320 bitrate vs FLAC (distinguishable differences)?

[lovinthehd]

yepimonfire:
I am interested, how does one get involved or get to participate in dbt tests for xxx reason in audio?
I'm not sure I could pass a dbt on any topic, much less a sketchy one like sampling rates and bit depth.
Nobody has ever invited me to participate in a dbt. I'm still quite sure most folks still don't understand what a dbt is compared to an a-b test. Not the same. Not the same at all.

I can agree with your entire last paragraph: "i wouldn't recommend". I wouldn't recommend any of those items either. There are folks who will do them anyway. There are folks who will recommend them anyway too. That's why this subject generates discussion.

I suspect the discussion will continue.........................for a long, long time..............

You can do some simple tests with minimal gear with software like Foobar2000's abx comparator mode, and believe Audacity also has such. Large scale tests with many subjects are relatively rare due to time/expense involved....

 

[3db]

I'm not sure where lossless would make a difference, but I don't care to find out. As for orchestras, they make a difference because there are many tens (often a hundred) discrete sound sounds playing at once. In the fourth movement of Beethoven's 9th Symphony it's more like 250 sources.

You really have done several DBTs on compression schemes? Why? What prompted you to torment yourself like that? Or do you actually like being tested? How many trials per test did you do, roughly?

I know how lossy compression schemes work, and they are calculated, educated gambles, but gambles biased towards data reduction. MP3 320Kbps is dropping about 60% of the incompressible information. (Compressed lossless CD data rate is about 750Mbps.) You can argue with me all you want, but we're just going to agree to disagree, because I think DBTs as proof point for equality in this case are inconclusive at best.

If one cannot reliably choose one format over the other ina DBT, this would indicate that the difference between lossless and high bit lossy in audio quality is very very close rather then fault the DBT test based on theoretical knowledge of lossy formats. The next thing your going to tell me is that cables make a difference because DBT tests are some how flawed?

 

[yepimonfire]

If one cannot reliably choose one format over the other ina DBT, this would indicate that the difference between lossless and high bit lossy in audio quality is very very close rather then fault the DBT test based on theoretical knowledge of lossy formats. The next thing your going to tell me is that cables make a difference because DBT tests are some how flawed?

And a good portion of those sounds are going to be inaudible to humans at any given time, which is why we simply can't pass a dbt at high bit rates. If one were to measure the frequency response of a speaker that is considered accurate +- 3dB, one would end up with a graph that was actually more like +- 10dB before applying a psychacoustic curve to the graph. If a speaker is - 6dB at 100hz, and +7dB at 101 hz, the human ear is going to hear it as flat at 100hz because we can't perceive a difference in pitch less than 3% at 100hz and we can't perceive a difference in loudness less than 3dB. We are just not as good at hearing as we'd like to thing we are.

I'm not sure where lossless would make a difference, but I don't care to find out. As for orchestras, they make a difference because there are many tens (often a hundred) discrete sound sounds playing at once. In the fourth movement of Beethoven's 9th Symphony it's more like 250 sources.

You really have done several DBTs on compression schemes? Why? What prompted you to torment yourself like that? Or do you actually like being tested? How many trials per test did you do, roughly?

I know how lossy compression schemes work, and they are calculated, educated gambles, but gambles biased towards data reduction. MP3 320Kbps is dropping about 60% of the incompressible information. (Compressed lossless CD data rate is about 750Mbps.) You can argue with me all you want, but we're just going to agree to disagree, because I think DBTs as proof point for equality in this case are inconclusive at best.

Exactly. This is science we are talking about. It's objective, not subjective. If I can't perceive something and both of us have perfect hearing then it's extremely likely you won't be able to perceive it either.

Sent from my SM-G360T1 using Tapatalk

 

[Alex2507]

If one cannot reliably choose one format over the other ina DBT, this would indicate that the difference between lossless and high bit lossy in audio quality is very very close rather then fault the DBT test based on theoretical knowledge of lossy formats. The next thing your going to tell me is that cables make a difference because DBT tests are some how flawed?

