Blank Error Messages

[MyKart]

Paul,
I'm getting blank error messages, like the one attached. This one popped-up when I was doing a cut(x) in Cleanup Audio. I've also gotten these when Normalising from the Equalization tab which I've take to mean that Normalisation is not needed.
VS 9.03, Mac OS 10.13.4

[Paul Sanders (AlpineSoft)]

Hi,

Are you working with a FLAC file by any chance?  We have another report of a similar problem with those.

In any case, could you please upload a copy of the recording you were working with when this happened and then let me know.  I'll then take a look.

[MyKart]

Yes. There are other anomalies too such as, the levels that Normalization reports do not coincide with the levels that are indicated in the Graphic Equalizer tab. In one instance Normalization reported that the levels were -12 when, in fact, they were -1. As far as I can tell it is not impacting the FLAC files in a negative way, it just seems like an interface bug.

[Paul Sanders (AlpineSoft)]

OK, thanks.  Can you upload one of these files to somewhere like Dropbox so I can take a look?  Thanks.

Also, is this happening on a Mac or PC please?

[MyKart]

VS 9.03, Mac OS 10.13.4

Do you mean one of the working FLACs (a whole side) or one of the saved-out tracks?

[Paul Sanders (AlpineSoft)]

Hi,

A whole side please.  One (or more) that shows as many of these problems as possible.  Thanks.

[Paul Sanders (AlpineSoft)]

Hello again,

We are now on the trail of this and should have a fix soon (it's a previously undetected bug in our FLAC library).

HOWEVER, this should only affect scanning for clicks.  Could you please provide a bit more input on your Normalisation issue (perhaps with the a file that shows the problem) and anything else you have noticed.

I don't need a file for the blank error message anymore - I can reproduce that here now.

Many thanks.

[MyKart]

Sorry it's taken so long. I only have a free Dropbox account so I had to make room and it took forever to upload.
I'm not sure the file will show you what you're looking for since the reporting anomaly only shows itself immediately after the Normalization process completes. It is not apparent, at least to me... when I reopen the file(s), it appears normal.

https://www.dropbox.com/s/rhq64bhfn4rb59l/Radiohead~A_Moon_Shaped_Pool~side1.flac?dl=0

[Paul Sanders (AlpineSoft)]

You need to get fibre, it's fantastic.  Anyway, thanks very much for doing this, I'll take a look and post back here.  I have downloaded the file so you can delete it from Dropbox if you need the space.

[Paul Sanders (AlpineSoft)]

Hello again,

I took a look at this and the file normalises correctly here, in all versions of VinylStudio I tried.  The result is "gain = 1.24" = +2dB on the sliders when normalising to -1dB, which looks about right from what I see in the waveform display.  I also normalised after running a click scan - same result.

Do you have any other filters defined for this album?  If so, and if there is any problem with those, then that might explain the results you are getting.  If you want to follow up on this we can pursue that angle.

Incidentally, I would say that recording at 192kHz is overkill.  It's entirely up to you of course but I would recommend recording at 96 kHz / 24 bit, which will roughly halve your file sizes.  That gives excellent results worthy of even the best equipment ('studio quality').  Me, I always like smaller files.

Those blank error messages will be fixed in the forthcoming update (V9.0.4), by the way.  That is all in hand and should go out in the next few days.

[MyKart]

Yes, the key to happiness is more fibre.    (note to self: buy more Wheaties)

I'm not using any other filters.

Generally, I save tracks at 96kHz and delete the working 192kHz/24bit FLACs after a couple of weeks of proof-listening. Is that folly? You're most probably right but disk space seems cheap to me and I can always use VS to batch-convert if I run low, right? I can hear a marked difference between 44100kHz and 96k & above but, very little or no difference between 96 and above in my "blind" tests. Unless... converting from 192 to 96 is problematic?
Thanks.

[Paul Sanders (AlpineSoft)]

Hi,

If you're saving your tracks at 96k then I would record at 96k.  That avoids a sample rate conversion later on, which, although not the end of the world, might impair quality slightly and will slow things down a bit.  Your existing recordings are fine, but that would be my recommendation for any new ones.

As for the Normalising business, could you possibly email me a screenshot when normalisation is complete?  To do this:

1.  Press Print Screen.

2.  In your email program, select Paste from the Edit menu (or press Ctrl+V) and then email it in.

I will then see if I can make any sense of what's happening.

[Paul Sanders (AlpineSoft)]

Update:

MyKart was kind enough to send me screenshots which explain the confusion.  These are too large to post here but the essence of the problem was that telling VinylStudio to normalise to (say) -3dB does not result in a setting of -3dB on the preamp sliders when normalisation is complete

Nor should it.  When you tell VinylStudio to normalise to -3dB, that's the level it is aiming for in the resulting audio (i.e. the peak level in the normalised audio will be -3dB).  The resulting setting on the sliders is not related to this figure at all.  It is merely the adjustment that needs to be applied to the input signal (boost or cut) to get the result you want.

So, to take a concrete example, if I have source material that peaks at -6dB and I normalise to -3dB then the resulting setting on the sliders would be +3dB.

I hope that clarifies things.  One question remains in my mind though.  Winding up with -12dB on the sliders sounds very wrong.  If you want to follow up on that please let me know.

[MyKart]

Paul,
I was not expecting a Normalise setting of -3dB to result in a 1/1 lowering of the Preamp setting in the Graphic EQ to also be -3dB. The discrepancy I was referring to was in the 'reporting window' which appears after the normalize process is completed. In the screenshots I sent to you, the reporting window reports a gain of "1.51" on the left channel and "1.48" on the right (after normalizing) but back in the Graphic EQ, the adjusted Preamp levels are, respectively, "4" and "3".
In one instance the discrepancy was as high as "-12" and "-1" between what was reported after the process and the levels that are actually implemented within the Graphic EQ. I don't recall on which album that occured and was unable to find it as the 'reported' level alterations are not retained.
At the time of that occurrence, I merely selected another side and then back again to the side I was processing... and the "-12" had changed itself to "-1" on the Preamp slider. I did a reset and tried again with the same result. I wish I had thought to take a screen shot but did not. The processing, itself, seems to have been done correctly. It was only just the 'reported result' that was in error.

[Paul Sanders (AlpineSoft)]

Hmmm, that's strange, I'll take a look at the code.  If you rediscover the offending album / album side, please let me know and we can look into it.

A gain of 1.51 mapping to +4dB whereas 1.48 maps to +3dB is a result of the granularity of the preamp sliders.  It probably doesn't help that we report the gain as a factor whereas the sliders work in dB.  I should probably revisit that.

Sorry I misinterpreted your original report.  I thought I was on the right track there, but alas no.