Combining Virtual Bass Array and Alternative to VBA Pre-filters

[Ecwl]

So even though many people have had great results using the Alternative to VBA pre-filters alone, I have always found in my system, that the Virtual Bass Array pre-filters sound better. This is regardless of where I set the cut-off amplitude for the Alternative to VBA pre-filters.

I finally did some more extensive listening to test as to what the issues were and what I found was that with VBA, it really addresses my primary longitudinal resonance frequency really well. If I were to use Alternative to VBA pre-filters, what would happen is that I can set the cut-off amplitude very low (and 1-2dB above final target) in which case the primary longitudinal resonance frequency still resonates but the bass also sounds slightly too anemic. But if I set the cut-off amplitude a little higher above the final target (e.g. 10dB), I actually no longer get anemic bass and the non-primary longitudinal resonance bass peaks do sound better but the primary longitudinal resonance frequency peak would resonate even more. And since the dominant resonance in the room is the longitudinal frequency, the VBA pre-filter always sounds better.

After thinking about it for a while, I realized I can create an Alternative to VBA pre-filter after the VBA pre-filter so that the VBA pre-filter would address the primary longitudinal resonance frequency and and the additional Alternative to VBA pre-filter would address the other bass peaks but less aggressively.

The process is more tedious because first I have to convolute Pulse48L & Pulse48R with the VBA pre-filters first in order to create the Alternative to VBA pre-filter. Moreover, I ended up setting the cutoff amplitude for the Alternative to VBA pre-filter at about 5dB above the target amplitude as I did not want excessive bass peak removal from the pre-filter.

It's getting late in the evening so I probably won't be able to do further extensive testing tonight. But it seems the combination of properly constructed VBA prefilters and additional Alternative to VBA prefilters did achieve better bass response than either approaches alone.

I think this is perhaps pretty esoteric. Moreover, I'm not sure if I have optimally set the cutoff amplitude of the Alternative to VBA prefilters yet. But I thought I would mention this approach to people in case they are interested in trying it.

[Ecwl]

I just noticed that adding an Alternative to VBA broad bass peaks pre-filter actually increases the RT60. I was wondering if Uli can explain why that is.

Fortunately, the increase in RT60 didn't seem to affect the sound significantly enough in my system and the benefits of the Alternative to VBA filter seems to be worth the compromise.

In the end, after quite a bit of experimentation, I'm currently using a VBA pre-filter to reduce my primary longitudinal room resonance (that a simple Alternative to VBA broad bass peaks pre-filter alone could not adequately take care of) and then I create an Alternative to VBA pre-filter to take care of the remaining bass peaks <200Hz. By setting the Alternative to VBA pre-filter at about 2dB above the target frequency response, the combination of these two pre-filters seems to give my system the best sound I can achieve so far.

I'm really appreciating that all these filters and settings are about what compromises I want to live with. I think if people find that the Alternative to VBA pre-filter is not adequately addressing their primary longitudinal room mode resonance, they may want to try what I'm doing by combining a VBA pre-filter with a new Alternative to VBA pre-filter. On the other hand, I can see how in some systems, the Alternative to VBA pre-filter might increase the RT60 to an unacceptable level so they would have to rely on no pre-filter or just the VBA pre-filter.

[whitesilo]

I just noticed that adding an Alternative to VBA broad bass peaks pre-filter actually increases the RT60.

I have tried quite a few variations using the Alternative to VBA method and also found significant increases to RT60 (mostly on longitudinal node frequencies).

Following Uli's suggestion in the original thread of adding a FDW before step 7, I found these increased decay times were mostly tamed. The cost is a smoothing of the inverse resulting in slightly less flat bass FR.

In a listening comparison of a prefilter without an added 15/15 FDW versus one with, I found I preferred the one with the FDW but only slightly.

[Ecwl]

Following Uli's suggestion in the original thread of adding a FDW before step 7, I found these increased decay times were mostly tamed. The cost is a smoothing of the inverse resulting in slightly less flat bass FR.

Thanks for pointing this out to me. I’ve actually gone back to a VBA filter alone. But I’ll try using FDW on the pre filter to smooth it out and see if I can get a better sonic result from Alt to VBA.

