What is tone and limit for?

[Tibor]

Hi, 

I am not very audiophile.... I understand what the equalizer is for, but I don't fully understand tone and limiter. Tone boost bass and treble? Are EQ and tone been used together or shouldn't? And limiter? 

Thanks

[andrewilley]

Tone is a bass/treble boost control (so anti-clockwise to zero, rather than midway, is the 'flat' setting). You can use these along with the EQ, but personally I'd suggest one or the other.

Limiter stops high peak levels from oversaturating and thus distorting.

Andre

[blaubär]

On 4/16/2018 at 7:06 PM, andrewilley said:

Limiter stops high peak levels from oversaturating and thus distorting. 

Is it correct to assume that in the absence of EQ and Tone settings these high peak levels won't occur ? So that the limiter may be safely deactivated in this scenario ?

Could one correctly describe the limiter as a dynamic range compression ? Thta operates only on those parts of the music which would otherwise exceed the valid volume level ?

[NickIst]

19 hours ago, blaubär said:

So that the limiter may be safely deactivated in this scenario ?

No, AFIK it helps against clipping

[blaubär]

2 hours ago, NickIst said:

No, AFIK it helps against clipping 

Yes, sure. But clipping couldn't occur without eq and tone . Or could it ?

[NickIst]

3 hours ago, blaubär said:

Yes, sure. But clipping couldn't occur without eq and tone . Or could it ?

If you don't use RG tags, it'll likely occur

[blaubär]

2 hours ago, NickIst said:

If you don't use RG tags, it'll likely occur

So far I had rather thought of replay gain as a mechanism to make music louder if it had been created with too much headroom.

Without replay gain and EQ I would expect the sound processing not to exceed 0 db. For the minimal effects with music which has been created very near this threshold there is headroom available ( if not manually disabled ), so I don't see the necessity of replay gain. Of course you could use replay gain to reduce the volume if you know that you'll otherwise get problems due to your EQ enhancements, but then if you don't use the EQ, there wouldn't be any need for that.

Or so I think - is there an error in my reasonig ? What part of the sound processing would further enhance a signal that is loud, meaning near the absolute 0db threshold, if not the EQ, Tone or RG (when enhancing rather than reducing)?

[blaubär]

I tried it out. Result : it seems there is no limiter, and DVC doesn't need one, see the bug :

 

[NickIst]

15 hours ago, blaubär said:

had rather thought of replay gain as a mechanism to make music louder if it had been created with too much headroom.

It's an instrument of volume normalization, and in most cases it makes music lower. hat's why tags can be in negative numbers. So the normalization means an equal target volume for differrent tracks and albums. But another implication if its use is the elimination of soft clipping but this is relevant only to lossy files which don't contained an original audio signal and thus every player decodes it with approximations. This approximate decoding can cause clipping.

As for me I didn't test it on the current PA, but v2  with limiter had the clipping. PC software players foobar, MusicBee, Aimp have RG to prevent clippping. However the use of RG can cause clipping if you manually increase volume. But usually reasonable man will tick "apply RG preventing clipping according to peak"

RG does the following: 

1. Prevent clipping (perform peak normalization to 0 dBFS of those tracks / albums that have level of the peak sample higher than one). 
2. Perform a gain by a certain amount (in dB). In this case, a preliminary analysis of the tracks is not required. 
3. Perform a track / album gain to a predetermined average perceived volume (the default target is 89 dB, but it can be increased / decreased within +/- 20dB). 
4. Perform amplification of the track / album to a predetermined average perceived volume and, if as a result, the peak sample level is out of tolerance, adjust the level to 0 dBFS (full scale).

[andrewilley]

9 minutes ago, NickIst said:

It's an instrument of volume normalization, and in most cases it makes music lower.

Quite. One of the problems of audio that has been encoded at 100% digital maximum level is there is no headroom left for EQ boosts etc. By default, the spec for ReplayGain suggests that peak track (or album) levels should play back at about -14dB below full level. PA has a variable override control (+/-16dB) to change that behaviour though, if you are the sort of person who just wants more volume. To be honest though, it's generally better to leave some headroom in the whole audio processing chain, and only "turn it up to 11" in the final output stage.

Andre

[blaubär]

56 minutes ago, NickIst said:

But another implication if its use is the elimination of soft clipping but this is relevant only to lossy files which don't contained an original audio signal and thus every player decodes it with approximations. This approximate decoding can cause clipping.

As for me I didn't test it on the current PA, but v2  with limiter had the clipping.

Thank you ! So the limiter probably won't prevent soft clipping ... that leaves hard clipping. Do you know whether this worked in V2 ?

[NickIst]

7 hours ago, blaubär said:

Do you know whether this worked in V2 ?

Maybe the last sentence was not clear, sorry. In V2 limiter really decreased clipping, but RG tags did it far more better. In the current version right now I didn't manage to catch clipping both with limiter or RG turned off while listening to a couple of tracks. Although it's just matter of finding an appropraite track.

Plus the whole clipping thing is depend on decoding process that's why it's considered that floating point players are better. Though I don't remember if the current PA has this. I forgot It really has DSP float.

[Eckless]

So a dumb question...what settings do people usually set for Poweramp for general listening? Lately I've just been putting bass and treble at 25%, and increase the pre-amp a bit so it's not so quiet. Are there any other "must have" settings?

[NickIst]

26 minutes ago, Eckless said:

So a dumb question...what settings do people usually set for Poweramp for general listening? Lately I've just been putting bass and treble at 25%, and increase the pre-amp a bit so it's not so quiet. Are there any other "must have" settings?

Except RG I don't recall anything. You can turn DVC it's far more better, but you'll lose factory made effects, if they're important for you. For me yes. Also you can change sample rate in order to optimize hardware's work to 48 and its multiples or to 44.1.and its multiples. Also you can turn off all transitions, crossfades etc.In fact most of the options are a matter of taste, hardware limitations etc. And the last two are more tweaks than a decisive improvement 

[blaubär]

1 hour ago, Eckless said:

So a dumb question...what settings do people usually set for Poweramp for general listening? Lately I've just been putting bass and treble at 25%, and increase the pre-amp a bit so it's not so quiet. Are there any other "must have" settings?

My personal opinion, history and practice : I decided some years ago that I would rather like to hear music as it is and get out of the habit of drowning out the subleties with bass, bass and yet more bass. It took a while to get used to the sound, at first it was shallow, bland, dismal ( add further negative adjectives as you see fit ), now it's normal, and I can't stand unnaturally boosted bass and treble any longer.

How to achieve that ? Buy decent headphones with a flat frequency response curve, I use AKG K702, they sell for 123 Euros at amazon.de. Deactivate Equalizer and Tone and all os-sound-effects. I you have to listen with other headphone, find their frequency response curve and counteract with the equalizer.

And then use High-Res only for technical reasons ( if e.g. Samsung on Pie works better with OpenSl High-Res driver for whatever reason ). Use standard sampling rate 44.1 kHz or 48 kHz if possible. Unless you're a bat or a dog you won't hear the difference. Everything to be understood as IMHO !

As for stereo with headphones : until very recenty I used to apply StereoX at 30 %. Now I found out that a moderate crossfeed sounds more naturally. Sadly I cannot use Poweramp for that.

[andrewilley]

15 hours ago, blaubär said:

. Use standard sampling rate 44.1 kHz or 48 kHz if possible.

From what Max was saying before, on many modern Android devices the physical output stages don't go down to 44.1 anyway, and the hardware's minimum operating rate is 48k. So at some stage (whether it's in PA or the device's output stages) 44.1 files will get converted to 48 anyway.

Andre