REW EQ Filter creation and export issue

[lexicon]

Hi
I have just started using the REW EQ function to create manual EQ filters to export into my Storm processor.
The problem I am encountering is that having set a slope down from say 300hz, the highest filter it creates is 6980k meaning that the final slope down to say -X db or more at 20k is not being replicated. As a result it sounds way too bright.

If I add a 13th filter manually to pin it down say 12db at 20k set at 2 octaves that creates the slope but surely it should be doing this automatically?]

What am I missing?

Thanks!

 

[John Mulcahy]

Filter assignment depends on the measurement, which isn't shown in your images and you haven't attached the measurement data, so it isn't really possible to comment.

 

[lexicon]

Thanks you for the response John - I have not used this function before so please bear with me while I try to understand the context of what the EQ function is doing.
The mesaurement I was using (I think!) is attached which was an average of L&R mains using trace arithmetic A+B/2.
I took full range measurements of all speakers before starting this exercise and from what I can see they are all exhibiting the same behavior in that they do not implement the slope down to 20k at the high end.
I note that on the LF settings you have a rise start and end value but not on the HF one - perhaps that accounts for this in some part?
I have tried adding a low pass filter at 1000 but the options are all fixed filter types and start at 6db BT per octave which may be a bit steep. Applying a bell at -12db set at 20khz 2 octave seems to be closer but I'm not sure that's advisable?
A couple more questions if I may also.
1. I see that it is applying a 15db increase and decrease to level the curve, presumably because the max boost is set at 18db. Is this a wise thing for me to be leaving at that level do you think or am I in danger of pushing it too hard? A large low frequency dip appears at about 80hz when I combine L&R in that way as you can see and your EQ has to apply big adjustments to correct this.
2. I note that only 12 filters have been created and that the Storm holds 20 slots. Is your system limited to 12 or does it do more for some configurations than others? I presume it's a case of the more it can do the better result it can achieve? If it could do more is there a possibility that the ability to specify the number required be included in a future update? I can see that some users may wish to leave a few slots free to manually apply their own on top, while the majority would probably just want to let it take full advantage of all 20 slots.
Thank you for your kind response - what a brilliant piece of software engineering - clearly a lifetime's labour of love!
Jon

 

[John Mulcahy]

I took full range measurements of all speakers before starting this exercise and from what I can see they are all exhibiting the same behavior in that they do not implement the slope down to 20k at the high end.

A downward slope is often a feature of measurements at the listening position, it depends on the distance from the speakers, the polar response of the speakers and the materials and surfaces in the room. It isn't typically something you should impose on a measurement, since that would also affect the direct signal from the speakers, which would ideally be flat. The room curve controls are there mainly to allow the target to be a better match to the measurement before any EQ is applied.

1. I see that it is applying a 15db increase and decrease to level the curve, presumably because the max boost is set at 18db. Is this a wise thing for me to be leaving at that level do you think or am I in danger of pushing it too hard? A large low frequency dip appears at about 80hz when I combine L&R in that way as you can see and your EQ has to apply big adjustments to correct this.

You have set very aggressive limits for individual and overall filter boost, resulting in a large boost. That is pretty much always a bad idea, a boost that fills a dip at one measurement position will typically result in a peak at another position, even a slight distance away.

2. I note that only 12 filters have been created and that the Storm holds 20 slots. Is your system limited to 12 or does it do more for some configurations than others? I presume it's a case of the more it can do the better result it can achieve? If it could do more is there a possibility that the ability to specify the number required be included in a future update? I can see that some users may wish to leave a few slots free to manually apply their own on top, while the majority would probably just want to let it take full advantage of all 20 slots.

8 filter slots are reserved for the crossover filters. EQ is best applied sparingly, and below a few hundred Hz.

 

[lexicon]

A downward slope is often a feature of measurements at the listening position, it depends on the distance from the speakers, the polar response of the speakers and the materials and surfaces in the room. It isn't typically something you should impose on a measurement, since that would also affect the direct signal from the speakers, which would ideally be flat. The room curve controls are there mainly to allow the target to be a better match to the measurement before any EQ is applied.

You have set very aggressive limits for individual and overall filter boost, resulting in a large boost. That is pretty much always a bad idea, a boost that fills a dip at one measurement position will typically result in a peak at another position, even a slight distance away.

8 filter slots are reserved for the crossover filters. EQ is best applied sparingly, and below a few hundred Hz.

Thanks John - it's all a big learning curve - I'll have another go. As regards the settings it's more a case of me leaving anything I don't understand what effect it will result in as default. Many thanks for the explanations - that will help me cut down the trial and error.
Jon