Help - Search - Members - Calendar
Full Version: lossyWAV 1.1.0 Development Thread.
Hydrogenaudio Forums > Lossy Audio Compression > Other Lossy Codecs
Pages: 1, 2, 3, 4, 5, 6
Nick.C
May 21 2008, 06:46
I've been thinking about the difference between the replaygain / wavegain values calculated for the .lwcdf.wav file with --shaping 0 and --shaping 1 (spreadsheet attached).

There seems to be an average 12dB difference between the two (--shaping 1 = higher gain), although how this relates to bits to remove, I still need to determine.
halb27
May 21 2008, 07:44
I think that' s very natural and shows that your kind of noise shaping is effective.
The replaygain technique computes the perceived loudness. So replaygain says the perceived loudness of the error file is considerably lower when using noise shaping. So it confirms in an objective way what we can hear subjectively: noise shaping reduces the perceived noise - when listening to the noise.
Unfortunately it doesn't tell about masking, and so a noise shaped version isn't necessarily better when listening to the music.
With a high quality setting I feel pretty safe to use noise shaping, that's why I do it. The lower the quality setting however the more the risk IMO that noise concentrated in certain frequency regions can hurt (though it may well be that the benefits outweigh the risk also with lower quality settings).
Mardel
May 21 2008, 11:27
There is a somthing up.
I'm hearing heavily distorted sound at -q0.

Nick.C
May 21 2008, 11:40
QUOTE(Mardel @ May 21 2008, 18:27) *
There is a somthing up.
I'm hearing heavily distorted sound at -q0.
Thanks for the sample, at what position in the audio are you hearing the distortion?

Also, what is the significance of the images?
sauvage78
May 21 2008, 12:19
I confirm,
I can hear it too sec 01, sec 07 & sec 12,
between sec 01 & sec 02 it is very very noisy in the back. I can ABX it 100%.
halb27
May 21 2008, 12:29
QUOTE(Nick.C @ May 21 2008, 19:40) *

QUOTE(Mardel @ May 21 2008, 18:27) *
There is a somthing up.
I'm hearing heavily distorted sound at -q0.
Thanks for the sample, at what position in the audio are you hearing the distortion? ...

I just abxed sec. 6.4-8.6 8/10 (and I'm sure those who know this track will do better). I wouldn't call this spot heavily distorted though, but that's a matter of taste.
We know -q 0 isn't a real quality mode (though usually it's quite okay). My personal favorite for low bitrate is -q 1.5. It's not significantly higher in bitrate (312 kbps on average for my regular music test set) but to me quality does a little jump there. The spot I mentioned with your track, Mardel, is ok to me using -q 1.5.

Mardel and sauvage78, do you mind trying -q 1.5?
Mardel
May 21 2008, 12:43
QUOTE(Nick.C @ May 21 2008, 19:40) *

QUOTE(Mardel @ May 21 2008, 18:27) *
There is a somthing up.
I'm hearing heavily distorted sound at -q0.
Thanks for the sample, at what position in the audio are you hearing the distortion?

Also, what is the significance of the images?

I hear some sizzle in the high pitch range.
The essence of the picture, that 12 khz full is with sounds like that what they were not originally there.
halb27
May 21 2008, 12:47
QUOTE(Mardel @ May 21 2008, 20:43) *

... The essence of the picture, that 12 khz full is with sounds like that what they were not originally there.

Sure. lossyWAV does add noise. The essential question is: is it audible? We can't read that from a picture.
But with your sample we can hear it when using -q 0.
sauvage78
May 21 2008, 12:50
tried quickly v1.0.1f -q 1.5, the artefact is much reduced but still easyly ABXable IMHO
I wouldn't have thought only a 1.5 increase would have reduced it so much, so I think it should go away on higher setting ... futher test is required.

I was using -q 0 too, all I can say is that I don't want to use -q 0 anymore, for me it was as easy to ABX as MP3 64Kbps on castanets sample, a nightmare wink.gif

Typo: I typed -q 1 intead of -q 0 several times, it's fixed now
Nick.C
May 21 2008, 12:56
QUOTE(halb27 @ May 21 2008, 19:47) *
QUOTE(Mardel @ May 21 2008, 20:43) *
... The essence of the picture, that 12 khz full is with sounds like that what they were not originally there.
Sure. lossyWAV does add noise. The essential question is: is it audible? We can't read that from a picture.
But with your sample we can hear it when using -q 0.
Maybe now is the time to revisit the -spf string (again) and maybe tighten up its response in the 12kHz to 16kHz region. I'll get to work re-introducing the -spf parameter (as --spf <string>) and post v1.0.1g this evening.

Thanks for the sample and listening input sauvage78 and Mardel (and, of course halb27).
halb27
May 21 2008, 13:11
QUOTE(sauvage78 @ May 21 2008, 20:50) *

tried quickly v1.0.1f -q 1.5, the artefact is much reduced but still easyly ABXable IMHO ...

Thanks for your test. Glad to hear that -q 1.5 is better for you though still easily ABXable. Your ears are better than mine.
We should keep in mind that general transparency is expected at -q 5, at least for people with good hearing. Due to a series of internal precautions it's not necessarily -q 5 which is to use, and we can stay a bit lower.

It would be interesting to learn at what quality level this sample becomes transparent to everybody who tries.


QUOTE(Nick.C @ May 21 2008, 20:56) *

Maybe now is the time to revisit the -spf string (again) and maybe tighten up its response in the 12kHz to 16kHz region. ....

Please keep in mind that all the tests so far have been performed with very low quality settings.
When introducing a more restrictive -spf string for the highest frequency zone (you did that already with your latest spf change) bitrate will go up, and we must compare the results to the results of the current spf setting at a correspondingly higher quality setting.

Perhaps it would be wise to find out first what current quality setting is necessary for generally accepted transparency of this sample.
In case transparency can be achieved with a rather low current quality setting I think we shouldn't mind and just call it a problem sample we should be willing to accept with very low quality settings.
In case transparency is achieved not before ~ -q 5 I think it's worth investigating whether a change in the machinery can bring a progress.
sauvage78
May 21 2008, 13:46
I agree with halb27, I have spend the last 20 min doing serious ABX & it happened that -q1.5 was much harder to ABX than what I first thought ... I can ABX -q 0 100% without serious listening, I was forced to listen carefully to ABX -q 1.5, it was placebo when I said it was easy ... I was finally able to ABX with near 100% success but it was not easy at all in the end ... I can ABX -q 0 with only the 2 first seconds ... I was forced to concentrate & listen up to second 7 to ABX -q1.5 ... & in the end it was very little added noise to an already noisy original sample for -q 1.5 while it's LOTS of noise for -q 0 ... so I was not self-confident even if I finally succeeded.

I think it is very specific to -q 0 & already very noisy source. -q 0 is not safe anymore for me but everything is not wasted. I am 100% sure that this sample is transparent for me far before -q 5 now as I doubt I can ABX even -q 2 ... but I won't test I am bored now wink.gif
halb27
May 21 2008, 14:03
QUOTE(sauvage78 @ May 21 2008, 21:46) *

I agree with halb27, I have spend the last 20 min doing serious ABX & it happened that -q1.5 was much harder to ABX than what I first thought ... I can ABX -q 0 100% without serious listening, I was forced to listen carefully to ABX -q 1.5, it was placebo when I said it was easy ... I was finally able to ABX with near 100% success but it was not easy at all in the end ... I can ABX -q 0 with only the 2 first seconds ... I was forced to concentrate & listen up to second 7 to ABX -q1.5 ... & in the end it was very little added noise to an already noisy original sample for -q 1 while it's LOTS of noise for -q 0 ... so I was not self-confident even if I finally succeeded.

I think it is very specific to -q 0 & already very noisy source. -q 0 is not safe anymore for me but everything is not wasted.

Glad to hear this. Do you mind going up a bit in quality level to learn when this track is transparent for you?
sauvage78
May 21 2008, 14:10
Sorry, I just edited that I wouldn't test -q 2 now ... maybe I will try tomorow cause it's bedtime for me now.

Don't expect miracle, I am french but I am no Guruboolez wink.gif
halb27
May 21 2008, 14:12
QUOTE(sauvage78 @ May 21 2008, 22:10) *

Sorry, I just edited that I wouldn't test -q 2 now ... maybe I will try tomorow cause it's bedtime for me now.

Don't expect miracle, I am french but I am no Guruboolez wink.gif

Sure, but your testing is very welcome. Thank you, and have a good night.
Nick.C
May 21 2008, 14:13
QUOTE(sauvage78 @ May 21 2008, 21:10) *
Sorry, I just edited that I wouldn't test -q 2 now ... maybe I will try tomorow cause it's bedtime for me now.

Don't expect miracle, I am french but I am no Guruboolez wink.gif
Thanks for all of the input - I am very interested in your -q 2 ABX results.

Nick.
Nick.C
May 22 2008, 01:54
I've been thinking about -q 0 and 1 ("my" quality levels as my ears are a bit shot....). As I have been refining the reference_threshold > threshold_index > bits_to_remove calculations, I had a few kbit/s to play with which I have recycled into the quality of the output of the lower quality presets.

I have re-implemented the --lowpass parameter to allow user with better ears to raise the upper limit of the range of frequencies lossyWAV uses in its analyses.

I have tweaked the --spf and minimum_bits_to_keep:
CODE
  spreading_function_string         : string[precalc_analyses*(spread_zones+2)-1]='22222-22222-22223-12224-12234-12345';
  quality_minimum_bits_to_keep      : array[0..Quality_Presets] of double = (3.000,3.000,3.000,3.125,3.250,3.375,3.500,3.625,3.750,3.875,4.000);
and get the following:
CODE
|-----------|-------|-------|-------|-------|-------|-------|-------|-------|-------|-------|-------|-------|
| lossyWAV  | FLAC  | -q 10 | -q 9  | -q 8  | -q 7  | -q 6  | -q 5  | -q 4  | -q 3  | -q 2  | -q 1  | -q 0  |
|-----------|-------|-------|-------|-------|-------|-------|-------|-------|-------|-------|-------|-------|
|1.0.1b     |784kbps|637kbps|607kbps|577kbps|545kbps|513kbps|480kbps|449kbps|427kbps|390kbps|349kbps|306kbps|
|-----------|-------|-------|-------|-------|-------|-------|-------|-------|-------|-------|-------|-------|
|1.0.1g     |784kbps|639kbps|610kbps|580kbps|548kbps|516kbps|484kbps|453kbps|430kbps|396kbps|354kbps|320kbps|
|-----------|-------|-------|-------|-------|-------|-------|-------|-------|-------|-------|-------|-------|
|1.0.1g --LC|784kbps|656kbps|628kbps|600kbps|569kbps|538kbps|506kbps|474kbps|450kbps|414kbps|369kbps|334kbps|
|-----------|-------|-------|-------|-------|-------|-------|-------|-------|-------|-------|-------|-------|

lossyWAV beta 1.0.1g attached to post #1 in this thread.
sauvage78
May 22 2008, 02:45
This morning I tested the problem sample at V1.0.1f -q 0/-q 0.5/-q 1/-q 1.5 with my speaker (Logitech Z3-I) instead of my headphone (Philips SHP1900) it was easier to ABX but I dunno if that is because I was tired yesterday or because of the cheap headphone (Mainboard is Asus A7N8X-Deluxe) this time I have keep my log to backup my claims as I done it very seriously.

foo_abx 1.3.3 report
foobar2000 v0.9.5.2
2008/05/22 09:56:09

File A: C:\Documents and Settings\Mes documents\02- Test\Nouveau dossier\b.lossless.flac
File B: C:\Documents and Settings\Mes documents\02- Test\Nouveau dossier\b.lossyQ0.0.flac

09:56:09 : Test started.
09:56:37 : 01/01 50.0%
09:56:50 : 02/02 25.0%
09:57:02 : 03/03 12.5%
09:57:18 : 04/04 6.3%
09:57:31 : 05/05 3.1%
09:57:45 : 06/06 1.6%
09:57:58 : 07/07 0.8%
09:58:09 : 08/08 0.4%
09:58:20 : 09/09 0.2%
09:58:33 : 10/10 0.1%
09:58:47 : 11/11 0.0%
09:58:58 : 12/12 0.0%
09:59:30 : Test finished.

----------
Total: 12/12 (0.0%)

foo_abx 1.3.3 report
foobar2000 v0.9.5.2
2008/05/22 10:01:16

File A: C:\Documents and Settings\Mes documents\02- Test\Nouveau dossier\b.lossless.flac
File B: C:\Documents and Settings\Mes documents\02- Test\Nouveau dossier\b.lossyQ0.5.flac

10:01:16 : Test started.
10:01:38 : 01/01 50.0%
10:01:53 : 02/02 25.0%
10:02:06 : 03/03 12.5%
10:02:19 : 04/04 6.3%
10:02:38 : 05/05 3.1%
10:02:51 : 06/06 1.6%
10:03:50 : 07/07 0.8%
10:04:01 : 08/08 0.4%
10:04:23 : 09/09 0.2%
10:04:36 : 10/10 0.1%
10:04:48 : 11/11 0.0%
10:04:59 : 12/12 0.0%
10:05:02 : Test finished.

----------
Total: 12/12 (0.0%)

foo_abx 1.3.3 report
foobar2000 v0.9.5.2
2008/05/22 10:06:22

File A: C:\Documents and Settings\Mes documents\02- Test\Nouveau dossier\b.lossless.flac
File B: C:\Documents and Settings\Mes documents\02- Test\Nouveau dossier\b.lossyQ1.0.flac

10:06:22 : Test started.
10:06:50 : 01/01 50.0%
10:07:20 : 02/02 25.0%
10:07:36 : 03/03 12.5%
10:07:47 : 04/04 6.3%
10:08:00 : 05/05 3.1%
10:08:24 : 06/06 1.6%
10:08:45 : 07/07 0.8%
10:09:06 : 08/08 0.4%
10:09:21 : 09/09 0.2%
10:09:52 : 10/10 0.1%
10:10:17 : 11/11 0.0%
10:10:36 : 12/12 0.0%
10:10:46 : Test finished.

----------
Total: 12/12 (0.0%)

foo_abx 1.3.3 report
foobar2000 v0.9.5.2
2008/05/22 10:20:41

File A: C:\Documents and Settings\Mes documents\02- Test\Nouveau dossier\b.lossless.flac
File B: C:\Documents and Settings\Mes documents\02- Test\Nouveau dossier\b.lossyQ1.5.flac

10:20:41 : Test started.
10:21:01 : 01/01 50.0%
10:21:24 : 02/02 25.0%
10:21:54 : 03/03 12.5%
10:22:34 : 04/04 6.3%
10:24:09 : 05/05 3.1%
10:25:48 : 06/06 1.6%
10:26:20 : 07/07 0.8%
10:27:01 : 08/08 0.4%
10:27:21 : 09/09 0.2%
10:28:15 : 10/10 0.1%
10:29:33 : 11/11 0.0%
10:30:18 : 12/12 0.0%
10:30:22 : Test finished.

----------
Total: 12/12 (0.0%)

Will test -q 2 later as it took me 10min just for -q 1.5 but it becomes hard. -q 0 & -q 0.5 were easy (no hesitation), -q 1 medium (a few hesitations) & -q 1.5 hard (several hesitations), I was forced to re-listen several times at -q 1.5 but unlike yesterday I was able to ABX all with only the first 2 second & was not forced to listen up to second 7.
sauvage78
May 22 2008, 03:41
I can ABX -q 2 too but it takes 23min with a pause in the middle ... I will not try to ABX further it's too hard.

foo_abx 1.3.3 report
foobar2000 v0.9.5.2
2008/05/22 11:08:32

File A: C:\Documents and Settings\Mes documents\02- Test\b.lossless.wav
File B: C:\Documents and Settings\Mes documents\02- Test\b.lossyQ2.0.wav

11:08:32 : Test started.
11:10:25 : 01/01 50.0%
11:12:44 : 02/02 25.0%
11:13:49 : 03/03 12.5%
11:14:53 : 04/04 6.3%
11:16:12 : 05/05 3.1%
11:19:18 : 06/06 1.6%
11:22:21 : 07/07 0.8%
11:23:00 : 08/08 0.4%
11:23:57 : 09/09 0.2%
11:24:57 : 10/10 0.1%
11:28:47 : 11/11 0.0%
11:31:24 : 12/12 0.0%
11:31:28 : Test finished.

----------
Total: 12/12 (0.0%)

In the end I would describe it like this:
-q 0.0 & -q0.5 added noise/distortion, clear artefact/flaw
-q 1.5 & -q2.0 tearing in the original distortion, I don't ABX an added artefact, I ABX either the lenght of the distortion or the depht of the tearing (it's slightly longer/deeper)
-q1.0 is in the middle: slightly added noise, slighly longer distortion.
Mardel
May 22 2008, 03:53
QUOTE(Nick.C @ May 22 2008, 09:54) *
I have re-implemented the --lowpass parameter to allow user with better ears to raise the upper limit of the range of frequencies lossyWAV uses in its analyses.
THX. -q 2.5 --lowpass 17000 is a very good sound quality for my sample. I tried q 1; q 1.5; q 2; q 2 lowpass 16k, 17k, 18, 19, 20k; q 2.5; q2.5 lowpass16k and 17k. I haven't heard any sound artifacts at -q 2.5 --lowpass 17000.
Nick.C
May 22 2008, 03:58
QUOTE(sauvage78 @ May 22 2008, 10:41) *
I can ABX -q 2 too but it takes 23min with a pause in the middle ... I will not try to ABX further it's too hard.

In the end I would describe it like this:
-q 0.0 & -q0.5 added noise/distortion, clear artefact/flaw
-q 1.5 & -q2.0 tearing in the original distortion, I don't ABX an added artefact, I ABX either the lenght of the distortion or the depht of the tearing (it's slightly longer/deeper)
-q1.0 is in the middle: slightly added noise, slighly longer distortion.
Many thanks for your efforts. As you can see above, I have modified the HF end of the spreading function in 1.0.1g. Also, you might be interested in using the --lowpass parameter to increase the upper limit of the range of frequencies lossyWAV uses in its analyses.

When your ears have recovered, I would appreciate it if you could listen to 1.0.1g -q 0 (it probably will still not require ABX, however I would be interested to hear if there is any improvement). Maybe slowly increasing --lowpass from 16000 to, say, 18000 will improve this particular sample (at the expense of FLAC bitrate).

Thanks again,

Nick.
2Bdecided
May 22 2008, 04:33
Would it be rude to request a reality check here? No lossy codec should be transparent at the very bottom of its quality scale.

I think it's very useful to have at least one non-transparent setting at the bottom of the range - it lets people hear what kind of artefacts lossyWAV is adding. It also confirm to the average user that the quality scale works, and stops people from saying "even q0 is transparent - what happens if you lower the bitrate still further".

Cheers,
David.
collector
May 22 2008, 04:38
CODE
-s, --scale <n>  scaling factor from WaveGain, etc; default=1.000000; n<>0!

LossyWav doesn't permit n to be a positive scale factor either, but many of my classical and easy listening piano albums require it.
My question: can it be changed in a future version ? My Wavegain.exe can do it but that's another run in the process.
Nick.C
May 22 2008, 05:04
QUOTE(2Bdecided @ May 22 2008, 11:33) *
Would it be rude to request a reality check here? No lossy codec should be transparent at the very bottom of its quality scale.

I think it's very useful to have at least one non-transparent setting at the bottom of the range - it lets people hear what kind of artefacts lossyWAV is adding. It also confirm to the average user that the quality scale works, and stops people from saying "even q0 is transparent - what happens if you lower the bitrate still further".

