24:45 : Test finished.
----------
Total: 10/10 (0.1%)
foo_abx v1.2 report
foobar2000 v0.8.3
2006/06/13 23:28:22
File A: file://E:\test\Vorb_Q0\Liszt_in_B.flac
File B: file://E:\test\Vorb_Q0\Liszt_in_B.ogg
23:28:24 : Test started.
23:29:10 : 01/01 50.0%
23:29:20 : 02/02 25.0%
23:29:27 : 03/03 12.5%
23:29:37 : 04/04 6.3%
23:29:49 : 05/05 3.1%
23:29:56 : 06/06 1.6%
23:30:16 : 07/07 0.8%
23:30:24 : 08/08 0.4%
23:30:34 : 09/09 0.2%
23:30:41 : 10/10 0.1%
23:30:43 : Test finished.
----------
Total: 10/10 (0.1%)
foo_abx v1.2 report
foobar2000 v0.8.3
2006/06/13 23:32:29
File A: file://E:\test\Vorb_Q0\orion_ii.flac
File B: file://E:\test\Vorb_Q0\orion_ii.ogg
23:32:31 : Test started.
23:32:54 : 01/01 50.0%
23:32:58 : 02/02 25.0%
23:33:01 : 03/03 12.5%
23:33:05 : 04/04 6.3%
23:33:08 : 05/05 3.1%
23:33:11 : 06/06 1.6%
23:33:15 : 07/07 0.8%
23:33:18 : 08/08 0.4%
23:33:22 : 09/09 0.2%
23:33:28 : 10/10 0.1%
23:33:29 : Test finished.
----------
Total: 10/10 (0.1%)
Exactly, at 128 kbps modern codecs (like newest LAME, Vorbis AoTuV, Nero AAC) are quite good.
(Exception: Harpsichord lover? Ditch lossy! ;-) )
The major giveaway with 128 kbps MP3 for me is the lowpass at ~16 kHz, hard to detect. Removing it brings other artifacts. Of course not everyone can hear that and especially not on all kinds of music.
Easier to hear for some are typical underwatery artifacts, warbling, sometimes drop-outs. (sounds like a pop)
While testing Vorbis listen to very slight high frequency boost, metallic quality of sound, roughness. Also look for stereo field distortions (wrong sound positioning, easier to detect with headphones) at low bitrates.
AAC artifacts similarly to MP3, but of course much less. There, the usual most standing out quality is pre-echo (smoothed out sound, ploppy), especially with HE-AAC - low bitrates. Not really bad. No very audible lowpass anymore. At low bitrates with HE-AAC you should easily be able to detect the difference in high frequencies due to SBR. They sound artificial.
1. "Practice, the master of all things."
2. "Ignorance is bliss."
I did very little execise so far, but when I do I start with the lowest bitrate. I only use vorbis so this is q=-2. I identify the problems and increase the rate until I cannot find any difference. Generally I can ABX up to q=2 or q=3. Sometimes with problem samples I can even ABX q=4. So it is only down to execise I think.
Regardless I decided that q=4 will be enough for me :-)
Triza
Heh, I got a real confidence boost when someone on the vorbis-dev mailinglist told me
Secondly, for someone with golden ears like yourself, I wouldn't consider q5 to be "high bitrate".
after I reported a problem sample (that was two years ago). I rip my CDs to Monkey's Audio nowerdays, and store them on my PC as Vorbis -q6 (I already knew, though, that this was probably overkill for me).
A few weeks ago I tried to determine what bitrates I really need and ABXed a sample at several bitrates encoded with Lame 3.98a3 and AoTuV b4.51. ABXing 64kbps Vorbis was OK, but 80kbps left me guessing... I didn't try that hard, though. With mp3, the limit with this sample was between 112 and 128kbps. Not so golden, I guess, but it's not too bad considering many people used 128kbps mp3 for years with older encoders.
I dont know if I could tell lossy from lossless or anything, but I want as much lossless as I can hold on my hard drive. I dont know, I just like knowing all the data is there.
...I COULD NOT EVEN ABX BETWEEN THE FLAC FILE AND AOTUV VORBIS AT 64kbps!!!!
Welcome to the humbling world of true perception, my brave friend...
(http://66.49.140.133/assets/icon/bow.gif)
I think I'm the winner here, in our last codec test lab session at uni, I could not tell a higher bitrate gsm codec thingy from the original... Though, must admit that massive hangovers and Britney Spears don't exactly help with ABX
i'd get in on this action, but i don't have the patience for foobar and its abx comparator.
anybody know of a gui abx program for dumbasses?
i'd get in on this action, but i don't have the patience for foobar and its abx comparator.
anybody know of a gui abx program for dumbasses?
foobar2000
Though, must admit that massive hangovers and Britney Spears don't exactly help with ABX :D
Thats normal. It cannot sound much worse, even if you encode it lossy.
dpaint4, what's your laptop's soundcard? A better external soundcard may help, unless the existing one is from m-audio or something.
I'm absolutely sure that my laptop has some kind of integrated sound thing. It's seriously not high end. But I don't think I can hide behind that excuse. Certainly my laptop sounds as good as my iPod or my X5. And it's better than my last computer which had awful ambient noise. This one is at least silent when it's supposed to be.
I think the reason is in large due to the common misconception that audio compression heavily alters the sound. Less dynamics, weaker bass and all those other descriptions "audiophiles" like to throw around
My 'favourite' descriptive term is "watery", as in, "all MP3s sound watery". Not only is this a generalisation, but I have no idea what "watery" means.
Don't forget also that lossy codec have greatly improved during the last years.
A fact that is constantly ignored. Some think just because a file has an MP3 extension it must mean the encoder used to create those files was automatically the same, and dates from about 1998.
My 'favourite' descriptive term is "watery", as in, "all MP3s sound watery". Not only is this a generalisation, but I have no idea what "watery" means.
That's funny, I use the word watery to refer to warbling.
That's funny, I use the word watery to refer to warbling.
Woudln't just saying warbling be better? The problem with "watery" is that it means different things to different people, it isn't a description of an artifact.
@ dpaint4, welcome to the real world
I always thought I could distinguish easily between bitrates and quality but when I really tried to ABX a few tracks, I fell flat (not to mention the severely deflated ego)
I realised that my ears are not all that "golden" after all and I depend on the (established) golden ears at HA for any quality comparison. Enjoy the music, friend, and if a lower bitrate is alright for you, that's good 'cos you can take more of that wonderful thing with you!!
audiomars
(He who enjoys his music more now that he does not try to find artifacts in every music sample)
That's funny, I use the word watery to refer to warbling.
Woudln't just saying warbling be better? The problem with "watery" is that it means different things to different people, it isn't a description of an artifact.
It would be, if I the primary language in Mexico was english. A "watery" equivalent in spanish carries more meaning than warbling. At least to common people.
Enjoy the music, friend, and if a lower bitrate is alright for you, that's good 'cos you can take more of that wonderful thing with you!!
My findings have pushed me towards a kind of unexpected decision, and that's that I plan to keep my lossy library in LAME 3.97b2 at V5, which I previously considered to be 'intollerable'. But now that I know my actual limitations I'm going to enjoy them by having this rather comact library in a very compatible format.