[Ecwl]

I’m glad I re-tried Alternative to VBA filter with and without FDW. I also re-tried my VBA + Alternative to VBA filter. I have been experimenting with Macro 1’s FDW selection lately and I ended up building a spreadsheet to calculate for each LF/HF setting pair, what the FDW would be at 20, 100, 200, 500, 1000, 5000 and 10000Hz as I found that higher FDW in the bass provides in general better correction until it’s overdone and it sounds weird while lower correction in the trebles tend to give more air back into the music with more indirect sound (even if it worsens the IACC).
Anyway, I suddenly realized that if we are happy with a particular macro 1 FDW setting without any pre filter, say 20/10, by adding more pre-filters, particularly alternative to VBA, there could be too much low frequency correction as a result. So the macro 1 FDW settings may have to be adjusted.
I was able to create a new VBA + Alternative to VBA filter (without additional FDW) and then find a new optimal macro 1 FDW setting to improve my system’s sound.
Moreover since it sounds like I’m the only person who uses two pre filters, I suspect my room acoustics is unique.

[whitesilo]

That's really interesting. Thanks for sharing your analysis and findings. Its great to find someone else who is as perfectionist about this as I am!

However, I am wondering if there is no FDW on the prefilter, doesn't that mean all the bass in the impulse, right through to the diffuse tail, is being corrected by the prefilter, so perhaps too much correction? I guess it depends on how the prefilter is combined with the main filter.

If my thinking here is not totally haywire then perhaps another approach might be to use FDW values in macro1 which work best for the frequencies not targeted by the prefilter and then use FDW values in the prefilter that work best for the bass frequencies (based on the amount of room that offers the best compromise between bass flatness and increased decay at modal frequencies).

So, the prefilter FDW values would be focused on the bass and the macro1 FDW values on the mids/highs.

This might mean for example a macro1 FDW of 8/2 because 2 is best for the highs and you want to limit correction on the lows after the prefilter but need a low value high enough to get the right slope for the mids/highs. Then for the prefilter an FDW of 12/2 which provides the right window slope for the bass frequencies being corrected. Unfortunately this makes things quite complicated.

Perhaps my understanding of how the prefilter works in combination with the main filter is wrong here. Any education on this would be really helpful.

What do you think?

Its really good to be able to discuss this stuff with someone as the more I learn the more I find I don't know!

Regards, Tim

[Ecwl]

When I first got Acourate in September 2020, I was under the impression that there is an "optimal" filter for my system that I can mathematically figure out or I can just look at the measurements (RT60/IACC/peak value on Step response in Macro 5/Shape of step response on Macro 5). Partly, it's me wishing that I can overcome room acoustics using Acourate, partly it's because there are people online who make it a business to sell their remote DSP services who promote this idea.
But obviously, we all have different preferences for different frequency response targets. Moreover, we have different sensitivities to different DSP artifacts. For example, you can read online/in print that excess group delay below 100Hz in Macro 5 is not audible but that certainly has not been my experience. Moreover, I used to love EPW in Macro 4 to be higher for LF & lower for HF because of something I read online but over time, I realized it's just a difference in sound and I came around to Uli's recommendation of keeping HF/LF for EPW in Macro 4 the same.
Also, the measurements are helpful if it correlates to what I can hear, e.g. an odd frequency boom that correlates to the RT60 measurement. But if say the RT60 looks a little wonky and I can't hear it, or the IACC looks worse but the overall sound is better, I have slowly accepted that my ears just prefer the sound over the measurements. To prove to yourself this is true, just use no pre filters, set your FDW in Macro 1 to 15/15 first and then keep increasing it to 20/20 then 30/30, you'll see the IACC look better but I strongly suspect the sound will get worse at some point.
I also fully acknowledge that it seems from reading everybody else's posts, only I run into these issues but I clearly notice that other people's room setup and acoustics are far more regular and far superior to mine. So my experience may be in the extreme minority. And most people in my shoes would have just moved the system to a dedicated room with treatment.
This is why I've avoided posting specifics since I know what applies to my system would almost definitely not apply to others. But just for your reference, with no pre filter, the best sound I can get that I prefer is using Macro 1 FDW at 25/4. If I use a VBA pre-filter, the best sound I can get is Macro 1 FDW at 22/4. If I use a VBA and an Alt to VBA pre filters with no FDW, best sound for Macro 1 FDW is 10/5. And if I use a VBA and an Alt to VBA filter with FDW 15/15 applied, the best sound has a Macro 1 FDW setting of 12/5. And after further listening, and going back to verify on RT60, Alt to VBA without FDW is still causing resonances at certain frequencies that I didn't detect earlier but after I play a track like Peer Gynt Suite No. 1 4th movement (In the Hall of the Mountain King), the problem becomes more audible.
My personal priority right now, is to explore whether adding Alt to VBA pre filter with FDW is truly superior to just using VBA filter alone. And at a minimum, I need 10 different musical samples to hear whether one filter is truly superior to another. Sometimes, it's super obvious where a few musical samples are enough. But for things that are already reasonably optimized to my taste, I find that I simply need a lot more samples to hear and appreciate the difference. Because sometimes a single musical selection simply would not highlight the sonic abnormalities that might crop up in other samples.
With all that said, most of the time, even when my settings are not "optimal" to my taste, they still in general sound way better than without any correction or just with parametric EQ correction using REW. So I have always wondered if I should be less obsessive compulsive and just relax and simply enjoy the music more.

[whitesilo]

Thanks @Ecwl that is really helpful.

Yes, there does seem to be some conflicting advice online and the only way to know for sure seems to be to try it out. I've only been using Acourate for about 7 weeks but have clocked up 76 different sessions trying out different variations of various parameters. I have nailed down many of these now to values that are to my taste and have also gained an idea of how they effect the sound but trying out an Alt to VBA prefilter has opened up many new permutations.

Like you I sometimes wonder if I should just accept the "best" filter I have created so far and spend more time on the music but when I think I might still be able to get some further improvement with a new idea, in this case the Alt to VBA prefilter, I think it is worth spending the time on it. I am time-rich and money-poor so this is a better approach for me rather than spending money on the speakers/room.

I think I am going to try out a range of prefilter and macro1 FDW values like you have and see where this takes me.

Once again, I appreciate the information and advice you have offered.

Tim

[UliBru]

So I have always wondered if I should be less obsessive compulsive and just relax and simply enjoy the music more.

Each one of us has to find his own way.
I have followed the discussion but not stepped in. IMO you are on the right path so far.
But I just like to add some thoughts:

After a while of tweaking the system you may notice that the result is quite good. But you feel unsafe if it is good enough.
A good sign for the progress is the recognition that music tracks sound different. Some sound really good, by other tracks you think again about fine adjustment of e.g. the target curve etc. But there is no one correction for all the music. Simply because the tracks are different recordings with different musicians, different sound engineers, different recording setups incl. studio monitors and different mastering. So at the end we would need to apply a remastering with our correction, e.g. target curves. So when you reach this state you need to relax.

One approach to learn more is also to create slightly different targets and corrections. Simply change the target by lifting/lowering the usual slope by 0.1 dB steps. You will learn that you can recognize this by listening. And you may (hopefully) find a setting where a tracks sounds magic.

And then try to relax and to pay attention to the music itself. Here are two exceptional examples:
Listen to
- Armando's Rhumba - Chick Corea/Bobby McFerrin on Rendezvous in New York
- Musica Nuda - You're The One That I Want on Live à Fip
Try to grasp how the musicians have fun by themselves and how they interact to each other.
Try to understand WHY THEY MAKE MUSIC !
You will notice that suddenly it is no longer important if they play in your room or if you are there. Maybe you will wish you would have a chance to participate the original event. But now you can hear it everytime when you want. That's the good part of music replay.
And if your system sounds good enough you will no longer think about the target curve. You will enjoy the music.
And you can start to search for the music which touches you most. The only downside: there is so much bad music out there. It's like looking for the needle in the haystack. But It is worth to do it.

[whitesilo]

Uli, thanks for taking the time to write this. It is great advice to help recognise when you are reaching the final stages of refining filters. I think I am close now to this point as, for the many of the tweakable settings, I have found myself oscillating back and forth between a couple of values and having to choose one which is the best compromise for my setup and preferences. As you say, even then it is only "best" for some recordings and genres.

@Ecwl, I thought you might want to know the outcome of my testing of different FDW values. I have found the "best" for my setup and tastes are 18/18 for the Alt to VBA prefilter and 18/3 for macro1. Compared to default settings of 15/15 and 15/15, the decreased high value of 3 for macro 1 significantly opens out the high end providing greater instrument separation, clarity and purity - easily noticeable on all the music genres I listen to. The slightly increased lower value of 18 provides excellent bass quality while only slightly extending the RT60 at room mode frequencies. There is no further audible improvement above 25, just greater RT60 decay. I am very pleased with these two FDW settings and don't believe any further tweaking will yield significant improvements

To anyone else reading this and thinking about testing out FDW values, I think it should be a profitable exercise. If you are looking for a methodology, I started with the default low value of 15 and compared the results first with 8 and 4, neither of which I liked. I then tried 25, 35 and 45. 15 and 25 were the "best" and so I then compared 18, 21 and 23 and found 18 to give the best compromise of bass/low mids quality versus RT60 degradation. For the high value, I again started with the default of 15 and tried out 20 which resulting in a "breathing" sound in the mids so I then tried 12, 9, 6 and 3. 6 and 3 were close but 3 was marginally better.