SACD hell

[ArmyOfQuad]

Anyone come up with a simplified workflow for going from SACD/SACD iso to properly leveled multichannel flac?

Been experimenting, been reading threads, been consulting with AI (which has a hilarious habit of responding to accusations of not being able to tell time by inserting incorrect time stamps verified by time.gov to every response going forward), but coming up empty.

I am impossibly behind on ever ripping everything I've purchased in the past few years, and need to come up with efficient workflows, and then automate them the best I can.

Not to mention, if I never get caught up on ripping, I'll never build things to use my rips.

 

[steelydave]

If you convert everything with zero boost (so no chance of anything clipping and having to be redone), apply ReplayGain data in foobar, and then play back in foobar, kodi, or another app that supports ReplayGain, everything will come out sounding roughly the same volume.

 

[ArmyOfQuad]

replaygain keeps saying the levels are in the range I want already, but all conversions give me .flac files at -5. Sure, I can manually boost, but I need to automate. Yet to find a way to get a reliable read and setting set. It seems the only way is to convert, check each file, manually set - this is way too many steps and way too time consuming. We can land a man on the moon, but we can't read an accurate level of an audio file and set a gain automatically to rip a decent flac?

 

[steelydave]

The flac files will be at -5 or whatever - replaygain adds metadata to the file telling the playback software to either boost or cut the playback volume to meet a target specified in the standard:

"The target loudness is specified as the loudness of a stereo pink noise signal played back at 89 dB sound pressure level or −14 dB relative to full scale.[3] This is based on SMPTE recommendation RP 200:2002, which specifies a similar method for calibrating playback levels in movie theaters using a reference level 6 dB lower (83 dB SPL, −20 dBFS).[note 1]"

So it doesn't actually change the files, your file will still digitally peak at -5dB, but ReplayGain will tell your playback software to increase the volume so that it sounds just as loud as a track that peaks at -0.1dB.

 

[ArmyOfQuad]

That's not what I want though, I want properly ripped flac files. Not improperly ripped quiet files that rely on tech to level, I don't want to rely on software to be automatically levelling everything.

Another weekend wasted on tech nightmares trying to do something getting nothing done finding out tech has failed me again while project after project sits undone.

Sometimes I feel like the world is conspiring to stab me in the back.

 

[steelydave]

Ripping with a 0dB boost is "properly ripped" FLAC files, and you're 'relying on tech' to both do the conversion, and have them digitally peaking at a uniform level inside the file itself.

Fast / Cheap / Good: pick two, and you want cheap and good - ReplayGain actually gives you all three, ask anyone who's used it, it's a game changer for digital music playback. Feeling like the world is conspiring to stab you in the back will always become a self-fulfilling prophecy if your approach is so myopic that you won't consider good solutions just because they don't tickle your OCD just right.

 

[ArmyOfQuad]

I don't even know how I'm going to play these back or where replaygain will fit in with that. And even if I do sort that out today, I don't know if that's relevant tomorrow.

Finally found a DR reading app for foobar - not as easy to find as you would think, I just love spending 30 minutes searching down an add in that many refer to with dead links, finally finding the add in, installing it, using it - and it telling me that it expired and to download the latest version from a website that doesn't load anymore. I hate assholes that embed time bombs into software. But I eventually found a working one, so now at least I have a simple method of getting a relevant number which I can plug into a gain.

I think a workflow is shaping up, after some refinement next step will be asking AI if I could script it. At least I get some amusement from that round of stupidity - yesterday it was walking me through steps to add a stripped down interface to an outdated fire cube, and kept giving me dead links, which I kept calling it out on, and it kept apologizing. At some point I asked it to verify all links before providing them, and it responded by stating it verified it, with a timestamp of verification, verified by time.gov. The timestamp was 45 minutes off (it included the time zone, so that wasn't a part of it either). Which devolved into a conversation about it's ability to tell time. Which - since it keeps all info in the chat relevant to further response, devolved into a discussion about modifying a fire cube, with every step verified with incorrect time stamps verified by time.gov. That I found this behavior amusing was about the only value - the end result was that amazon has fucked with the fire cube so much, all efforts to customize it are blocked and extremely difficult to work around, if possible at all. Absolutely infuriating the effort that amazon goes to to prevent one from using their physical property that they own, bought, and paid for, how they see fit. Well - I can see fit to use it as an object to smash to bits in the driveway I suppose.

Anyways - back to.....I dunno what anymore. Something with SACDs maybe?

