Bit Ten Install and Tuning

pentoncm

Yeah I need to check the frequencies around it, especially since EQ in the crossover region can cause phase changes.

pentoncm

When you guys level match, how close do you try to get the frequencies? Do you try to match up each frequency exactly or do you just get it within .5 to 1 db of each other?

MacLeod

2db is the most you want between the 2 frequencies. 3db is the point where its pretty audible so 2 and below is where you want to be. Also dont forget to use your ears as the final judge. Your SPL meter wont read reflections accurately so you can have 800 Hz blasting your ear drums from the right side but your SPL meter might read it the same level as the left.

pentoncm

On the input side on your Bit ten, do all of the channels clip at the same point on the sensitivity scale. I forgot to mention this before, on the Aux in on mine my left channel clips earlier than the right channel input. A while ago I thought that I should just find where the left clipped, back it off a tad, and then match up the right channel to the same marking. Nope this results in the right channel being -2db lower on all of the frequencies. The best way that I've come up with is to take it to where the left channel clips, take it back a tiny bit. Now measure the output with the SPL meter at a set distance like 1 inch away from the speaker grill. Next set the right channel so that the right speaker outputs at the same intensity.

I don't know if this is helpful but thought I would put it out there.

arun1963

Well, if the bit10 is clipping, you'll hear it in the sound. Typically in your mids and highs. Haven't really had an issue of one side being louder. What triggered your deep dive?

pentoncm

My guess is that the input channels on the Aux L+R aren't perfectly calibrated so 1 on the dial on the left does not equal the same intensity as 1 on the right. On the test disc my left channel clips just above 1, the right channel doesn't clip until about half way until 1 and 2. So if I match the right to the left it will not be as sensitive and result in a little loss in output.

MacLeod

Shoulda bought a P99. Wouldn't have those problems. Bazing! :cheesygrin: :razz:

pentoncm

Bit One does it automatically. That's one of the things they skimped on when making the bit Ten to make it cheaper I guess.

pentoncm

Here is a graph of my current Frequency Response and what I am going to try by modeling after FM curves.

arun1963

How does it look below 80 and above 5khz? Is that only for the MB or all drivers? How does it sound?

pentoncm

I didn't measure below 80 with the curve. That represents the mids and tweeters playing together.

I'm not as concerned with the >5kHz area, i will look at that later but not for now.

Most of the issues are in the 3 -5kHz region which were way too bright. Also 1 and 1.2kHz were too bright on right side causing some pulling. Will try out new settings in my car here soon.

arun1963

Try keeping the 20-60 range about 8-10db louder than the 80-125 range. Try and flatten out the 80-125 range. Leave 160 where you have it now. Cut 300-400 on the near side and boost a bit on the far side.A slight bump at 300khz will help with clarity on vocals. bring 200-250hz down a bit.

Agan a slight bump at 1khz is good for adding dynamics to fundamentals. 2.5-3 khz is where it should dip the most. Both 6.3khz and 10khz will also add dynamics to the harmonics. 12 khz adds air. But 12-13khz is also the beaming range for a 3/4-1" tweet. If you have them in the doors, the one on the far side should be a lot hotter than the near side. Check for that. The FM curve is a guide. You have to factor in your environment, where the speakers are mounted etc.

pentoncm

Sounds Awesome! That was definitely a good idea. Midbass is much fuller and realistic, much more neutral sounding now and the harshness is gone.

There are different frequency and equal loudness curves floating around out there. I know that the original FM curves differ than this curve here that I modeled my EQ after from the independent recording network.

Here the most sensitive part of the spectrum is 3-5kHz.

pentoncm

i still have to work on the midbass a little but it sounds AWESOME now! Stereo imaging is insane, and it is really starting to sound like my towers in my living room. The stage moved farther out too, before it would go to the middle of my side mirror, now its on the edge and maybe an inch or two beyond.

pentoncm

So you know how when i measured 400Hz it was 2db hotter on near side. Not the real case, I tried increasing the far side at 400Hz and it really stood out on that side. I mean increasing by .5db really made it stand out. Turns out that even though the far side reads -2db compared to the left, there must be a reflection the spl meter isn't reading because after i cut the far side another -.5 to -1db as compared to before it sounded much better. Because of this i tried increasing 800Hz by .5db and it sounds much more natural on the far side similar to the near side now.

arun1963

Most cars with door mounted mids, will have a suck out in the 300-500hz range. The dip will also be higher from the far side. Not sure if its purely on account of reflections though. If your spl meter tells you one thing and your ears tell you otherwise, go with the ears.

