IPB

Welcome Guest ( Log In | Register )

> Hydrogenaudio Forum Rules

- No Warez. This includes warez links, cracks and/or requests for help in getting illegal software or copyrighted music tracks!


- No Spamming or Trolling on the boards, this includes useless posts, trying to only increase post count or trying to deliberately create a flame war.


- No Hateful or Disrespectful posts. This includes: bashing, name-calling or insults directed at a board member.


- Click here for complete Hydrogenaudio Terms of Service

5 Pages V  < 1 2 3 4 5 >  
Closed TopicStart new topic
Anyone else use 320 kbps mp3's?, n00b here.
320 kbps mp3's
Do You ever use 320 kbps mp3's?
Yes, rarely. [ 117 ] ** [23.12%]
Yes, often. [ 93 ] ** [18.38%]
No, never. [ 296 ] ** [58.50%]
Can you tell the difference between 256 kbps VBR or CBR mp3's and 320 kbps mp3's?
Yes. [ 48 ] ** [9.49%]
No. [ 434 ] ** [85.77%]
Only between 256 CBR and 320 CBR [ 24 ] ** [4.74%]
Total Votes: 643
  
j7n
post Aug 2 2008, 09:26
Post #51





Group: Members
Posts: 813
Joined: 26-April 04
Member No.: 13720



In comment to the very first post:

I'd use 320 kBit/s MP3 in case when absolutely best possible quality is required, but lossless or hybrid formats are unavailable, or would cause confusion. Archival is definitely not a use for high bitrate lossy. Possible application could be a radio retranslator in a LAN.
Go to the top of the page
+Quote Post
sld
post Aug 2 2008, 09:32
Post #52





Group: Members
Posts: 1017
Joined: 4-March 03
From: Singapore
Member No.: 5312



QUOTE (Synthetic Soul @ Aug 2 2008, 13:52) *
If you didn't - even now - continue to use patronising name-calling then perhaps I may give a damn. It's taken days from your initial post for you to actually take our advice. Kudos for doing so, but I for one am still pissed with the effort that had to go into this thread to get a sensible post from you.

Sorry for being an angry nerd, but we see this too many times; you're not a special case: just another noob wasting my time.

Maybe I shouldn't post so early in the morning...

Whoa there, yes it's too early in the morning for you.

There's a small community of us here who know how to deal with audio objectively. That means there are many others out there who do not give audio the scientific treatment it deserves. When they come to HA.org they are definitely going to bring in their old mindsets and prejudices. Give them a chance; look, our patience in this thread has paid off already.

And Jake, CBR is old-fashioned and outdated. Use VBR if you want quality, the technology has matured a few years back. Sorry for being the 5th or so guy to repeat this. Perhaps this is baggage from Fraunhofer or r3mix, but 256 and 320 are no longer magic numbers. In contrast, 128 is still a magic number.

This post has been edited by sld: Aug 2 2008, 09:36
Go to the top of the page
+Quote Post
JAKE196
post Aug 5 2008, 03:01
Post #53





Group: Members
Posts: 10
Joined: 12-July 08
Member No.: 55662



Synthetic Soul,

I knew I shouldn't have said that last remark, don't take it seriously.

Your right, if you've seen this many times before you have the right to be annoyed.


Anyway I've done some more tests now and I've found out the following:

I can't tell the difference between 256 CBR and 320 kbps. (took two tests and failed) (fair enough)

I can't tell the difference between 224 CBR and 256 CBR. (took a test and failed) (hmm. . . )

I can tell the difference between a VBR averaging 235 and a VBR averaging 277 kbps. (took a test and got 11/12)

Weird how I can tell the difference between 235 VBR and 277 VBR, yet not between 224 CBR and 256 CBR.

I might use VBR's averaging a bit higher than 256 just to be on the safe side, I've got enough memory to fit my library 3 or 4 times over anyway. I'm sure 320 kbps is overkill now, but I'm still a bit skeptical about VBR's at averages a bit below 256. I might also try ABR's around 272. . .

I'm a bit OCD about some things.
Go to the top of the page
+Quote Post
timcupery
post Aug 5 2008, 05:16
Post #54





Group: Members
Posts: 780
Joined: 19-December 01
From: Tar Heel country
Member No.: 683



Jake, I'm just reading over this thread and was thinking how nicely and patiently people were treating you and hoping you'd take to it. Glad you did. I remember before I tried ABX-ing I would do blind tests on Winamp playing two tracks back-to-back on repeat so I didn't remember which one I was on, and seeing if I could tell any differences. ABX makes blind testing a lot easier.

