Two REW FR EQ'd sweeps - which would you choose

[muslhead]

Without adding enough to trigger reader bias, below are two REW plots of the same EQ'd speaker pair. The difference between the two I will reveal after comments.
Subjectively, listening to the same songs with either option was indistinguishable. There may have been a slight difference but due to the fact it took me about 15 seconds to switch between the two options could have been what my memory may have heard (or didnt hear)

What would you pick and why?

 

[sigbergaudio]

Somewhat difficult, but probably the top one. Less problematic through the midbass area (100-300hz). The lower one looks more tidy below 100hz, but the narrow dips in the top one isn't necessarily problematic.

 

[fpitas]

The top one.

 

[fpitas]

FWIW, in my experience the best results are obtained by doing sweeps at 3rd octave, 6th octave, 12th octave and 24th octave, while maintaining a reasonably flat looking response. Sometimes I even tweak at 50th octave. Nothing beats listening but that gets me close to what I like.

 

[muslhead]

Does the dropoff from 20-26hz bother anyone on the second plot. Not an issue because our minds will fill it in or just little music that will have content at those levels?

 

[fpitas]

Does the dropoff from 20-26hz bother anyone on the second plot. Not an issue because our minds will fill it in or just little music that will have content at those levels?

Very few tracks have anything that low.

 

[muslhead]

Where is @Purité Audio when you need him?

 

[sigbergaudio]

Does the dropoff from 20-26hz bother anyone on the second plot. Not an issue because our minds will fill it in or just little music that will have content at those levels?

It's an early roll-off, but that happens in some rooms, there's likely a dip there. May be fixed by placement. But very little music has content this low. But in the extent it worries you, that's yet another reason for going for the top one.

 

[muslhead]

It's an early roll-off, but that happens in some rooms, there's likely a dip there. May be fixed by placement. But very little music has content this low. But in the extent it worries you, that's yet another reason for going for the top one.

True but does it bother you? Your the speaker maker Would you keep working to refine one of your creations to not have that dip?

 

[sigbergaudio]

True but does it bother you? Your the speaker maker Would you keep working to refine one of your creations to not have that dip?

How something measures by design and anechoically is quite different from what happens in a random room, so that's not immediately comparable. Our subwoofers don't typically roll off that early, but depending on the room it may still happen.

Seeing it in an in-room measurement like that doesn't necessarily bother me. If it was a subwoofer/speaker that usually could go below 20hz (as is apparently the case here if it's true that both measurements are from the same speaker), I would speculate why it happened, and likely try alternative placements to see if it went away.

 

[ozzy9832001]

Without adding enough to trigger reader bias, below are two REW plots of the same EQ'd speaker pair. The difference between the two I will reveal after comments.
Subjectively, listening to the same songs with either option was indistinguishable. There may have been a slight difference but due to the fact it took me about 15 seconds to switch between the two options could have been what my memory may have heard (or didnt hear)


What would you pick and why?

I don't think either would sound all that different. The dips are high quality and I doubt our ears could really distinguish that at any one moment in time unless only that note was played.

The bass on the bottom looks better but the high bass/low mids looks worse. So, to me it would depend on what music you are listening to and how sensitive you are to those frequencies.

To my ears, the bottom one would probably sound better since I typically lower from 200-300hz anyway.

However, the overall smoothness is probably with the top.

 

[Ruhled]

Are you trying to correct the deeper dips found in the second one via the first one?

 

[muslhead]

Are you trying to correct the deeper dips found in the second one via the first one?

no, just trying different methods to EQ my speaker/room combo and get the best result. I did not post the pre EQ plot. if you had that you would have seen the messy room modes. Both of my attempts to EQ turned out pretty good, look fairly similar and sound close, very close

 

[boxerfan88]

Looking at the two graphs, looks like you're going through the same journey that I went through.

My initial EQ result was similar to the first graph, flattish, which I found sounded a bit thin and lacked bass oomph.

I'll choose the lower/second graph ... because that's where I landed with my setup ... and it sounded nicest to my ears for different types of genres.
(Quite interesting that despite all those dips in the lower mid-range it can sound nicer. Somehow our brain can hear through all those peaks & dips.)
[My equalization approach was to add a +2.5dB/octave rise from 200Hz to 50Hz in the REW target setting; before asking REW to calculate the PEQs.]

 

[sam_adams]

below are two REW plots of the same EQ'd speaker pair.

You should have applied variable smoothing to the plots as it has less of an effect in the lower octaves. Applying single octave smoothing-1/24, 1/12—obscures critical detail in that region. Looking past that error, measuring both speakers at the same time has resulted in comb filtering in the 100 Hz to 400 Hz region of the plot. Additionally, if you are using a subwoofer(s), you have a timing issue with the sub(s) and the mains that is very apparent in the second plot.

 

[muslhead]

You should have applied variable smoothing to the plots as it has less of an effect in the lower octaves. Applying single octave smoothing-1/24, 1/12—obscures critical detail in that region. Looking past that error, measuring both speakers at the same time has resulted in comb filtering in the 100 Hz to 400 Hz region of the plot. Additionally, if you are using a subwoofer(s), you have a timing issue with the sub(s) and the mains that is very apparent in the second plot.

RE the potential timing issue w/subs - what tells you this and how would one fix it (can you point me to some literature for me to learn if you dont want to take the time with what may be a lenghty response)?

 

[Sokel]

That's sub+ mains,right?

 

[Vladimir Filevski]

no, just trying different methods to EQ my speaker/room combo and get the best result. I did not post the pre EQ plot. if you had that you would have seen the messy room modes.

But we need to see exactly that pre EQ plot! How can we judge the quality of equalization without seing the pre EQ plot? What if you are trying to equalize the room depended dip (rhetorical question)?
Please show plots with 1/6 and 1/12 octave smoothing.

 

[muslhead]

That's sub+ mains,right?

Since the answer to your question is very relevant. To respond will require i uncover the difference between the two plots which i was going to wait on but since you asked ....

The speakers being used are the Dutch 8c's.
The bottom plot is the 8cs after EQ with REW.
The top plot is after EQ with REW of the 8cs that include a sub (using the sub out signal available on the back of the 8c's and no high or low pass filters being used within REW. All adjustments to EQ were done with the REW filters and tweaking the available sub external controls - phase, xover, and volume)

 

[muslhead]

But we need to see exactly that pre EQ plot! How can we judge the quality of equalization without seing the pre EQ plot? What if you are trying to equalize the room depended dip (rhetorical question)?
Please show plots with 1/6 and 1/12 octave smoothing.

i wasnt asking you to judge quality of EQ. I asked which would you choose. I dont want a critique of the EQ improvments (or not) but rather which you would choose based solely upon the plots. If you cant or dont want to do that, no problem. I am trying to learn what and why people would pick one plot over another. Just like i found @sam_adams response very interesting. He clearly sees something i dont and as such i want to find out what that is and why. I knew latency may be an issue but you have the ability to control some of that with the selections the 8c's provide you. I made those selections with the assumptions (since i did not know how to measure it) it would minimize any latency issues. But, based upon his response, I could be wrong.
Its ok to be wrong, staying wrong is not