MacLeod

That's why I strongly recommend level matching by ear and using the SPL meter just for a 2md opinion but always let your ears be the final judge. Your ears are a much better mic than a $40 SPL meter from Radio Shack and its not going to be able to measure reflections as accurately as your ear.

pentoncm

I have now been tweaking my 400 to 1.6kHz region to get it to sound better. It had sounded a little thin and so I've been tuning based on fundamentals and harmonics. So whenever I try to level match I check that an octave below and an octave above sound ok. Example is 800Hz standing out on the right side because it is really too hot or is it standing out because 400Hz is too low ? Also I check the harmonic up at 1.6kHz to see if that is the issue. This method has helped me a lot. I tried raising my bands 500Hz through 800Hz by .5db and lowered the 1kHz to 1.6kHZ region by .5db. This has made a big improvement in making the sound more full and realistic.

I noticed that it sounded really thin during Judith by A Perfect Circle. It sounded fine but lacked impact. Making the switch as i mentioned above got me much closer to what its supposed to sound like. Also, I have now turned my 80Hz band on my sub all the way down to -12db, this really does sound better. The kick drum now has that nice impact similar to a rock concert.

MacLeod

800 Hz was always a problem area for me in my Accord. It was always resonating big time from the right side so voices would pull to that side all the time. I think thats just a problem with the speaker locations and relative to your seating position. The only way to really get rid of it is to move the speaker itself. One thing Im really learning these last 2 years moving out of Street class that uses all stock speaker locations, is how moving speakers around REALLY makes a huge difference. I guess you can have all the processing power in the world but no EQ can help a frequency once its left the speaker. Speaker placement is everything.

pentoncm

I'm am really starting to get things balanced now and tuning by ear. Things would pull to the right sometimes and i couldn't figure out how to fix it. I now figured it out that i had to bring down 1.6kHz while raising 400Hz and 800Hz on the right side. If I only messed with 800 by itself it never worked right. Also on the left side I had to drop 400Hz by a bit. It just makes everything so much more complicated because if a note has a fundamental at 400 it will have a 1st harmonic in the 800 band and a 2nd harmonic at 1.6. The extra 400 on my near side hid the harmonics above it, while the reduced 400 on the far side made 800 and 1.6 seem brighter because there was less fundamental.

arun1963

Not sure how you're measuring impact of harmonics, unless you're using an RTA. Get L/R balanced as good as you can and then level match for accurate tonality using your ears. Don't get caught up by the fundamental/harmonics bit. Instead focus on how ech instrument / vocals sound. Does the piano sound like a piano at all notes? Does the shimmer on the high hats sound accurate or is it diffused. By doing this you're automatically correcting for fundamentals and harmonics. It's very easy to get twisted out of shape in this hobby.

Typically some of the frequencies I have problems with in terms of affecting the overall balance and accuracy are:

1. 160hz - Damn thing is much hotter on the right and sometimes there are resonance issues here. If I cut this too much it really hurts the MB. If it's too hot it pulls to the right a bit.

2. 400hz - Too much here and the vocals are bloated. Too little and they are thin and you lose a chunk of MB presence.

3. 630-800hz - The SR 6.5 mid has a nasty peak in this region. So This range needs to correct for that while maintaining balance with whats lower and higher.

4. 1-1.25khz - Which side is hotter? The meter tells me one thing and my ears tell me something else. Effect of reflections starting to kick in. I'm about 90% sure here but still.......1khz is great for making the fundamentals more dynamic. 1.25khz adds texture to vocals but can make them too grainy.