Cable arguments aside, anything short of CD quality is for the plebs.

 

[3db]

Cable arguments aside, anything short of CD quality is for the plebs.

I gotta couple concerts on Blu Ray and a Dire Strates on VHS .

 

[killdozzer]

Lossless is a safety feature for me as well. I may be going the wrong way about this, but I am updating my collection to lossless even though I think 320 should be enough if everything was done properly. I think of lossless as almost no change, like EAC and this is why I prefer it. I don't care for inaudible but for no changes in audible. As I said earlier; I had two different sounding compressed albums of a same recording (same mastering, same edition, same everything). This lead me to believe that there is something there to be messed up in the process of compressing and this is why I avoid it. Although even this is no safe bet, I had 320 sounding better than lossless, it's just that this is with fewer examples.

 

[Irvrobinson]

If one cannot reliably choose one format over the other ina DBT, this would indicate that the difference between lossless and high bit lossy in audio quality is very very close rather then fault the DBT test based on theoretical knowledge of lossy formats. The next thing your going to tell me is that cables make a difference because DBT tests are some how flawed?

That's ridiculous. My point is simply that I don't think DBTs are good proof that well-designed cables sound alike, even though all cables certainly are nearly identical in electrical performance. When I was testing Google's 320Kbps service I wasn't sure if I did or didn't hear a difference, but sometimes I thought I did, so I just gave up and swore off lossy compression. I just don't want to even think about degradation.

 

[cel4145]

That's ridiculous. My point is simply that I don't think DBTs are good proof that well-designed cables sound alike, even though all cables certainly are nearly identical in electrical performance.

I don't see why that's ridiculous at all. We know that human ears with DBT are more reliable than human ears in a sighted test. That's not disputable.

That reminds me. I've always wanted to gut a Schiit Yggdrasil, put in an ODAC, and take it to a Head-Fi.org GTG. Tell them it's special modded Yggy, and I have no doubt they'll be proclaiming how great it is. LOL

 

[Irvrobinson]

I don't see why that's ridiculous at all. We know that human ears with DBT are more reliable than human ears in a sighted test. That's not disputable.

I was referring to his statement about how I think cables sound different.

I am disputing the value of audio DBTs, so it obviously is disputable.

 

[PENG]

I stick with 24b/96kHz and DSD because I believe the odds are better that they are made from good quality masters. Recording/mastering is key to the eventual playback quality, not so much the format if our lowest reference point is 320 kbps.

 

[3db]

That's ridiculous. My point is simply that I don't think DBTs are good proof that well-designed cables sound alike, even though all cables certainly are nearly identical in electrical performance. When I was testing Google's 320Kbps service I wasn't sure if I did or didn't hear a difference, but sometimes I thought I did, so I just gave up and swore off lossy compression. I just don't want to even think about degradation.

I understand your point of the safety issue even though you weren't sure of audible differences between the two formats. However, that doesn't diminishe in any way the validity of DBT or blind listening tests, something that your trying to sell to justify your personal preferences.

 

[Irvrobinson]

I understand your point of the safety issue even though you weren't sure of audible differences between the two formats. However, that doesn't diminishe in any way the validity of DBT or blind listening tests, something that your trying to sell to justify your personal preferences.

I'm not trying to sell anything, I'm just stating a preference, and an opinion about audio DBTs. I don't feel I have to justify my preferences either. It does seem like the advocates of lossy compression are trying to justify a choice they made by falling back on the DBT crutch. Audio DBTs don't reveal better or worse, they only reveal preference, if they reveal anything at all, which they usually don't in my experience. In the case of lossy compression schemes it could be that a loss is perceived by the listener as an improvement, because the compression schemes are just algorithms. They only know about frequencies and amplitudes. This isn't like cables; cables can be proven through measurement that they don't affect audio performance. With 320Kbps MP3 we know it's throwing away ~60% of the information compared to CDs. The question is, does that 60% matter? Maybe not, maybe so, I'm just on the side of why bother, and I don't care to depend on arbitrary DBTs (which music, which system, which room, which volume level) to make a decision. Just ignore me if that sounds stupid to you.

