Hello,

I've wondered if it is possible in any way to get the SoundID profile of my speakers into FabFilter Pro-Q with EQ matching. It would be very useful to then give the Pro-q preset to clients who use my studio without need to have a license for soundID.
Any help is much appreciated.

 

Create a track with a white noise generator on in your DAW and send it to one track with SoundID+Pro Q on and another track with just Pro Q on. Capture the profile with the Pro Q instance that is by itself. I haven't tested this, but it shouldn't be harder than that.

 

thanks!

 

isn't soundid also changing the phase etc and not just the gain of frequencies ?

 

I have done this before and the result is almost there, just serviceable enough and better than no curve at all. It's a better idea to use pink noise instead of white noise, though.

Edit: worth saying that I have done this on headphones, not speakers. Speakers might have different processing applied by SoundID that won't translate well by capturing the curves. Still worth trying.

 

yes and no. it changes the left and right signal independently but in my case they a nearly identical. It should be more than enough for my clients though. I may have a difference on some frequency of 0,5 or 1 db

 

It depends on what your settings are. If you use linear phase mode it just works like a linear phase EQ. There is also "listening spot" correction that creates a short delay on one of the speakers, but this is only necessary if your speakers are not placed correctly.

 

You capture SoundID's correction curve with Pro Q's EQ match function and then save that as a preset or whatever. It won't work exactly like SoundID, but you get close to it. If you want it to be closer I think you need to do one capture for each speaker and use two instances of Pro Q, one for L and one for R, if you want to use it instead of SoundID. You can probably mash together the two captures in some way so that you only need to use one instances of Pro Q, but I don't know how you would do that.

EDIT: Tried doing the one capture for each speaker and it works fine. It sounds very close to SoundID doing it this way. You have to make sure that only the channel you want to capture is playing on both the SoundID track and the Pro Q capture track when you do this, otherwise the capture will just be a huge mess. When you use these two captures together afterwards you need to make sure that the two Pro Q instances are placed in parallel and are processing the correct channels in relation to the captures.

 

Last edited: Apr 26, 2022

keep us posted about the results, let me try it with my arc3 either

 

When I capture white noise through ARC3 it gives me this as captured curve result:


The ARC3 Curve looks however like this:




Bertom Curve analyzer shows this:

 

I just created an impulse response with studioone IR maker of my arc3, here the the result using khz convolved (or even 0 latency with open air):



The difference to am MatchEQ is that it is Stereo.

 

Last edited: Apr 26, 2022

Just like every EQ dating back to actual hardware. The only EQ that isn't also changing the phase is a linear phase one, and that is a digital-only thing that introduces latency and pre-ringing, instead of phase shift. There's no way to "change the gain of frequencies" without changing the phase, at least not if you want it done in real-time as opposed to processing an audio file offline.

 

just tried this with pink noise as suggested above and captured an approximation of my soundID headphone calibration curve. the two sound pretty close A/B'd, especially when putting pro-q into linear phase mode. not sure i'd mix thru this but it could be really useful for general referencing. thanks for posting!

 

thast why I wrote etc. I am not sure what else is changed.
__
What is the difference between a a learned EQ curve with a match EQ and a captured Impulse Response beside being Stereo ?

 

A IR can be both stereo and mono, and additionally there are true stereo IRs where you have the sample of a sound coming from the Left being reverberated in stereo, and another one of the sound from Right being reverberated in stereo. Not all convolution reverbs support this, but it is a thing. There are also surround impulse responses.

A match EQ curve cannot reproduce the changing frequency response over time, which an IR will capture. Though this is irrelevant for room correction purposes.

 

Hi, could you explain this in more detail, please? My experience in audio is not much.

I am not sure if I am doing it right, as the curve I am getting has nothing to do with the Sound ID's. My experience in audio is not much

On Ableton, I am running pink noise in one channel and sending the audio to another channel with Sound ID + Pro Q (I am not sure what the function of this eq is) and this audio is going to another channel with just the Pro Q as you mentioned above, the one I am using to capture the correction... with no success.

I do not understand why I need 2 Pro Q instances either, I think I am missing something.

And not sure how I would do that for each speaker, Do you mean switching L-R to L or R on Sound ID plugin?

I would really appreciate some help. Thanks

 

I can't remember exactly what I did - as it was so long ago - but you need two Pro-Q 3 (needs to be Pro-Q 3, not v1 or v2, for this to work) instances because in one of the instances you will be using the match EQ function, and the other one will be your reference. You are basically capturing the difference between these two Pro-Q instances.

No, you are not changing anything in SoundID, you will only be panning the track with the white noise generator, first to one side and then the other.
Since SoundID/Reference are doing individual corrections for your left and right speakers (or simply your left and right channels), you need to capture the left correction curve by itself and then the right correction curve. If you capture both channels at once, the resulting curve will only be useable in mono.


To do this the easiest way that works in every DAW, you need to create three tracks. The first track needs to make stereo sends to both the second and the third track (how you do this is different in every DAW, but I hope you know how to create a simple stereo send from one track to another). The FX chains on each track will look like this:

1. White Noise Generator
2. SoundID + Pro-Q
   
3. Pro-Q

To capture the left channel:

• Pan track 1 a 100% to left
• Open the EQ match feature in the Pro-Q instance on track 3
• In the window that pops up you should be able to select the other Pro-Q instance as your reference (will be called "track 2" if you haven't named your tracks)
• Click on the "EQ match" button
• Where it says "Reference", click on it and select "External: Track 2" from the drop-down menu
  
• Pro-Q now starts to match the other Pro-Q instance's curve
• After 5-10 seconds, click on the "Match" button
• Drag the slider (for the amount of bands) up to 24, and then click "Finish"
• Save the finished EQ curve as a preset in Pro-Q (e.g "SoundID Genelec 8030 Left")

To capture the right channel:

• Pan track 1 a 100% to right
• Do the same steps as above again
• Save the finished EQ curve as a preset in Pro-Q (e.g "SoundID Genelec 8030 Right")

After you are done - and you want to utilize Pro-Q as your room correction - you have to use two separate Pro-Q instances in your monitor chain (one processing only the left channel and the other one processing the right channel). How to do this is a bit different in every DAW (maybe not even possible in all DAWs now that I think of it), but in Reaper you can simply insert the two Pro-Q instances afterwards each other, use the Plugin Pin Connector to disable processing of the right channel in the first instance, and disable processing of the left channel in the second instance. It also requires you to have "Pass through unmapped channels" enabled in the plugin I/O routing for both instances.

Good luck!

 

Thank you very much for the reply! I have done this using sends and return tracks on ableton and the curve I got was pretty similar to the sound id one, slightly different but a very good approximation to be honest, much much better than nothing, now I need to figure out how to do the last step, using 2 instances of pro Q 3 one for each channel.

Many thanks mate, appreciate your time taken to reply in a detailed way