Dirac Live standalone for PC and Mac

[kolestonin]

You can see my frequency response before and after Dirac in the screenshots. 1/6 smoothing was used in the graphs.

 

[PeterNL]

You can see my frequency response before and after Dirac in the screenshots. 1/6 smoothing was used in the graphs.

Looks good.
I would move last dot downward or move curtain left to avoid boosting highest frequency and let it roll off naturally.
How does it sound ? Not too bright ? You can try also steeper curve (+6dB to -3dB shelfs and convert to dots and adjust last dot or more to follow natural roll off).

 

[kolestonin]

No, it is not at all bright.
To be honest it sounds amazing.
It is too good to be true given this was my first attempt and the first filter I ever created. Today I started reading the manual and I realized I did many things, maybe not wrong but not as I should have done them. And this makes my surprise even bigger.
I now bought the license so I have all the time to further investigate web guides, videos etc. and I think the best is yet to come.

As per my Dirac curve, you can check the screenshot but I seriously doubt it's the one I use. Given the control points are completely different from the ones I had when I first created the curve as I took notes and I can remember them. You can see I created other curves as well, but I loaded my initial project and on the left panel I selected the initial curve. No matter what I chose I always get the curve I see on the screenshot. I really struggle with that part of the software and it seems I cannot understand what's going on. I only see one curve no matter what I choose, and no matter I can see and select all 4 filters in Dirac live processor.
I will of course read and make some measurements from scratch but feel free to advice what I could change on the curve you see if you think it can get any better.

 

[DWPress]

Try a filter where you draw the curtains closed to only correct up to 350-400Hz. If your midrange and tweeter are behaving as your graphs suggest it may sound (and measure) better without full 20-20kHz correction.

 

[levimax]

No, it is not at all bright.
To be honest it sounds amazing.
It is too good to be true given this was my first attempt and the first filter I ever created. Today I started reading the manual and I realized I did many things, maybe not wrong but not as I should have done them. And this makes my surprise even bigger.
I now bought the license so I have all the time to further investigate web guides, videos etc. and I think the best is yet to come.

As per my Dirac curve, you can check the screenshot but I seriously doubt it's the one I use. Given the control points are completely different from the ones I had when I first created the curve as I took notes and I can remember them. You can see I created other curves as well, but I loaded my initial project and on the left panel I selected the initial curve. No matter what I chose I always get the curve I see on the screenshot. I really struggle with that part of the software and it seems I cannot understand what's going on. I only see one curve no matter what I choose, and no matter I can see and select all 4 filters in Dirac live processor.
I will of course read and make some measurements from scratch but feel free to advice what I could change on the curve you see if you think it can get any better.

I know you bought the DIRAC software but since you know how to use REW you may want to let it take a shot at creating filters. I find REW's filter interface to be easier to understand and it can easily generate FIR convolution filters that can be used in many different player apps. For FR filters you can use a MMM measurements at the LP as the starting point. I always find it interesting to see how different programs work and some would claim FIR filters are better than the mixed filters DIRAC uses. If you did this it would be interesting to see the differences. If not I understand and good luck, you have made a lot of progress in a short time.

 

[kolestonin]

Try a filter where you draw the curtains closed to only correct up to 350-400Hz. If your midrange and tweeter are behaving as your graphs suggest it may sound (and measure) better without full 20-20kHz correction.

You mean I need to take new measurements and create a filter from scratch?
Or I just take my existing curve, I drag the left curtain to 350-400hz, I save target as '..', I export the filter and try it on Dirac live processor?

 

[PeterNL]

You mean I need to take new measurements and create a filter from scratch?
Or I just take my existing curve, I drag the left curtain to 350-400hz, I save target as '..', I export the filter and try it on Dirac live processor?

Existing curve. Drag RIGHT curtain to the left 350-400 and export filter.

Also look how Dirac managed to correct impulse response. Check Left and Right channel. Select spectrum corrected to see before and after correction. I would set curtain around 380-390Hz after the peak (peak still will be corrected).

Right mouse button click on Group 1 (yellow circle) You will be able to Load Target Curve, Restore default target etc. Two squares will switch between shelves or dotted target Curve.
Try different Low frequency target shelf like +6,5dB for low level listening, or lower it for loud levels. Adjust it to your needs but remember that we are trying to bring peaks down and avoid boosting too much. Export many filters and switch between them listening music.