Author Topic: Advice for a Non-Audiophile  (Read 4556 times)

phred

  • Global Moderator
  • Sr. Member
  • *****
  • Posts: 9305
I always run the 'analyze volume' tool when I'm adding a new album to MB. I have the replay gain always set to -1db and I analyze the leveling on track basis and on album basis. I've done this over the years to my roughly 35,000 tracks (2,700 albums.) All tracks are MP3.

Today I bought an MP3 album from Amazon and it sounded great. I was surprised to see that it had a VBR of about 286 on average as I thought Amazon only sold 320s.  I ran it through the volume analyzer and the vocals virtually disappeared and the overall volume was substantially lower. I removed the volume leveling and it once again sounded great.

That caused me to remove the leveling from a few test albums and every one of them sounded much better without the leveling.

Is this typical? I'm not an audiophile, but I sure can tell the difference before and after. And the tracks without leveling always sound better. Am I doing something wrong with the leveling? There aren't that many settings that I can see other than setting it to -1db.

Is it worth removing the leveling from all the tracks and not look back?

To my ears enabling or disabling seems to make no difference in the overall sound level/quality. The only DSP Plulgin I use is the Stereo Enhancer. The difference I hear is that the sound is 'brighter' with it enabled and somewhat 'muddy' with it off.

Suggestions and comments please.

Thanks.

Download the latest MusicBee v3.5 or 3.6 patch from here.
Unzip into your MusicBee directory and overwrite existing files.

----------
The FAQ
The Wiki
Posting screenshots is here
Searching the forum with Google is  here

The Incredible Boom Boom

  • Sr. Member
  • ****
  • Posts: 1269
Most likely the tracks "sound better" without the volume normalization because they're louder.

What's your signal flow from MB to the output source? Your ears are used to whatever is happening in-between the two, so something like what you think of as "somewhat muddy" may be how the track you're listening to was actually intended to be heard.

phred

  • Global Moderator
  • Sr. Member
  • *****
  • Posts: 9305
Most likely the tracks "sound better" without the volume normalization because they're louder.
Would a louder volume make the sound clearer?

Quote
What's your signal flow from MB to the output source? Your ears are used to whatever is happening in-between the two, so something like what you think of as "somewhat muddy" may be how the track you're listening to was actually intended to be heard.
Typically I use WASPI (Shared) to my Primary Sound Driver and from there to a USB BlueTooth transmitter to SkullCandy Grind headphones. Since the headphones can also use a 1/8" cable, I will try that tomorrow removing the BT from the mix. I will also try direct to desktop speakers.

Thanks.
Download the latest MusicBee v3.5 or 3.6 patch from here.
Unzip into your MusicBee directory and overwrite existing files.

----------
The FAQ
The Wiki
Posting screenshots is here
Searching the forum with Google is  here

frankz

  • Sr. Member
  • ****
  • Posts: 3834
I agree with IBB that this is perception.  Listen to something and then turn your output down 7 or 8 dB and it will sound worse because you're hearing less detail, but the "quality" is the same.

Higher frequencies cut through less than lower, and as we age we hear less and less high-frequency, and that's where the detail is.  When you listen to something that's 7 or 8 dB louder than you're used to, you're reacting to that extra energy in the range that's normally harder to hear and reading it as detail.

phred

  • Global Moderator
  • Sr. Member
  • *****
  • Posts: 9305
Higher frequencies cut through less than lower, and as we age we hear less and less high-frequency
Was that a crack about my age???   :-)

So if this is all about perception, and I don't disagree, then I should use whatever makes it sound the best or enjoyable to me. And if that means stripping the leveling from all my tracks, I should do it.

Agree? Yes or no?
Download the latest MusicBee v3.5 or 3.6 patch from here.
Unzip into your MusicBee directory and overwrite existing files.

----------
The FAQ
The Wiki
Posting screenshots is here
Searching the forum with Google is  here

sveakul

  • Sr. Member
  • ****
  • Posts: 2438
phred try using WASAPI Exclusive instead of Shared.  The latter still ends up in the Windows Mixer, just like DirectSound, and that "can" do who-knows-what to the sound.  The only clear path from the file to the drivers is WASAPI Exclusive--or ASIO, but the latter seems to be more problematic for people from what I see here.  Spoken by a non-audiophile (yes, I use a VST compressor/limiter, Sonic Anomaly's "Unlimited.")

psychoadept

  • Global Moderator
  • Sr. Member
  • *****
  • Posts: 10691
I should use whatever makes it sound the best or enjoyable to me. And if that means stripping the leveling from all my tracks, I should do it.