Despite how much AoTuv amazed me (it's now my undisputed favorite), I think I prefer to have one lossless and one lossy library. Keeping three librarys is a bit much. And LAME was pretty darn impressive too.
Of course, it does hurt pretty bad that it has taken me a couple years of exploring new codecs and settings, and building up this huge image of what is acceptable compression and what isn't, and subscribing and unsubscribing to Audiophile magazine (they rarely talked about compression or anything much more than analog and compact disc) to arrive at the one codec and setting that most 'ignorant consumers' use without thinking.
But at the end of the day I guess I'm glad!
dpaint4, what's your laptop's soundcard? A better external soundcard may help, unless the existing one is from m-audio or something.
I'm absolutely sure that my laptop has some kind of integrated sound thing. It's seriously not high end. But I don't think I can hide behind that excuse. Certainly my laptop sounds as good as my iPod or my X5. And it's better than my last computer which had awful ambient noise. This one is at least silent when it's supposed to be.
// Take my note only as a personal reflection, not a statement //
I have a hard time abx:ing 128 kbps in both recent mp3 encodings and in vorbis using my laptop integrated Soundmax chip. It gets a lot easier when I use the M-Audio Firewire 410 (both with AKG 240 M headphones). The best description I can give, is that the Soundmax sound is kind of "blurry". So, do this test of yours again using a better soundcard!
I think the reason is in large due to the common misconception that audio compression heavily alters the sound. Less dynamics, weaker bass and all those other descriptions "audiophiles" like to throw around, and that in fact are nothing more than just placebo. But in reality, the artifacts are much more subtle, and often require actual training for an inexperienced user to be able to hear them. So when somebody tries an ABX test for the first time, without previous training the results are most of the time surprising.
But I would say that sometimes it's not really a good idea to train for compression artifacts. I guess in this case you could really say that ignorance is a bliss. Enjoy your music and forget about the golden ears.
Ignorance IS bliss. It used to be that only changes in tonality bothered me - as a classical and jazz musician very familiar with timbres of all types of instruments and ensembles, audio compression easily stuck out in ideal (quiet) listening environments. Then I did some training on pre-echo, stereo, etc. on web sites mentioned here on AH and I regret it because those sometimes now bother me too. My advice? Just enjoy your compressed music as it is and don't worry about ABXing anything.
this topic forced me to do my first abx test:
LAME -V8 was easy but -V7 was already indistinguishable
I have to admit that my equipment is rather poor (integrated SoundMAX chip, cheap Phillips headphones), so i'm still sticking with -V2 for future use
I remember when i did my first encodings with Lame 3.92 at 128kbps CBR i was quite satisfied with the results and never worried about the quality, but mp3s from different sources (downloads, friends) even at high 320 kpbs were (and still are) suspicious, not to mention the possible ripping errors.
However the most annoying artifact was always the missing gap-less encoding
maybe i'll get bladeenc and try to abx it at 128kbps, just to compare a modern and old codec
// Take my note only as a personal reflection, not a statement //
I have a hard time abx:ing 128 kbps in both recent mp3 encodings and in vorbis using my laptop integrated Soundmax chip. It gets a lot easier when I use the M-Audio Firewire 410 (both with AKG 240 M headphones). The best description I can give, is that the Soundmax sound is kind of "blurry". So, do this test of yours again using a better soundcard!
Indeed, some AC97 chips have cheap resamplers built in.
You should test your AC97, Windows resampler (in case it's done in the software) and M-Audio in ABC/HR+ABX.
Some far from perfect resamplers have too low lowpasses.
E.g. Sensaura resampler removes high frequencies. (ABXed vs PPHS foobar2000 resampler on Ultra)
Though, must admit that massive hangovers and Britney Spears don't exactly help with ABX
Thats normal. It cannot sound much worse, even if you encode it lossy.
I have the nagging suspicion that in your case, lossy encoding improves it considerably...
erm,
the truth is different for different listening envirements and people.
Though I am meanwhile in my 30s even going 40 and luckily don't need help yet to walk over a street ,
lossy formats ABXing at 128k vbr target bitrate are still no problem for me,
"even" not with headphones, but with 3 way speakers.
1 fact is, "even" mp3-Lame made huge improvements with time, especially at 128k vbr -V5 --vbr-new.
The "artifact" though, which let find me lossy material easily, in normal music, which had been recorded/mixed/mastered in god quality: highs, attacks.
So, in past, abxing was easier to me, which might be due to my younger age at that time, and probably by the progress of lossy formats. But it's still easily possible in normal music, no "problem samples" !, at 130k vbr.
I add, that I visited loud concerts very rarely, and after my 1st Jethro Tull, I bought ear plugs to protect my ears in concerts.
I listen sometimes to headphones at home, but mostly via good broad HiFi-speaker setup, source since pc days lossless or hq-mpc 220k vbr and higher, digitally fed to amp.
at topic starter:
your laptops 44.1->48k sampling will already destroy a lot, or even more, the cheapo/crappy analogue outputs of your laptop. Your laptop might improve, if you use digital out, maybe let foobar2000 even resample to 48k on your laptop.
I carry out my fine abxings on stereo hifi speakers, cd-player, with burnt cd-r contaning the tracks in different formats/codecs and the lossless/original track to have an A comparison anchor.
Switching inside 1 track helps, where the same track in another format runs with few seconds difference, so that a passage is repeated.
so, you can carry out some A/B and X/Y listening and ABX if needed.
If I listen to headphones, I used always lower volumes than my friends. see my need to wear ear protections at concerts.
btw., my brother and his friend, who visited some concerts eg. guns n roses, have both tinnitus sometimes...
and yes, my brother thinks, it's just my imagination to need more than 128k bitrates to be transparent.
Though, at certain evenings he was able to listen same differences in music at certain effects with my setup.
He has personally a way smaller/simpler hifi setup.
But he brought me into compressed/pc audio, though at his 1st party, where he played only music as mp3 (most 128k cbr stuff, at very early times in the 90s...), I was hurt by that analogue music output..., without abxing... (probably the cheap soundcard together with very crappy mp3s at party volumes via broad stereo setup, excellent amp and good speakers)
So, lossy audio can be satisfying, transparent even for guys with fine ears and good hifi, if bitrate is roughly doubled to contain nearly full spectrum and make no comprises regarding stereo content, ie. >220k vbr, (imo mpc preferred, as it was especially tuned for this, and this will not change, even as it is some months ago )
To put this into relativismn:
I like to listen to 128k vbr lame-mp3 nowadays, in non-hifi environments,
ie. car stereo, or running outdoors with Koss KSC 75 headphones (owning for hifi: AKG K 500)
and: abxing results have nothing to do with ego !
ego = style of music u listen too...
Ego certainly is related....because he perceived his hearing was better than it actually was, therefore deflating his ego.
(imo mpc preferred, as it was especially tuned for this, and this will not change, even as it is some months ago smile.gif )
Yes very good choice for this bitrate but there are no improvements being made either. Have you tried ABX'ing the latest versions of Vorbis, MP3 and MPC at that bitrate range? I'd love to see your results. (I have problems ABX'ing 128 kbps, so I wouldn't be a good candidate to conduct such a test).
Okay.. uhmm.....
I just did a little abx test.
First I tried to ABX between lame -v2 & v3.
I had kinda figured that would be hopeless.
So next I added in a lame v6 file to the mix. Rrrroight. No cookies.