Cheers,
David.
Not at all - it's a requirement of development.... I agree with what you're saying, though maybe there is some scope for improving the performance at -q 0 a bit....

QUOTE(collector @ May 22 2008, 11:38) *
CODE
-s, --scale <n>  scaling factor from WaveGain, etc; default=1.000000; n<>0!

LossyWav doesn't permit n to be a positive scale factor either, but many of my classical and easy listening piano albums require it.
My question: can it be changed in a future version ? My Wavegain.exe can do it but that's another run in the process.
At present you would have to wavegain the file before processing with lossyWAV. What upper limit do you think would be reasonable, as it's an easy fix (1 value changed in the nparameters unit)?
2Bdecided
May 22 2008, 05:07
If you wavegain at more than unity, you can introduce clipping, but more importantly the efficiency could be reduced (probably not by much in practice).
sauvage78
May 22 2008, 05:10
V1.0.1g -q 0 Vs. Original

foo_abx 1.3.3 report
foobar2000 v0.9.5.2
2008/05/22 13:00:52

File A: C:\Documents and Settings\b.lossy.wav
File B: C:\Documents and Settings\b.wav

13:00:52 : Test started.
13:01:09 : 01/01 50.0%
13:01:21 : 02/02 25.0%
13:01:32 : 03/03 12.5%
13:01:44 : 04/04 6.3%
13:02:02 : 05/05 3.1%
13:02:17 : 06/06 1.6%
13:02:32 : 07/07 0.8%
13:02:52 : 08/08 0.4%
13:03:04 : 09/09 0.2%
13:03:15 : 10/10 0.1%
13:03:28 : 11/11 0.0%
13:03:42 : 12/12 0.0%
13:04:01 : Test finished.

----------
Total: 12/12 (0.0%)

For me V1.0.1g -q 0 = V1.0.1f -q 0.5 it's better but the artefact is still obvious & easy to spot. (3 min to ABX)
Nick.C
May 22 2008, 05:27
QUOTE(sauvage78 @ May 22 2008, 12:10) *
For me V1.0.1g -q 0 = V1.0.1f -q 0.5 it's better but the artefact is still obvious & easy to spot. (3 min to ABX)
Thanks for that - taking into account what David said about reality checking and that the lowest preset *should* introduce noticable artifacts, I think that I will step back a bit from the settings I put in place for 1.0.1g.

If I keep the spreading_function string at '22222-22222-22223-12224-12234-12345' but reduce the minimum_bits_to_keep to 2.333 and 2.667 respectively for -q 0 and -q 1, the resultant bitrate for -q 0 is 312kbps/325kbps and 352kbps/367kbps (default / --linkchannels) for my 54 problem sample set (to which I must add Mardel's latest problem sample).

I feel that this is a reasonable compromise as the HF regions are given a little more influence in the result at a small change in bitrate.

lossyWAV beta 1.0.1h attached to post #1 in this thread.
lvqcl
May 22 2008, 11:56
With this sample and -q 0 setting, it looks like level of added noise changes very frequently, and this modulation is more noticeable than the noise itself.
collector
May 22 2008, 12:07
QUOTE(Nick.C @ May 22 2008, 03:04) *

At present you would have to wavegain the file before processing with lossyWAV. What upper limit do you think would be reasonable, as it's an easy fix (1 value changed in the nparameters unit)?

Up till now, with dozens of tracks to go, the biggest album difference is +6.23 dB. Don't know what scale factor that is.
Nick.C
May 22 2008, 12:28
QUOTE(collector @ May 22 2008, 19:07) *
QUOTE(Nick.C @ May 22 2008, 03:04) *
At present you would have to wavegain the file before processing with lossyWAV. What upper limit do you think would be reasonable, as it's an easy fix (1 value changed in the nparameters unit)?
Up till now, with dozens of tracks to go, the biggest album difference is +6.23 dB. Don't know what scale factor that is.
About 2.0488, I reckon.
halb27
May 22 2008, 12:50
QUOTE(2Bdecided @ May 22 2008, 12:33) *

... I think it's very useful to have at least one non-transparent setting at the bottom of the range - it lets people hear what kind of artefacts lossyWAV is adding. It also confirm to the average user that the quality scale works, and stops people from saying "even q0 is transparent - what happens if you lower the bitrate still further". ...

This is a great argument IMO. We should really stop to expect -q 0 being transparent. Good quality as a rule should be enough.


QUOTE(sauvage78 @ May 22 2008, 11:41) *

I can ABX -q 2 too but it takes 23min with a pause in the middle ...

Thank you, sauvage 78. Your results show that -q 2 isn't transparent with this sample, but (please correct me if I'm wrong) seems to be a setting which is 'close enough to transparency' to be useful for very high quality DAP use. To a slightly minor degree this can be said also of -q 1.5.
Remains the question: What quality setting brings full transparency?

QUOTE(Mardel @ May 22 2008, 11:53) *

... haven't heard any sound artifacts at -q 2.5 --lowpass 17000.

Thank you, Mardel.


QUOTE(Nick.C @ May 22 2008, 13:27) *

If I keep the spreading_function string at '22222-22222-22223-12224-12234-12345' but reduce the minimum_bits_to_keep to 2.333 and 2.667 respectively for -q 0 and -q 1, ...

I am having trouble to keep up with the changes.
What's exactly the difference between 1.0.1h and 1.0.1f?
1.0.1h spf string = '22222-22222-22223-12224-12234-12345'? What was it with 1.0.1f?
1.0.1h minimum_bits_to_keep to 2.333 and 2.667 respectively for -q 0 and -q 1? What was it with 1.0.1f?
Any other changes?

Anyway thank you for your restless struggle for improvement.
Nick.C
May 22 2008, 13:13
QUOTE(halb27 @ May 22 2008, 19:50) *
QUOTE(Nick.C @ May 22 2008, 13:27) *
If I keep the spreading_function string at '22222-22222-22223-12224-12234-12345' but reduce the minimum_bits_to_keep to 2.333 and 2.667 respectively for -q 0 and -q 1, ...
I am having trouble to keep up with the changes.
What's exactly the difference between 1.0.1h and 1.0.1f?
1.0.1h spf string = '22222-22222-22223-12224-12234-12345'? What was it with 1.0.1f?
1.0.1h minimum_bits_to_keep to 2.333 and 2.667 respectively for -q 0 and -q 1? What was it with 1.0.1f?
Any other changes?

Anyway thank you for your restless struggle for improvement.
1.0.1f had the existing --spf string, 1.0.1g & h have the modified one, although the minimum_bits_to_keep was:

2.000 and 2.500 for -q 0 and -q 1 respectively in 1.0.1f.
3.000 and 3.000 for -q 0 and -q 1 respectively in 1.0.1g;
2.333 and 2.667 for -q 0 and -q 1 respectively in 1.0.1h;

I increased the "resolution" of the threshold index so that fewer bits are potentially lost due to the requirement to floor the spreading_function result to read the corresponding bits_to_remove value from the threshold_index array. By widening the threshold_index array by a factor (now 64) the flooring of (spreading_function result * 64) has less of an impact.

I will leave --lowpass implemented as it will give those with more acute hearing some scope to include frequencies in the analyses above the existing 16kHz limit.
halb27
May 22 2008, 13:33
QUOTE(Nick.C @ May 22 2008, 21:13) *

1.0.1f had the existing --spf string, 1.0.1g & h have the modified one, although the minimum_bits_to_keep was:

2.000 and 2.500 for -q 0 and -q 1 respectively in 1.0.1f.
3.000 and 3.000 for -q 0 and -q 1 respectively in 1.0.1g;
2.333 and 2.667 for -q 0 and -q 1 respectively in 1.0.1h;

I increased the "resolution" of the threshold index ...

Thank you, Nick. So we have 1.0.1f's spf string with 1.0.1h - I appreciate this very much.
The other things make lossyWAV more defensive without a remarkable bad impact on bitrate as I understand it. So that's fine.
Nick.C
May 22 2008, 13:34
QUOTE(lvqcl @ May 22 2008, 18:56) *
With this sample and -q 0 setting, it looks like level of added noise changes very frequently, and this modulation is more noticeable than the noise itself.
I'll have a look at the bits-to-remove detail for the sample at -q 0 and see if it is large stepping in the bits-to-remove (i.e. added noise) - from what you say, it would seem to be.
sauvage78
May 22 2008, 13:55
halb27:
you're right -q 2 was near transparent with 1.0.1f & it is even more with 1.0.1h.
I tried to quickly ABX 1.0.1h -q2 vs original but I gave up after 5min as it was a boring task, I think it is still ABXable but it will take me 40 min just to prove to others something I am already sure for myself & even if -q2 is ABXable the fact that I can hear a difference doesn't mean that this difference sounds ugly.

So yes -q2 was listenable for DAP with this specific sample, specially in a noisy environnement, but my personal interest for lossywav is not in DAP even if I like the flexibility of using any lossless format.

Nick.C:
I was thinking lossywav was just removing unearable random noise in the back that takes much space to losslessly encode, but obviously it also add its own noise. I was thinking lossywav was as a kind of "lossless audio cleaner" by optimizing wav before encoding to lossless, I was thinking lossywav was not introducing "quantization like noise", but I was wrong it seems that even without quantization it add lossy-like additionnal noise. So I haven't followed all the lossywav developments, but now I am interested in knowing if lossywav applies some encoding strategy at some settings & not at some other settings. Even without understanding in deep the used techniques, just to understand the logical step within the scale.

Why is Q0 to Q4 for DAP, Q5 magically called "transparent", Q6 & Q7 as good as lossless & Q8 to Q10 for transcoding ??? My technical knowledge is very limited but I would like to know if these claims are rationnal or just arbitrary. Is it just me or lossywav have become more lossy than it was supposed to be in the beginning ? If there is no different techniques at different settings isn't the scale a little too wide ? Wouldn't it be logic to have only 2 recommended settings within the scale covering each specific use:

"-near lossy" for DAP, covering Q0 to Q5 use, targetting transparency.
"-near lossless" for archiving, covering Q6 to Q10 use, targetting transcodable archiving.