Not necessarily. Replay gain is about equalizing the volume between tracks. If they generally sound better to you without it, you might just need to turn the volume up. :)

Also, check what the replay gain is doing on some tracks that sound better to you without it. If the adjustment is a negative number (which I find is the case more often than not) then the volume change is almost certainly the issue.
Last Edit: October 06, 2020, 11:46:41 PM by psychoadept
MusicBee Wiki
Use & improve MusicBee's documentation!

Latest beta patch (3.5)
(Unzip and overwrite existing program files)

frankz

  • Sr. Member
  • ****
  • Posts: 3834
Higher frequencies cut through less than lower, and as we age we hear less and less high-frequency
Was that a crack about my age???   :-)

So if this is all about perception, and I don't disagree, then I should use whatever makes it sound the best or enjoyable to me. And if that means stripping the leveling from all my tracks, I should do it.

Agree? Yes or no?

:)

I'm all for doing whatever makes it sound best to you, but you could mostly get the same effect by turning your output up by about the same amount on average as replaygain takes it down.  Or adding some gain via preamp in MB (be careful to avoid clipping doing this though) to compensate.

It depends on if you value the enhanced gain over the leveled volumes.  I listen a lot in headphones and with other people around, so I try to keep things level so I can manage it without blowing my ears up or annoying everyone in the other room, but leveled volumes aren't a requirement to enjoy listening.

The Incredible Boom Boom

  • Sr. Member
  • ****
  • Posts: 1269
Would a louder volume make the sound clearer

Quieter volumes are generally perceived as less clear, even though that isn't necessarily the case. The SkullCandy's sound profile and the Bluetooth transmitter might be also a factor here too. My headphones sound awful through the audio jack and turned off, but drastically better when turned on and even better using the USB output.

Audio enthusiasts all differ, but I like my signal flow to be as flat as possible. I use only a single equalization VST to slightly make up for the frequency "deficiencies" of my output source, but that's it. Not to brag (I'm actually super sentimental about this - I loved running sound  :(), but in college when I would run audio at various well-known venues in Nashville, the house engineers (seasoned audio vets) would go from "Great, another young punk who wants to run the board" to "It sounds incredible tonight, let's exchange contact info." I don't think I was any special except for I always followed my mentors' advice of just loud enough (never too loud) and cutting frequencies instead of adding.

I would do two things. First, equalize your headphones properly. The benefits of doing this are self-explanatory. Next, start normalizing your volume fairly low (I use -6dB across the board, no matter the type of music), playback using WASAPI (Exclusive) (not sure if your cans have USB output, but if they do, MB can output directly to the headphones and you won't have to worry about stopping playback to play something on YouTube or whatever) and disable VSTs. At first you'll think your music is unlistenable, but lower volumes without enhancement are necessary so you can re-train your ears to become accustomed to how your music sounded in the mastering room. Often times, what people think of as "sounding good" is actually (as @frankz describes below) over-excitement (a.k.a brightness) of frequency ranges our brains are "tuned" to pay more attention to. Over a period of weeks (more likely months,) you can start adding back VSTs, but you might find you don't need to.

All of what I suggest is a lot of "work" just to enjoy music, but I think the benefits are tenfold once it becomes something that you can do without thinking about it. I'll sometimes come across songs that give me literal goosebumps just from the sound alone (Motown records do this to me a lot.) Big warning: as your ears adjust, you're going to start noticing clipping like crazy, especially if you listen to a lot of post-'90s music.

I agree with IBB that this is perception.  Listen to something and then turn your output down 7 or 8 dB and it will sound worse because you're hearing less detail, but the "quality" is the same.

Higher frequencies cut through less than lower, and as we age we hear less and less high-frequency, and that's where the detail is.  When you listen to something that's 7 or 8 dB louder than you're used to, you're reacting to that extra energy in the range that's normally harder to hear and reading it as detail.

The above was posted as I was typing/working/eating/typing some more, but basically yes. I always recommend listening to music at lower volumes, because you're forcing/training your ears to pick out details, which saves your ears as you age and lessens the need to listen loudly to enjoy.
Last Edit: October 06, 2020, 11:34:17 PM by The Incredible Boom Boom

phred

  • Global Moderator
  • Sr. Member
  • *****
  • Posts: 9305
Lots of information here to digest and experiment with. I thank you all for the advice.

My first thought is to get rid of the BT transmitter to the headphones and replace it with a direct cable from the PC to phones. But of course I can't find the cable that came with the phones, so a search will take place later today.

