Vorbis 1.0 Quality Settings

[Emp3r0r]

what quality setting do you use? I personally make my video about ~745 megs to match up with a ~50 meg vorbis file, resulting in an 800 meg XCD. This usually means I use quality settings from .001 to .220 on the vorbis file. The resulting movie is usually excellent quality when around 2 hours.

So, do you have a sweet spot that you use for your vorbis audio files? What is the absolute lowest quality setting you would use?

PS: attention quality nut's only

[kxy]

You asked for a quality NUT reasoning, so I hope I can answer your question.

If you don't care about the importance in keeping the Dolby prologic I & II surround encoding accuracy, then q2 or q3.

Keep in mind that at the lower -q settings, say around the 64kbps region, it is relatively untested. 128kbps has being very solid tuned for a while now. The 64kbps mode is essentially the same one as from rc3 with some noise normalization and more aggressive stereo pump into it.

About your q0, because I haven't done an extensive testing, so I can't tell you much. But you can wait a couple weeks till ff123 listening test is done and draw your own conclusions. Please keep in mind that lower the q value, the more stereo image needs to be audibly reduced, thus the stereo image will shift from an elliptical cardoid toward a sonic gravity level. That is the reason why Monty disabled the channel coupling switch in 1.0 I think. Because I suspect that he doesn't want newbies to put a hold value on lossless coupling with q0, because at a lower q, you need to reduce the stereo image to compensate more bits for the audio.

This is what I do personally. I tweak the values of my movie encodes(usually takes me 5 to 10 encodes, in the span of a week) till I get a very good visual with the lowest size that is allowed. Then I use highest q value that is allowed for audio. For the SFX movie, I use q5 q6 in order to keep the DPII downmix accuracy.

Hey, if you are going to do it, then do it right the first time.

[metrom]

Originally posted by kxy
Keep in mind that at the lower -q settings, say around the 64kbps region, it is relatively untested.

I do not agree with you, alot of the devolopment/work from RC3 to 1.0 has been around the "64kbps area". The diffrence from a 64kbps RC3 file and 1.0, is quite big. So I wouldn't call is "untested".

Listen to the great quality here:
http://www.xiph.org/ogg/vorbis/listen.html

Make your own decision, do not let ff123's test decide for you (I don't like his test-setup :))

I think you can be happy with a 64kbps Ogg Vorbis in most films, but if there is some action and effects, you better go up to around -q 4.

[Emp3r0r]

@kxy: i think you are a NUT if you are encoding from AC3 using q6. I could understand q6 if you had an uncompressed source, but AC3 is already lossy, you might as well keep the AC3 soundtrack. Correct me if I'm wrong but isn't the point of vorbis to have better quality at lower bitrates. I thought this was vorbis strong point. I mean if your going with bitrates as high as 128, you might as well use MP3.

So 64kbps isn't sufficient in most cases?

[kxy]

Should I bust out a quote?
Quoting Monty:
" For what it's worth, I've excused mysefl from the test (64kbps is too easy to tell which codec is which) but sent my predictions of outcome to ff123 before it started.

I find it odd, actually, that folks are considering Ogg the codec to beat when 64kbps is relatively very untested. 128-ish kbps has seen about two and a half years of solid tuning now. The 64kbps mode is essentially the same one as from rc3 (which no one used because it was unlistenable ;-) plus noise normalization and somewhat more aggressive stereo to plow bits back into treble. How good it would turn out in the big bad world was a huge unknown; it sounded good on my primary test sample suite, but that's only about thirty samples. I do think Ogg has at least an edge over the competition, even not considering that 64kbps has alot of tuning room left, but I also am not going to pretend Ogg wins every one of these samples. Hell, I even picked WMA as winner for one, co-winner for another (classical solo instruments *should* be easy to encode... they're the only thing WMA doesn't run out of bits on and close its metallic lowpass like a rocket-powered garage door).