 

[PENG]

Has anyone purchase/read this Paper? Reading the paper may reinforce my belief that CD quality is all we need. I am not sure about 320 kbps MP3 but like Irv I see no point to find out and I focus my effort and time on finding digital files derived from high quality recording/mastering master copies of anything.

 

[Alex2507]

I'm just on the side of why bother ...

Why bother is right. That's well worded. The further up the food chain you go with playback equipment, the better your chances are of hearing differences. Maybe on bookshelves < $3k, 320 and CD do sound alike but on > $20k worth of towers and subs, it becomes a different story?

Next you have to try bi-amping again. Night and day difference. Even my wife could hear it.

 

[gzubeck]

Has anyone purchase/read this Paper? Reading the paper may reinforce my belief that CD quality is all we need. I am not sure about 320 kbps MP3 but like Irv I see no point to find out and I focus my effort and time on finding digital files derived from high quality recording/mastering master copies of anything.

I can tell if something is 44 vs 96 most of the time. Ive always felt theres a sense of ease and a blacker background with a higher bitrate. After 96 its pretty much undetectable.

Quote from top of paper " The noise of the CD-quality loop was audible only at very elevated levels."

thats enough for me to want the higher bitrate version.

 

[Irvrobinson]

Next you have to try bi-amping again. Night and day difference. Even my wife could hear it.

Oh yeah, sorry, I never bothered to respond. My system has been passively bi-amped for a while now, after you told me to, and there's no difference I could discern whatsoever. It does look cooler though, but not as cool as when I was using dedicated monoblocks per speaker.

 

[cel4145]

Audio DBTs don't reveal better or worse, they only reveal preference, if they reveal anything at all, which they usually don't in my experience. In the case of lossy compression schemes it could be that a loss is perceived by the listener as an improvement, because the compression schemes are just algorithms.

Yes. Some DBT can be used to determine preference.

However, typically ABX is recommended for comparing lossy to lossless, which is not a test for preference. An ABX test answers the simple question, "Can you reliably hear a difference." If one can't hear a difference, then one can't hear a difference. It's not a preference or choice for one or the other. It's a fact.

Note that ABX is also often used for comparing cables. Since you don't seem to understand ABX, I'd would imagine you have never tried it. I'd recommend giving it a shot. You might be surprised.

 

[Irvrobinson]

Yes. Some DBT can be used to determine preference.

However, typically ABX is recommended for comparing lossy to lossless, which is not a test for preference. An ABX test answers the simple question, "Can you reliably hear a difference." If one can't hear a difference, then one can't hear a difference. It's not a preference or choice for one or the other. It's a fact.

Note that ABX is also often used for comparing cables. Since you don't seem to understand ABX, I'd would imagine you have never tried it. I'd recommend giving it a shot. You might be surprised.

Actually, I do understand it, and yes I've been in multiple A/B/X tests, for amplifiers, at a dealer in SoCal a long time ago. I was fascinated by the testing process back then. And I agree, audio DBTs can be used to detect differences in addition to preferences.

 

[cel4145]

Actually, I do understand it, and yes I've been in multiple A/B/X tests, for amplifiers, at a dealer in SoCal a long time ago. I was fascinated by the testing process back then. And I agree, audio DBTs can be used to detect differences in addition to preferences.

Then I'm confused by your post 92 then since it insists DBT "only reveals preference" and your argument about advocates of lossy compression is based upon that.

 

[Irvrobinson]

Then I'm confused by your post 92 then since it insists DBT "only reveals preference" and your argument about advocates of lossy compression is based upon that.

Because I answering in the context of this discussion and misspoke. Obviously having a preference involves the detection of difference. Next time I'll be more careful with my choice of words.