One thing for you: I wouldn't recommend comparing 256 to 320 or 224 to 256 etc. as you report in your most recent post.
Rather, I assume what you're really interested in is, at what quality level of encoding can you tell a difference from the original (pre-encoding) .wav file?

So, you should compare 256 to the wav file, and 224 to the wav file, and see at what level you can start to notice differences.

Or better yet, compare Lame -V4 to the original wav file, and if you can tell differences there, try -V3, and if you can tell diffs there, try -V2, etc.
People have strongly made the case (and correctly) that you should use VBR encoding with Lame's -V settings.

There is really no reason anymore to use CBR encoding. It used to be that certain players had problems decoding VBR or didn't correctly display the time-length of a VBR file, but those days are past.


--------------------
God kills a kitten every time you encode with CBR 320
Go to the top of the page
+Quote Post
kornchild2002
post Aug 5 2008, 06:15
Post #55





Group: Members
Posts: 2080
Joined: 8-April 05
From: Cincinnati, OH
Member No.: 21277



It is nice that you are conducting blind ABX tests but you are still doing them wrong. You should compare a lossless file to a lossy one. You should also use the different presets (ie -V values) that Lame offers. Start off with something low such as -V 6 or -V 5 and ABX it against the lossless file. If you hear a difference, move onto the next higher -V value. Do this until you fail your ABX test.

There is no reason to start at such high bitrates. You also must compare to the source lossless file. Never compare two lossy files directly unless one lossy file is a source and you transcoded to a lower bitrate (or different format). Being able to hear a difference between 235kbps VBR and 277kbps VBR (what -V settings where those?) means nothing other than you were able to hear a difference. It doesn't signify which one was better as both are lossy files. Using a lossless file and being able to hear a difference means that you were able to pick out the lossless file as being better than the lossy. Being able to hear two different lossy files means nothing as you don't know which file had the better sound quality, you just know that you could hear a difference between the two. For all you know the 235kbps VBR file could have produced better results than the higher bitrate one (it has been known to happen).

So start off with a lossless file and transcode it down to -V 6 or -V 5 using Lame mp3. ABX the lossless and lossy file. Fail the test and you know that -V 6 or -V 5 would be perfectly fine for you. Pass the test and you know that you need to go higher. No offense but it seems that you are still a little stubborn when it comes to using lower bitrate files. I am giving up now unless you start off with lower bitrate files and compare them to lossless ones. So feel free to post back if you do that. Otherwise, enjoy yourself.
Go to the top of the page
+Quote Post
Soap
post Aug 5 2008, 12:30
Post #56





Group: Members
Posts: 1019
Joined: 19-November 06
Member No.: 37767



QUOTE (kornchild2002 @ Aug 5 2008, 01:15) *
Using a lossless file and being able to hear a difference means that you were able to pick out the lossless file as being better than the lossy.

I agree with you throughout this thread, but feel I must be pedantic on this one.

A successful ABX test does not mean he finds the lossless sounds better than the lossy. It means he finds the lossless sounds distinguishably different than the lossy. Nothing more.


--------------------
Creature of habit.
Go to the top of the page
+Quote Post
timcupery
post Aug 5 2008, 14:13
Post #57





Group: Members
Posts: 780
Joined: 19-December 01
From: Tar Heel country
Member No.: 683



QUOTE (Soap @ Aug 5 2008, 07:30) *
I agree with you throughout this thread, but feel I must be pedantic on this one.

A successful ABX test does not mean he finds the lossless sounds better than the lossy. It means he finds the lossless sounds distinguishably different than the lossy. Nothing more.

This is technically true. But for most people, the point of lossy compression is to sound indistinguishable to the original, but at a much lower bitrate.
So you're right, ABX testing means being able to hear a difference, not being able to hear which one sounds better. But if we assume the whole point of the endeavor is that the lossy sounds like the lossless, the lossy sounding unlike the lossless is worse.

[you're probably on the same page with me here, now that I think about your vs. kornchild's phrasing more carefully, but I figured this is worth elucidating]


--------------------
God kills a kitten every time you encode with CBR 320
Go to the top of the page
+Quote Post
Soap
post Aug 5 2008, 15:51
Post #58





Group: Members
Posts: 1019
Joined: 19-November 06
Member No.: 37767



QUOTE (timcupery @ Aug 5 2008, 09:13) *
[you're probably on the same page with me here, now that I think about your vs. kornchild's phrasing more carefully, but I figured this is worth elucidating]

Yes I am, I had just come back to clarify my position but you replied first.
I was just trying to be absolutely clear for our self-professed "noob" original poster that ABX testing determines differences, not quality. That quality is subjective and something determined by personal preference.

Clearly in the case of lossless vs lossy there is an ability to ABX (lack of) transparency (and therefore a reasonable claim of "better"), but in the case of lossy vs lossy better is completely subjective.

This post has been edited by Soap: Aug 5 2008, 15:52


--------------------
Creature of habit.
Go to the top of the page
+Quote Post
JAKE196
post Aug 6 2008, 03:11
Post #59





Group: Members
Posts: 10
Joined: 12-July 08
Member No.: 55662



Am I doing ABX tests the wrong way? Well this is the first time I've done any, someone should have told me.

I don't think comparing .mp3's to .wav's would work because I can always tell between them simply by listening to the cymbals. cymbals often sound like a jingly mess on mp3's but they sound good on .wav's.

. . . I guess that's because the frequency cutoff, before I decided to use mp3's I also tried AAC files with frequency cutoff's at 18khz, 20khz, etc. The cymbals on those files sounded better, but I decided the higher frequency cutoff wasn't worth the extra memory because the difference wasn't that big. Then I eventually decided to stick to mp3's because the frequency cutoff doesn't really effect the music that much. (cymbals, pffft)

I know there has been numerous debates about frequency stuff in the past, and i don't want to start another one, but also note that I am under 20 which means i should be able to hear higher frequencies than middle aged people. (I've at least done my homework on this one) (yay wiki) (also, you know those annoying ring tones that only young ppl can hear? Well if your old you don't, but yeah.)

I have tried comparing lower bitrate files and higher ones and i can tell the difference. . . I don't think comparing mp3's to .wav files would help much. I guess I'll try it just to make sure though, I will compare a 320 kbps mp3 (just to be safe) and a .wav file and see if I can beat an ABX test.

And of course remember that I have so much memory I can use slightly overkill bitrates if i want. smile.gif I'm not going to use 320 kbps any more but i might as well use the highest bitrate for which I can tell the difference between lower ones. If quality is subjective like one of you said, I'll use the highest bitrates I deem fit, ABX tests or not.

Seriously, I got almost 8 gigs and the player wasn't even that expensive. Cheaper than those dam ipods, and better sound output. My Library isn't that big either.

You guys aren't going to convince me to use a lot lower bitrates, but I appreciate the continued input.

. . . and one more time, sorry, I guess you guys were fairly patient considering I was rambling on about something that isn't true. I was quite ignorant.

And all right, I get the point, CBR isn't as good as VBR, I knew that. But it isn't terrible.

That's all my thoughts.

Damn, that has to be one of the longest posts I've ever made on a forum. top 5.
Go to the top of the page
+Quote Post
lesswire
post Aug 6 2008, 03:25
Post #60





Group: Members
Posts: 37
Joined: 4-July 08
Member No.: 55275



QUOTE
I don't think comparing .mp3's to .wav's would work because I can always tell between them simply by listening to the cymbals.


If you can always tell the difference between the lossless and lossy files you should post your ABX results here because you could help LAME developers improving it. As mentioned before, statements here are worthless if you can't back them up.

Cheers
Go to the top of the page
+Quote Post
greynol
post Aug 6 2008, 03:44
Post #61





Group: Super Moderator
Posts: 10339
Joined: 1-April 04
From: San Francisco
Member No.: 13167



For the record, I believe JAKE196. His ability to determine high-bitrate lossy encodings from one another was enough for me.

Let's assume I cannot ABX 320 from lossless for a given sample, we can conclude that for that sample 320 is transparent for me. If I were to test -V5 against 320 of that same sample it would be no different than if I were to test -V5 from the lossless version of that same sample. Let's also say that I cannot ABX -V0 or -V2 the original. It is perfectly acceptable to conclude that I cannot ABX -V0 from -V2. If I were able to ABX -V0 from -V2, then it is safe to say that I would be able to ABX at least one of these from the original, if not both.

Anyhow, this pretty much sums it up...
QUOTE (JAKE196 @ Aug 5 2008, 19:11) *
note that I am under 20

Jake, if you have the time post your logs to shut these guys up, please do. It would be great to post some 30 second clips as well. Mp3directcut may be a useful tool if you're posting mp3 clips, though I think people usually like a clip of the lossless original to convert themselves.

Also, as has been said already, please stick around and perform some listening tests on behalf of the developers. Sebastian Mares is a good person to contact.


--------------------
Your eyes cannot hear.
Go to the top of the page
+Quote Post
kornchild2002
post Aug 6 2008, 03:47
Post #62





Group: Members
Posts: 2080
Joined: 8-April 05
From: Cincinnati, OH
Member No.: 21277



QUOTE (JAKE196 @ Aug 5 2008, 20:11) *
Am I doing ABX tests the wrong way? Well this is the first time I've done any, someone should have told me.


We were told that someone was helping you with your ABX tests. How were we to know that they didn't tell you (or maybe they did but you just didn't see) how to properly set them up.

QUOTE (JAKE196 @ Aug 5 2008, 20:11) *
I don't think comparing .mp3's to .wav's would work because I can always tell between them simply by listening to the cymbals. cymbals often sound like a jingly mess on mp3's but they sound good on .wav's.


I said it before and I say it again: placebo, placebo, placebo, placebo, placebo! Switching back and forth between tracks is NOT a good way to judge audio quality. The whole point of ABX tests are to compare lossless with lossy tracks. Otherwise we shouldn't even take ABX tests and just use lossless for everything.

QUOTE (JAKE196 @ Aug 5 2008, 20:11) *
. . . I guess that's because the frequency cutoff, before I decided to use mp3's I also tried AAC files with frequency cutoff's at 18khz, 20khz, etc. The cymbals on those files sounded better, but I decided the higher frequency cutoff wasn't worth the extra memory because the difference wasn't that big. Then I eventually decided to stick to mp3's because the frequency cutoff doesn't really effect the music that much. (cymbals, pffft)


The "decrease" in sound quality that you are hearing cannot be justified unless you perform ABX tests. Again, your comments are meaningless unless you have your ABX tests to back them up.

QUOTE (JAKE196 @ Aug 5 2008, 20:11) *
I know there has been numerous debates about frequency stuff in the past, and i don't want to start another one, but also note that I am under 20 which means i should be able to hear higher frequencies than middle aged people. (I've at least done my homework on this one) (yay wiki) (also, you know those annoying ring tones that only young ppl can hear? Well if your old you don't, but yeah.)


There is actually a difference between hearing a certain frequency and hearing a certain frequency while playing music. Hell, I can hear certain tones up to the extent of the human ear but, when those tones are playing inside of a song, I can't hear them at all.

QUOTE (JAKE196 @ Aug 5 2008, 20:11) *
I have tried comparing lower bitrate files and higher ones and i can tell the difference. . . I don't think comparing mp3's to .wav files would help much. I guess I'll try it just to make sure though, I will compare a 320 kbps mp3 (just to be safe) and a .wav file and see if I can beat an ABX test.


I have said it before and I will say it again: placebo, placebo, placebo, placebo, placebo!

QUOTE (JAKE196 @ Aug 5 2008, 20:11) *
And of course remember that I have so much memory I can use slightly overkill bitrates if i want. smile.gif I'm not going to use 320 kbps any more but i might as well use the highest bitrate for which I can tell the difference between lower ones. If quality is subjective like one of you said, I'll use the highest bitrates I deem fit, ABX tests or not.


Again, you don't compare lossy formats directly to lossy formats like that. You compare a lossless file to lossy files. You can use the highest bitrate that you want. We are just trying to show you the light, take the blindfold off your eyes, and let you know that you don't need to keep using these 256kbps+ bitrates that you keep insisting on using.


QUOTE (JAKE196 @ Aug 5 2008, 20:11) *
Seriously, I got almost 8 gigs and the player wasn't even that expensive. Cheaper than those dam ipods, and better sound output. My Library isn't that big either.

You guys aren't going to convince me to use a lot lower bitrates, but I appreciate the continued input.

. . . and one more time, sorry, I guess you guys were fairly patient considering I was rambling on about something that isn't true. I was quite ignorant.

And all right, I get the point, CBR isn't as good as VBR, I knew that. But it isn't terrible.

That's all my thoughts.

Damn, that has to be one of the longest posts I've ever made on a forum. top 5.


You know what? I don't even know why I bother. I am sorry but if I knew that this was going to be your end response then I just would have told you to bugger off (no, I am not English) from the very get go. There have been so many people here on these forums trying to help you and many of them are experienced. You then take their experience, spit on it, and throw it in the garbage. The whole point of blind ABX tests are to determine the right lossy encoder and setting for your ears. You keep insisting that you must use high bitrates even though you can't hear a difference. Why? What is the point to carrying around 256kbps+ music if you can't hear the difference between 128kbps files and the source lossless ones? That is like buying a car and only driving it at speeds of 100 mph or higher. Again, don't take anything the wrong way but you have all this experience and advice about this topic and you continue to ignore it. Why?

timecupery and Soap, you are both right. I was a little simplistic in my wording though as I didn't want to complicate things for Jake. But yes, all an ABX test does it tell if you can hear a difference or not, it doesn't give you any notion of sound quality. That being said, not being able to hear the difference between a lossy file and lossless file means that to you, the lossy file produces the same perceptual sound quality. This is just looking at ABX tests from a basic standpoint.

This post has been edited by kornchild2002: Aug 6 2008, 03:49
Go to the top of the page
+Quote Post
Tahnru
post Aug 6 2008, 04:20
Post #63





Group: Members
Posts: 325
Joined: 17-October 05
From: United States
Member No.: 25178



QUOTE (JAKE196 @ Aug 5 2008, 21:11) *
Am I doing ABX tests the wrong way? Well this is the first time I've done any, someone should have told me.

From what I have seen, you aren't doing them the wrong way. An ABX test is conducted to determine whether or not a difference can be reliably detected between two files. The important thing to keep in mind with ABX tests, is what your test hypothesis is.

If your aim is to test whether or not you can tell the difference between two differing bitrate MP3's, then you use 2 differing bitrate MP3's. To determine if you can tell a difference between a lossless copy of the original and a particular bitrate MP3, you use those two files as your test material.

QUOTE (JAKE196 @ Aug 5 2008, 21:11) *
I don't think comparing .mp3's to .wav's would work because I can always tell between them simply by listening to the cymbals. cymbals often sound like a jingly mess on mp3's but they sound good on .wav's.

Noted above by Greynol, you would do us a great favor to generate a battery of tests and post the results. For example, grab a 30 second clip of your choosing in a self-ripped lossless format, and test it against different bitrate MP3's encoded with LAME. Post the results here for collaboration with the developers.

QUOTE (JAKE196 @ Aug 5 2008, 21:11) *
I have tried comparing lower bitrate files and higher ones and i can tell the difference. . . I don't think comparing mp3's to .wav files would help much. I guess I'll try it just to make sure though, I will compare a 320 kbps mp3 (just to be safe) and a .wav file and see if I can beat an ABX test.

You're not trying to beat an ABX test. Don't take an ABX test personally - the test is set up to provide you with evidence. Collect and interpret, young scientist!

QUOTE (JAKE196 @ Aug 5 2008, 21:11) *
That's all my thoughts.

Damn, that has to be one of the longest posts I've ever made on a forum. top 5.

Long posts are fun!

This post has been edited by Tahnru: Aug 6 2008, 04:22
Go to the top of the page
+Quote Post
timcupery
post Aug 6 2008, 04:31
Post #64





Group: Members
Posts: 780
Joined: 19-December 01
From: Tar Heel country
Member No.: 683



QUOTE (JAKE196 @ Aug 5 2008, 22:11) *
I don't think comparing .mp3's to .wav's would work because I can always tell between them simply by listening to the cymbals. cymbals often sound like a jingly mess on mp3's but they sound good on .wav's.

You're very confident on this. And I'm not ready to write you off yet.
But remember how confident you were before that it was easy for you to tell differences - and then you couldn't tell the difference when you took an ABX test?
Well, just to say with kornchild, it wouldn't surprise me if that happens again. Greynol may be right that you actually ARE hearing differences and could reproduce them in a blind-test environment. And if so, you and your ears would be very valuable to this forum. Though if I were to bet, my money would be with kornchild, and you only THINK you're hearing differences.

Anyway, it's not a matter of comparing mp3's to wav's "not working" because you can easily hear a difference in the cymbals. If you can easily hear this difference, then the test will simply show that you ACTUALLY CAN hear the difference. Which would be useful information.
Just remember, if you make a statement about how well you can hear something - back it up. Give evidence. People will get tired of you pretty quickly around here if they have to remind you to provide ABX results every time you say you can hear a difference. Lots of smart and knowledgeable people have taken time to try to explain stuff to you in this thread. Don't be hardheaded about it. (for example, CBR "isn't bad" - it's true - but that doesn't mean there's any reason to use it instead of using VBR)

Good luck. I'm interested to hear just how good your ears actually are.