I don't mean that the actual scale is bad, I like it, freedom is always good. But it is missleading for newbies as I ended using Q0 for archive thinking lossywav was overkill anyway no matter the setting. I dunno but

"-near lossy" could be the same as let's say -q 3 (ABXing can help here)
"-near lossless" could be the same as ... well I don't know but the least agressive possible while still optimizing the lossless encoding. (ABXing is useless here, only technical knowledge so I let the codecs developpers decide together, not only you Nick.C ... Monty, Ivan, Gabriel may have an hintfull opinion here)

I know this is still very arbitrary but it would help newbies using lossywav as it should, & not lowering its reputation ...

for me for there is only 4 uses for hybrid:
100% lossless for Archive + "-near lossy" for DAP.
"-near lossless" for Archive + 100% lossy for DAP.
"-near lossless" for Archive + "-near lossy" for DAP.
hybrid lossless part for Archive + hybrid lossy part for listening.

so in the end two well tuned settings can cover it all.
Nick.C
May 22 2008, 14:32
QUOTE(sauvage78 @ May 22 2008, 20:55) *
halb27:
you're right -q 2 was near transparent with 1.0.1f & it is even more with 1.0.1h.
I tried to quickly ABX 1.0.1h -q2 vs original but I gave up after 5min as it was a boring task, I think it is still ABXable but it will take me 40 min just to prove to others something I am already sure for myself & even if -q2 is ABXable the fact that I can hear a difference doesn't mean that this difference sounds ugly.

So yes -q2 was listenable for DAP with this specific sample, specially in a noisy environnement, but my personal interest for lossywav is not in DAP even if I like the flexibility of using any lossless format.

Nick.C:
I was thinking lossywav was just removing unearable random noise in the back that takes much space to losslessly encode, but obviously it also add its own noise. I was thinking lossywav was as a kind of "lossless audio cleaner" by optimizing wav before encoding to lossless, I was thinking lossywav was not introducing "quantization like noise", but I was wrong it seems that even without quantization it add lossy-like additionnal noise. So I haven't followed all the lossywav developments, but now I am interested in knowing if lossywav applies some encoding strategy at some settings & not at some other settings. Even without understanding in deep the used techniques, just to understand the logical step within the scale.

Why is Q0 to Q4 for DAP, Q5 magically called "transparent", Q6 & Q7 as good as lossless & Q8 to Q10 for transcoding ??? My technical knowledge is very limited but I would like to know if these claims are rationnal or just arbitrary. Is it just me or lossywav have become more lossy than it was supposed to be in the beginning ? If there is no different techniques at different settings isn't the scale is little too wide ? Wouldn't it be logic to have only 2 recommended settings covering each specific use:

-near lossy for DAP, covering Q0 to Q5, targetting transparency.
-near lossless for archiving, covering Q6 to Q10, targetting transcoding.

I don't mean that the actual scale is bad, I like it, freedom is always good. But it is missleading for newbies as I ended using Q0 for archive thinking lossywav was overkill anyway no matter the setting. I dunno but

-near lossy could be the same as -q 3 (ABXing can help here)
-near lossless could be the same as ... well I don't know but the least agressive possible while still optimizing the lossless encoding. (ABXing is useless here, only technical knowledge so let the codecs developpers decide together, not only Nick.C)

I know this is very arbitrary but it would help newbies using lossywav as it should, & not lowering its reputation ...
lossyWAV has always added noise to the output, see the original lossyFLAC post started by David (2Bdecided) - the method used is his.

The Delphi transcode of the Matlab original started as just that - an attempt to duplicate David's method in a form that more users could make use of - Matlab is after all rather expensive....

Once agreement between both versions was achieved (about 0.3.0 roughly from memory), other ideas blossomed and added to the method.

The preset scale started as -1, -2 and -3. -1 = best, -2 = transparent, -3 = more aggressive. This evolved over time into -1 to -5, -1 to -7 and finally -q 10 to -q 0. As it states in the help:
CODE
Quality Options:

-q, --quality <n>   quality preset (10=highest quality, 0=lowest bitrate;
                    -q 5 is generally accepted to be transparent)
                    default=-q 5.
On the subject of transparency, as many problem samples as have been pointed out to halb27 and me have been included in the refinement of the quality presets. v0.6.4 RC1 was thought to be good until the day after release Alex B found a problem sample that halb27 also ABX'ed.

Transparency at -q 5 is not an arbitrary claim - although not every possible sample has been tested, therefore it is still "only" a claim (it should be noted that a very small group of users has been involved in the developmental ABX'ing process).

However, unless the optional (and still beta) noise shaping is selected, the actual bit removal process only rounds off the least significant <bits-to-remove> bits of each sample in a codec-block, thus adding white noise. If noise shaping is selected, the added noise is modified and a lot of it is pushed much higher in the spectrum (thanks to SebastianG's noise shaping coefficients).

Another, more advanced, method of noise shaping is in the early stages of development at present - but don't hold your breath.

Personally, I only use lossyWAV & FLAC as a DAP transcode - my lossless FLAC archive will remain intact.

The bottom line on the quality presets: 10 = best, 0 = worst - use one which you are comfortable with.

Thanks for taking the time to post. smile.gif

Nick.

QUOTE(Nick.C @ May 22 2008, 20:34) *
QUOTE(lvqcl @ May 22 2008, 18:56) *
With this sample and -q 0 setting, it looks like level of added noise changes very frequently, and this modulation is more noticeable than the noise itself.
I'll have a look at the bits-to-remove detail for the sample at -q 0 and see if it is large stepping in the bits-to-remove (i.e. added noise) - from what you say, it would seem to be.
I had a look:
CODE
Z:\WAV\tmp>lossywav "..\_swav\01 - Ginnungagap - The Black Hole.original.wav" -q 0 --detail -f
lossyWAV beta 1.0.1h, Copyright © 2007,2008 Nick Currie. Copyleft.
This is free software under the GNU GPLv3+ license; There is NO WARRANTY, to
the extent permitted by law. <http://www.gnu.org/licenses/> for details.
%lossyWAV Warning% : Detailled output mode enabled
Processing : 01 - Ginnungagap - The Black Hole.original.wav
Format : 44.10kHz; 2 ch.; 16 bit.
Progress :
Block Time 00 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 Tot.
====================================================================
0 0.00s. 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
16 0.19s. 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
32 0.37s. 5 7 10 10 10 9 9 10 9 9 9 8 7 7 9 8 136
48 0.56s. 9 11 10 10 10 10 9 10 9 9 9 9 9 9 8 9 150
64 0.74s. 7 9 8 8 6 7 6 8 7 7 7 7 7 7 8 10 119
80 0.93s. 10 10 10 9 9 9 9 8 8 8 8 9 8 8 9 11 143
96 1.11s. 10 11 11 11 11 11 11 11 11 10 11 10 10 10 10 10 169
112 1.30s. 10 10 10 9 10 10 9 9 10 10 10 10 9 9 9 10 154
128 1.49s. 9 10 9 9 10 10 10 9 9 10 10 9 8 9 9 8 148
144 1.67s. 9 9 9 9 9 8 9 9 9 9 9 9 9 9 8 8 141
160 1.86s. 8 9 8 9 9 9 9 9 9 8 9 9 9 9 9 8 140
176 2.04s. 8 8 9 8 8 8 9 9 8 9 8 8 9 8 8 8 133
192 2.23s. 7 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 127
208 2.41s. 7 8 7 7 8 8 7 8 8 7 8 8 7 7 7 8 120
224 2.60s. 7 8 7 7 6 8 7 8 7 7 7 7 7 7 7 6 113
240 2.79s. 7 7 8 6 7 6 7 6 7 7 6 7 7 7 7 7 109
256 2.97s. 7 7 7 7 7 7 6 7 7 7 7 6 6 7 7 7 109
272 3.16s. 6 6 7 7 7 7 7 8 8 7 8 8 8 8 8 8 118
288 3.34s. 7 8 7 8 7 7 8 7 7 7 7 7 7 7 7 7 115
304 3.53s. 7 7 7 8 8 8 7 7 7 7 7 8 7 7 8 6 116
320 3.72s. 7 7 7 7 7 7 7 7 7 7 7 7 7 6 6 7 110
336 3.90s. 7 7 5 7 7 6 7 6 6 6 7 6 6 7 7 7 104
352 4.09s. 7 7 7 7 7 7 6 7 6 6 7 7 6 7 7 7 108
368 4.27s. 6 7 6 6 6 6 6 7 7 7 7 6 7 6 7 6 103
384 4.46s. 6 7 6 6 6 7 6 6 6 6 6 6 6 6 7 6 99
400 4.64s. 6 6 7 7 6 6 7 6 6 7 7 7 7 6 6 6 103
416 4.83s. 6 6 5 6 6 6 6 6 6 6 6 6 6 5 6 5 93
432 5.02s. 6 6 5 6 6 6 6 6 6 6 6 5 6 5 6 6 93
448 5.20s. 5 6 6 6 6 6 6 6 5 5 5 5 6 5 5 6 89
464 5.39s. 6 6 6 5 5 6 6 6 6 6 5 6 6 5 5 5 90
480 5.57s. 5 6 5 5 6 6 6 6 6 6 5 5 6 6 6 6 91
496 5.76s. 6 5 5 5 6 5 5 6 6 6 6 6 5 6 5 6 89
512 5.94s. 6 5 5 5 5 5 6 6 10 10 10 10 9 9 9 10 120
528 6.13s. 9 8 8 8 9 9 8 10 10 10 9 10 10 9 9 9 145
544 6.32s. 9 9 8 8 8 7 8 7 8 8 8 7 7 7 7 7 123
560 6.50s. 7 6 7 7 7 10 10 10 10 10 9 8 9 9 9 9 137
576 6.69s. 8 8 8 8 8 11 11 11 11 11 10 11 10 11 10 11 158
592 6.87s. 10 11 11 10 10 10 10 10 10 10 10 9 10 10 9 9 159
608 7.06s. 9 9 9 10 10 10 9 9 9 10 10 10 10 10 10 9 153
624 7.24s. 10 9 9 9 9 9 9 9 10 10 9 9 10 9 9 9 148
640 7.43s. 9 9 10 9 9 9 7 9 9 9 9 9 8 9 9 8 141
656 7.62s. 9 9 9 8 9 9 9 8 9 8 9 8 8 8 8 8 136
672 7.80s. 9 8 8 8 8 8 7 7 8 9 9 8 8 9 8 7 129
688 7.99s. 7 7 8 8 8 8 8 8 9 7 7 8 7 7 8 7 122
704 8.17s. 8 8 8 8 7 7 7 7 8 6 7 7 6 7 8 6 115
720 8.36s. 7 7 8 8 7 7 7 7 7 8 7 7 6 6 7 6 112
736 8.54s. 7 7 6 7 5 6 7 7 6 6 5 7 6 7 7 7 103
752 8.73s. 7 7 6 6 6 7 7 7 7 7 7 8 8 7 8 8 113
768 8.92s. 8 8 8 8 7 7 7 7 7 7 7 6 6 7 8 7 115
784 9.10s. 8 7 7 7 7 6 8 7 7 7 7 7 7 8 8 7 115
800 9.29s. 7 7 7 7 7 7 7 8 8 7 7 7 7 7 7 7 114
816 9.47s. 7 7 6 6 7 7 7 7 8 8 7 7 8 7 7 7 113
832 9.66s. 7 7 7 7 7 7 6 7 7 8 7 7 7 6 6 7 110
848 9.85s. 7 6 7 5 7 7 6 6 7 7 7 7 7 7 7 7 107
864 10.03s. 6 6 6 7 6 7 7 6 7 5 6 6 4 6 7 7 99
880 10.22s. 7 6 6 7 8 7 8 7 8 7 8 8 8 6 8 8 117
896 10.40s. 8 8 8 7 7 7 8 7 8 7 7 7 7 8 7 7 118
912 10.59s. 7 7 6 7 7 7 7 7 5 7 6 6 7 6 6 7 105
928 10.77s. 7 7 7 6 6 7 7 7 7 7 7 7 7 7 8 7 111
944 10.96s. 7 8 7 7 7 7 8 8 7 7 8 8 7 7 7 7 117
960 11.15s. 7 7 7 7 7 7 7 7 7 7 7 7 8 7 7 7 113
976 11.33s. 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 112
992 11.52s. 7 7 7 6 6 7 7 7 7 6 7 7 7 6 8 11 113
1008 11.70s. 10 10 10 10 10 9 9 9 8 8 8 8 8 9 10 9 145
1024 11.89s. 10 9 9 10 10 9 9 9 9 8 8 9 8 8 9 7 141
1040 12.07s. 7 7 8 8 7 8 7 7 7 6 7 8 10 10 10 9 126
1056 12.26s. 10 9 9 9 9 8 8 8 8 8 8 8 11 11 11 11 146
1072 12.45s. 11 11 10 11 10 11 11 10 11 10 11 10 10 10 10 10 167
1088 12.63s. 10 10 10 9 9 9 7 8 9 10 9 10 9 9 9 10 147
1104 12.82s. 10 8 10 10 9 10 10 10 9 10 9 9 9 10 9 10 152
1120 13.00s. 7 9 9 8 9 9 8 9 9 9 9 9 9 8 9 9 139
1136 13.19s. 8 9 9 9 9 9 9 9 7 9 9 9 9 9 8 7 138
1152 13.37s. 9 7 8 8 8 8 8 8 8 8 8 8 8 8 8 8 128
1168 13.56s. 7 8 8 7 8 5 8 8 8 8 8 7 7 7 8 8 120
1184 13.75s. 7 8 8 8 8 8 8 8 8 8 7 8 7 7 8 7 123
1200 13.93s. 7 7 8 7 7 7 7 7 6 8 8 7 7 7 7 6 113
1216 14.12s. 6 7 6 7 7 7 7 6 7 8 6 6 7 7 7 7 108
1232 14.30s. 7 7 7 7 7 7 6 7 7 7 7 7 7 7 7 7 111
1248 14.49s. 7 7 7 8 8 8 8 8 7 8 6 8 7 6 6 7 116
1264 14.68s. 8 7 8 7 6 8 7 8 8 8 7 8 6 7 8 7 118
1280 14.86s. 7 8 8 7 7 7 7 7 7 7 7 7 6 6 7 6 111
1296 15.05s. 7 7 7 7 7 6 7 7 6 6 7 7 7 7 7 7 109
1312 15.23s. 7 6 7 7 7 7 7 7 7 7 6 6 7 6 7 6 107
1328 15.42s. 7 7 6 6 7 7 6 7 7 7 7 7 7 7 7 7 109
1344 15.60s. 7 5 6 6 6 7 7 6 6 6 6 7 7 6 6 5 99
1360 15.79s. 6 7 6 7 6 6 6 6 6 6 6 6 6 7 7 7 101
1376 15.98s. 7 6 7 5 5 6 6 6 6 6 6 6 6 6 6 6 96
1392 16.16s. 7 6 7 5 6 6 5 7 6 5 5 6 6 5 6 6 94
1408 16.35s. 6 6 6 6 6 6 6 5 6 6 6 6 6 6 5 5
Average : 2.78MB; 7.4112 bits; 32.93x; 0.50s;
%lossyWAV Warning% : 0.0112 bits not removed due to clipping.
%lossyWAV Warning% : 17 sample(s) clipped to limiting amplitude.