What I've found out so far: Ogg noise normalization can 'wheedle' when there's a single dominant note with lots of harmonics and no background noise (ala electronic piano/organ, solo string instrument or a single almost solo vocal with reverb). This is due to noise normalization restoring a harmonic that should probably best be ignored. Also, the same basic sample with the instrument far to one side of the stereo image with reverb bouncing back and forth is causing midrange noise to crop up in the center where it's not masked. "

You can read more here...

http://www.hydrogenaudio.org/forums/showthread.php?s=&threadid=2776&perpage=25&pagenumber=4



Originally posted by metrom


I do not agree with you, alot of the devolopment/work from RC3 to 1.0 has been around the "64kbps area". The diffrence from a 64kbps RC3 file and 1.0, is quite big. So I wouldn't call is "untested".

Listen to the great quality here:
http://www.xiph.org/ogg/vorbis/listen.html

Make your own decision, do not let ff123's test decide for you (I don't like his test-setup :))

I think you can be happy with a 64kbps Ogg Vorbis in most films, but if there is some action and effects, you better go up to around -q 4.

[kxy]

Originally posted by Emp3r0r
@kxy: i think you are a NUT if you are encoding from AC3 using q6.


Please see your first post.

Originally posted by Emp3r0r
PS: attention quality nut's only


Didn't you ask for a nut quality?


Originally posted by Emp3r0r
I could understand q6 if you had an uncompressed source, but AC3 is already lossy, you might as well keep the AC3 soundtrack.


Uncompressed source, how do you define uncompressed source? Movie audio is usually 7.1 which is compressed. So you mean pre recording with all the noises prior to the final sound editing?

You will save about half of file size from a 5.1 ac3. Let say it is 213mb, the final size will be around 120mb.

Originally posted by Emp3r0r
Correct me if I'm wrong but isn't the point of vorbis to have better quality at lower bitrates. I mean if your going with bitrates as high as 128, you might as well use MP3.[/B]

Okay, I will correct you. I will not consider 128 high bitrate. If you believe that 128 is CD quality then you are wrong. In the past, some people suggested to use r3mix as 2 cd high quality setting. I don't consider r3mix high quality, I certainly won't even use it for one cd. There are too much pre-echo for my taste, and the way he set it up it still has a lot of high-frequency distortions. To conclude, vorbis beats mp3 at 128, hands down. If I really have to use mp3s, I would use --alt-preset, not to mention vbr mp3 is a nightmare with avi, but that is another issue.

[Emp3r0r]

Nice reply, I listen to you because you are a NUT ;) and I don't have my stereo setup for listening. Actually, I only have one speaker set up so I can't do any Stereo quality comparisons. I almost rode down to walmart (or RadioShack) to pick up some nice headphones but I'm not certian as to which headphones are best quality for the money.

I'm tired of all this work to get a 1 CD encode to look and sound good. I realize now why I used to go the 2 CD AC3 route as it is so much easier and less time consuming for a quality outcome. Then again, Enemy at the Gates has been the hardest movie I've ever attempted. Neither XVID nor Divx5 has been giving me correct filesize on this movie, and I'm sure they can't be saturated.

Anyway, before I risk going too far off topic, I think I'm going to resort to using Vorbis for comedies and drama's only and stick to AC3 for the rest.

[Peters]

Originally posted by kxy
Please keep in mind that lower the q value, the more stereo image needs to be audibly reduced, thus the stereo image will shift from an elliptical cardoid toward a sonic gravity level. That is the reason why Monty disabled the channel coupling switch in 1.0 I think. Because I suspect that he doesn't want newbies to put a hold value on lossless coupling with q0, because at a lower q, you need to reduce the stereo image to compensate more bits for the audio.



From Hydrogen Audio Forum, JohnV Comment

http://www.hydrogenaudio.org/forums/showthread.php?threadid=1265&highlight=coupling+channel



Vorbis is totally flexible VBR, you could use full stereo (or lossless channel coupling in Vorbis' case) even with -q 0. All it would do is, that VBR scales the bitrate higher.
There's no quality issues with Vorbis why you couldn't use lossless channel coupling even with -q 0. Of course when you use lossless coupling, the bitrate will be higher.

[canadian_fbi]

yeah, theoretically there's no reason in vorbis why you couldn't use lossless coupling in -q 0. but vorbis most definitely does not do this by default, and there's no way that i know of of modifying this. so it's true that a low q value does not imply that the stereo effects must be destroyed, but it happens to be that way how the stereo coupling is implemented now in vorbis 1.0.