Plugin to keep volume constant when doing A/B comparisons?

And the question is how good do they work, is it a simple auto gain like in StudioEQ, or a more advanced algorithm that take the human perception into account. BTW would love the auto gain from the StudioEQ to find its way into the regular EQ
I think I will just use my platinum ears, for now

Samplemagic A/B has RMS, don’t know about Perception.

In GearSlutz the co developer of Perception (Ian something) said that it worked with ebu128 loudness plus frequency content to copy the way our ear works

There is another option (free) for those that are willing to go to a small amount of trouble:

Install ReaPlugs (Reaper’s free, ugly, but very powerful fx suite)
http://www.reaper.fm/reaplugs/index.php

Download TB Audio’s AB Compare
Direct link to zip file: http://www.tb-software.com/TBProAudio/download/AB%20Level%20Matching%20JSFX.zip
Link to overview page: http://www.tb-software.com/TBProAudio/ab_lm.html

After installing ReaPlugs you should have a ReaPlugs folder in your VST2 folder.
Unzip TB Audio’s zip file into your directory: VSTFolder\ReaPlugs\JS\Effects\TBAudio

Place an instance of reajs in slot 1 of your inserts
Place another instance of reajs in the last slot of your inserts (probably just use slot 6 which is pre-fader)

Within the reajs window, click on Load, go to the TBAudio folder and click on the following:
first slot: AB_LM_src
last slot: AB_LM_cntrl

The cntrl window is the one you’ll use to monitor and bypass the comparisons so you can close the src window. There is a detailed PDF instruction manual in the TBAudio folder so refer to that for help.

I’ve noticed a little bugginess with reajs inside Cubase 8 so be sure to test it before using it on an important project.

Side note: There are several other gems within the reajs folders, such as the Stillwell plugs (similar to Stillwell’s excellent VST plugs but w/out the GUI and lacking features in some instances).

Great tip lachrimae, I will try it on WaveLab for Mastering purposes

Thank you for pointing out this option, lachrimae! BTW - what kind of bugginess are you seeing in CPro8 using that plug-in?


Here’s another potential option:
Melda MCompare, http://www.meldaproduction.com/plugins/product.php?id=MCompare .

Per the developer, one instance goes at the start of the processing chain, and one at the end, it sounds similar to how the TB one is used.

Thanks again!

OK, I did some checking of this. Using the Cubase LUFS meter, I found the reajs TB volume-matching was only accurate to within 1-1.5 dBLUFS. In other words, when “BYPASS ON” was toggled with “BYPASS OFF” in the plug-in playing a loop of audio, the volume changed by that much.

The “BYPASS OFF” was always a bit louder during my testing, making the eq’d/comp’d audio that much louder.

How acceptable that is is a judgement call, of course, but I just wanted to pass along my observations.





Also - on 1 or 2 occasions there was some phasiness as though the TB PDC wasn’t working properly. I had to close down Cubase and reopen the project to make it go away. I notice there is a PDC toggle described in the TB manual, but it doesn’t appear in my GUI.

Also noticed that the equal volume comparison of comp eq active/bypassed was made (via “Autogain RMS”) at the volume of the bypass. It would have been much sweeter if it had chosen to present the comparison of the two at the volume of the comp/eq being active … since that’s what our ears have been listening to as we mix.

Finally, the ergonomics of using the reajs/TB didn’t seem optimal. First of all, that plugin stole the qwerty keyboard from Cubase, so every time I wanted to start/stop Cubase, I had to click outside the plugin. Also, the A/B comparison needed to be click in one spot, drop down menu, then click in a different spot … a toggle would have been awesome for eyes closed/blind testing (I didn’t see a keyboard shortcut option). Finally, I found that it took a really long time for the plugin to “wake up” when coming back in from clicking outside its GUI unless one of its controls was toggled (I used the RMS Window MS), so more click/click.

Having said all that, it is free! And I think it is really really cool that it’s not just limited to bypassing inserts on one track - I had a chain of effects in different groups, and I was able to put the source before the start of the chain/the control at the end, and then could check out the true effect (within that 1-1.5 dBLUFS) of each of the inserts in the chain, individually and “summed”. :smiley: I will probably trial the Melda one, MCompare I think it’s called, when it comes out, and see how much if anything extra one gets for $60 or so.

Until then - I really appreciate your pointing the way to this thing, lachrimae!

Thanks for the detailed report. The issues you’re reporting are exactly the kind of bugginess I’ve experienced.
Regarding the volume matching, it does seem that different material requires a different rms value (or peak setting) in order to be more level matched (kind of defeats the purpose of having an automated solution…). But yeah, I should have said in my original post that I really appreciate TB making the tool available, for free no less :slight_smile:.

You mentioned MCompare which reminds me of another tool: http://lacinato.com/cm/software/othersoft/abx
This one is stand alone but I think it works great for comparing mix downs, or comparing to reference tracks. It’s also very handy when comparing reverb, preamp, and other shootout files.

Looked some more at it, trying to get it more accurate and less fiddly.

I noticed there was a choice in the loading to choose a different TB insert plugin. Instead of this one from lachrimae’s post:

last slot: AB_LM_cntrl

I selected this one:
last slot: AB_LM**LT**_cntrl .

Though it fixed the phasing TDC issue, and a lot of the ergonomics (toggles, yeah!), … the volume matching still was off - up to 3dBLUFS this time (I guess it probably depends on the source material). That is a lot, IMO. I guess the “Autogain” function isn’t as accurate as I was hoping for.

