|
Post by Blackdawg on Dec 29, 2018 1:34:33 GMT -6
Normalizing to peaks for outboard I can see making sense.
But still, given that clip adjustment these days I'd just do it case by case.
|
|
|
Post by keymod on Dec 29, 2018 6:40:02 GMT -6
Ok . . . a collosal waste of energy, because it is a destructive process that accomplishes nothing that a non-destructive infinitely adjustable and changeable process better achieves. It's non-destructive in reaper.. Mixbus lets you undo Normalize, and you can Normalize any or all tracks to different levels. So, I'm not sure I understand the difference between Normalize and Clip Gain / Trim / etc. . I am in a hybrid situation, where everything ends up going out through the console where I can also access hardware normally through send/returns, auxes and busses that then can come back on channel strips. And, of course, the channel strips on the console have their own trims which I balanced by using a 1k tone from my tape deck, going out at 0db. Did this so everything was as equal as possible before mixing. Setting all faders on the console to zero, I adjusted inputs to read -6. Seems to work quite well, and sounds good to me.
|
|
|
Post by popmann on Dec 30, 2018 19:42:27 GMT -6
I've considered this as a time saving measure...but, it's not as simple as it appears on the surface. A snare drum 57 might peak at -1dbfs....and register -20vu...what is your level normalization going to do for that? Meanwhile, the least dynamic thing in the world--heavy gain guitars, to get the in proper VU range you probably DO want to "normalize" down to calibration level (so -18dbfs generically)...all to get everything set to what is basically the -18dbfs=0VU calibration that analog modeling plug ins will expect.
Point being, while it's not rocket science...there's no peak level you can change it all to. I will tend to use some logic for shortcuts like if the drummer is clipping...I'll highlight all the drum tracks and pad them 10db. But, that's another part--you want the whole kit to be balanced and move together, so even though the high hat mic doesn't have the peak level, you wouldn't want to peak normalize it to the snare peak level, else you end up burying the hat fader...
I HAVE always wished that a DAW maker would make a kind of "set initial levels" where it goes through and analyzes all the tracks...and moves everything to a calibrated VU level..."virtual assistant engineer"...the VAE could be saving me a lot of time with automation of chores that aren't super subjective but aren't simple enough for "macro" creation.
End of the day, a 40 track production takes me a solid 15min of level setting....and frankly I have to hear what on each track anyway--I just put the one VU up full time while I'm soloing tracks and adjusting trim.
|
|
|
Post by M57 on Dec 30, 2018 20:07:43 GMT -6
It's non-destructive in reaper.. Mixbus lets you undo Normalize, and you can Normalize any or all tracks to different levels. That doesn't sound like normalization; it sounds more like clip gain to me. ..though possibly minus the 'clip' part? Do Reaper and Mixbus let you "normalize" just a section of a track?
|
|
|
Post by jampa on Dec 30, 2018 20:14:36 GMT -6
I use the REAPER add-in sws script "normalise to x LUFS" if I want to normalise
The x is user defined
I find for file transfers - normalising to a fixed point bit depth requires dither, no?
|
|
|
Post by christopher on Dec 30, 2018 20:53:07 GMT -6
I use control+Shift+n quite a bit in Reaper. It normalizes the clip gain. I use it a ton for editing, esp on things like a quiet Tom hit. I’ll slice the clip before and after the hit, normalize and then pull the clip gain down to match the other hits.
Actually in Reaper there’s an option to split wherever a transient crosses a threshold. When I was trying to do impossible things like full band 4 song demo tracked/mixed/mastered all in one day, it came in handy. It allows settings for before and after the transient. I would run that on the kick and snare and it would turn every hit into its own clip. Then right click and draw a box around all the clips, control+shift + n. Now every drum hit was normalized to peak. Because I’m not much into drum replacement, I hoped it would work. The results were meh.. ok, but kinda choppy. So I’d blend it into the original. Much more useful, but really drum replacement probably still would impress people more. I also did this with vocals since there wasn’t time to automate. Again, fine for rap but not a great solution for most things.
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Jan 5, 2019 1:30:44 GMT -6
I use normalization using loudness LUFS in Mixbus after importing the audio tracks on everything. Mixbus shows clipped tracks anyway, blazing red in the waveforms. What this does for me is, that I can re-use whole mixer settings and only have to adjust balances that are song-related, it really is a time saver for me in the end. Seeing normalization as a destructive process is kind of misleading. If the audio format of the normalized track is the internal audio format of the DAW, let's say 32Bit FLOATing point format, it is as non-destructive as any gain. Gain doesn't cost anything in floating point and normalization like in Reaper or Mixbus just does this. The interesting part is the automation you have with it. You have reliable levels to start with and can use mixer templates/settings that often instantly work, depending on genre maybe and on comparability of the imported tracks sonical qualities. For me, it already was a kind of time saver in a few of the last projects. I did not expect it would help as much as it actually did... I see it *somehow* like the "mix to pink noise reference" thing, that some producers let their assistants do before they even start their own magic. Well, I have no assistant, an Normalization does part of the trick for me - getting a better starting point right away. It is different to "just set good recording levels", it is based on measured values of the recorded performance. Really, I try to do both anyway...
