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.

Poll

For use at home:

The same quality N, because I want things to sound the same as on the CD but I don't want to waste more space than necessary.
[ 9 ] (25%)
The next higher quality setting (e.g. -q N+1 for Vorbis, -V N-1 for LAME, etc.), because I want to take a safe margin. After all I didn't check all music in the world.
[ 11 ] (30.6%)
Even higher quality than the previous option, because I'm paranoid or because I want to accomodate for people with better ears (please post: how much higher?).
[ 6 ] (16.7%)
The next lower-quality setting than from my tests (e.g. -V N+1 for LAME, -q N-1 for Vorbis, etc.), because the few audible differences I hear at that setting don't annoy me and I like to shave off 16-32kbps.
[ 1 ] (2.8%)
Even lower quality than the previous option, because I find efficiency more important than perceived quality (please post: how far down would you go?).
[ 1 ] (2.8%)
My choice is not listed above (please post).
[ 8 ] (22.2%)

Total Members Voted: 58

Topic: Congratulations, you found your lowest transparent setting! Now wh (Read 27807 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

Congratulations, you found your lowest transparent setting! Now wh

Quoting myself from the end of the thread:
Jplus has posted the new version.

I’m going to lock this to avoid any more votes being placed here instead of in the revised poll. If anyone has any more to say about subjects from this thread, I think everything should be OK to post in the new location.

[original post below]


I assume the poll to be self-explanatory. Please feel welcome to discuss, even if the option of your choice doesn't explicitly ask for it!

For the first poll I have a hard time choosing between -V N and -V (N-1). I find it reassuring that -V N will sound excellent to my ears, even in quiet environments. On the other hand -V (N-1) won't sound any worse than a casette tape or a slightly dusty vinyl record (perhaps better, in a way), and back in those days I never had a problem listening to tape or vinyl. Then again, back in those days I had never done an ABX test. By searching for transparency I've sensitized my ears to hear problems that I was never aware of before.

For the second poll my choice is much easier because I'll use the same setting as for home use. While I carry my portable music almost everywhere, I tend to listen to it in reasonably quiet environments only so I see no reason to make a separate version.

I'm not asking for advice, just wondering how other people make their choice. I'm looking forward to your reactions!

Congratulations, you found your lowest transparent setting! Now wh

Reply #1
I use lossless because I am too lazy to find my transparent setting, and storage is so cheap (just ordered a 2.5 TB external drive for $70).

Congratulations, you found your lowest transparent setting! Now wh

Reply #2
I have no idea where my general transparency limit is, though there are tracks I can ABX at -V2, and there are tracks I cannot ABX at -V5.

lossless at home, V3 on portable (which sometimes gets connected to a stereo).

Congratulations, you found your lowest transparent setting! Now wh

Reply #3
I feel safe with -V63 (AAC TVBR with qaac), I transcode to -V73. My vote goes to -V (N+1) and same for portable use.

I didn't fully get the first one though, home use is lossless really. <- I didn't realize "My choice is not listed above (please post)." meant Lossless as well. Can someone please change my vote (again) from "-V (N+1)"? Thanks.

Congratulations, you found your lowest transparent setting! Now wh

Reply #4
I feel safe with -V63 (AAC TVBR with qaac), I transcode to -V73. My vote goes to -V (N+1) and same for portable use.

I didn't fully get the first one though, home use is lossless really. <- I didn't realize "My choice is not listed above (please post)." meant Lossless. Can someone please change my vote from "-V (N+1)" (again)? Thanks.

At first I included the lossless option, but I decided to remove it because I realized most people probably do that and I thought it might be more interesting to know what people would do (hypothetically) if they were to use lossy instead. Perhaps that was a mistake.

Either way, thanks for sharing!

Congratulations, you found your lowest transparent setting! Now wh

Reply #5
Oops, I didn't read carefully enough: I voted for -V N in both cases, but the reason for -V N in the portable case is that I don't want to encode twice which was an extra choice as I saw right now but after voting.
Can my vote be put there, please?
lame3995o -Q1.7 --lowpass 17

Congratulations, you found your lowest transparent setting! Now wh

Reply #6
I use FLAC -0 + lossyWAV xtraportable on portable devices due to the reasonable bitrates (generally well below 400kbps) and ridiculous decoding efficiency.  I also have yet to encounter an artifact using lWAV, and so while I keep certain albums archived losslessly, I have no qualms about processing other albums which I'm not planning to use for sample testing right out of the starting gate.  The processed tracks also seem to hold up just as well as the lossless originals when transcoding to a further lossy codec (be it Vorbis, mp3, Opus, or anything else.)