Next step. I threw a FRIGGEN v8 encode of the same track into the mix.
...
I cannot tell any difference between all four. Wow.
Finally I dropped the v2 and added a v9.
Interesting. I can pick out the v9 everytime, but would just be guessing to pick between the v3, v6, and v8. Man.
SOOOOOOOO, conclusions: obviously, this was just ONE track and so.. doesn't necessarily mean.. well, anything. But somehow I don't think it would be premature to join the I've-been-encoding-my-music-at-way-higher-bitrate-than-necessary-for-my-ears club.
Hear, hear.
/me joined the club too. Although this time I use Vorbis.
-q 2 & -q 3 = Hopeless
-q 1 & -q 2 = I thought I heard something. But still no go.
-q 0 & -q 1 = FINALLY I heard something. If I pay extra attention, that is (1st 5 trials easy, but the next trials up to #16 exponentially more difficult)
-q -1 & -q 0 = Rather easy
BUT
Played the -q -1 file in its entirety... very acceptable. (Not to mention still (sliiiightly) smaller than AAC 48 kbps).
Whoa.
I've a feeling that 48 kbps test by Sebastian is the 'just right' bitrate... and many testers will be blown out of the water when rating the quality.
For me personally, I think I'll lower the -q for my mobile devices.
why not lossy vs original/lossless...?
A while ago I tried ABX-ing my mp3's (-V4) with the lossless originals and I seemed to be able to tell them apart about 75% of the time. The funny thing though was that I couldn't really say what I noticed that was different, each time it seemed very like a guess but somehow I got a lot more right than wrong. I still think Lame -V4 sounds good enough to me. I also only want one lossy copy and Lame @ -V4 is a filesize/quality/compatability compromise that works ok for me.
this topic forced me to do my first abx test:
LAME -V8 was easy but -V7 was already indistinguishable
My ears also quite suck. I'm using -V6 with lame and -q0.35 with NeroAAC. But -V7 is for me easy to ABX : indeed V7 is at 32kHz whereas V6 is still at 44kHz, so the difference in high frequencies is quite important.
And I think i'll try to compress a bit more with NeroAAC, because 0.35 is very transparent to me.
And I think i'll try to compress a bit more with NeroAAC, because 0.35 is very transparent to me.
Transparancy is binary.
So, here are my Vorbis lossy-lossless ABX results. I used the first seconds of "Shallow" by Poets of the Fall, an acoustic guitar solo, and aoTuV 4.51.
Q 1: Quite easy, a "warbling" sound and a lowpass (at least I think): 12/12
Q 2: The warbling slowly fades away, lowpass still present: 12/12
Q 3: Warbling nearly non-existent, lowpass seems higher: 13/14
Q 4: No more warbling, lowpass nearly non-existent. Ears getting tired, I stop testing here: 8/8, but then 10/13.
I think that, depending on the source material, for me Q 4 or Q 5 are indistinguishable from the original. Well, I did expect that.
Well, you made me ABX the usual codecs again !
I chose a track that should be hard to encode : Depeche Mode - Some Great Reward - Something to Do (old mastering).
I begain with Lame 3.97b2. I checked the first encoding for clipping and offset, it's all right (I decoded with Foobar2000).
-V5 --vbr-new : 131 kbps, ABX 16/16. Easy (noise)
-V4 --vbr-new : 151 kbps, ABX 16/16. Easy (noise)
-V3 --vbr-new : 163 kbps, ABX 16/16
-V2 --vbr-new : 192 kbps, ABX 16/16. More difficult. The noise is partially masked and stands out as a slight resonance.
Since I thought that I heard a stereo problem (low frequencies seem out of phase), I tried
-V2 --vbr-new -ms : 221 kbps. ABX 16/16
Then I abxed -V2 --vbr-new versus -V2 --vbr-new -ms (the first one gives 17.7 % LR and 82.3 % MS) : 15/16
However, this is not the stereo that made the difference, it's the fact that the V2 -ms encoding sounds as bad as the -V3 one (http://perso.numericable.fr/laguill2/smileys/lol.gif)
Next, I will have to try Musepack and Vorbis (http://perso.numericable.fr/laguill2/smileys/ouf.gif)
EDIT : oops ! Sorry, I scored only 15/16 ABXing the -ms option, not 16/16 !
Good lord, Pio......
No, no, no, it's just a killer track ! Or I think so...
Case in point, in the multiformat 128 kbps test of Sebastian Mares, for Lame MP3, I could only ABX 3 of the 8 samples that I tried !
I totally agree in the ignorance is bliss thing. It took me a long time to abx even a few of the -abr 128 files I encoded.
But I did learn to pick some out so I now resort to overkill to help prevent some future revelation forcing me to re-encode my cds.
Thats why I use -V1 --vbr new --lowpass 20. Hard to ever say that these are bad files.
Q 4: No more warbling, lowpass nearly non-existent. Ears getting tired, I stop testing here: 8/8, but then 10/13.
Just a heads up, but when you ABX, you need to decide how many trials you'll do in advance. If you stop when you're ahead, you defeat the purpose of the test.
I abx'd one 320kbps lossy MP3 (more chances of mistake with Sennheiser HD650/Meier Audio Corda Ha-1Mk2/Audigy 2. I wanted to hear if there really are differences between Audigy's and software resampling. I converted the same mp3 to two different wav's, one with 48KHz Secret Rabbit Code converter (best sinc interpolation), the other without any dsp's. I used normal Directsound output plugin with 16bit output, without any dsp's.
I got 7/7. I guess that's enough, altough some page suggested 16 is considered as a good compromise... well, it wasn't a scientific test anyways, and if I really just guessed them all right accidentally, I might as well go to fill some lotto coupons this week...
sinc interpolation ? Given the awful results that it gives for picture, I don't think that this is the best way to resample ! It causes a tremendous amount of ringing at the cutoff frequency. 22050 Hz is usually inaudible, but in case of non-linear behaviour from an element of the hifi system, it is safer to use a more classic filter.
The standard filters for CD, for example, go from 0 dB at 20 kHz to -140 dB at 22049 Hz (not sure about the -140, but it is something well below -96 dB). Sinc go from 0 dB at 22049 Hz to minus infinite at 22050 Hz.
Just a heads up, but when you ABX, you need to decide how many trials you'll do in advance. If you stop when you're ahead, you defeat the purpose of the test.
That really depends on the purpose of your test. If the purpose is to create statistically based scientific evidence, then you are correct. You cannot prove anything by using the "continue until you succeed" strategy. That is similar to doubling your bet at the casino until you win -- while it works in theory (but not in practice -- even Bill Gates isn't rich enough to keep doubling indefinitely) that fact does not prove that the odds on an individual bet are in your favor.
However, an ABX tool can also be used to gauge for yourself whether you can hear the difference between two tracks. In that case, you are trusting your own judgement about what you can hear, and the cumulative results are simply one more piece of information you can take into account. In that case, you are
not performing a true ABX experiment. You are simply using an ABX tool to assist yourself in comparing tracks.
In fact, I was doing just that last night. I wanted to determine what quality level to use with the Nero aac encoder. What I found was somewhat surprising, and right in line with what the original poster discovered. With four tracks picked out of my collection ("The World Turned Upside Down" by Karan Casey, "The Sunbed Song" by Level 42, and a Prelude & Fugue from the Well-Tempered Clavier performed by Ton Koopman on harpsichord) I was unable to distinguish from the original at very low quality levels.