 

[Mandrake2011]

I have 77 SACDs ripped to FLAC with the levels preserved. I use Foobar2000's ReplayGain scanner plugin to analyze the loudness (using EBU R128) and tag the gain levels per track and per album in the files. I then play the FLAC files through Foobar2000 with replay gain enabled. This gives me a uniform perceived loudness for anything I play.

See https://wiki.hydrogenaud.io/index.php?title=ReplayGain

The ReplayGain scanner plugin has some settings that you can tweak. It's important to use the same settings across your entire collection. Then any playback software or device that supports replay gain will play back your music with a uniform loudness level. (There are some nuances to this depending on whether you are playing back an album, or a playlist with tracks from various albums.)

Dynamic Range and the DR plugin are not involved in the replay gain process. I do use the DR plugin, but only for comparing different releases and masters of the same album/track. The DR plugin is informational only. It does not affect the sound or level.

 

[ArmyOfQuad]

That's useful information.

However, I don't care about playback level right now, I care about creating ideally ripped files that don't require worrying about playback level later.

The DR plugin is for determining the level needed for the flac creation process.

I'm currently working with Grok to build a python script that will automate the whole process into 1 step to extract .dsf files, determine the peak value, and convert to flac with that. Not sure if it's possible, but I'm finding that grok can help someone with no python coding experience code a script by making a broken version, testing it, copy paste the error output, and get an update script....lather rinse repeat....until it works.

So far I've got the thing automatically extracting .dsf files to a folder structure - the part of the script is running into elevation being required to run the process for checking levels. I fucking hate security - it seems to not actually prevent what it's meant to prevent, and just makes everything we need to do harder. But that's the world we live it, the bad guys have won, and have made the world a much worse place for it, so instead of efficiently utilizing our tech, we waste our hours away working around the restrictions we're forced to accept, in the name of security.

 

[DuncanS]

I've never bothered ripping my SACDs, but DSD can clip/go unstable if signals are too high when converting to PCM because there is feedback and noise shaping used in the process/algorithm (I have had my player or AVR will blank the audio if that happens - its why I have occasionally thought discs were bad). So analogue audio signals are converted to DSD at (I think) -3dB below the maximum it could be, to limit this.

 

[ssully]

Anyone come up with a simplified workflow for going from SACD/SACD iso to properly leveled multichannel flac?

Been experimenting, been reading threads, been consulting with AI (which has a hilarious habit of responding to accusations of not being able to tell time by inserting incorrect time stamps verified by time.gov to every response going forward), but coming up empty.

I am impossibly behind on ever ripping everything I've purchased in the past few years, and need to come up with efficient workflows, and then automate them the best I can.

Not to mention, if I never get caught up on ripping, I'll never build things to use my rips.

The ony channel that might need to be 'properly leveled' is the LFE. This is because LFE is basically the wild west of mixing and decoding. It's tragic that we even have to deal with it, 99.9% of music does not need a Low Frequency Extension/Effects channel. It's there because consumers are mostly ignorant, and complain when they 1) don't use bass managment and 2) don't hear their overly priced subwoofers doing 'something'.

Otherwise, 'proper levelling' can be done with your volume control when you play the file. Or a tool like Replaygain, though was designed for two-channel and I'm not convinced it really is smart enough to do multichannel right given the extreme level differences between channels in multichhannel mixes. Replaygain is used in your file player, e.g. foobar2000, and it can only be used with decoded output to your AVR, i.e., not raw bitstreamed data that needs decoding in your AVR.

All that said, I don't understand the problem you are having. How/why are all your SACDs rips coming in at '-5'? What does that even mean?

 

[ArmyOfQuad]

It's always been in my training that if you leave a lot of headroom in the level of your digital files, you're not taking advantage of the full bitdepth. But converting DSD to PCM doesn't seem to follow a consistent level which leaves many significantly quieter than others. I was hoping maybe someone had come up with a quick workflow to check levels and set gain - but apparently the world has moved in a different direction of ripping quiet files and utilizing level detection and gain on the fly when playing back. Which seems ass backwards to me - but whatever floats your boat.

Now that I've determined there isn't a workflow that one has come up with yet, I'll get back to my python project.

 

[mkt]

I have sort of stopped converting from ISO to flac.

I was doing a two-step process using saracon (but I don't think the converter is critical). First convert the ISO with 0 dB gain into a single flac file. Use sox to get the peak gain (single file means the peak for the entire album). Convert a second time with the adjusted gain, trim and split with sox. All in bash script with lots of seds and grips. And worse, saracon under wine and the rest in linux

So if you come up with a python workflow, I'd give it a spin.
EDIT: there might be a mostly sox way to script it.