Opus is on my watchlist, but it uses far more resources to decode than does my setup with FLAC + lWAV, and storage space is not a concern for me at the moment.  I'd use Opus for applications such as audiobook encoding, most likely - lower complexity decoding, and higher resilience to audible artifacts.
FLAC -2 w/ lossyWAV 1.3.0i -q X -i

Congratulations, you found your lowest transparent setting! Now wh

Reply #7
FreaqyFrequency: that's an interesting choice, thanks for sharing. What makes decoding efficiency such an important consideration for you, if I may ask?

Congratulations, you found your lowest transparent setting! Now wh

Reply #8
Before anyone else votes, can I ask: Many people may think, due to the terminology used, that this is geared specifically towards LAME, so do you realise that your pluses and minuses are the wrong way around in terms of how that encoder presents its settings? Some others use the same type of scheme, too.

Put another way, if I determined my threshold of transparency with LAME to generally be -V3 (I just made this up; I don’t worry enough to test it) but wanted a bit of extra peace of mind, I would choose -V2, i.e. N minus 1; if I wanted to take a risk and save some space, I would use -V4, which is N plus 1.

The distinction between plus and minus in terms of quality vs. the actual setting could do with being made crystal-clear in the poll itself. I acknowledge people already requesting changes, but before I do anything, I need to make sure that future voters are certain of what they’re selecting, too. I acknowledge that the wording in the options provides context, but it’d be nice to clarify slightly just in case.

Congratulations, you found your lowest transparent setting! Now wh

Reply #9
Improved battery life, and the personal satisfaction I get from the fact that -0 on FLAC rather often compresses more effectively than -8 when encoding/decoding lWAV-processed content while simultaneously being the fastest/lightest setting for the processor.  I feel like I'm cheating entropy every time I process/encode a new file. 
FLAC -2 w/ lossyWAV 1.3.0i -q X -i

Congratulations, you found your lowest transparent setting! Now wh

Reply #10
I have no idea where my general transparency limit is


Put another way, if I determined my threshold of transparency with LAME to generally be -V3 (I just made this up; I don’t worry enough to test it)


+1

I don't know, don't care and don't worry enough to bother testing.

Congratulations, you found your lowest transparent setting! Now wh

Reply #11
Before anyone else votes, can I ask: Many people may think, due to the terminology used, that this is geared specifically towards LAME, so do you realise that your pluses and minuses are the wrong way around in terms of how that encoder presents its settings? Some others use the same type of scheme, too.

Put another way, if I determined my threshold of transparency with LAME to generally be -V3 (I just made this up; I don’t worry enough to test it) but wanted a bit of extra peace of mind, I would choose -V2, i.e. N minus 1; if I wanted to take a risk and save some space, I would use -V4, which is N plus 1.

The distinction between plus and minus in terms of quality vs. the actual setting could do with being made crystal-clear in the poll itself. I acknowledge people already requesting changes, but before I do anything, I need to make sure that future voters are certain of what they’re selecting, too. I acknowledge that the wording in the options provides context, but it’d be nice to clarify slightly just in case.

I know LAME does things the other way round but I didn't realise it might pose a problem. I'm fine with some additional clarification in the poll. I'd suggest the following: change

"For home use:"

(question field for the first poll) into this:

"For home use (greater N means better quality at larger file size):"

Thanks in advance.

Congratulations, you found your lowest transparent setting! Now wh

Reply #12
No problem. I chose to replace all mentions of “-V N” with a generic “Quality N” and all plus/minus comparatives with more precise references to higher or lower quality.

I have reallocated the two votes as requested earlier in the thread.

Editing votes in polls is never easy, especially when one must check constantly in another tab to see whether anyone else has voted… so these edits are to the best of my ability and knowledge, without promises.

Congratulations, you found your lowest transparent setting! Now wh

Reply #13
I really think the "I use the same settings" and "I use a different codec" options should be split from the question about portable use. Whether they use the same settings as on the desktop or not, whether they use a different codec or not, we want to find out how whatever setting they do use for portable compares to their transparency threshhold, and leaving out the settings from those who chose those two options makes the answers useless by biasing the remaining data.

It would make make more sense to ask two separate questions: for portable use, do you use the same settings as for home use, lower settings with the same codec, or a different codec? How does your portable setting compare to your transparency threshhold?