At LC q=0.1 the difference was obvious. The high frequencies were muted in the compressed track.
At LC q=0.2 I just couldn't tell when listening to the whole track and flipping from one to the other.
Then I paid a visit to www.pcabx.com and downloaded some of Arny Kruger's test samples which are designed to make life difficult for perceptual encoders. Castanets didn't do it. I still couldn't tell LC q=0.2 apart from the original. But then I tried the nc-2 "harmonic test tone" sample that's designed specifically to break encoders, and could tell the difference at LC q=0.2 easily. LC q=0.3 was harder, but I could still tell the difference. Next I'll have to try LC q=0.4. The interesting thing I noticed was that the average bitrate for the nc-2 compressed sample was about half the average bitrate for the music tracks. So apparently, the encoder's method of assessing the required bitrate doesn't think that sample needs a high bitrate.
I am not sure, but I think that the Level 42 track that I used was the same one that provoked LAME's CBR 128kbps encoder into very easily heard pre-echo when I tried it a year ago. So I am very impressed with the Nero compression, since it isn't nearly as easy to distinguish between Nero at around 85 kbps average and the original.
The question is, If music doesn't display any obvious artifacts, then do I really need to increase the quality parameter until the specially engineered test tone is indistinguishable from the original? I know, that's a personal decision, but one that a lot of us have to make. It sure would be nice if I could use a higher level of compression and fit my entire collection onto my portable player.
I've got a lot of listening to do -- maybe I can find music tracks that are as critical as Arny's test tone. Or maybe I can't. Perhaps the easiest & laziest plan would be to encode everything at LC q=0.2 and wait to see if I hear something that's unacceptable.
sinc interpolation ? Given the awful results that it gives for picture, I don't think that this is the best way to resample ! It causes a tremendous amount of ringing at the cutoff frequency. 22050 Hz is usually inaudible, but in case of non-linear behaviour from an element of the hifi system, it is safer to use a more classic filter.
The standard filters for CD, for example, go from 0 dB at 20 kHz to -140 dB at 22049 Hz (not sure about the -140, but it is something well below -96 dB). Sinc go from 0 dB at 22049 Hz to minus infinite at 22050 Hz.
I don't know anything about resampling Maybe I should read about different resamplers (PPSH, SSRC, SRC) and their settings.. from somewhere.
Just a heads up, but when you ABX, you need to decide how many trials you'll do in advance. If you stop when you're ahead, you defeat the purpose of the test.
That really depends on the purpose of your test. If the purpose is to create statistically based scientific evidence, then you are correct. You cannot prove anything by using the "continue until you succeed" strategy. That is similar to doubling your bet at the casino until you win -- while it works in theory (but not in practice -- even Bill Gates isn't rich enough to keep doubling indefinitely) that fact does not prove that the odds on an individual bet are in your favor.
However, an ABX tool can also be used to gauge for yourself whether you can hear the difference between two tracks. In that case, you are trusting your own judgement about what you can hear, and the cumulative results are simply one more piece of information you can take into account. In that case, you are not performing a true ABX experiment. You are simply using an ABX tool to assist yourself in comparing tracks.
Because he posted the results, I think its obvious the purpose of his test was to see if he could ABX it. You're right, if you don't care about the results, you can do whatever you like, but thats not really what I was talking about.
Just a heads up, but when you ABX, you need to decide how many trials you'll do in advance. If you stop when you're ahead, you defeat the purpose of the test.
I know that stopping at 8/8 doesn't prove I "passed" the test, but I wanted to point out that I
didn't pass, and for that purpose I thought 10/13 were sufficient.
Anyway - since I still have the encodings on my HD, I did a second test today, going to 16 no-matter-what-happens, and it ended 11/16. No surprise for me.
Though I have to confess that considering the 10/13 result, the comments I made about the warbling and lowpass in the same line are obsolete, as I couldn't distinguish the Q 4 sample anymore. Sorry about misleading you.
Well this thread inspired me to test myself... I successfully ABX'd fatboy, castanets and the harpsicord samples using Lame and Nero AAC thusly:
Harpsicord:
Lame -V2 --vbr-new
Nero -q 0.4
Fatboy:
Lame -V7 --vbr-new
Nero -q 0.6
Castanets:
Lame -V9 --vbr-new
Nero -q 0.3
First thing to note is that on Harpsicord, and no previous training, I managed to defeat --preset fast standard! That was suprising, actually (figured I would be a poser).
On fatboy I defeated Nero's standard profile (0.5), and the one above it as well. Lame shut me down hard, which is surprising given MP3's supposedly inferior low-bitrate encoding.
Castanets I just had trouble with... I know what to listen for, but it just gets transparent too easily on both encoders.
This was fun, and I'll probably do it again some time! For now, it looks like Nero -q 0.5 should be satisfactory for my iPod use (I don't listen to a lot of Fatboy ).
My ABX tests were the one reason I started ripping to -V 4 --vbr-new, Nero AAC at -q 0.35, and iTunes AAC at 128kbps VBR (depending on what I feal like). I conducted a ABX test with my equipment and I could somewhat spot -V 5 --vbr-new (it was hit or miss with some samples but nothing to get in a bunch over).
I stopped fooling myself and now have a lower bitrate library. Funny how ABX testing can do this.
I tried the same Depeche Mode track with Vorbis (aoTuV b4.51) and Musepack (1.15v).
I managed to ABX successfully musepack thumb preset (86 kbps) and radio preset (126 kbps), and for the first time of my life, Vorbis at q4 ! (126 kbps)
Funny how codecs sounds. Mp3 sounds Skrotchgloosploof, Musepack sounds kwishwikwilwi, and Vorbis just shhhh !
----snip----
Funny how codecs sounds. Mp3 sounds Skrotchgloosploof, Musepack sounds kwishwikwilwi, and Vorbis just shhhh !
eh?
My experience with ABX-ing (and listening tests in general for that matter) is that the equipment, the music, and the environment make all the difference.
In a quite room on a pair of studio monitors, I could easily ABX .wav from .mp3 at all levels (classical piano sample... I could tell anything below -V2 when it was rock music, -V0 and -V1 were transparent to my ears).
On my laptop, in a coffee house, using Sennheiser PX100's (which how I listend to music 75-80% of the time) I couldn't ABX the same .wav from -V5 or 128 AAC VBR (iTunes).
As has been mentioned before and I'm sure above... know where, how and what you listen to and encode accordingly.
Thanks to my Sennheiser PX100's I can ABX Lame -V5, Vorbis -q4 and iTunes 128kbps VBR, from the original file
I still use -q4 Vorbis for my iPod, I'll compromise a tiny loss in quality so I can fit more music on it
In general I tend to not believe claims of successful abx tests unless actual logs are posted -- exceptions being for known killer samples, bitrates lower than 128, or from members well-known for their abxing contributions.
----snip----
Funny how codecs sounds. Mp3 sounds Skrotchgloosploof, Musepack sounds kwishwikwilwi, and Vorbis just shhhh !
eh?
That's what I can hear and ABX in the compressed files versus the original ones. There are some noises in them.