CODE
Z:\WAV\tmp>lossywav "..\_swav\01 - Ginnungagap - The Black Hole.original.wav" -q 0 --detail -f --minbits 3
lossyWAV beta 1.0.1h, Copyright © 2007,2008 Nick Currie. Copyleft.
This is free software under the GNU GPLv3+ license; There is NO WARRANTY, to
the extent permitted by law. <http://www.gnu.org/licenses/> for details.
%lossyWAV Warning% : Detailled output mode enabled
Processing : 01 - Ginnungagap - The Black Hole.original.wav
Format : 44.10kHz; 2 ch.; 16 bit.
Progress :
Block Time 00 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 Tot.
====================================================================
0 0.00s. 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
16 0.19s. 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
32 0.37s. 5 7 10 10 10 9 9 10 9 9 9 8 7 7 9 8 136
48 0.56s. 9 10 10 10 10 10 9 10 9 9 9 9 9 9 8 9 149
64 0.74s. 7 9 8 8 6 7 6 8 7 7 7 7 7 7 8 10 119
80 0.93s. 10 10 10 9 9 9 9 8 8 8 8 9 8 8 9 10 142
96 1.11s. 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 160
112 1.30s. 9 10 10 9 10 9 9 9 10 10 10 10 9 9 9 10 152
128 1.49s. 9 10 9 9 9 10 10 9 9 10 10 9 8 9 9 8 147
144 1.67s. 9 9 9 9 9 8 9 9 9 9 9 9 9 9 8 8 141
160 1.86s. 8 9 8 9 9 9 9 9 9 8 9 9 9 9 9 8 140
176 2.04s. 8 8 9 8 8 8 9 9 8 9 8 8 9 8 8 8 133
192 2.23s. 7 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 127
208 2.41s. 7 8 7 7 8 8 7 8 8 7 8 8 7 7 7 8 120
224 2.60s. 7 8 7 7 6 8 7 8 7 7 7 7 7 7 7 6 113
240 2.79s. 7 7 8 6 7 6 7 6 7 7 6 7 7 7 7 7 109
256 2.97s. 7 7 7 7 7 7 6 7 7 7 7 6 6 7 7 7 109
272 3.16s. 6 6 7 7 7 7 7 8 8 7 8 8 8 8 8 8 118
288 3.34s. 7 8 7 8 7 7 8 7 7 7 7 7 7 7 7 7 115
304 3.53s. 7 7 7 8 8 8 7 7 7 7 7 8 7 7 8 6 116
320 3.72s. 7 7 7 7 7 7 7 7 7 7 7 7 7 6 6 7 110
336 3.90s. 7 7 5 7 7 6 7 6 6 6 7 6 6 7 7 7 104
352 4.09s. 7 7 7 7 7 7 6 7 6 6 7 7 6 7 7 7 108
368 4.27s. 6 7 6 6 6 6 6 7 7 7 7 6 7 6 7 6 103
384 4.46s. 6 7 6 6 6 7 6 6 6 6 6 6 6 6 7 6 99
400 4.64s. 6 6 7 7 6 6 7 6 6 7 7 7 7 6 6 6 103
416 4.83s. 6 6 5 6 6 6 6 6 6 6 6 6 6 5 6 5 93
432 5.02s. 6 6 5 6 6 6 6 6 6 6 6 5 6 5 6 6 93
448 5.20s. 5 6 6 6 6 6 6 6 5 5 5 5 6 5 5 6 89
464 5.39s. 6 6 6 5 5 6 6 6 6 6 5 6 6 5 5 5 90
480 5.57s. 5 6 5 5 6 6 6 6 6 6 5 5 6 6 6 6 91
496 5.76s. 6 5 5 5 6 5 5 6 6 6 6 6 5 6 5 6 89
512 5.94s. 6 5 5 5 5 5 6 6 10 10 10 10 9 9 9 10 120
528 6.13s. 9 8 8 8 9 9 8 10 10 10 9 10 10 9 9 9 145
544 6.32s. 9 9 8 8 8 7 8 7 8 8 8 7 7 7 7 7 123
560 6.50s. 7 6 7 7 7 10 10 10 10 10 9 8 9 9 9 9 137
576 6.69s. 8 8 8 8 8 11 11 11 10 10 10 10 10 10 10 10 153
592 6.87s. 10 10 10 10 10 10 10 10 10 10 10 9 10 10 9 9 157
608 7.06s. 9 9 9 10 10 10 9 9 9 10 10 10 10 10 10 9 153
624 7.24s. 10 9 9 9 9 9 9 9 10 9 9 9 9 9 9 9 146
640 7.43s. 9 9 10 9 9 9 7 9 9 9 9 9 8 9 9 8 141
656 7.62s. 9 9 9 8 9 9 9 8 9 8 9 8 8 8 8 8 136
672 7.80s. 9 8 8 8 8 8 7 7 8 9 9 8 8 9 8 7 129
688 7.99s. 7 7 8 8 8 8 8 8 9 7 7 8 7 7 8 7 122
704 8.17s. 8 8 8 8 7 7 7 7 8 6 7 7 6 7 8 6 115
720 8.36s. 7 7 8 8 7 7 7 7 7 8 7 7 6 6 7 6 112
736 8.54s. 7 7 6 7 5 6 7 7 6 6 5 7 6 7 7 7 103
752 8.73s. 7 7 6 6 6 7 7 7 7 7 7 8 8 7 8 8 113
768 8.92s. 8 8 8 8 7 7 7 7 7 7 7 6 6 7 8 7 115
784 9.10s. 8 7 7 7 7 6 8 7 7 7 7 7 7 8 8 7 115
800 9.29s. 7 7 7 7 7 7 7 8 8 7 7 7 7 7 7 7 114
816 9.47s. 7 7 6 6 7 7 7 7 8 8 7 7 8 7 7 7 113
832 9.66s. 7 7 7 7 7 7 6 7 7 8 7 7 7 6 6 7 110
848 9.85s. 7 6 7 5 7 7 6 6 7 7 7 7 7 7 7 7 107
864 10.03s. 6 6 6 7 6 7 7 6 7 5 6 6 4 6 7 7 99
880 10.22s. 7 6 6 7 8 7 8 7 8 7 8 8 8 6 8 8 117
896 10.40s. 8 8 8 7 7 7 8 7 8 7 7 7 7 8 7 7 118
912 10.59s. 7 7 6 7 7 7 7 7 5 7 6 6 7 6 6 7 105
928 10.77s. 7 7 7 6 6 7 7 7 7 7 7 7 7 7 8 7 111
944 10.96s. 7 8 7 7 7 7 8 8 7 7 8 8 7 7 7 7 117
960 11.15s. 7 7 7 7 7 7 7 7 7 7 7 7 8 7 7 7 113
976 11.33s. 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 112
992 11.52s. 7 7 7 6 6 7 7 7 7 6 7 7 7 6 8 11 113
1008 11.70s. 10 10 10 10 10 9 9 9 8 8 8 8 8 9 10 9 145
1024 11.89s. 10 9 9 10 9 9 9 9 9 8 8 9 8 8 9 7 140
1040 12.07s. 7 7 8 8 7 8 7 7 7 6 7 8 10 10 10 9 126
1056 12.26s. 10 9 9 9 9 8 8 8 8 8 8 8 11 11 11 10 145
1072 12.45s. 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 160
1088 12.63s. 10 10 10 9 9 9 7 8 9 10 9 10 9 9 9 10 147
1104 12.82s. 10 8 9 10 9 10 10 10 9 10 9 9 9 10 9 10 151
1120 13.00s. 7 9 9 8 9 9 8 9 9 9 9 9 9 8 9 9 139
1136 13.19s. 8 9 9 9 9 9 9 9 7 9 9 9 9 9 8 7 138
1152 13.37s. 9 7 8 8 8 8 8 8 8 8 8 8 8 8 8 8 128
1168 13.56s. 7 8 8 7 8 5 8 8 8 8 8 7 7 7 8 8 120
1184 13.75s. 7 8 8 8 8 8 8 8 8 8 7 8 7 7 8 7 123
1200 13.93s. 7 7 8 7 7 7 7 7 6 8 8 7 7 7 7 6 113
1216 14.12s. 6 7 6 7 7 7 7 6 7 8 6 6 7 7 7 7 108
1232 14.30s. 7 7 7 7 7 7 6 7 7 7 7 7 7 7 7 7 111
1248 14.49s. 7 7 7 8 8 8 8 8 7 8 6 8 7 6 6 7 116
1264 14.68s. 8 7 8 7 6 8 7 8 8 8 7 8 6 7 8 7 118
1280 14.86s. 7 8 8 7 7 7 7 7 7 7 7 7 6 6 7 6 111
1296 15.05s. 7 7 7 7 7 6 7 7 6 6 7 7 7 7 7 7 109
1312 15.23s. 7 6 7 7 7 7 7 7 7 7 6 6 7 6 7 6 107
1328 15.42s. 7 7 6 6 7 7 6 7 7 7 7 7 7 7 7 7 109
1344 15.60s. 7 5 6 6 6 7 7 6 6 6 6 7 7 6 6 5 99
1360 15.79s. 6 7 6 7 6 6 6 6 6 6 6 6 6 7 7 7 101
1376 15.98s. 7 6 7 5 5 6 6 6 6 6 6 6 6 6 6 6 96
1392 16.16s. 7 6 7 5 6 6 5 7 6 5 5 6 6 5 6 6 94
1408 16.35s. 6 6 6 6 6 6 6 5 6 6 6 6 6 6 5 5
Average : 2.78MB; 7.3919 bits; 38.17x; 0.43s;
%lossyWAV Warning% : 0.0056 bits not removed due to clipping.
%lossyWAV Warning% : 3 sample(s) clipped to limiting amplitude.

