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: foo_truepeak True Peak Scanner (Read 16674 times) previous topic - next topic - Topic derived from Re: Resampling Hi-Res...
0 Members and 1 Guest are viewing this topic.

Re: foo_truepeak True Peak Scanner

Reply #200
You can now write RMS. In settings you can choose what method ... whether normal or AES +3dB mode. The tag you write for both methods is the same, so there is no way to determine afterwards which method was used.
You can change the tag field name for your own knowledge. Other than that no.
No sweat. I'm not going to use this field anyway. Same for Max LUFS-S/M. I don't see the point in knowing these maxvalues in an audiofile while already having all the other info. Unless somebody can explain me the extra merit of these values/tags.

Re: foo_truepeak True Peak Scanner

Reply #201
Let's say someone, me perhaps, already scanned 1,000's of files with this and I want to go back and add the new RMS values.  I can turn off all scanning aside from the True Peaks, Peak Timestamp, and clipping..  If I just want to add the RMS scanning it will scan True Peaks, Peak Timestamp, and clipping too.... along with RMS. Can they (True Peaks, Peak Timestamp, and clipping) be toggled off or are they required to generate RMS? 

Thanks for the reply on clipping and its on-hold status.

I'll let you know if anything is broke in 0.6.12, haven't broken it yet.  :)

Re: foo_truepeak True Peak Scanner

Reply #202
Let's say someone, me perhaps, already scanned 1,000's of files with this and I want to go back and add the new RMS values.  I can turn off all scanning aside from the True Peaks, Peak Timestamp, and clipping..  If I just want to add the RMS scanning it will scan True Peaks, Peak Timestamp, and clipping too.... along with RMS. Can they (True Peaks, Peak Timestamp, and clipping) be toggled off or are they required to generate RMS? 

Thanks for the reply on clipping and its on-hold status.

I'll let you know if anything is broke in 0.6.12, haven't broken it yet.  :)
You only save around 40% of the time if you disable the LUFS-S/M scan. All the other scans don't add any significant time to the scan at all.

Re: foo_truepeak True Peak Scanner

Reply #203
Without AES +3dB there's a fairly big difference between the values reported by TPS and DR Meter, scanned album is Queen I


Re: foo_truepeak True Peak Scanner

Reply #204
That is expected. I believe I have said it many times, the DR Meter's RMS method is nonstandard. For some reason the mean value is doubled before square root is taken.

Re: foo_truepeak True Peak Scanner

Reply #205
That is expected. I believe I have said it many times, the DR Meter's RMS method is nonstandard. For some reason the mean value is doubled before square root is taken.
Does that in your opinion mean it would make more sense to display RMS (without the +3dB) instead of DR when available grid space for display is limited?

Re: foo_truepeak True Peak Scanner

Reply #206
Personally I have never trusted DR numbers, it's quite arbitrary and its measurement is easy to fool. But as many people said they want it for knowing which mastered version they have of a certain record, and for that it works well. If your use case is similar, DR is probably more useful.

Re: foo_truepeak True Peak Scanner

Reply #207
Personally I have never trusted DR numbers, it's quite arbitrary and its measurement is easy to fool. But as many people said they want it for knowing which mastered version they have of a certain record, and for that it works well. If your use case is similar, DR is probably more useful.
I noticed that in your minibar you show Peak and RMS, not DR. I tend to switch to displaying RMS instead of DR in my skin.

As a more general point for discussion (and I've asked this question in different places, but never an answer by anyone).
Because of TPS a myriad of audiotrack keyvals is available, but space in a grid is limited.

I'm showing something that represents clipping and I know it's flawed because it doesn't quantify how far the clipping is over 0dB, but it actually is a very good indicator anyway.

I show LRA and I think that one is quite unique compared to the other keyvals.

I also show both PLR and LUFS-I, but is that really useful since I always play audio with RG (track, apply gain and prevent clipping), and PLR is just LUFS-I corrected by peak. So I think in my usecase PLR makes more sense to display than LUFS-I, I guess.