This thread and the latest multi-format test has caused me to change my library to 128 itunes vbr. I cannot tell the difference between lossless and 128 vbr aac itunes. I was born with a slight deafness. I'm not hard of hearing but I can't hear certain frequencies. I remember taking audio tests in school and I'd be embarrassed because I wouldn't hear certain tones. Sometimes I'd just press the button in certain intervals just in case.
I'm kind of bummed I can't tell the difference because I feel like I'm missing out on something as great as lossless audio. I have AKG K701s with a Total BitHead amp and that's helped me bring me closer to experiencing better audio but to me 128 kbps vbr sounds really good and I can hear "all" the details just the same as the lossless.
Maybe my ear is untrained but I'm guessing its mostly my slight deafness.
I'm kind of bummed I can't tell the difference because I feel like I'm missing out on something as great as lossless audio. I have AKG K701s with a Total BitHead amp and that's helped me bring me closer to experiencing better audio but to me 128 kbps vbr sounds really good and I can hear "all" the details just the same as the lossless.
If you are enjoying your music at 128 itunes vbr (i'm guessing you use the AAC encoder) then don't be bummed, just enjoy it. It's better than obsessing on what kind of encoding you should be using.
Maybe i'm listening to the wrong kind of music to hear more things. Can anyone suggest something more complex? I listen to a lot of rock and punk and I guess things get drowned out in the distortion of the guitars.
I guess I'm not obsessing over what encoding to use, I guess I'm just worried that I'm not hearing what I'm suppose to hear.
Maybe i'm listening to the wrong kind of music to hear more things. Can anyone suggest something more complex? I listen to a lot of rock and punk and I guess things get drowned out in the distortion of the guitars.
I guess I'm not obsessing over what encoding to use, I guess I'm just worried that I'm not hearing what I'm suppose to hear.
Trumpets. Or harpsichords. Both are killers.
sinc interpolation ? Given the awful results that it gives for picture, I don't think that this is the best way to resample ! It causes a tremendous amount of ringing at the cutoff frequency. 22050 Hz is usually inaudible, but in case of non-linear behaviour from an element of the hifi system, it is safer to use a more classic filter.
The standard filters for CD, for example, go from 0 dB at 20 kHz to -140 dB at 22049 Hz (not sure about the -140, but it is something well below -96 dB). Sinc go from 0 dB at 22049 Hz to minus infinite at 22050 Hz.
Sorry for answering to such an old post, and OT as well :-)
I'm currently reading up on a few audio technology topics to better understand what's going on (and maybe to contribute to some audio project later). So I'm asking just to make sure I understood correctly. Shouldn't sinc interpolation give (ideally) perfect results when upsampling? As I understood it, when you only add the sinc values up to a certain distance from the sample you're calculating, you get inaccuracies of course, but these should only be noticable when there are frequencies close to half the sampling rate of the original signal, since the contribution from lower frequencies mostly cancels out.
Did I misunderstand something there?
The sinc interpolation gives a perfect brickwall. Try it at 11 kHz if you can and see by yourself. It introduces a lot of ringing. A perfect brickwall is usually not wanted.
The sinc interpolation gives a perfect brickwall. Try it at 11 kHz if you can and see by yourself. It introduces a lot of ringing. A perfect brickwall is usually not wanted.
What do you mean by "try it at 11 khz"? Upsampling from 11 khz? upsampling to 11 khz? I was only referring to upsampling in my last post, of course downsampling without using a lowpass before causes aliasing.
The sinc is not an upsampling or downsampling algorithm. It is a lowpass filter. Also called antialias filter when used together with upsampling or downsampling. I suggested to remove frequencies above 11 kHz using a sinc lowpass.
In ff123's sample page (http://www.ff123.net/samples.html), the Mustang samples, aimed at testing low-pass audibility, used to be filtered with sinc. I complained that the ringing was so obvious that it could be ABXed separately from the low pass itself. He then replaced the samples by new ones, filtered more progressively. They now sound as lacking treble, with no audible artifacts.
The sinc is not an upsampling or downsampling algorithm. It is a lowpass filter. Also called antialias filter when used together with upsampling or downsampling. I suggested to remove frequencies above 11 kHz using a sinc lowpass.
In ff123's sample page (http://www.ff123.net/samples.html), the Mustang samples, aimed at testing low-pass audibility, used to be filtered with sinc. I complained that the ringing was so obvious that it could be ABXed separately from the low pass itself. He then replaced the samples by new ones, filtered more progressively. They now sound as lacking treble, with no audible artifacts.
Thanks for your replies, I learn a lot by being contradicted. I did not know the sinc function could also be used for lowpass filtering. My understanding is the following: When you sample a function, you can reconstruct the original perfectly as long as the function is sampled with more than twice its highest frequency component (Nyquist's sampling theorem). I read that the way to do this reconstruction was by using sinc interpolation (Whittaker-Shannon interpolation formula (http://en.wikipedia.org/wiki/Whittaker%E2%80%93Shannon_interpolation_formula)).
So if you sample this reconstructed function with a higher rate, you should have a perfectly upsampled signal.
Where does the ringing come in?
Where does the ringing come in?
It's in the time domain. A sinc filter will transform an impulse (think one sample at 1 surrounded by 0:s) into a sinc shape. So you will spread the energy from being infinitely narrow to be infinitely wide - sure, after the impulse the ringing is probably masked, but you will have something similar to pre-echo before the impulse which can be easily heard.
Where does the ringing come in?
It's in the time domain. A sinc filter will transform an impulse (think one sample at 1 surrounded by 0:s) into a sinc shape. So you will spread the energy from being infinitely narrow to be infinitely wide - sure, after the impulse the ringing is probably masked, but you will have something similar to pre-echo before the impulse which can be easily heard.
I see your point. A single pulse like you describe would, after upsampling, be heard as a signal of half the original sampling frequency. But I still think that this interpolation is correct. An impulse of no width is not a bandlimited signal. I'm always assuming that the original signal has been filtered before recording to exclude any frequency above half the sampling frequency. If the sampled signal then looks like 000000010000000, then the filtered input signal did have a sinc shape.
edit: Maybe sinc gives bad results because actual recording equipment cannot lowpass an input signal ideally?
So if you sample this reconstructed function with a higher rate, you should have a perfectly upsampled signal.
Where does the ringing come in?
It comes from the resctrictions needed to apply the sampling theorem : a given sample rate allows you to reproduce frequencies inferior to half the sample rate. In real life, analog signals are not filtered in such a mathematical way. They have a small amount of energy outside their main frequency range.
When you deal with digital, you assume that your signal has a strictly null amplitude above half the sampling frequency. That's your "perfect signal". It is not the original one. It is the original one minus the peanuts above half the sample rate. Taking back these peanuts introduces ringing.
If you work at 44100 Hz, and your signal have significant content near 22050 Hz, you define a "perfect signal" as having a brutal discontinuity in frequency response. There is something near 21 kHz, then, absolute zero at 22050 Hz exactly. This discontinuity in the frequency domain equals a lot of 22050 Hz oscillations in the time domain.
They are not a problem, since 22.05 kHz is usually an inaudible frequency. But if you do the same at 11 kHz, then you will get a disturbing 11 kHz resonance in addition to your music. This is not an inaccuracy or imperfection of the process. This is a signal with a brutal discontinuity in frequency response. This is how it is supposed to sound.