As for switching to WASAPI exclusive, when trying to play a track I get the error
Code
Unable to initalise the output device (error=BASS_ERROR_FORMAT)
when using Speakers (Realtek High Definition Audio). If I change the sound device to my only other option (VX2452 Series Display Audio) it appears to play out of the crummy monitor speakers. I have always had the monitor speakers disabled, so I don't really know if audio is coming out, but it appears to be. WASAPI shared uses the primary sound driver, which works and is what I've been using for quite some time.

Any idea what to do about the error for WASAPI exclusive? Driver update? BASS update? Are the speakers not capable of using WASAPI?

I'm off to search for my cable. More later.


Download the latest MusicBee v3.5 or 3.6 patch from here.
Unzip into your MusicBee directory and overwrite existing files.

----------
The FAQ
The Wiki
Posting screenshots is here
Searching the forum with Google is  here

frankz

  • Sr. Member
  • ****
  • Posts: 3834
In windows (10) try settings->Devices and then Sound settings (at the right)  -> Sound Control Properties (at the right) -> Advanced tab and make sure "Allow applications to take exclusive control..." is on.

MTVhike

  • Full Member
  • ***
  • Posts: 127
What is the purpose of "Analyze Volume" and "Replay Gain"? Is it to reduce the dynamic range of the track? I know that some of the tracks I've ripped from CDs have such a large dynamic range that I can't play them in my car without manually fiddling with the car's volume control. On the other hand, when I'm playing them at home, I like the wide dynamic range. Maybe I should have two versions!

phred

  • Global Moderator
  • Sr. Member
  • *****
  • Posts: 9305
In windows (10) try settings->Devices and then Sound settings (at the right)  -> Sound Control Properties (at the right) -> Advanced tab and make sure "Allow applications to take exclusive control..." is on.
WHOA!!! The difference between exclusive and shared was immediately apparent. Thanks frankz for getting the error resolved. Even to the point of greater stereo separation. Or is that just what I'm perceiving? Can simply switching the output from shared to exclusive improve the stereo separation?

I'm still using the BT transmitter because the 3.5mm cable is nowhere to be found. I have ordered one from the world's largest superstore and it will be here sometime Thursday. I still have a feeling that the transmitter to headphones may be contributing to (lack of) sound quality. 
Download the latest MusicBee v3.5 or 3.6 patch from here.
Unzip into your MusicBee directory and overwrite existing files.

----------
The FAQ
The Wiki
Posting screenshots is here
Searching the forum with Google is  here

psychoadept

  • Global Moderator
  • Sr. Member
  • *****
  • Posts: 10691
What is the purpose of "Analyze Volume" and "Replay Gain"? Is it to reduce the dynamic range of the track?

Replay gain tries to minimize the relative difference between tracks. It basically tells the player "turn this track up by x amount" or "down by x amount". So if the dynamic range within a track is the issue it won't help much.
MusicBee Wiki
Use & improve MusicBee's documentation!

Latest beta patch (3.5)
(Unzip and overwrite existing program files)

vincent kars

  • Sr. Member
  • ****
  • Posts: 444
Quote
Or is that just what I'm perceiving?

I'm afraid the answer is yes but let me try to explain the difference between Exclusive and Shared mode.

In shared mode you can have multiple audio streams e.g. you watch a video, hear the sound track but if an email comes in, you will hear the notification too.

This means mixing.
You can only mix if the streams have the same sample rate.
This rate is set in the Win audio panel.
However, if you play Redbook ( 44.1 kHz) and you set the Win audio to 44.1, no resampling is needed.
In this scenario it is not very likely that there will be a difference in sound quality.

However, mixing is also calculating. Basically sum the streams and average them.
This is done by converting the audio to float, calculate, dither (add noise) and convert back to integer (DAC needs 16 or 24)
I can imagen that if you set Win audio to 16 there might be an audible effect.
If the audio path is 24, the dither (if done at all) is at -144 dBFS.
This is way below the noise floor of your gear.

In case of WASAPI Exclusive none of the above is happening.
So WIN cannot meddle with the sound quality.
You don't have email notifications full blast over the speakers either :)

I won't rule out audible differences but when configured right I do think the differences between Exclusive and Shared marginal.

If you do use some Win Audio enhancements (loudness, etc) the differences can be substantial as the moment you use Exclusive, you get rid of these enhancements as well.


Bit more about WASAPI: https://www.thewelltemperedcomputer.com/KB/WASAPI.htm