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: Here\'s an easy sample to ABX for rev8 (Read 5854 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

Here\'s an easy sample to ABX for rev8

http://members.rogers.com/evil-ned/lame/ting2_original.wav

This sample is easy to ABX with lame dm rev8 using --alt-preset standard.

wav was ripped from Bon Jovi - Blaze of Glory using EAC in secure mode on my TEAC 24xcdrw

Here\'s an easy sample to ABX for rev8

Reply #1 –
Can you please be a little more specific about the problem?  I'm going to download this clip and take a look at it shortly but it would help me quite a bit if you could give a more detailed account of exactly what you believe you are hearing wrong. Thanks

And out of curiousity, why no abx results, since you mention it being easy in the title?

Here\'s an easy sample to ABX for rev8

Reply #2 –
Hmmm, I just tried a quick ABX only to find out that I couldn't fix my attention towards anything wrong.

I used Rev8b with --alt-preset standard.

Regards stoff

Here\'s an easy sample to ABX for rev8

Reply #3 –
Rev8d is already available.

Here\'s an easy sample to ABX for rev8

Reply #4 –
What is rev 8d?  I only see an announcement from Dibrom about rev 8b, and that's all that appears at http://static.hydrogenaudio.org/extra/
It's is not, it isn't ain't, and it's it's, not its, if you mean it is.  If you don't, it's its.

Here\'s an easy sample to ABX for rev8

Reply #5 –
The so called "c" and "d" are actually modifications to the alpha (in other areas) by other people, not by myself..

I do have another version coming up later tonight though.

Here\'s an easy sample to ABX for rev8

Reply #6 –
Just did a quick ABX.

20/20 correct.

The hammer strike sounds about the same but the rustling/clicking sound is dead easy to tell apart. It doesn't seem to have the same upper frequencies of the original.

I'm using the rev8b ICL. Is it better on c/d?

Here\'s an easy sample to ABX for rev8

Reply #7 –
Ok, thanks, I'll take a look shortly.

To clear up confusion, there is no "c" or "d" revisions from me.. I believe there may have been confusion from updates to CVS that others may have though were from me, but they weren't.  Gabriel made an update last night and today Mark made another one, both small adjustments to abr bitrates and completely unrelated to my work.

Oh, and btw, do you still hear the problem with the extreme or insane mode?  How about "--alt-preset standard --no-preset-tune" ?

Here\'s an easy sample to ABX for rev8

Reply #8 –
Still got 8/8 of correct with --alt-preset insane

Here\'s an easy sample to ABX for rev8

Reply #9 –
I just took a quick listen to this sample, and I believe the problem is pre-echo.  I believe I hear it most clearly in the left channel.  I haven't abx'd yet since I'm still working on a few other things (I'll be releasing a new revision in an hour or so if all goes well), but if this is what I think it is, I don't believe there will be much I can do to fix it... especially if you still hear it in the insane mode.

MP3 just has pre-echo problems really, and there's really no way to completely eliminate it.  The --alt-presets usually sound better than most, but samples like death2, castanets, and others are still not pre-echo free.  For that matter though AAC and Vorbis aren't perfect in that regard either.. (though usually much better.. at least with the case of AAC and higher bitrate vorbis or RC3).

Here\'s an easy sample to ABX for rev8

Reply #10 –
I listened to this sample a little more, and I cannot abx a problem (translation -- there is none that I can tell, for me that is).  I also discussed this with ff123 who said he could not hear a problem either, and apparently Wombat couldn't either.  So I don't really know what to say.  If there is an audible difference it may be a slight pre-echo issue, but I'm usually good at picking up on that.

Here\'s an easy sample to ABX for rev8

Reply #11 –
Here's the total method I'm using.

Encode ting2_original.wav with razorlame GUI which uses the lame.exe (178,688 bytes) rev8b ICL version. It's set to use

--alt-preset standard

with the "use only custom options" checked.

When that's done I use winamp diskwriter to convert back to .wav

Same method when doing --alt-preset insane


The artifact I'm hearing is clearly audible in the clicking/rustling sound before the hammer strike.

28/28 trials correct and it takes me about 3 seconds to pick the correct one.