Cutting frequencies above a given threshold is required for the sampling theorem to work. In practice, we don't want to cut them suddenly, the resulting signal is not very interesting. We prefer recording signals with softly limited frequency response.
Given the excellent results of Vorbis in recent tests, and given that Vorbis was the only codec that I could not ABX at all in Sebastian Mares 128 kbps multiformat test, I consider switching from Musepack to Vorbis.
I have just tested Vorbis 2.83 aotuv beta 4.51 -q5 with my collection of killer samples. Here are the results :
Amnesia : transparent
Astral : transparent
Drone : transparent
Fsol : transparent
Ravebase : transparent
Rebecca : transparent (not a killer sample)
Short : transparent
Spahm : ABX 16/16. Not very good.
Transwave : ABX 7/8, then I decided to go on to 24. Total 23/24. Some slight artifacts.
Vilbel : 16/16, awful !
I tried vilbel at -q6, same thing. ABX 16/16, awful.
At q7, it is transparent for me.
I'm currently trying Transwave and Spahm at -q6, but its quite hard.
I have some serious hearing problems.. From birth to age 10 or so, I had tubes in my ears, was partially deaf, and they'd always puss... Disgusting..
I'm 20 now, and I still have to turn the TV way up to hear it.. The interesting thing is that I can pick out artifacts very easily. I had no formal training with ABX testing, had some cheap $20 sony headphones, and ABX'd 128kbps AAC+ 10/10 (I think, I'll post the logs if need be... I have them saved on my server.. And it wasn't a formal test with perticular encoders, it was some quick test using Winamp's encoder... This isn't the point I'm trying to make though).. The point, however, is that audio codecs are subjective. Every encoder has its own method of hiding its shortcomings. So you think your ears suck? Honestly, it's highly doubtful. Lots of people can be trained to hear whats wrong with audio encoding schemes.. Some people are pickier than others.
Audio encoders are perceptual (obviously). Most people, when they are looking for flaws, they'll find them. If you know what to look for, it's a lot easier than blindly looking for problems. While I don't have golden ears, I'm good at finding something and grabbing onto it, which is the only reason I believe I passed the ABX test I mentioned above so well. My hearing sucks horribly, probably worse than most people here. However, I believe that audible tests are more mental than physical, and it can be proven by my TV volume that almost always sits at 20 (out of 34), and my constant "What??" over and overs.
When people tell you you need training to ABX, they are absolutely right. I believe almost ANYBODY can ABX something if they know what to look for. While I didn't have formal training, I did a lot of experimenting to figure out what to look for. My brain tells me whats wrong, not my ears. If you can't do 64kbps, then do some experimenting (actually, you know what? Don't. It'll ruin music for you... Ignorance is bliss when you can fit 2x-3x more music than someone who knows what to look for... That's one thing I miss... When downloading a 64kbps mp3 simply meant 'faster download' to me, I was much happier)
There is an upside and a downside to everything. Golden ears are good when I want to help my friend (he writes a media player) test new psychoacoustic tunings in the decoder, I'm happy that I know what I'm doing to some extent. Do I wish that I was one of the people who had no problem downloading 96kbps mp3 files? Yeah, sometimes I do.
I'm glad that his media player got good reviews for sound quality, but I'm mad that I can't fit 20,000 songs instead of 8,000 on my harddrive.
To end this long post, if you want to ABX, you need to be aware of the fact that although it may be fun for a second to brag about how you can hear certain things that others can't, you can say goodbye to being blissfully ignorant of the little things, like being able to play a song without having to 'approve' it before it goes into your media library. I lurk HA all the time, and the original poster struck a chord with me... Some may think its great to hear a bunch of things in audio, but I can honestly say I got bored with it very quickly. I'll go back to lurking now, since I'm apparently in a 'downer' mood right now... But I did feel that I needed to state my true opinions on sound quality... And when I get my setup fixed (damaged due to flooding last week), I'll probably be happy again... I guess I'm just mad that I didn't make lossless backups of my music that I lost in the flood.. When I post something else, pretend this post didn't happen. Thanks.
However, this is not the stereo that made the difference, it's the fact that the V2 -ms encoding sounds as bad as the -V3 one (http://perso.numericable.fr/laguill2/smileys/lol.gif)
wow...that's a difference almost 60kbps that was saved using Joint Stereo
When people tell you you need training to ABX, they are absolutely right. I believe almost ANYBODY can ABX something if they know what to look for. While I didn't have formal training, I did a lot of experimenting to figure out what to look for. My brain tells me whats wrong, not my ears. If you can't do 64kbps, then do some experimenting (actually, you know what? Don't. It'll ruin music for you... Ignorance is bliss when you can fit 2x-3x more music than someone who knows what to look for... That's one thing I miss... When downloading a 64kbps mp3 simply meant 'faster download' to me, I was much happier)
Welcome in our strange world, where searching for the best quality cause the perceived quality to actually decrease!
That is why many people, while still interested in lossy audio, are starting to use lossless encoding.
Yeah, I'm pretty much the same in that regard. I still encode mostly to q4-q6 vorbis out of habit (and because I lack in harddrive space), but every encode feels like a waste to me. I still love playing with lossy encoders, even if they make me hate what I listen to.. I can't tell you how many times I encoded "Blackout" by (hed) pe and listened to it in the quest for finding the magical settings at each bitrate. It's still extremely interesting to me, but at the same time, I still wish I was fine with 128kbps mp3s.. Even ones not encoded by your lovely encoder (which makes mp3s tolerable, so thank you!).. I'm talking early fhg encodes with no joint stereo, the ones that warble and hiss, with nasty preecho and dull highs... I still find some of my early files and wonder what was wrong with me for getting those songs, but neglect to remember that I used to not notice and/or care about it.
Not to discourage any people who want to be good at ABX testing, but there is a downside to it that I'm not so sure is worth it yet... At least until I get more harddrive space... Then I wont care.
my pleasure to meet you all.(newby here)
@ Gabriel
That is why many people, while still interested in lossy audio, are starting to use lossless encoding.
i'm AC-3(lossy) user because i use dvdplayer(DD 5.1 decoder built in) with 6 discrete amplifiers.(no HT or receiver/decoder)
i think(not sure) that don't need ABX for lossless formats and of course i want to use.(flac for example)
(maybe don't will run in my system but) what lossless format can be used for dvds or mini-dvds?
thanks.
@ token
i read the whole thread and your posts/argumments.
very cool.
regards all.
what lossless format can be used for dvds or mini-dvds?
PCM.
all right but if i use 6 discrete amplifiers(2*3RCA cables) like i wrote in my last post will work?
(i think that i only hear 2 channels...hard doubt )
thanks!
6ch PCM on DVDV is theoretically possible, however you will have trouble finding the right authoring procedure (i dont know one) and also trouble finding a standalone player that will actually play that....
you're right.
i only can do is 5.1 wav 44100 16 bit and burn as cda in nero( a trick).
the cd play in the HT of my friend very fine using digital output or coaxial but,poor me,in my discrete channels i only have "stereo".
i think that "no way out" with my audio system,for now i have to stay with AC-3.
but thanks so much (and sorry to be off topic here,cool thread).
Welcome to "The Club", dpaint4
I guess this is the thread to come clean for us tin-ear types. I recently started using EAC/Foobar instead of iTunes to store and listen to my music on the computer. The biggest reason for the change was to be able to try some true ABX tests, which I'd always been interested in.