FreaqyFrequency, I really don't think FLAC -0 will increase your battery life; quite the contrary. Increased storage accesses cost quite a bit of time and power. Your FLAC -0 battery life may be better than 320kbps MP3s, but 190kbps MP3s would be considerably better on most platforms. For instance, on a Rockbox'd Clip Plus, with volume etc the same, battery life was as follows: 320kbps MP3 10h50m, FLAC -5 11h50m, 192kbps MP3 16h15m.

Even just switching to higher FLAC compression levels would probably save you a slight amount of battery life. (Just not -8, which greatly increases the max LPC order and thus has an impact on decode time, but has only a vanishingly small effect on compressed size.)

Congratulations, you found your lowest transparent setting! Now wh

Reply #14
I really think the "I use the same settings" and "I use a different codec" options should be split from the question about portable use. Whether they use the same settings as on the desktop or not, whether they use a different codec or not, we want to find out how whatever setting they do use for portable compares to their transparency threshhold, and leaving out the settings from those who chose those two options makes the answers useless by biasing the remaining data.

I don't agree about the "I switch to a different codec" option because I think the way it's formulated already implies that you switch to a setting that is (significantly) lower than your transparency treshold, for reasons of efficiency.

However, I'm afraid that you're right about the "Same settings as for home use" option. It would not be a problem if we could assume that the distribution of home use poll choices of the people who select that option conforms to that of the total population, but unfortunately we can't assume that.

So ideally the question "Do you use the same settings for portable and home use" should be moved to a third poll. Unfortunately I have no idea what to do with all votes that have already been cast on the "Same settings" option sofar.

Congratulations, you found your lowest transparent setting! Now wh

Reply #15
I don't agree about the "I switch to a different codec" option because I think the way it's formulated already implies that you switch to a setting that is (significantly) lower than your transparency treshold, for reasons of efficiency.
Well, you make it sound as though switching to a different codec is just like the "Even lower than [one step less than transparent]: usually I can't hear my music very well" answer, only even more so; kinda a "three steps down." But someone could, for instance, decide that they find 112kbps AAC and 96kbps Opus to both be transparent, use 96kbps Opus on their portable, but use 160kbps AAC at home because with the looser space constraints they prefer to use a less experimental encoder and allow plenty of room for paranoia/safety/&c. They're "switching to a codec that is specifically meant for efficiency" on their portable player, but it's not at all like "three steps down from transparency."

There could of course be reasons besides efficiency to switch codecs for portable use, e.g. portable player format compatibility. You might want to consider how you'd want such people to respond. Also, while those who only listen to lossless have no reason to participate in this poll, I think a fair number of people simply use their lossless copy at home ("I have the FLACs saved here for archival anyways, might as well use them") and use lossy encoders for portable. One way to address this would be to give lossless its own option in the "home use" poll.

Quote
Unfortunately I have no idea what to do with all votes that have already been cast on the "Same settings" option sofar.
Well, rather few of us have voted so far, and I think most of those who have voted would happily vote again in a revised poll if you decided to push the reset button.

Congratulations, you found your lowest transparent setting! Now wh

Reply #16
I really don't think FLAC -0 will increase your battery life; quite the contrary. Increased storage accesses cost quite a bit of time and power. Your FLAC -0 battery life may be better than 320kbps MP3s, but 190kbps MP3s would be considerably better on most platforms. For instance, on a Rockbox'd Clip Plus, with volume etc the same, battery life was as follows: 320kbps MP3 10h50m, FLAC -5 11h50m, 192kbps MP3 16h15m.


Interesting, thanks for the data on that.  However, you should know that a number of these files have a bitrate lower than 320 by the time lWAV is through with them, so the files which are being retrieved from storage actually aren't terribly huge.  This in conjunction with the ridiculous decode efficiency would probably gain an edge over even the 192CBR mp3 time you quote (though I would need to do some of my own testing to see precisely how much that would be.)

Even just switching to higher FLAC compression levels would probably save you a slight amount of battery life. (Just not -8, which greatly increases the max LPC order and thus has an impact on decode time, but has only a vanishingly small effect on compressed size.)


This may well surprise you, but as I mentioned earlier, -0 often achieves phenomenal, and even occasionally the best compression on a number of these samples.  FLAC behaves in a highly nonlinear fashion with respect to compression modes here.

For example: Paul Simon's "Father and Daughter" produces the following after being processed with lWAV, syntax -q X -i :

WAV      43,083kB
FLAC -0  10,723kB
-1          11,107kB
-2          9,428kB
-3          12,150kB
-4          12,235kB
-5          10,895kB
-6          10,891kB
-7          10,782kB
-8          10,764kB

Nothing else about the codec's behavior (encode/decode time for each level of compression, etc.) is anomalous like this.  Clearly -2 is the best for this sample, but that isn't always the case.  -0 still beats out -8 here, and this is not atypical for the files I've processed thus far.