And I show DR, which I might swap for RMS.

Keyvals available are:
LRA
PLR
LUFS-I
Max LUFS-S
Max LUFS-M
RMS (or RMS+3)
DR

So the question is ... what combination of 4 of the above available keyvals represents the nature of the audiotrack best.
Anyone has an opinion and is willing to share?

Re: foo_truepeak True Peak Scanner

Reply #208
Hi @Case,

I have very strange measurements of RMS with your TPS.
If I scan with DRMeter the RMS gives different values for different tracks, albums, pressings/masters, genres. Ex. Sade or Dires Straits or Pink Floyd give bigger values than some modern or remastered metal - that is to be expected.
I know that RMS in TP is somewhat "custom" and those values are not real RMS but they show what is expected - more RMS gives more DR, more dynamic music gives bigger RMS that modern compressed noise.
But using TPS with +3 mode the measurements always (really always) give me readings of 9,xx to 10,xx no matter what music style, pressing, codec and everything else. So ex. SADE from 1980s with DR15 and LUFS 20 gives me RMS 9-10 and newest metal recordings with DR6 and LUFS7-8 also give me RMS 9-10. And within the same album it does not matter if there is some pretty dynamic ballad that begins with soft passages and ends with metal or if there is straight power metal one tempo song...all give me RMS 9-10.
That seems very strange.

Re: foo_truepeak True Peak Scanner

Reply #209
Thanks for noticing. Horrible and idiotic bug. In the first internal version the scanner converted the track RMS results to dB and the +3dB mode just added the offset to the value. But before release I had changed the RMS values to be stored in linear floating point for simpler calculations. But I had forgotten to modify the +3dB mode offset shift to take this into account. That caused values in different units to be mixed incorrectly causing the results to be entirely wrong. Sorry about that.
Fixed version uploaded.

Re: foo_truepeak True Peak Scanner

Reply #210
@Case

When you select either one of the first two context menu items for a SACD iso, the 'Peak timestamp', 'Clipped samples in track' and 'Clipped samples in album' fields still get written to the xml file, albeit without any values. So, you end up with empty entries. This did not happen when I tried it with MP3's (verified by looking at an MP3 with a hex editor).

I tried the same with External TAGs and this changed the picture overall, here too the same fields get written to the SQLite database, verified with DB Browser for SQLite. Both for the SACD iso and for regular MP3's. Tried again with External TAGs but his time using a folder.tags file, and again, those same three fields get written without values for both formats. And with both External TAGs options, these empty fields show up in the Properties dialog for both formats.

Looks like this only happens when an external target is used to store the tag information. Like XML for SACD iso's, SQLite or folder.tags with External TAGs installed.
I've not tried m-Tags... trying to wean myself off from using that one.

Re: foo_truepeak True Peak Scanner

Reply #211
@Case

Two questions:
1) I have the foo_hdcd plugin installed, so when I play HDCD content the bits per sample will be 24bit (20bit actually) and also Peak Extension, Transient Filter and HDCD Gain will be applied when present.
When TPS is run on a HDCD, does it use the standard CD 44/16 content or the HDCD runtime content to calculate it's values?

2) When starting fooBar upon playing first track I notice that fooBar will use dBpoweramp/RetroArch for automatic resampling.
I tried to find the setting in Preferences to change this to SoX but did not find it. How can I set automatic resampling to SoX?

Re: foo_truepeak True Peak Scanner

Reply #212
2) When starting fooBar upon playing first track I notice that fooBar will use dBpoweramp/RetroArch for automatic resampling.
I tried to find the setting in Preferences to change this to SoX but did not find it. How can I set automatic resampling to SoX?
It's in Preferences/Tools/ReplyGainScanner:

Re: foo_truepeak True Peak Scanner

Reply #213
2) When starting fooBar upon playing first track I notice that fooBar will use dBpoweramp/RetroArch for automatic resampling.
I tried to find the setting in Preferences to change this to SoX but did not find it. How can I set automatic resampling to SoX?
It's in Preferences/Tools/ReplyGainScanner:


Never thought this option was hidden under Replaygain Scanner, which I disabled because I include Replaygain in TPS.