Here\'s an easy sample to ABX for rev8

Reply #12 –
It is possible like said that there's a problem here, but at the very least, I can't hear it with this equipment at the moment (very noisy sound source).  I won't have a "clean" source to listen from for awhile still.  I can try listening again tonight but I'm sure if I'll have anymore luck.

If someone else wants to give this a go and verify there is a problem, you are more than welcome.  Again, that being said.. I don't think there's much I can do about it, especially if the problem is still there in the insane mode.

I think maybe I heard something which sounded like pre-echo, but I wasn't able to abx this.. but pre-echo is an issue I can't really "fix" with MP3.

I'd imagine that if you are hearing this problem, and it is still there with the "insane" mode, then it's basically not transparent with anything you have tried with LAME, correct?

At any rate, I think it's a very safe bet to say that this isn't an "easy" sample to ABX.

Here\'s an easy sample to ABX for rev8

Reply #13 –
Here's my wav>mp3>wav files in case there's anything wrong with my process.

standard

insane

Here\'s an easy sample to ABX for rev8

Reply #14 –
Do you have EQ on or something? I downloaded this standard version you posted and it definitely sounds worse than it should.
I didn't have rev 8 anymore, but here is version encoded with rev10. Compare to this: http://www.saunalahti.fi/~cse/ting2_dm10.wav

Here\'s an easy sample to ABX for rev8

Reply #15 –
Yes, there is something wrong with your process.  I just downloaded your file, and I can instantly hear a huge difference.. it sounds very muffled.

As case asked, are you using some sort of EQ or post-processing before you listen?  Because the file definitely isn't supposed to sound like that..

Here\'s an easy sample to ABX for rev8

Reply #16 –
I sent the ABX program to my friend and he was able to identify it easily 10/10 tries with his Sennheiser headphones. I'm just using mediocre quality Cambridge Microworks PC speakers and it's clearly audible even on those.

I'd like to note that although it's easy to identify each in ABX with these short samples, it's much harder to tell when just listening to the full song.

Here\'s an easy sample to ABX for rev8

Reply #17 –
Quote
Originally posted by Ned
I sent the ABX program to my friend and he was able to identify it easily 10/10 tries with his Sennheiser headphones. I'm just using mediocre quality Cambridge Microworks PC speakers and it's clearly audible even on those.

I'd like to note that although it's easy to identify each in ABX with these short samples, it's much harder to tell when just listening to the full song.


There is something wrong with your encoded sample, that's why it is easy to ABX.  I can tell the difference between the original .wav and the encoded sample you provided without even a need for ABX.

You must either be using some sort of EQ, or normalization, or other processing on the file because encoded on my system (with no pre/post-processing at all) the file does NOT sound like the one you provided at all.

Case seems to have verified this as well in his post above.

Can you explain in specific detail exactly the process you have used to create this file?  Something is very wrong with the process you are using, but I don't know what..

Here\'s an easy sample to ABX for rev8

Reply #18 –
I don't use the eq or any DSP plugins. Just using the MAD input plugin set to 16-bit (sblive which doesnt apparently work in 24).

Dunno why it would alter the output using diskwriter.

Here\'s an easy sample to ABX for rev8

Reply #19 –
Quote
Originally posted by Ned
I don't use the eq or any DSP plugins. Just using the MAD input plugin set to 16-bit (sblive which doesnt apparently work in 24).

Dunno why it would alter the output using diskwriter.


You probably had the auto-attenuation (clipping prevention) turned on in MAD, and when you output the file via diskwriter it altered the gain on the file, accounting for the change in sound.  I'm willing to bet that's exactly what is going on..

Never use any sort of clipping prevention or anything which can alter gain or perceived volume or anything like that when comparing files, cause it can be very misleading.  The same thing happens with MPC all the time when people claim to hear problems then later find out it was because they had clipping prevention on, which made the file sound slightly "dull" (which is exactly how your file sounds compared to the original, or a file decoded simply via lame at the command line).

If you want to hear how this file should sound absolutely clean, use lame --decode <infile> <outfile>.  You should hear a huge difference, at which point you probably shouldn't be able to abx the encoded file anymore.