CODE
Z:\WAV\tmp>lossywav "..\_swav\01 - Ginnungagap - The Black Hole.original.wav" -q 0 --detail -f --minbits 3.5
lossyWAV beta 1.0.1h, Copyright © 2007,2008 Nick Currie. Copyleft.
This is free software under the GNU GPLv3+ license; There is NO WARRANTY, to
the extent permitted by law. <http://www.gnu.org/licenses/> for details.
%lossyWAV Warning% : Detailled output mode enabled
Processing : 01 - Ginnungagap - The Black Hole.original.wav
Format : 44.10kHz; 2 ch.; 16 bit.
Progress :
Block Time 00 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 Tot.
====================================================================
0 0.00s. 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
16 0.19s. 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
32 0.37s. 5 7 10 10 10 9 9 9 9 9 9 8 7 7 9 8 135
48 0.56s. 9 9 10 10 10 9 9 9 9 9 9 9 9 9 8 9 146
64 0.74s. 7 9 8 8 6 7 6 8 7 7 7 7 7 7 8 10 119
80 0.93s. 10 10 10 9 9 9 9 8 8 8 8 9 8 8 9 10 142
96 1.11s. 10 10 10 9 10 9 9 10 9 10 9 9 9 9 10 9 151
112 1.30s. 9 9 9 9 9 9 9 9 9 9 9 10 9 9 9 9 145
128 1.49s. 9 9 9 9 9 9 9 9 9 9 9 9 8 9 9 8 142
144 1.67s. 9 9 9 9 9 8 9 9 9 9 9 9 9 9 8 8 141
160 1.86s. 8 9 8 9 9 9 9 9 9 8 9 9 9 9 9 8 140
176 2.04s. 8 8 9 8 8 8 9 9 8 9 8 8 9 8 8 8 133
192 2.23s. 7 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 127
208 2.41s. 7 8 7 7 8 8 7 8 8 7 8 8 7 7 7 8 120
224 2.60s. 7 8 7 7 6 8 7 8 7 7 7 7 7 7 7 6 113
240 2.79s. 7 7 8 6 7 6 7 6 7 7 6 7 7 7 7 7 109
256 2.97s. 7 7 7 7 7 7 6 7 7 7 7 6 6 7 7 7 109
272 3.16s. 6 6 7 7 7 7 7 8 8 7 8 8 8 8 8 8 118
288 3.34s. 7 8 7 8 7 7 8 7 7 7 7 7 7 7 7 7 115
304 3.53s. 7 7 7 8 8 8 7 7 7 7 7 8 7 7 8 6 116
320 3.72s. 7 7 7 7 7 7 7 7 7 7 7 7 7 6 6 7 110
336 3.90s. 7 7 5 7 7 6 7 6 6 6 7 6 6 7 7 7 104
352 4.09s. 7 7 7 7 7 7 6 7 6 6 7 7 6 7 7 7 108
368 4.27s. 6 7 6 6 6 6 6 7 7 7 7 6 7 6 7 6 103
384 4.46s. 6 7 6 6 6 7 6 6 6 6 6 6 6 6 7 6 99
400 4.64s. 6 6 7 7 6 6 7 6 6 7 7 7 7 6 6 6 103
416 4.83s. 6 6 5 6 6 6 6 6 6 6 6 6 6 5 6 5 93
432 5.02s. 6 6 5 6 6 6 6 6 6 6 6 5 6 5 6 6 93
448 5.20s. 5 6 6 6 6 6 6 6 5 5 5 5 6 5 5 6 89
464 5.39s. 6 6 6 5 5 6 6 6 6 6 5 6 6 5 5 5 90
480 5.57s. 5 6 5 5 6 6 6 6 6 6 5 5 6 6 6 6 91
496 5.76s. 6 5 5 5 6 5 5 6 6 6 6 6 5 6 5 6 89
512 5.94s. 6 5 5 5 5 5 6 6 10 10 10 10 9 9 9 9 119
528 6.13s. 9 8 8 8 9 9 8 10 10 10 9 10 9 9 9 9 144
544 6.32s. 9 9 8 8 8 7 8 7 8 8 8 7 7 7 7 7 123
560 6.50s. 7 6 7 7 7 10 10 10 10 10 9 8 9 9 9 9 137
576 6.69s. 8 8 8 8 8 10 10 11 10 10 9 10 9 10 10 10 149
592 6.87s. 10 10 10 9 9 10 9 10 9 9 9 9 9 9 9 9 149
608 7.06s. 9 9 9 9 10 9 9 9 9 9 9 9 9 9 9 9 145
624 7.24s. 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 144
640 7.43s. 9 9 9 9 9 9 7 9 9 9 9 9 8 9 9 8 140
656 7.62s. 9 9 9 8 9 9 9 8 9 8 9 8 8 8 8 8 136
672 7.80s. 9 8 8 8 8 8 7 7 8 9 9 8 8 9 8 7 129
688 7.99s. 7 7 8 8 8 8 8 8 8 7 7 8 7 7 8 7 121
704 8.17s. 8 8 8 8 7 7 7 7 8 6 7 7 6 7 8 6 115
720 8.36s. 7 7 8 8 7 7 7 7 7 8 7 7 6 6 7 6 112
736 8.54s. 7 7 6 7 5 6 7 7 6 6 5 7 6 7 7 7 103
752 8.73s. 7 7 6 6 6 7 7 7 7 7 7 8 8 7 8 8 113
768 8.92s. 8 8 8 8 7 7 7 7 7 7 7 6 6 7 8 7 115
784 9.10s. 8 7 7 7 7 6 8 7 7 7 7 7 7 8 8 7 115
800 9.29s. 7 7 7 7 7 7 7 8 8 7 7 7 7 7 7 7 114
816 9.47s. 7 7 6 6 7 7 7 7 8 8 7 7 8 7 7 7 113
832 9.66s. 7 7 7 7 7 7 6 7 7 8 7 7 7 6 6 7 110
848 9.85s. 7 6 7 5 7 7 6 6 7 7 7 7 7 7 7 7 107
864 10.03s. 6 6 6 7 6 7 7 6 7 5 6 6 4 6 7 7 99
880 10.22s. 7 6 6 7 8 7 8 7 8 7 8 8 8 6 8 8 117
896 10.40s. 8 8 8 7 7 7 8 7 8 7 7 7 7 8 7 7 118
912 10.59s. 7 7 6 7 7 7 7 7 5 7 6 6 7 6 6 7 105
928 10.77s. 7 7 7 6 6 7 7 7 7 7 7 7 7 7 8 7 111
944 10.96s. 7 8 7 7 7 7 8 8 7 7 8 8 7 7 7 7 117
960 11.15s. 7 7 7 7 7 7 7 7 7 7 7 7 8 7 7 7 113
976 11.33s. 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 112
992 11.52s. 7 7 7 6 6 7 7 7 7 6 7 7 7 6 8 10 112
1008 11.70s. 10 10 10 9 10 9 9 9 8 8 8 8 8 9 10 9 144
1024 11.89s. 10 9 9 9 9 9 9 9 9 8 8 9 8 8 9 7 139
1040 12.07s. 7 7 8 8 7 8 7 7 7 6 7 8 10 10 10 9 126
1056 12.26s. 9 9 9 9 9 8 8 8 8 8 8 8 10 10 10 10 141
1072 12.45s. 9 10 9 10 9 10 10 9 9 9 9 9 9 9 9 9 148
1088 12.63s. 9 10 9 9 9 9 7 8 9 9 9 9 9 9 9 9 142
1104 12.82s. 9 8 9 9 9 9 9 9 9 9 9 9 9 9 9 9 143
1120 13.00s. 7 9 9 8 9 9 8 9 9 9 9 9 9 8 9 9 139
1136 13.19s. 8 9 9 9 9 9 9 9 7 9 9 9 9 9 8 7 138
1152 13.37s. 9 7 8 8 8 8 8 8 8 8 8 8 8 8 8 8 128
1168 13.56s. 7 8 8 7 8 5 8 8 8 8 8 7 7 7 8 8 120
1184 13.75s. 7 8 8 8 8 8 8 8 8 8 7 8 7 7 8 7 123
1200 13.93s. 7 7 8 7 7 7 7 7 6 8 8 7 7 7 7 6 113
1216 14.12s. 6 7 6 7 7 7 7 6 7 8 6 6 7 7 7 7 108
1232 14.30s. 7 7 7 7 7 7 6 7 7 7 7 7 7 7 7 7 111
1248 14.49s. 7 7 7 8 8 8 8 8 7 8 6 8 7 6 6 7 116
1264 14.68s. 8 7 8 7 6 8 7 8 8 8 7 8 6 7 8 7 118
1280 14.86s. 7 8 8 7 7 7 7 7 7 7 7 7 6 6 7 6 111
1296 15.05s. 7 7 7 7 7 6 7 7 6 6 7 7 7 7 7 7 109
1312 15.23s. 7 6 7 7 7 7 7 7 7 7 6 6 7 6 7 6 107
1328 15.42s. 7 7 6 6 7 7 6 7 7 7 7 7 7 7 7 7 109
1344 15.60s. 7 5 6 6 6 7 7 6 6 6 6 7 7 6 6 5 99
1360 15.79s. 6 7 6 7 6 6 6 6 6 6 6 6 6 7 7 7 101
1376 15.98s. 7 6 7 5 5 6 6 6 6 6 6 6 6 6 6 6 96
1392 16.16s. 7 6 7 5 6 6 5 7 6 5 5 6 6 5 6 6 94
1408 16.35s. 6 6 6 6 6 6 6 5 6 6 6 6 6 6 5 5
Average : 2.78MB; 7.3322 bits; 32.50x; 0.51s;
%lossyWAV Warning% : 0.0014 bits not removed due to clipping.

The issue might be the 11 bits removed around 1.1 seconds. They are removed because the signal is considered by the algorithm to be strong enough to mask the noise - in this instance obviously not. More analysis of the spectral makeup of the sample around the 1.1 second mark is required.
halb27
May 22 2008, 14:34
QUOTE(sauvage78 @ May 22 2008, 21:55) *

...So yes -q2 was listenable for DAP with this specific sample, specially in a noisy environnement, but my personal interest for lossywav is not in DAP even if I like the flexibility of using any lossless format. ...

If keeping bitrate very low isn't your primary concern but quality is (like for me) you're better off using a higher quality setting like -q 4 or higher.

Nick C. may go into more detail, but I think I can answer some questions you addressed at Nick (hope you don't mind, Nick - oh, you were faster anyway).
lossyWAV definitely adds noise by rounding the wave samples in such a way that the least significant bits become zero, and a series of lossless codecs can make good use of this. Roughly speaking this is not done with quiet spots in the music but it is done when the music is loud. The added noise is controlled - it should be rarely audible with the lowest quality settings, it should be very rarely audible and - if it's audible - not annoying with the low quality settings, it should never be audible with the medium quality settings (though the focus here is also on bitrate and we don't allow for a safety margin), and it should be always inaudible with an ever increasing safety margin at the high quality settings.
Nick's description addresses this with emphasis on lossyWAV applications like DAP use or archiving/transcoding.
-q 5 is special as it takes exact care of the original lossyWAV idea originating from 2Bdecided: make some frequency analyses and use the number of bits to remove (zero out) depending on the signal's frequency region with lowest volume.
From experience so far the idea is correct and -q 5 is really transparent.
Below -q 5 2Bdecided's requirements are more and more softened, but NickC.'s other methods of controlling noise obviously do a very good job in keeping noise low. They can't work miracles however: the noise control of -q 0 has so low requirements that it's more remarkable that quality usually is still good than that problem samples exist.
Above -q 5 all quality requirements are fulfilled with an ever increasing added safety margin.
Originally there were just 3 quality settings: apart from the -q 5 related quality setting a setting with an additional safety margin for archiving quality, and another one with a softened quality requirement for DAP use.
This is clearer than the current settings. Movement came in when a certain demand came up for very low bitrate usage. It's all a matter of taste - I think it's a good thing to have a fine grained quality scale. Below -q 5 we're exactly in the same situation as the usual lossy codecs like mp3 for which everybody likes a quality scale to use according to personal quality and bitrate demands. Above -q 5 we have no analogue to the usual lossy codecs, but for the safety margin it's pretty much the same thing as for the lowering of the quality demands: I think it's good to be able to choose the degree of added security.
What has happened with these many quality settings is that the 'standard quality' has gotten a bit out of focus, though for good reason as I think -q 5 is overkill for the most of us, but obviously the feeling has gone that the more we lower the quality requirements compared to -q 5, the more natural is an increasing chance for audible noise.
sauvage78
May 22 2008, 15:06
I may look dumb, but with my little knowledge it sounds like lossywav is cleverly removing hard to compress inaudible noise & then cleverly dithering easy to compress inaudible noise ? no.
so the problem I heard was not a "quantization like distortion" but more a audible "bad dithering" ?

also where does low, medium & high setting starts & ends ... that's a big problem for a noob, with the old 3 settings best/transparent/more agressive it was easier not to make mistake, I think both a scale for freedoom & recommended settings for noobs are usefull & complementary.

not everyone is willing to annoy the dev with stupid questions or to read plenty of technical threads just to know how to use correctly the codec. I know where to find the original thread & wiki help, the wiki were too general, the original discussion too technical. (& I was too lazy wink.gif )

finally I don't get the use of any quality setting above Q5, if the original idea was proven correct so far what will a "more transparent than transparent" bring ? if it doesn't bring anything for transparency then it must bring at last some added safety for transcoding, is it technically the case at last on paper ? Is there 11 settings just to have 5 in the middle ? -q0 & -q1 were proven ABXable, you could as well remove -q9 & -q10 & have a 7 settings scale with -q3 in the middle being the actual -q5 ... it would be clearer. Anyway it's just my thoughts, I know I don't understand all the logic behind & I am not the dev afterall, in the end hybrid audio is a real headhache wink.gif

Thks for taking the time to answer.
Nick.C
May 22 2008, 15:17
QUOTE(sauvage78 @ May 22 2008, 22:06) *
I may look dumb, but with my little knowledge it sounds like lossywav is cleverly removing hard to compress inaudible noise & then cleverly dithering easy to compress inaudible noise ? no.
so the problem I heard was not a "quantization like distortion" but more a audible "bad dithering" ?
No question is dumb - ignoring the answer would be wink.gif. By default lossyWAV does not dither at all and does not know if any noise is easily compressed or not.