I changed both settings under Replaygain Scanner to SoX, exited fooBar, restarted fooBar, checked if both settings are set to SoX, and start playing a normal 44/16 flac file.

Still I get:
Opening track for playback: "D:\+T\Combichrist - 2016 - Live at Summer Breeze 2015 [FLAC] (16bit-44.1kHz)\03 - Zombie Fist Fight (Live at Summer Breeze).flac"
Automatic resampling: using Resampler (dBpoweramp/SSRC): 48000 Hz, Resampler (RetroArch): 48000 Hz
Device: Speakers (Realtek(R) Audio)
Mix format: 48000 Hz / 32-bit float / 2 channels (0x3)
Sending stream: 48000 Hz / 32-bit float / 2 channels (0x3)
Opening track for playback: "D:\+T\Combichrist - 2016 - Live at Summer Breeze 2015 [FLAC] (16bit-44.1kHz)\04 - Can't Control (Live at Summer Breeze).flac"

I don't get it ...

Re: foo_truepeak True Peak Scanner

Reply #214
2) When starting fooBar upon playing first track I notice that fooBar will use dBpoweramp/RetroArch for automatic resampling.
I tried to find the setting in Preferences to change this to SoX but did not find it. How can I set automatic resampling to SoX?
It's in Preferences/Tools/ReplyGainScanner:


Never thought this option was hidden under Replaygain Scanner, which I disabled because I include Replaygain in TPS.

I changed both settings under Replaygain Scanner to SoX, exited fooBar, restarted fooBar, checked if both settings are set to SoX, and start playing a normal 44/16 flac file.

Still I get:
Opening track for playback: "D:\+T\Combichrist - 2016 - Live at Summer Breeze 2015 [FLAC] (16bit-44.1kHz)\03 - Zombie Fist Fight (Live at Summer Breeze).flac"
Automatic resampling: using Resampler (dBpoweramp/SSRC): 48000 Hz, Resampler (RetroArch): 48000 Hz
Device: Speakers (Realtek(R) Audio)
Mix format: 48000 Hz / 32-bit float / 2 channels (0x3)
Sending stream: 48000 Hz / 32-bit float / 2 channels (0x3)
Opening track for playback: "D:\+T\Combichrist - 2016 - Live at Summer Breeze 2015 [FLAC] (16bit-44.1kHz)\04 - Can't Control (Live at Summer Breeze).flac"

I don't get it ...

There isn’t a setting to change the resampler.  True Peak should just detect Sox being installed and use it. 

Re: foo_truepeak True Peak Scanner

Reply #215
2) When starting fooBar upon playing first track I notice that fooBar will use dBpoweramp/RetroArch for automatic resampling.
I tried to find the setting in Preferences to change this to SoX but did not find it. How can I set automatic resampling to SoX?
It's in Preferences/Tools/ReplyGainScanner:


Never thought this option was hidden under Replaygain Scanner, which I disabled because I include Replaygain in TPS.

I changed both settings under Replaygain Scanner to SoX, exited fooBar, restarted fooBar, checked if both settings are set to SoX, and start playing a normal 44/16 flac file.

Still I get:
Opening track for playback: "D:\+T\Combichrist - 2016 - Live at Summer Breeze 2015 [FLAC] (16bit-44.1kHz)\03 - Zombie Fist Fight (Live at Summer Breeze).flac"
Automatic resampling: using Resampler (dBpoweramp/SSRC): 48000 Hz, Resampler (RetroArch): 48000 Hz
Device: Speakers (Realtek(R) Audio)
Mix format: 48000 Hz / 32-bit float / 2 channels (0x3)
Sending stream: 48000 Hz / 32-bit float / 2 channels (0x3)
Opening track for playback: "D:\+T\Combichrist - 2016 - Live at Summer Breeze 2015 [FLAC] (16bit-44.1kHz)\04 - Can't Control (Live at Summer Breeze).flac"

I don't get it ...