|
|
|
Post by keymod on Jan 5, 2019 6:08:15 GMT -6
I use normalization using loudness LUFS in Mixbus after importing the audio tracks on everything. Mixbus shows clipped tracks anyway, blazing red in the waveforms. What this does for me is, that I can re-use whole mixer settings and only have to adjust balances that are song-related, it really is a time saver for me in the end. Seeing normalization as a destructive process is kind of misleading. If the audio format of the normalized track is the internal audio format of the DAW, let's say 32Bit FLOATing point format, it is as non-destructive as any gain. Gain doesn't cost anything in floating point and normalization like in Reaper or Mixbus just does this. The interesting part is the automation you have with it. You have reliable levels to start with and can use mixer templates/settings that often instantly work, depending on genre maybe and on comparability of the imported tracks sonical qualities. For me, it already was a kind of time saver in a few of the last projects. I did not expect it would help as much as it actually did... I see it *somehow* like the "mix to pink noise reference" thing, that some producers let their assistants do before they even start their own magic. Well, I have no assistant, an Normalization does part of the trick for me - getting a better starting point right away. It is different to "just set good recording levels", it is based on measured values of the recorded performance. Really, I try to do both anyway... This workflow is what I'm trying to get to as well. Do you normalize the entire project, regardless of number/types of tracks, to the same value?
|
|
|
Post by EmRR on Jan 5, 2019 7:43:40 GMT -6
Seeing normalization as a destructive process is kind of misleading. . It IS destructive in many DAW: The original file is overwritten.
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Jan 5, 2019 9:16:14 GMT -6
I use normalization using loudness LUFS in Mixbus after importing the audio tracks on everything. Mixbus shows clipped tracks anyway, blazing red in the waveforms. What this does for me is, that I can re-use whole mixer settings and only have to adjust balances that are song-related, it really is a time saver for me in the end. Seeing normalization as a destructive process is kind of misleading. If the audio format of the normalized track is the internal audio format of the DAW, let's say 32Bit FLOATing point format, it is as non-destructive as any gain. Gain doesn't cost anything in floating point and normalization like in Reaper or Mixbus just does this. The interesting part is the automation you have with it. You have reliable levels to start with and can use mixer templates/settings that often instantly work, depending on genre maybe and on comparability of the imported tracks sonical qualities. For me, it already was a kind of time saver in a few of the last projects. I did not expect it would help as much as it actually did... I see it *somehow* like the "mix to pink noise reference" thing, that some producers let their assistants do before they even start their own magic. Well, I have no assistant, an Normalization does part of the trick for me - getting a better starting point right away. It is different to "just set good recording levels", it is based on measured values of the recorded performance. Really, I try to do both anyway... This workflow is what I'm trying to get to as well. Do you normalize the entire project, regardless of number/types of tracks, to the same value? I do it for the entire project. So the mixer template reflects the entire mix balance. Say I have an album to do, I normalize all tracks of every song and then can try the balance of the first one on all following with the mixer template. Depending on how different the songs are, this works very well for me.
|
|
|
Normalize
Jan 5, 2019 10:11:17 GMT -6
via mobile
Post by stormymondays on Jan 5, 2019 10:11:17 GMT -6
|
|
|
Normalize
Jan 5, 2019 15:46:09 GMT -6
via mobile
Post by keymod on Jan 5, 2019 15:46:09 GMT -6
This workflow is what I'm trying to get to as well. Do you normalize the entire project, regardless of number/types of tracks, to the same value? I do it for the entire project. So the mixer template reflects the entire mix balance. Say I have an album to do, I normalize all tracks of every song and then can try the balance of the first one on all following with the mixer template. Depending on how different the songs are, this works very well for me. What level do you normalize to?
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Jan 5, 2019 18:34:02 GMT -6
I do it for the entire project. So the mixer template reflects the entire mix balance. Say I have an album to do, I normalize all tracks of every song and then can try the balance of the first one on all following with the mixer template. Depending on how different the songs are, this works very well for me. What level do you normalize to? Oh, sorry, I messed it up yesterday! I initially WANTED to normalize to a LUFS loudness but ended up using Russel Cottier's lua script which normalizes a batch of regions for -18dB RMS. It is exactly for this purpose, preparing regions before actually mixing. Reason was, that i realized, that it would take quite some time to export audio track by track, and realized the fast way to normalize regions in Mixbus is to peak value or RMS. Then got hold of the lua script here: www.3ravensrecords.com/russellcottierblog/en/downloads/summary/2-free/2-russ-region-normaliser-mix-prep-lua-scriptIt worked so well for me, that after I just let it run over all regions for an album initially, I never thought about it again. Actually, RMS and LUFS normalization results are not SOOO different for well balanced tracks, although, I would like to see LUFS normalization for regions additionally to Peak and RMS, it would accomplish the task even better obviously ... Nevertheless, Russells script is perfectly written for the RMS normalization, so I just used it. It worked well anyway and forgot about that I initally wanted to LUFS normalize but actually didn't and it was RMS. Sorry for the misleading post above. Still, I was thinking about writing a script for the LUFS normalization, which exports tracks normalized and then reimports them to replace the original regions. Maybe I come back to this later, when I have more time, maybe with my next mix project ...
|
|
|
Post by sirthought on Jan 5, 2019 20:15:25 GMT -6
I just purchased their new plugin "TheNormalizer" yesterday while they have it discounted. For something like $5, I can see if it's useful. It's using the same tech for some of their other autogain and VU plugs, but supposed to have less DSP usage.
|
|
|
Post by junior on Jan 6, 2019 1:21:59 GMT -6
I just purchased their new plugin "TheNormalizer" yesterday while they have it discounted. For something like $5, I can see if it's useful. It's using the same tech for some of their other autogain and VU plugs, but supposed to have less DSP usage. I usually use Hornet TrackUtility for this type of thing. I'll have to check out TheNormalizer, though. Seems like it adds LUFS to the options, whereas TrackUtility and VU Meter only auto-gain to RMS or Peak, right? Anyone using both that can comment?
|
|