like employing an adaptive lowpass, with a lower nominal cutoff, but increasing it beyond 19khz in critical moments. A more agressive adaptive ath might help here also.
Originally posted by mp3fan Currently --ath-lower x results in an unrecognizable error while using razorlame. Anyone know why this is?
Originally posted by kxywhen I use -alt-preset standard -b96 -Y --lowpass 17.5, it got a following warning msg.... *** WARNING *** the meaning of the experimental -Y has changed! now it tells LAME to ignore sfb21 noise shaping (VBR) I am not really sure how sfb21 noise shaping works, could someone please explain that to me?
Also by using the lowpass would it have any effect on the determination of bits in joint stereo and nspsytune?
What about and the -V3 switch, would it have any effect on the determination of bits in joint stereo and nspsytune OR conflict with the internal switch of alt-preset?
Originally posted by kxy To recap, the -Y or lowpass 16, will disable sfb21, in effect not encoding anything above 16kHz and --lowpass 17.5 is just lowpass 17.5. Correct?
Now the question is should one raise ns-bass value in relation to the lowpass?
Thank you for being so patient with a newbie.
By using -Y, and disabling noise shaping in sfb21, it has the effect I specified... where it will not "go out of it's way" to encode this material if it requires a signficant jump in bitrate.
Originally posted by mp3fan So by using a --lowpass 17.5 & -Y, I'm telling the encoder that when it does "bother to encode" freqs above 16k, it won't go beyond 17.5, right?
You're former description seems to indicate that it won't encode ANYTHING beyond 16k while your prior statement indicates that some material beyond 16k will get encoded. Please explain.
Also, when I use an --athlower number up to 8, --athlower 8, there is no change in bitrate and I expected to see a filesize increase with this setting.
Also, what is the effect of using -Y with ABR or CBR? Will it still work the same way in those modes?
Originally posted by jkeating You are forgetting about stereo settings!Right now i use --nsmsfix 5.0 to make LAME use much less stereo frames, and reduce bitrate.I use 5.0 value because 5.5 sometimes makes LAME use more joint frames than "normal" settings.
Originally posted by Dibrom The reason for this is as follows. sfb21 is the scalefactor band which is assigned to the frequencies beyond 16khz. However, the problem is that it doesn't actually have a scalefactor, or basically a way to efficiently compress these frequencies, while all other sfbs do. Apparently, this was an oversight in the design process of the format, because either the developers thought frequencies beyond 16khz were inaudible (or irrelevant), or as robert suggested, maybe the format was designed for 32khz PCM as is used on DAT, which resultingly goes up to this frequency. In any case, to actually encode these frequencies then, you have to increase the resolution of all of the sfbs, so that you can "in effect" also control the resolution of sfb21 (which, because of a lack of a scalefactor, you cannot control independently). Make sense?
Originally posted by ErikS No. Sorry for beeing a bonehead, but i don't understand. Why isn't it possible to lower the other scalefactors (<21), when you add to the global scalefactor, if the only purpose is to increase resolution in sfb21? Ok, i admit that i haven't read a single line of code. I'm sure there is a reason why it's not done, but it would be interesting for me to know./Erik
Also, when I use an --athlower number up to 8, --athlower 8, there is no change in bitrate and I expected to see a filesize increase with this setting. I'm not sure what the problem is there, it should work. What's the full command line you are using? And did you verify that for sure this switch is being passed to the encoder by trying it on the command line instead of through some sort of frontend?