There isn’t a setting to change the resampler.  True Peak should just detect Sox being installed and use it. 

Ehm. I wasn't asking because of TPS. I know it uses SoX automatically upon scanning when SoX is installed.

I want to set the automatic resampler to SoX too. Quite weird you apparently cannot do that within fooBar.



Re: foo_truepeak True Peak Scanner

Reply #218
When you select either one of the first two context menu items for a SACD iso, the 'Peak timestamp', 'Clipped samples in track' and 'Clipped samples in album' fields still get written to the xml file, albeit without any values. So, you end up with empty entries. This did not happen when I tried it with MP3's (verified by looking at an MP3 with a hex editor).
Thanks for noticing. All other tag fields had extra checks so that empty fields are not written but for some reason they were missing from clip stats. And as most tag types have internal prevention against such fields the bug went unnoticed. Fixed in 0.6.14.

1) I have the foo_hdcd plugin installed, so when I play HDCD content the bits per sample will be 24bit (20bit actually) and also Peak Extension, Transient Filter and HDCD Gain will be applied when present.
When TPS is run on a HDCD, does it use the standard CD 44/16 content or the HDCD runtime content to calculate it's values?
Postprocessors like DTS decoder and HDCD decoder are enabled by default and used automatically unless a decoder is opened with special flags asking to bypass those. No component of mine asks to bypass them, post processors are always allowed to be enabled.

Re: foo_truepeak True Peak Scanner

Reply #219
Wishlist item: Let it do the RG scan & tag too.
Because if you want to do both, RG scan & tag, and scan positions or DR, then that is the kind of job to wait for to complete, before starting the next.

Rather than having tons of submenu items (combinations of what to do, and include album scans?) then I propose the following:
* Top item in the submenu is to scan all and tag and open file properties immediately afterwards (like External Tags can do). Facilitates the following workflow: Start the job -> wait for it to finish -> delete the unwanted tags -> OK.

Re: foo_truepeak True Peak Scanner

Reply #220
Wishlist item: Let it do the RG scan & tag too.
Because if you want to do both, RG scan & tag, and scan positions or DR, then that is the kind of job to wait for to complete, before starting the next.

Rather than having tons of submenu items (combinations of what to do, and include album scans?) then I propose the following:
* Top item in the submenu is to scan all and tag and open file properties immediately afterwards (like External Tags can do). Facilitates the following workflow: Start the job -> wait for it to finish -> delete the unwanted tags -> OK.
As far as I know, that's exactly how I use it already. One scan does all in one go. I also cleaned up the rightclick menu.

Or do you mean something else?


Re: foo_truepeak True Peak Scanner

Reply #222
I have obviously missed something, and in particular this

I also cleaned up the rightclick menu.
Wait. How?!
See attachment.

Items in the context menu you deactivated, won't show up on a rightclick. Unless you rightclick while holding SHIFT, then all context menus including the disabled ones magically reappear.

...

Forgot to tell you in the previous post. The only thing that TPS does not do in a scan including all that stuff is writing the physical DR logfile. It does calculate and write DR tags though.

So if you really need the physical DR file as well you still need to do a separate foo_dr_meter scan. Which is quite unfortunate.


Re: foo_truepeak True Peak Scanner

Reply #224
Looks like @Defender replied to your comment, which is good. That Context Menu control is not a Columns UI thing. It is a built-in feature allowing you to hide commands you don't want to see. Back in the ancient days there was a similar config option for main menu commands too.

I don't think automatically opening properties with this component offers benefits.