At first I couldn't tell the difference in anything which I blamed on the AC97 or whatever it's called on my computer. So I installed a E-MU 0404 and rigged it to run on ASIO from Foobar with no resampling and what have you. Straight from the 0404 into my Portaphile and HD595. Nice sounding setup.
I still can't tell the difference in anything. I ripped my whole CD library to WavPack files on the computer and then encoded AAC (using Nero) at -q .55 for around 170-200kpbs typical rate. No need to ABX between WavPack and 190kpbs AAC because I can't hear a thing different between A and B. So I went as low as 82kpbs before I thought I heard something. However, I still ended up less than 50% correct on ABXing various samples of my favorite music.
Last night I finally found a distinguishable point. I encoded a few tracks at (IIRC) -q 0.25 or something silly like that, resulting in some approximately 60kbps lossy files. I haven't done any "real" tests yet but on a couple of tracks I tried I can get my first two or three trials in a row correct. So my basic conclusion is that Nero AAC is transparent w.r.t. WavPack down to at least 100kpbs and probably lower. And that's listening with my best gear. Most of the time I use an unamped PX200 straight from the iPod or listen in the car with an iPod and cassette adapter. Ye Gods!
I'm tempted to re-encoded my whole library at -q .40 and save a third of the space on my iPod.
LOL, congrats Brent, for your ABX experience!
The problem with training yourself for ABX, as token had said, is that once you know what to listen to, you'll start hearing artifacts everywhere.
I train myself to hear artifacts, and untrain myself after ABX. How? Basically it's kind of a self-hypnosis thingy, so I can't recommend the right way; y'all have to find how yourself heh heh
Happy un-training, all!
pepoluan,
I spent way too many afternoons hanging around a high-end stereo shop while playing hooky from school and/or work back in the early 80's. After a while, I got so good and picky that everything sounded bad!
Unfortunately, I never learned the self-hypnosis thing. I just had to spend 20 years listening to my Ford car radio and avoiding decent audio equipment. But now I can happily jam along with my mid-fi equipment and think I'm in heaven.
So I have no, no, no interest in learning how to ABX better. I'm quitting while I'm ahead.
dpaint4, what's your laptop's soundcard? A better external soundcard may help, unless the existing one is from m-audio or something.
I'm absolutely sure that my laptop has some kind of integrated sound thing. It's seriously not high end. But I don't think I can hide behind that excuse. Certainly my laptop sounds as good as my iPod or my X5. And it's better than my last computer which had awful ambient noise. This one is at least silent when it's supposed to be.
I was doing some listening of FLAC on my laptop this weekend and was suitably impressed with the sound (I was using some new headphones I bought). Then, I plugged in my external sound blaster "card" (USB) rather than using my internal soundcard and the sound difference was HUGE.
So perhaps you can hide a little bit behind that excuse - the internal sound cards in laptops can be horrible.
Unfortunately, I never learned the self-hypnosis thing. I just had to spend 20 years listening to my Ford car radio and avoiding decent audio equipment. But now I can happily jam along with my mid-fi equipment and think I'm in heaven.
Basically, unlearning is like that. You play the lower-quality audio,
forcing yourself to
believe that there is no difference with the higher-quality audio (but
do not play the higher-quality audio just to compare). It is imperative that you do at least 1 or 2 rounds of unlearning
before you sleep or nap. Your brain will somehow believe what you want it to believe when you sleep. If you failed to unlearn before you rest, then your brain will reprogram itself to always recognize the artifact(s) next time it hears an audio. That will make unlearning a royal PITA, like you experienced.
So I have no, no, no interest in learning how to ABX better. I'm quitting while I'm ahead.
Wise of you so, if you really want the smallest file, just transcode it to the smallest you
guess possible, and listen
without ABX-ing. If you find it inacceptable, raise it by an arbitrary amount.
I once tried transcoding with ABX-ing to find the 'perfect' low bitrate... and have to live in great disappointment for 2 weeks... while busily unlearning
Now I transcode to Vorbis -q -1 ... and not bother to ABX, just mumbling to myself, "Oh goshwow, this is perfect..."
Note that I'm not claiming, not even opinionating, that -q -1 is transparent, mind you
I just try not to care too much, and concentrate on the music rather than the sound .
Edit: That's probably easier for me because I don't have great equipment anyway... but sometimes when (at a party for example) the quality is bad, i just try to ignore the fact and enjoy the music anyway.
Now I transcode to Vorbis -q -1 ... and not bother to ABX, just mumbling to myself, "Oh goshwow, this is perfect..."
I'm with you on this, pepoluan.
If it was WMA from a few years ago, I would certainly not be making such a "goshwow" proclamation (the metallic smearing would be unmistakeable) but with the modern incarnation of Vorbis at the same bitrate, the most noticeable artifact (distortions in stereo separation) is simply undetectable without ABX'ing or side-to-side comparison at the very least.
(snip)
You have no idea what that did to me. I am so embarrassed to post it here, but on the other hand, I think I should because there are possibly many folks out there who, like me, ASSUME that they have golden ears, when really the truth is less than flattering.
Of course, AoTuv is fabulous. I already knew I loved it, but it still hurts when you're the type that assumes you need the latest LAME at high variable bitrates and then tell yourself that you still prefer the FLAC files. I'm the guy who wouldn't touch a 128kbps AAC file. Wake up call for me I guess.
Yeah, I have to add my $.02 to this thread... Hello, I'm Michael and I'm a recovering audiophile (Hi, Michael).
A few months back I encoded everything I had on CD or in FLAC format to lame 3.97b2 -V 1 --vbr-new, because I wanted no compromises in my music and never wanted to have to re-rip/re-encode, and because I just *knew* I could tell the difference between -V 1 and -V 2. God *forbid* I should try -V 3 or even worse, -V 4 or 5.
Well... I wasn't able to fit as much of my music on my 60GB iPod as I'd like. So now that iPods are gapless I started using iTunes (and stopped using Rockbox on it), I decided to see if perhaps I could save space using AAC instead of lame.
I decided to finally install the ABX plugin on foobar and tried a few samples using my Shure e2c's comparing a FLAC source and an iTunes AAC encode of the same file @ 128 VBR. I got 10/16 -- and honestly... half of those correct answers I know for a fact were lucky guesses.
So I've got some re-encoding to do... Now I have to convince my brother, who says he can *only* listen to ALAC on his girlfriend's iPod because he simply *can't* listen to compressed stuff of *any* bitrate, because he'd just *know* it was compromised.
I too have concluded that I can get no useful information from most ABX testing unless the circumstances are dire. Can one "train" oneself to hear certain items? Perhaps, though I doubt that most of HA's golden ears wound stand up to external scrutiny, just like the rest. I did too much academic work in acoustics to believe otherwise.
Suffice to say, the differences between well encoded 128kbps AAC, LAME -V 3 and FLAC (as examples) are all very, very small when compared with more glaring variations such as loudspeakers, headphones, and room acoustics. Tiny. Infinitesimal.
The broader point is: does it matter? What is one trying to achieve? What exactly are you listening for? Should I worry about a paint chip on my car when the bigger problem is that the engine is blown?
Given that stereo recordings never, ever sound convincingly like the source (for myriad reasons), what is the goal of "perfect" playback?
