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.
Topic: Is the --PNS switch (Petri Net Simulation) safe to use? (Read 2928 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

Is the --PNS switch (Petri Net Simulation) safe to use?

The --pns switch was introduced during the 1.01-1.02 era, I believe, and was labelled as an "early alpha" function. Now that mppenc has moved to 1.10 release, is the --pns switch safe to use?

I don't quite understand what the value is supposed to mean. I've tested out --pns 50 and the output wasn't terribly butchered, yet Frank's notes claim that <0.2 is the high quality range whereas 0.5-2.0 is the low quality range. There also seems to be a limit: --pns 50 and 10 produce files of roughly the same size, yet there is a big difference between --pns 1 and 0.5. I gather that this figure is some kind of a scaling factor?

I'm curious about this switch because it doesn't seem to affect perceptual output while it reduces bitrates by 3-5% or more. It also seems to degrade the sound less when paired with a high quality preset (like --standard) than a low one (like --thumb).

Is the --PNS switch (Petri Net Simulation) safe to use?

Reply #1 –
PNS (which actually stands for Perceptual Noise Substitution.. Frank's Petri.... was a joke ) isn't really suitable for use with high quality presets.  It uses approximation and recreation techniques which are similar in result to something like SBR (note, not the same thing, but they are both meant to "recreate" something based on "hints", not actually store the original sound).  I don't think the accuracy loss that this would necessarily bring about would be worth saving only a few kbps at the risk of causing some audible artifacts on a preset like --standard.  At 128kbps and less, where transparency is not going to be the main focus, then yeah it's a good idea.  But if you are shooting for transparency, these type of techniques are not a very good idea IMO.

Is the --PNS switch (Petri Net Simulation) safe to use?

Reply #2 –
Quote
Originally posted by Dibrom
PNS (which actually stands for Perceptual Noise Substitution.. Frank's Petri.... was a joke ) isn't really suitable for use with high quality presets.  It uses approximation and recreation techniques which are similar in result to something like SBR (note, not the same thing, but they are both meant to "recreate" something based on "hints", not actually store the original sound).  I don't think the accuracy loss that this would necessarily bring about would be worth saving only a few kbps at the risk of causing some audible artifacts on a preset like --standard.  At 128kbps and less, where transparency is not going to be the main focus, then yeah it's a good idea.   But if you are shooting for transparency, these type of techniques are not a very good idea IMO.

I don't know who has drier humor: Germans or the British.

PNS 0.20 - 0.30 is giving me about a 3-5% bitrate reduction with --standard so it sounds like its potential pitfalls overweigh the size benefits.
 

Is the --PNS switch (Petri Net Simulation) safe to use?

Reply #3 –
Quote
Originally posted by mithrandir

I don't know who has drier humor: Germans or the British.

PNS 0.20 - 0.30 is giving me about a 3-5% bitrate reduction with --standard so it sounds like its potential pitfalls overweigh the size benefits.


Code: [Select]
             without PNS   --pns 0.5       saves

---------------------------------------------------

thumb          82 kbps        73 kbps       11%

radio         110 kbps        98 kbps       11%

standard      153 kbps       132 kbps       14%

xtreme        186 kbps       167 kbps       10%

insane        217 kbps       207 kbps        4.6%

--quality 10  314 kbps       313 kbps        0.3%

PNS code is still in the alpha stage.
--  Frank Klemm