Feature Requests

Bentxu

New Member
Thread Starter
Joined
Aug 2, 2023
Posts
8
@John Mulcahy a couple of feature Requests that I find useful (and have no idea how difficult it would be to implement).

Background to these is that I use REW for car audio, which may result in a bit odd need from a home audio perspective.

i) have an undo feature for EQ optimizations

Background: sometimes I am not quite happy with the results of the EQ optimizations after hitting "Match Response to target" in the EQ Section. I would then hit "Optimize gains, Qs and frequencies" in the hope that the prediction will match the target better. Often times it does, but sometimes it does not. Would be great to have an undo feature that if the optimization leads to - in my subjective and potentially not quite right view - worse results, I would love to be able to roll back to the previous values. I know one could do this manually by saving the filters after each step so one could reload them. Simply rolling them back would be great though.

ii) have certain frequency bands that are excluded from the optimizations

Background: in car audio we often times have to deal with install specific cancellations that cannot be fixed with EQ work. No matter how much one boosts, it will not change the frequency response in the slightest. For these situations it would be nice if one could specify 3 to 5 frequency bands that are excluded from the optimization. One would define a "Start" and "End" frequency for those bands and the optimization would ignore these bands. No idea how much complication this would bring in the optimization algorithms but if such a thing could work, it would be marvellous. At the moment I would try to take out frequency bands that have no real impact on the frequency response, however, this may mean that i) the overall optimization is off and ii) these bands could have been used elsewhere to get the response closer to the target. Admittedly one could leave the bands as calculated, however, I also do not want to put unnecessary strain onto the speaker if it does not really have an effect.

Hope the above explanations are understandable. I am not the best at explaining myself.

In case anything is unclear, please do not hesitate to shout.
 

John Mulcahy

REW Author
Joined
Apr 3, 2017
Posts
7,996
Notches in the response are excluded automatically, the target match range can be used to tailor EQ to a specific band.
 

Bentxu

New Member
Thread Starter
Joined
Aug 2, 2023
Posts
8
Notches in the response are excluded automatically, the target match range can be used to tailor EQ to a specific band.
Thanks for the reply John, highly appreciated.

Let's say I boost with 12 dB individual and max boost and the suck out is 8 dB, then it would still try to boost the suck out?

As for target match range: yes, this can obviously be done. But let's assume one does a wideband with 2 suckouts. I could then run 3 optimizations: before first, between and after second suckout. However, I would need to decide how many EQ bands to allocate to the respective area and run 3 optimizations. I assume that the machine could do it much more efficiently.

Understand it may be a very niche problem though that is not really worth the development time it would take vs there other, bigger fish to fry!
 

John Mulcahy

REW Author
Joined
Apr 3, 2017
Posts
7,996
EQ can fill dips, but the response may end up with a bump in a different measuring position. If you don't want any areas to be boosted set the overall max boost to zero.
 
Top Bottom