This post has been edited by timcupery: Aug 6 2008, 04:35


--------------------
God kills a kitten every time you encode with CBR 320
Go to the top of the page
+Quote Post
Tahnru
post Aug 6 2008, 04:47
Post #65





Group: Members
Posts: 325
Joined: 17-October 05
From: United States
Member No.: 25178



QUOTE (kornchild2002 @ Aug 5 2008, 21:47) *
QUOTE (JAKE196 @ Aug 5 2008, 20:11) *

Am I doing ABX tests the wrong way? Well this is the first time I've done any, someone should have told me.


We were told that someone was helping you with your ABX tests. How were we to know that they didn't tell you (or maybe they did but you just didn't see) how to properly set them up.


I was the one helping Jake with his initial setup of the ABX test. His initial goal was to conduct a test to determine whether or not he could tell the difference between a 320 kbps CBR mp3 and a lower-bitrate VBR MP3. I would invite you to explain what was wrong with his methodolgy, given this test objective.

QUOTE (kornchild2002 @ Aug 5 2008, 21:47) *
The whole point of ABX tests are to compare lossless with lossy tracks. Otherwise we shouldn't even take ABX tests and just use lossless for everything.

This is a mistake. The point of an ABX test on audio files is to determine whether or not a difference can be identified between the two samples. The collected evidence must then be correctly interpreted for what it is. There is NO requirement that the samples be lossless versus lossy.

ABX tests can also be used to tell the difference between speaker cables, power cables, and a host of other things that may affect sensory stimuli. Please read for more information.

QUOTE (kornchild2002 @ Aug 5 2008, 21:47) *
QUOTE (JAKE196 @ Aug 5 2008, 20:11) *
And of course remember that I have so much memory I can use slightly overkill bitrates if i want. smile.gif I'm not going to use 320 kbps any more but i might as well use the highest bitrate for which I can tell the difference between lower ones. If quality is subjective like one of you said, I'll use the highest bitrates I deem fit, ABX tests or not.


Again, you don't compare lossy formats directly to lossy formats like that. You compare a lossless file to lossy files. You can use the highest bitrate that you want. We are just trying to show you the light, take the blindfold off your eyes, and let you know that you don't need to keep using these 256kbps+ bitrates that you keep insisting on using.


Again, why not? We are here to educate on proper ABX test methodology. We aren't here to "show the light" or "unblindfold" anyone. Once someone learns to properly collect and present evidence to back up any factual claims they make, they have what they need to constructively participate here.

-Rob
Go to the top of the page
+Quote Post
timcupery
post Aug 6 2008, 05:07
Post #66





Group: Members
Posts: 780
Joined: 19-December 01
From: Tar Heel country
Member No.: 683



Rob - I'd agree with you in technicalities.
But the main point of ABX testing in the context of HA is to see which encodes are distinguishable from the original. Since, presumably, the thing we're all aiming for is indistinguishibility from the original.
Now, I'd agree with you that if someone can tell a difference between 320 CBR and 128 CBR, it's pretty safe to assume that the 320 is closer to the original (so long as they're from the same or similar encoder, at least).

But figuring out which encodes can be differentiated from the original is closer to the point, even if not closer to the original question that Jake asked which started this thread.


--------------------
God kills a kitten every time you encode with CBR 320
Go to the top of the page
+Quote Post
Tahnru
post Aug 6 2008, 05:28
Post #67





Group: Members
Posts: 325
Joined: 17-October 05
From: United States
Member No.: 25178



As far as the context for HA is concerned, I am in total agreement.

However, presentation methodology has been sorely lacking at times in this thread. In this case where we have a willing student like Jake, it seems to me that telling him that he is wrong and that his methodology is wrong WITHOUT understanding what his original test was trying to achieve and without fully understanding ABX testing in general, is BAD teaching. (apologies for the run-on sentence)

Proceeding then to imply that he should "bugger off", without attempting to pose a new question and to explain why the new question is more valid than the old one, adds nothing to the discussion.

First, you learn the technique to measure your test of a hypothesis and interpret the results. Later, you learn how to ask increasingly meaningful questions.

-Rob

This post has been edited by Tahnru: Aug 6 2008, 05:29
Go to the top of the page
+Quote Post
shadowking
post Aug 6 2008, 05:46
Post #68





Group: Members
Posts: 1529
Joined: 31-January 04
Member No.: 11664



I'd like to know what these differences are in cymbals - subtle, annoying etc.. You say mp3 cymbals sound like a mess. What does 18-20k lowpass have to do with it ? cymbals swoosh comes much earlier than even 16 k

I am interested in seeing abx logs for V5 and ratings out of 5, then for V4..V3.. V2 if V5 is indeed a crappy mess. And good abx logs too like < 3 % pval. So crappy V5 should give 8/8 abx on any cymbal sample.

This post has been edited by shadowking: Aug 6 2008, 05:48


--------------------
Wavpack -b450s0.7
Go to the top of the page
+Quote Post
kornchild2002
post Aug 6 2008, 06:41
Post #69





Group: Members
Posts: 2080
Joined: 8-April 05
From: Cincinnati, OH
Member No.: 21277



QUOTE (timcupery @ Aug 5 2008, 22:07) *
Rob - I'd agree with you in technicalities.
But the main point of ABX testing in the context of HA is to see which encodes are distinguishable from the original. Since, presumably, the thing we're all aiming for is indistinguishibility from the original.
Now, I'd agree with you that if someone can tell a difference between 320 CBR and 128 CBR, it's pretty safe to assume that the 320 is closer to the original (so long as they're from the same or similar encoder, at least).

But figuring out which encodes can be differentiated from the original is closer to the point, even if not closer to the original question that Jake asked which started this thread.


That is my whole point and I guess I shouldn't have come off so negative. I just get frustrated that we have someone come on here and tell us that the sky is green when we say that the sky is blue and then tell them to run some tests. They do run some tests but still come back with the notion that the sky is green.

Part of my problem when comparing directly between lossy formats is that one cannot tell which was closer to the actual lossless file. There have been some cases (I would be happy to point them out and even provide a sample out of my library) where a -V 2 encoded Lame mp3 actually did better than a -V 0 Lame mp3. I am not sure what caused the problem but I could hear the artifact more easily at -V 0 than -V 2 and no, the artifact is not in the lossless song. So I myself could take an ABX test with said song, distinguish between the two lossy files, but then that means nothing as theoretically the -V 2 file should produce the better results. That is why I prefer and preach ABXing between lossy and lossless formats. That way one can easily narrow down the proper encoder, format, and setting rather than comparing two lossy formats and making some educated guesses.

Oh, as for my bugger off comment, it was provoked from Jake's comment of "You guys aren't going to convince me to use a lot lower bitrates..." That right there leads me to believe that Jake is still using the 256kbps or 320kbps mindset even though he is being told otherwise and we are asking for tests. So, to me, we keep telling Jake that the sky is blue or feel free to prove us wrong with tests, Jake performs some tests, says that the sky is blue-ish, and then makes comments like the above which entirely regress what we are saying.

So I am in agreement with shadowking here as I want to see some ABX logs for -V 5, -V 4, -V 3, -V 2, and -V 0. Jake might be able to hear the difference between those files and the source lossless one. The point is that no one can truly know until they conduct a blind ABX test and from the sound of it, Jake won't be doing that as they use bitrates below their unestablished standard.

That is all I am trying to say and I am sorry if I came off negative.
Go to the top of the page
+Quote Post
Synthetic Soul
post Aug 6 2008, 07:41
Post #70





Group: Super Moderator
Posts: 4887
Joined: 12-August 04
From: Exeter, UK
Member No.: 16217



QUOTE (JAKE196 @ Aug 6 2008, 03:11) *
I don't think comparing .mp3's to .wav's would work because I can always tell between them simply by listening to the cymbals. cymbals often sound like a jingly mess on mp3's but they sound good on .wav's.
QUOTE
...you do it to yourself, just you; you and no-one else...
You're setting yourself up for a TOS #8 warning here again. For us to continue we really are going to need to see ABX results.

Given your previous misconceptions I am surpised at such a statement. I agree wholeheartedly with shadowking, that you should provide ABX results from tests of various VBR quality levels, and even 320 CBR, given the generalistion (you can tell the difference between any MP3 and a lossless source?). Thankfully, I believe that you are interested enough in the subject to perform these tests, and honest enough to post accurate results (as you have proved).

QUOTE (JAKE196 @ Aug 6 2008, 03:11) *
note that I am under 20
No shit. wink.gif

QUOTE (JAKE196 @ Aug 6 2008, 03:11) *
My Library isn't that big either.
My wife always tells me that size doesn't matter.

QUOTE (greynol @ Aug 6 2008, 03:44) *
For the record, I believe JAKE196. His ability to determine high-bitrate lossy encodings from one another was enough for me.
I agree. He obviously has good hearing and could be a really useful member of the community, if he stops violating TOS #8 and is willing to have an open mind. However, if anyone states that they can tell a difference between (by inference, 320 CBR) MP3 and WAVE we must see ABX results. Them's the rules.

I would dearly love to see JAKE196 post ABX results of lossless/lossy tests, achieve spiritual enlightenment, start wearing a mu'umu'u, and become an active member in the Listening Tests forum. As I am a huge fan of irony I would then, months down the line, love to see him posting in threads like these, batting for the other side.

I miss Guru too BTW. sad.gif I wonder if Gabs still hears from him.


--------------------
I'm on a horse.
Go to the top of the page
+Quote Post
smok3
post Aug 6 2008, 08:12
Post #71


A/V Moderator


Group: Moderator
Posts: 1747
Joined: 30-April 02
From: Slovenia
Member No.: 1922



QUOTE
note that I am under 20

i belive ~ 25 is the age when hearing ability starts to drop, so welcome aboard.


--------------------
PANIC: CPU 1: Cache Error (unrecoverable - dcache data) Eframe = 0x90000000208cf3b8
NOTICE - cpu 0 didn't dump TLB, may be hung
Go to the top of the page
+Quote Post
shadowking
post Aug 6 2008, 08:36
Post #72





Group: Members
Posts: 1529
Joined: 31-January 04
Member No.: 11664



I think i've just about read enough of this messy cymbals, HF lowpass, i am young stuff on the net over the years. Even if its true I think its nothing to brag about . If normal masking which works on most people doesn't for you its NOT a good sign of ear health.

This post has been edited by shadowking: Aug 6 2008, 08:37


--------------------
Wavpack -b450s0.7
Go to the top of the page
+Quote Post
JAKE196
post Aug 6 2008, 23:25
Post #73





Group: Members
Posts: 10
Joined: 12-July 08
Member No.: 55662



QUOTE (shadowking @ Aug 6 2008, 01:36) *
I think i've just about read enough of this messy cymbals, HF lowpass, i am young stuff on the net over the years. Even if its true I think its nothing to brag about . If normal masking which works on most people doesn't for you its NOT a good sign of ear health.


I'm not bragging! Not what I meant to sound like! >_>

If anything it's a bad thing because it effects how much i enjoy mp3's, which is mainly what I listen to. . .
------------------------------------------------------
So many mixed replies!!! OK, from now on I won't make a single post about something without ABX test support, and I'll try to document exactly what I did, I should have done that in the first place like Tahnru said. I won't post again until I do a lot more tests involving lossy vs. lossless and lossy vs. lossy.


It could take a while. . . . I'll be back. . .
Go to the top of the page
+Quote Post
Slipstreem
post Aug 7 2008, 04:21
Post #74





Group: Members
Posts: 966
Joined: 7-July 06
Member No.: 32660



Good man! Nobody here is trying to tell you what to think. We're just advising you on what not to assume. Awaiting your ABX results eagerly. smile.gif

Cheers, Slipstreem. cool.gif
Go to the top of the page
+Quote Post
Synthetic Soul
post Aug 7 2008, 09:18
Post #75





Group: Super Moderator
Posts: 4887
Joined: 12-August 04
From: Exeter, UK
Member No.: 16217



QUOTE (JAKE196 @ Aug 6 2008, 23:25) *
So many mixed replies!!!
Yes, it is a shame that the issue has been clouded slightly; this whole thread is pretty fubar, but that's what always happens with TOS #8 violations. However there is really only one thing that you need to remember here: you've made a claim, you need to back it up. The more test data you can provide the happier we'll be. smile.gif

QUOTE (JAKE196 @ Aug 6 2008, 23:25) *
OK, from now on I won't make a single post about something without ABX test support ...
I have never posted results of an ABX test in my life. You just have to remember: if you state that you can hear a difference between one encoding and another you must be prepared to post results of an ABX to prove it. If you don't want to create FUD then really you should have performed the test before making such claims, to ensure confidence in your claim. TOS #8 is there to combat FUD, caused by members posting claims that simply are not true.

QUOTE (JAKE196 @ Aug 6 2008, 23:25) *
... and I'll try to document exactly what I did, I should have done that in the first place like Tahnru said. I won't post again until I do a lot more tests involving lossy vs. lossless and lossy vs. lossy.
Good man. Looking forward to it.


--------------------
I'm on a horse.
Go to the top of the page
+Quote Post

5 Pages V  < 1 2 3 4 5 >
Closed TopicStart new topic
1 User(s) are reading this topic (1 Guests and 0 Anonymous Users)
0 Members:

 



RSS Lo-Fi Version Time is now: 26th December 2014 - 01:35