Step by Step “Repro”: Mixing chain, with ReaJS powered down: -16.9 dBLUFS.

: With ReaJS powered up, but it’s green glowing “Bypass” and “Autogain” buttons off: -27.7 dBLUFS
(in other words, just turning on ReaJS dropped the volume 9dBLUFS - not a deal killer, but it is at a very different volume than the one we’ve been mixing at.
: Turning on “Autogain” - no change, good!
: Turning ReaJS’s glowing green “Bypass” button (and this is where it is critical for the volume not to change): Volume jumps up to -24.7 dBLUFS.

So to summarize, at the time that the plugin is supposed to have no change in volume at all (when the effect chain is bypassed), it is sometimes changing a lot, in the range of 3dBLUFS, louder in the Bypass state usually for me. Tantalizingly, at times it does seem to work within 0.2-0.3 dBLUFS. I have this feeling it has to do mainly with the “SYNC PDC” function, because it will sometimes display different values when I click it a few times without changing anything, and that’s when things seem awry a bit.

Can someone else check this out also please, maybe just do the 4 underlined steps above? I would love it if all this was operator error, and someone could show me how to make it work.

As it is now, I’m not sure it is the right tool for what I was looking for, reliable volume-matched comparisons of effects like EQ/Comp/Distortion - on vs. off.

Interesting, thanks. That lacinato looks like an alternative option for SampleMagic A/B and Perception maybe? My holy grail is tweaking a comp or EQ, and being able to hear the difference in a volume matched fashion, but without having to bounce down.

I don’t think Lancelot, or Python, ever did find their versions of the holy grail, I should probably take a clue from that … :open_mouth: :laughing:

I have both Perception and MagicAB. The two products were really designed for different purposes… if your goal is to be able to hear the effects of processing on your own material without volume bias, Perception is the tool to use. MagicAB is more for easily referencing against other tracks.

Perception is very good at what it does. The very first version of it was actually quite buggy with Cubase, but within a week or two they patched it and it has worked great for me since then (Windows Cubase 7.5, 64-bit). There might not be any real magic behind it, but it is so quick and easy to use (and it does what it’s supposed to) that I think it’s worth the price. I have used it both on the master bus, when checking out the effects of my master processing- but I also occasionally put it on a track or a group when I’m fine-tuning compression or saturation.

EDIT: I had lots of problems with the first version of MagicAB. I reported the issues and they eventually issued an update that greatly improved its support of MP3 files (I had found many MP3s that simply wouldn’t load in the first version), but last I checked it still had a problem where you couldn’t import files that had UNICODE characters in their filenames. ASCII filenames only.

Great to know, thanks Ultimate Outsider! Sounds like you are combining Perception and SampleMagic A/B? That Melda product is being marketed as a combination of the two plugins - Perception for intra-mix comparisons, and SampleMagic A/B for reference track comparisons. If that doesn’t pan out for me (the Melda product), it’s good to know that Perception can at least do the trick for what I’m looking to do now.

Do you happen to know, can multiple (as in 3) instances of Perception be placed on a single track?

Thanks again!

Not quite on topic but; Both Voxengo Elephant and Slate Digital FGX have a mode (switch) to keep the effect on but match the input level. Very handy for mastering…

The Melda one looks nice on paper (I need to try the trial first)

Hey distante, seen you around the same places I’ve been on this topic!

Please let us know what you find about how accurate the volume-matching is. I looked at MCompare’s .pdf document, I agree it looks great, but of course it has to be accurate, AND sound good (no weirdness, phasing, etc.).

I’m brain-melted from looking at that ReaJS/TB Audio thing, thankfully I won’t be needing anything for a week or two - look forward to your observations!

I’ve used MagicAB a couple times when I was at the beginning of the mixing stage in projects when I wanted to compare arrangement and EQ balance between my own track and other peoples’ tracks that had a general vibe I was aiming for. I would just load up my reference tracks in MagicAB and set up loops in the sections I wanted to compare my track to. Super easy.

Regarding multiple Perception instances: The product comprises two plugins: The source (pre-FX) and controller (post-FX). While you can have multiple instances of the plugins in a single project, you can only have one source and one controller ENABLED at a time. The controller plugin will even display an error if multiple source instances are detected.

This hasn’t been a problem for me, since I only use the plugins when I’m working on one specific task. I insert/enable them when I’m actively doing things on a given track/bus, and disable or remove them when I’ve adjusted the processing to taste.

Thanks again, UltimateOutsider, that is all really helpful information. :smiley:

Hi great topic and handling.

I have the Perception plug it is working fine, it is good value to me. I got It half price, keep watching the sales …

I had an isue that the levels on or off where not consistent,they should be the same, but It was another plug in that was making the problem, I wrote to the support, and it was solved fast.

But what I want to share is, when you need support conserning some issue with the Perception plug, Ian Shepherd’s is fast in replying, great support - that’s what I value a lot, haha what I need alot :stuck_out_tongue:


ok succes.

cubase 7.5.3 - I7 3770 Windows 8 64 bit - 32 gig ram - many ssd’s -
http://www.basles-rotterdam.nl

just port smart bypass from wavelab to cubase… we already have this tech…

smart bypass isn’t in Wave Lab elements neither :frowning: