Replay Gain, Flac Bit Rate, and upsampling questions.

[Gio-Gio]

Hello you all, I am new to dB old EAC user.

I am trying to understand dB better.
First of all, the Replay Gain. I used the settings provided in the Help File (EBU R128 with -18LUFS).
When listening in Foobar, which is set to "prevent clipping" and "db 88", Foobar needed to decreas the value set by dBpoweramp of other 6db.
So I would like to understand how do I convert LUFS into db?
how many LUFS are 100db and how many LUFS are 88db? Which negative number should I use in the RG settings of dB, to get to the 88 of Foobar?

Then, the Flac: in EAC, which as far as I know uses the same codec, I can choose a max of 1024kbps, although when I test the codec I am told that the max real performance is 839 and that one will be used.
In dB the Bit Rate cannot be set, it is fix, at 1411, and I would like to know if it is a real 1411 or a nominal one.

Last question, which kind of sense and purpose and effect (and possible side effects) would have, if when ripping I choose to upsample to a higher freq and bit depth? like, 96000, and 32 or 24bit? Would I have less danger of distortion? More headroom to apply EQ and other effect without clipping? Better sound? Just more mb?

Thanks!

[Spoon]

Re: Replay Gain, Flac Bit Rate, and upsampling questions.

I am not aware of the db88 value, normally db is measured in minus, with 0db being maximum, -6db is half the signal, -12db is quarter. You can directly translate lugs to db.

You should not up sample at time of ripping, nothing to be gained, if you ask me.

[Gio-Gio]

Re: Replay Gain, Flac Bit Rate, and upsampling questions.

I am not aware of the db88 value, normally db is measured in minus, with 0db being maximum, -6db is half the signal, -12db is quarter. You can directly translate lugs to db.

About the negative thing you are right, but people use the absolute number (without the -) most of the time, I guess for simplicity.
In fact, in Foobar and in Mp3Gain it is just written db, without any -.

I am confused because I have always used Mp3Gain (at my times Flac was a luxury which people with ton of CD could not afford, due to limited storage technology back then). And now Foobar. And they both suggest 89. I use 88 because i like symmetry, but, you got it.
So I think it is meant 89 of 100. Which means, -11db. Correct me if I am wrong.
If now you say that I can directly translate lufs to db, then -18 lufs is -18db which is 82 in Mp3Gain. This is, very quiet.
Are you sure? Or am I understanding something wrong?
EDIT: I have found this article which confirms what you've said, 1lufs is 1db. But now I do not understand why they even suggest -24lufs (which is indeed the value I have found as default in the settings of dBpoweramp, despite the -18 mentioned in the help guide). This is like -76 of Mp3Gain and it is VERY quiet.

I also don't get what you mean with half and quarter. First of all, if -6 is half, then -12 is none. -3 would be 3/4, -9 would be quarter.
But also, you seem to be either pointing to a -12 or a -24 as absolute 0 (=no signal). Which is not compatible with the -89 of Mp3Gain and Foobar...
Mumble mumble...

[mville]

Re: Replay Gain, Flac Bit Rate, and upsampling questions.

Without getting too technical, the 89dB level is a professionally calibrated reference level and allows for headroom, so to avoid clipping.

ReplayGain with Gain Calculation = EBU R128 and -18 LUFS Target Volume, are now generally considered as the standard settings.

In foobar2000 set the following:
Source mode: by playback order
Processing: apply gain and prevent clipping according to peak
Preamp with RG info: 0 dB
Preamp without RG info: -8 dB

and see how you get on with these settings. You can always experiment with these settings as you get used to them.

Reducing signal by -6dB, reduces amplitude/voltage by 50% i.e. half.

I also agree with Spoon, in that upsampling when ripping will add nothing, except you will get larger file sizes.

[Gio-Gio]

Re: Replay Gain, Flac Bit Rate, and upsampling questions.

Hi, thanks for the reply. I suppose my previous one (before yours) is now temporarily invisible because of the link I have just added in the last edit, with a TCelectronic website article which confirms that 1lufs is 1db.

So, when you say -18lufs is now standard, this means that in the classic replay gain the standard now should be 82?
Besides, that article (which you will see when my post has been moderated, or you can google "ebu r128 lufs to db" and the first result is an article titled "loudness explained" on that website) suggests -24lufs (-23 with ebu), which is indeed the default value in dBpoweramp when one opens the replay gain dsp the first time.
In any case, -18 or -24, is that not a bit too quiet?
I suppose people with wired speakers or headphones and with an amp would have no problems, but I had already some problems with the 89 in Mp3Gain, as I use Bluetooth Headphones so I am limited by the digital 0db and I cannot use an amp to get more volume...
What would you suggest?

Do you also have any clue about the difference in the kbps between the two rippers which use the same codec?
I wonder how can dBpoweramp get to 1411 if EAC is stuck with 839.
I suspect that dBpoweramp just writes 1411 but the effective is not that. But I am not sure.

[mville]

Re: Replay Gain, Flac Bit Rate, and upsampling questions.

So, when you say -18lufs is now standard, this means that in the classic replay gain the standard now should be 82?
Besides, that article (which you will see when my post has been moderated, or you can google "ebu r128 lufs to db" and the first result is an article titled "loudness explained" on that website) suggests -24lufs (-23 with ebu), which is indeed the default value in dBpoweramp when one opens the replay gain dsp the first time.
In any case, -18 or -24, is that not a bit too quiet?
I suppose people with wired speakers or headphones and with an amp would have no problems, but I had already some problems with the 89 in Mp3Gain, as I use Bluetooth Headphones so I am limited by the digital 0db and I cannot use an amp to get more volume...
What would you suggest?

-24 LUFS Target Volume is the EBU recommendation, but a lot of end users have reported that this too quiet for their systems and so have adopted the -18 LUFS Target Volume setting as a compromise, me included.

In respect to the headphone level problem, I can only suggest getting different headphones that can increase the volume OR try increasing the foobar2000 Preamp with RG info setting to +3dB OR try a different LUFS Target Volume.

Do you also have any clue about the difference in the kbps between the two rippers which use the same codec?
I wonder how can dBpoweramp get to 1411 if EAC is stuck with 839.
I suspect that dBpoweramp just writes 1411 but the effective is not that. But I am not sure.

I believe you can ignore the bitrates in flac. See this recent post: Plummeting bitrate.

[Gio-Gio]

Re: Replay Gain, Flac Bit Rate, and upsampling questions.

I can only suggest getting different headphones that can increase the volume OR try increasing the foobar2000 Preamp with RG info setting to +3dB OR try a different LUFS Target Volume.
I believe you can ignore the bitrates in flac. See this recent post: Plummeting bitrate.

Changing phones is not an option. I am allergic to wires.
I will do a few experiments with the ebu settings.

For the bitrate, it makes sense what you say, given that it is a loseless format and there should not be any difference in quality among its settings.
But I remember that the final bitrate was always 1411, both when I have chosen compression 8, or no compression. I'll do some tests and come back with the results.
Anyway, why there are 8 compression levels and a no compression one, if the quality is the same? Who and why should want the non compressed (or less compressed) larger file, if the quality is absolutely the same than the more compressed one?

[mville]

Re: Replay Gain, Flac Bit Rate, and upsampling questions.

why there are 8 compression levels and a no compression one, if the quality is the same? Who and why should want the non compressed (or less compressed) larger file, if the quality is absolutely the same than the more compressed one?

Maybe, some users have very slow PCs and don't want the CPU overhead of uncompressing.

[Gio-Gio]

Re: Replay Gain, Flac Bit Rate, and upsampling questions.

got it. An old, now useless option.
I set in 8 and bye.
But, before I need to solve that other thing in the other thread. It is driving me nuts.
thanks