QUOTE(sauvage78 @ May 22 2008, 22:06) *
also where does low, medium & high setting starts & ends ... that's a big problem for a noob, with the old 3 settings best/transparent/more agressive it was easier not to make mistake, I think both a scale for freedoom & recommended settings for noobs are usefull & complementary.
Why not just use the default value - it is included just this reason.

QUOTE(sauvage78 @ May 22 2008, 22:06) *
not everyone is willing to annoy the dev with stupid questions or to read plenty of technical threads just to know how to use correctly the codec. I know where to find the original thread & wiki help, the wiki were too general, the original discussion too technical. (& I was too lazy wink.gif )
Maybe the wiki needs more work, however I thought it was clear enough....

QUOTE(sauvage78 @ May 22 2008, 22:06) *
Thks for taking the time to answer.
No problem.

Nick.

[edit]
QUOTE(sauvage78 @ May 22 2008, 22:06) *
finally I don't get the use of any quality setting above Q5, if the original idea was proven correct so far what will a "more transparent than transparent" bring ? if it doesn't bring anything for transparency then it must bring at last some added safety for transcoding, is it technically the case at last on paper ? Is there 11 settings just to have 5 in the middle ? -q0 & -q1 were proven ABXable, you could as well remove -q9 & -q10 & have a 6 settings scale with -q3 being the actual -q5 ... it would be clearer. Anyway it's just my thoughts, I know I don't understand all the logic behind & I am not the dev afterall, in the end hybrid audio is a real headhache wink.gif
You are correct, the higher the quality settings give less likelihood of problems in transcoding. I explained already the evolution of the numerous quality presets - different people have commented on the project at different times with with different aims - this has pulled the quality presets wider and wider. I also wanted to have a highest quality level which still gave some bitrate reduction compared to fully lossless (-q 10 gives at least a 20% bitrate saving on average). I cannot tell you which quality preset to use - that's up to your own evaluation and compromise between perceived quality and resulting bitrate when subsequently encoded.
[/edit]
lvqcl
May 22 2008, 15:57
QUOTE(Nick.C @ May 23 2008, 00:32) *

The issue might be the 11 bits removed around 1.1 seconds. They are removed because the signal is considered by the algorithm to be strong enough to mask the noise - in this instance obviously not. More analysis of the spectral makeup of the sample around the 1.1 second mark is required.

Well, the issue I hear is (I think) quick changes in noise intensity. They're highlighted in this picture:
IPB Image
It's the difference between original and lossy signals.

(btw: if you delete first 16463 samples (about 0.38 sec) then Adobe Audition says that the rest signal has only 15 bits huh.gif )
Nick.C
May 22 2008, 16:02
QUOTE(lvqcl @ May 22 2008, 22:57) *
QUOTE(Nick.C @ May 23 2008, 00:32) *
The issue might be the 11 bits removed around 1.1 seconds. They are removed because the signal is considered by the algorithm to be strong enough to mask the noise - in this instance obviously not. More analysis of the spectral makeup of the sample around the 1.1 second mark is required.
Well, the issue I hear is (I think) quick changes in noise intensity. They're highlighted in this picture:
IPB Image
It's the difference between original and lossy signals.

(btw: if you delete first 16463 samples (about 0.38 sec) then Adobe Audition says that the rest signal has only 15 bits huh.gif )
How long are the steps in the signal - if it's a multiple of 11.6msec (512 samples) then the steps are caused by different bits-to-remove being applied to the output and adding different intensities of white noise.
lvqcl
May 22 2008, 16:57
QUOTE(Nick.C @ May 23 2008, 02:02) *

How long are the steps in the signal - if it's a multiple of 11.6msec (512 samples) then the steps are caused by different bits-to-remove being applied to the output and adding different intensities of white noise.

Yes, they are.
But I can't see direct relationship between amplitude of noise and numbers that lossyWAV prints with --detail switch. blink.gif I can't find pattern
CODE
...
   80    0.93s. 10 10 10  9  9  9  9  8  8  8  8  9  8  8  9 11 143
   96    1.11s. 10 11 11 11 11 11 11 11 11 10 11 10 10 10 10 10 169
  112    1.30s. 10 10 10  9 10 10  9  9 10 10 10 10  9  9  9 10 154
...
shadowking
May 22 2008, 19:05
Maye we should add a sticky somewhere. Inexperienced users should not touch any quality setting and will use the defaulted Q5 ??
Nick.C
May 22 2008, 23:40
QUOTE(shadowking @ May 23 2008, 02:05) *
Maye we should add a sticky somewhere. Inexperienced users should not touch any quality setting and will use the defaulted Q5 ??
That sounds like an eminently sensible suggestion - it's now on my to do list for the wiki / --help text.
shadowking
May 23 2008, 01:21
I had another idea: create a few 'presets' that are mapped to common Q settings. Maybe even hide the Q scale from the normal screen and document it with --longhelp or similar.

-normal [Default] = Q5
-medium = Q3
-Portable = Q2

- Extreme / Archiving / transcoding = Q6..10
Nick.C
May 23 2008, 01:34
QUOTE(shadowking @ May 23 2008, 08:21) *
I had another idea: create a few 'presets' that are mapped to common Q settings;

-normal [Default] = Q5
-medium = Q3
-Portable = Q2

- Archiving / transcoding = Q7
Good suggestion - also added to to-do-list. Suggested names and associated -q value:

--highest := -q 10;
--high := -q 7.5
--archive := -q 7.5
--default := -q 5;
--normal := -q 5;
--portable := -q 2.5.
--nasty := -q 0 wink.gif

When I make this modification, I will de-emphasize the -q <n> settings in the --help text.

This is a sensible compromise position which retains the current flexibility for advanced users while giving the less experienced user a big hint as to what setting to use. In light of recent conversation on the topic it is a required modification and will be posted today.
halb27
May 23 2008, 01:57
QUOTE(Nick.C @ May 23 2008, 09:34) *

QUOTE(shadowking @ May 23 2008, 08:21) *
I had another idea: create a few 'presets' that are mapped to common Q settings;

-normal [Default] = Q5
-medium = Q3
-Portable = Q2

-extreme = Q6..10
Good suggestion - also added to to-do-list. Suggested names and associated -q value:

--highest := -q 10;
--high := -q 7.5
--default := -q 5;
--normal := -q 5;
--portable := -q 2.5.
--nasty := -q 0 wink.gif

....


Yeah, I like this very much: shadowking's approach and your incarnation cause it looks like -q 2.5/5/7.5/10 are really those settings which IMO are of most practical use (7.5 or 10 for the main part as a substitute for lossless archivíng).
I'd just prefer one '-', so '-normal' instead of '--normal', I would drop '-default' as a synonym for '-normal' (of course '-normal' should be defaulted), and we shouldn't use a nasty mode IMO.
The '-q' parameter can remain as an advanced '--q' setting for the lovers of quality fine tuning.

The solution is so wonderful to me as we have seen a growing discussion on quality levels since we started having a finer quality granularity. This is covered by this new self-explaining quality granularity while still keeping up fine tuning abilities for those who like them.
sauvage78
May 23 2008, 02:07
That's exactly what I was suggesting, except that as I think -q 0 & -q 10 are not really recommandable I would not add them as recommended setting, & now that I understund the symbolic importance of -q 5 I would add it.

so ideally it would look like this for me

-q 2.5 = -portable (my "-near lossy")
-q 5 = -default
-q 7.5 = -archive (my "-near lossless")

-q0 & -q10 looks too much like flac -fast & -best which makes sense for flac but are sub-obtimal here.

I think that after some transcoding tests I would then either use the -default or -archive preset ...

Edit:
I agree with halb27 too, one - is enough, the easier the better.

If you really can't do without adding -q 0 & -q 10 to the presets then I would use maybe:
-q 0 = -noisy or -this-is-trash-metal
-q 10 = -overkill or -music-for-the-deaf wink.gif
names that naturally sounds bad, but my english is too weak here.
halb27
May 23 2008, 02:36
QUOTE(sauvage78 @ May 23 2008, 10:07) *

...
-q 2.5 = -portable (my "-near lossy")
-q 5 = -default
-q 7.5 = -archive (my "-near lossless")
...

This brings us back to the 3 quality levels approach, but with the advantage of a proper naming.
Having thougt about it I also think it's better to stay with three levels.

I personally have dropped my lossless archive and encoded everything to -q 7 (for the most part), but on more or less rare occasion - depending on the meaning the track under consideration has to me - also used a higher -q setting up to -q 10 (and in very rare cases I encoded to lossless wavPack which I can use on my DAP). But it is really enough to use -archive as a rule, and I can still use a higher -q setting if available as an advanced option on those rare occasion where I want it.

So I suppoort this IMO better differentiation standard quality options / advanced quality options.

As a detail I suggest to use --shaping 0.5 with the -archive quality (more clearly in terms of the advanced options: with every -q setting from -q 7 up.
It makes the noise so much less audible (often totally inaudible when listening only to the noise containing correction file with the same volume setting as with the real music), and IMO is safe to use with these very high quality settings.
halb27
May 23 2008, 02:46
QUOTE(sauvage78 @ May 23 2008, 10:07) *

...
If you really can't do without adding -q 0 & -q 10 to the presets then I would use maybe:
-q 0 = -noisy or -this-is-trash-metal
-q 10 = -overkill or -music-for-the-deaf wink.gif
names that naturally sounds bad, but my english is too weak here.

I really wouldn't like -q 0 among the standard options.
-q 7.5 should be overkill already, so it's hard to find a simple name for -q 10 (maybe -paranoid hits it best - but it's a pretty negative word. However if most members would appreciate a -nasty mode among the standard options a -paranoid mode would be a good counterweight).
Another argument for having just three standard quality options.
This is a "lo-fi" version of our main content. To view the full version with more information, formatting and images, please click here.
Invision Power Board © 2001-2008 Invision Power Services, Inc.