If your goal is to feel some deep inner satisfaction that you are listening to the "most perfect copy" of some music, then by all means knock yourself out. Fix that paint chip.
If you wish to concentrate on the emotional and artistic merits of music, then it may not matter one whit. Or more importantly, there may be other things that can serve this goal that have nothing to do with bitrates and compression. Things like decent speakers and headphones, both of which exhibit high degress of variability.
I am much older than most HA posters - 48, an ex-professional musician, and my ears are certainly used - and I recall many years of vinyl use. Good lord, the things we would do to make that miserable medium sound decent - custom tonearms, endless cleaners, weighted platters, turntables suspended from the ceiling, preamps galore - there was no end to the effort. And you know what? Even the most mild artifacts from LPs are 10 times worse that those exhibited by MP3/AAC. I mean bad, non-musical, annoying stuff - crackles, warps, inner-groove distortion, pre-echo. And you couldn't rip another copy to make it better, either. When the CD matured in the late 1980s, I was thrilled at the improvement over vinyl - it is simply a better medium for accurate data.
So given this life experience, I will listen happily to 128kbps AAC all day long. I used to hear defects in my LPs. Now I hear music.
So given this life experience, I will listen happily to 128kbps AAC all day long. I used to hear defects in my LPs. Now I hear music.
*high fives*
Okay, I'm not anywhere as old as you but I dig your attitude
Really, folks. If you want to know whether a certain encoding strategy (i.e. format, compression level) is good for you, ... just encode and enjoy.
Hmm... somewhere in the forums I posted my methodology of finding the highest level of compression with
acceptable result...
To summarize the methodology: Just compress. And enjoy. If it is not enjoyable, compress at slightly lower compression. And enjoy. Keep doing it (track-by-track) until all tracks you have is enjoyable.
Never even try to think about planning to compare your supercompressed tracks with the original. Now, that's the key for DigitalAudio Nirvana
So given this life experience, I will listen happily to 128kbps AAC all day long. I used to hear defects in my LPs. Now I hear music.
*high fives*
Okay, I'm not anywhere as old as you but I dig your attitude
Really, folks. If you want to know whether a certain encoding strategy (i.e. format, compression level) is good for you, ... just encode and enjoy.
Hmm... somewhere in the forums I posted my methodology of finding the highest level of compression with acceptable result...
To summarize the methodology: Just compress. And enjoy. If it is not enjoyable, compress at slightly lower compression. And enjoy. Keep doing it (track-by-track) until all tracks you have is enjoyable.
Never even try to think about planning to compare your supercompressed tracks with the original. Now, that's the key for DigitalAudio Nirvana
You know, I think it has a lot to do with how obssessive someone is. If you are an obssessive compulsive, you will not accept to convert your CDs to even very high bitrate lossy compression (with whatever high quality codec). You will always fear that some track you have encoded will exhibit some kind of artefacts 5 years from now when you will get to listen to it.
I AM an obssesive compulsive. I know what I am talking about. I *KNOW* I don't have the ability to detect an artefact, but the very possibility that there will be one in my whole collection makes it impossible for me to use a lossy codec.
If I wasn't mentally challenged, I would use a modern codec at a relatively high bitrate and be happy.
Absolute certainity doesnt exist. Neither your data is 100% secure, nor are your speakers perfect, nor the room, nor your ears. Live is a game of risk-management where you cannot get a perfect score but only try to loose the least with the lowest possible effort: Efficiency. Get used to it.
- Lyx
Absolute certainity doesnt exist. Neither your data is 100% secure, nor are your speakers perfect, nor the room, nor your ears. Live is a game of risk-management where you cannot get a perfect score but only try to loose the least with the lowest possible effort: Efficiency. Get used to it.
- Lyx
Regarding this, but not too diverging as to be off-topic, I was astounded recently to learn that lossless can actually be non-lossless, provided an error occurs in the RAM (or on HDD write, or in the cache) which cannot be corrected in real-time, or delayed-corrected... Then you'd get a signal stream which isn't the exact replica of the original. I found that funny.
Absolute certainity doesnt exist. Neither your data is 100% secure, nor are your speakers perfect, nor the room, nor your ears. Live is a game of risk-management where you cannot get a perfect score but only try to loose the least with the lowest possible effort: Efficiency. Get used to it.
- Lyx
Are you sure about that?
Well, I'm not... but who cares
' date='Sep 20 2006, 16:12' post='433192']
Absolute certainity doesnt exist. Neither your data is 100% secure, nor are your speakers perfect, nor the room, nor your ears. Live is a game of risk-management where you cannot get a perfect score but only try to loose the least with the lowest possible effort: Efficiency. Get used to it.
- Lyx
Regarding this, but not too diverging as to be off-topic, I was astounded recently to learn that lossless can actually be non-lossless, provided an error occurs in the RAM (or on HDD write, or in the cache) which cannot be corrected in real-time, or delayed-corrected... Then you'd get a signal stream which isn't the exact replica of the original. I found that funny.
Just as once in a while a ZIP file gets corrupted on your HDD for "no reason". How often does that happen? That's why I always inject MD5 signatures in the WavPack files I create. I can always verify if the data is intact.
Practically speaking, lossless is lossless. We could say the quantization & sampling process is lossy but then, we would open one stinky can of worms.
Regarding this, but not too diverging as to be off-topic, I was astounded recently to learn that lossless can actually be non-lossless, provided an error occurs in the RAM (or on HDD write, or in the cache) which cannot be corrected in real-time, or delayed-corrected... Then you'd get a signal stream which isn't the exact replica of the original. I found that funny.
Thats no different from the same sort of error with an uncompressed file or playback of an original CD. Most CDs do have bit errors, a couple of the early high end players even had LEDs or internal test points where you could watch the single bit, interpolate, and mute depending on the severity of the bit error. Ripping a CD typically will retry many times to correctly read bad spots, and the bit error rate for hard drives is MUCH smaller than for a CD, so in most cases the lossless compressed ripped file is superior to the original CD.
My wife ripped all of her music at something 128k or less a few years ago to use on a Rio portable (64mb and proud of it), and didn't change when I got an archos 20 for her. Listening with the archos in the car via cassette adapter most of it sounded ok, but some songs were ACK! It took a few weeks, but eventually she ripped everything again at 192k vbr (lame, not sure rev etc.) or better and all seems fine now, but I am wondering how bad the original rips were, and if it was the compression or maybe just a rip/setting error?
I also wonder about the other side of the situation, when nothing you can put your finger on is wrong, transparent via abx etc., but music over time is less appealing. I bought a new CD player once, planning on just using it as a transport for a external DAC, but when I hooked it up I couldn't tell the difference from the CD players output and the DAC, so I left the DAC turned off. After a month or so I noticed that instead of my usual 2 or 3 CDs of music every day that many days I wasn't even listening to one, or putting it on and just not getting into it. Thinking about putting on a CD, but without enough appeal to do it. It kind of put this seed of doubt into my mind that poisoned the listening experience, such that even with the old DAC it wasn't a welcome back feeling of relief, but a continuing wonder if its going to be ok thing. Like eating a bad oyster, its hard to enjoy good oysters for a bit afterwards.
Good music should be refilliing, not dentist office apprehension (even if the dentist does refilling), so be cautious with your destructive music experience testing.