Anyway, I'm happy to move this discussion into a sidethread and to conduct further investigation if we'd like the conversation to continue.
FLAC -2 w/ lossyWAV 1.3.0i -q X -i

Congratulations, you found your lowest transparent setting! Now wh

Reply #17
I don't agree about the "I switch to a different codec" option because I think the way it's formulated already implies that you switch to a setting that is (significantly) lower than your transparency treshold, for reasons of efficiency.
Well, you make it sound as though switching to a different codec is just like the "Even lower than [one step less than transparent]: usually I can't hear my music very well" answer, only even more so; kinda a "three steps down." But someone could, for instance, decide that they find 112kbps AAC and 96kbps Opus to both be transparent, use 96kbps Opus on their portable, but use 160kbps AAC at home because with the looser space constraints they prefer to use a less experimental encoder and allow plenty of room for paranoia/safety/&c. They're "switching to a codec that is specifically meant for efficiency" on their portable player, but it's not at all like "three steps down from transparency."

There could of course be reasons besides efficiency to switch codecs for portable use, e.g. portable player format compatibility. You might want to consider how you'd want such people to respond. Also, while those who only listen to lossless have no reason to participate in this poll, I think a fair number of people simply use their lossless copy at home ("I have the FLACs saved here for archival anyways, might as well use them") and use lossy encoders for portable. One way to address this would be to give lossless its own option in the "home use" poll.

Quote
Unfortunately I have no idea what to do with all votes that have already been cast on the "Same settings" option sofar.
Well, rather few of us have voted so far, and I think most of those who have voted would happily vote again in a revised poll if you decided to push the reset button.

There's such a thing as a reset button? That's interesting. Maybe I'd want to press that. But only if the mods agree of course.

If I were to do the poll all over I'd change the options to this (of course I'd describe the options much more elaborately than this, mostly in the same way as they're formulated now):

Poll 1 - home
lossless
q(>N+1)
q(N+1)
qN
q(N-1)
q(<N-1)
other, please specify

Poll 2 - portable
lossless
q(>N)
qN
q(N-1)
q(<N-1)
other, please specify

Poll 3 - formats
same format and same setting for both home use and portable
same lossy format but lower q for portable than for home
lossless at home and a different lossless format or setting for portable
lossless at home and lossy for portable
lossy at home and a different lossy format for portable (e.g. higher efficiency)

Mods, are you hating me already?

Congratulations, you found your lowest transparent setting! Now wh

Reply #18
I found useful transparency around 128kbps on ABX but encode FLAC for Archive and 256 kbps for music server and ipod because I'm paranoid. 

Congratulations, you found your lowest transparent setting! Now wh

Reply #19
FLAC at home, Vorbis -q2 for portable and LAME -V5 for uPnP to XBox360
Music: sounds arranged such that they construct feelings.

Congratulations, you found your lowest transparent setting! Now wh

Reply #20
I always rip lossless (ALAC).
For portable, iTunes transcodes everything to 128kbps, otherwise my iPod would be full already.
I don't know where my transparent settings are. I assume around 128-160 kbps from tests I did several years ago.


Congratulations, you found your lowest transparent setting! Now wh

Reply #22
Unfortunately I have no idea what to do with all votes that have already been cast on the "Same settings" option sofar.
Well, rather few of us have voted so far, and I think most of those who have voted would happily vote again in a revised poll if you decided to push the reset button.
There's such a thing as a reset button? That's interesting. Maybe I'd want to press that. But only if the mods agree of course.

If I were to do the poll all over I'd change the options to this […] Mods, are you hating me already?

There’s not literally a reset button! You are, of course, free to create a new poll to replace this if you want. The first post could be edited to redirect users to the new poll. There’s no reason for any staff to dislike you for this, unless you wanted us to type it all for you.  On which note, I would recommend sticking with my revised ‘higher/lower quality than in your transparent test’ naming scheme, so none of us will have to do extra typing due to possible confusion about the differing numerical directions of various encoders.

Congratulations, you found your lowest transparent setting! Now wh

Reply #23
All right, will do that later today!

Congratulations, you found your lowest transparent setting! Now wh

Reply #24
Jplus has posted the new version.

I’m going to lock this to avoid any more votes being placed here instead of in the revised poll. If anyone has any more to say about subjects from this thread, I think everything should be OK to post in the new location.