5. 1.6-3khz - If the sound is thin this is the range you should check first. Too much in this range and it will kill the presence of your lower end. If it starts sounding thin, this is what I check first. Plus this is the range where the tweets are starting to kick in as well. My near tweet mounted on the dash is much hotter in the 2-4khz range and I need to cut it by -8 to -10db in this range. The 6.5" mid would be beaming at 2.1khz so that needs to be factored in as well. Tons of reflections. I hate this range :evil:

pentoncm

The fundamental - harmonic method REALLY works for tuning by ear. So i was having the problem of having midbass that was too boomy I thought that 315Hz needed to be cut extra. It helped with the boominess, not perfect but it helped. Well the downfall of this was that it made vocals seem thin. So I thought what if I wasn't on the fundamental note, maybe 315 was the harmonic for the bass guitar notes that i was hearing. So i tried cutting 160Hz while raising 315Hz. The result = exactly what I wanted i reduced the boomy bass notes while making vocals thicker and more realistic. It sounds really good right now.

So you were right, the thinness was not associated with a hot 800 to 1.2kHz region. It was the 1.6kHz to 6 kHZ region. I cut 1.6 kHz on both sides by .5db. 2kHz and 2.5kHz needed additional cutting on my near side tweeter. I then brought down 3.15, 5, and 6.3 kHz by just a teeny bit. After this I even increased the 800 to 1.2kHz region and it really helped. After all of this tuning I am really beginning to believe the equal loudness curves.

arun1963

pentoncm wrote: »

After this I even increased the 800 to 1.2kHz region and it really helped. After all of this tuning I am really beginning to believe the equal loudness curves.

Just don't say this over at diyma :cheesygrin:

pentoncm

arun1963 wrote: »

Just don't say this over at diyma :cheesygrin:

Yeah I know man, I would be eaten alive. Cause you know all of those equal loudness curves are just wrong and we should trust car audio junkies over scientific experiments. I wonder if people who really believe flat is better just get used to their crappy sound and convinced themselves that is the way its supposed to sound?

Yeah the 800 to 1.2kHZ adds some really nice harmonics to the fundamentals underneath. I feel they are necessary to give weight to the 1.6 to 3kHZ section above.

MacLeod

Do the guys at DIYMA think totally flat is best or are they just meaning that there are no big dips or peaks and all frequencies are within a couple db of each other?

pentoncm

Arun has spent way more time on those forums, but my interpretation is that bunches of them want a true flat response. I saw the forum post that Arun posted on (I'm assuming he is SQNUT) and they jumped all over him.

there are problems with many of the forums, I've had issues over at the soundman forum. I tried opening a tuning thread just for people to discuss settings and things. Instead nobody was interested and instead they just keep listening to the dealers that are telling them all they need to do is put in Hertz Mille three ways in and a bit one and just let the installer set it up. The installers keep pushing ported boxes and crossing at 80 or 100Hz for the sub, and crossing tweeters as low as possible. One dealer was crossing at 1kHz:eek:. The soundman shop tunes for a flat response, I've seen it on their web episodes using an RTA with a flat line.

MacLeod

That's kinda why I've almost stopped posting on audio forums. I post on DIYMA but almost all are in the 12V Event forum. All of this is subjective for the most part and it just gets old with "this sounds better" and "no that sounds better" and on and on.

This one though, I just can't see how a ruler flat FR can sound good. I might buy it from midrange up but definitely not from midbass down. I guess I could agree that that way would be the most "neutral" but it could never be the best sounding.

pentoncm

MacLeod wrote: »

That's kinda why I've almost stopped posting on audio forums. I post on DIYMA but almost all are in the 12V Event forum. All of this is subjective for the most part and it just gets old with "this sounds better" and "no that sounds better" and on and on.

This one though, I just can't see how a ruler flat FR can sound good. I might buy it from midrange up but definitely not from midbass down. I guess I could agree that that way would be the most "neutral" but it could never be the best sounding.

One of the reasons why flat is bad is that the ear canal naturally amplifies the 3 to 5kHz region. This is why the equal loudness curve has this scooped out a bit. In my experience if this region is as hot as the 400 to 800 region, the sound is way too thin and bright.

pentoncm

This is a nice review of the literature detailing the amplification of the ear canal.
http://www.freetopessays.com/content/integrative-function-auditory-pathway-perception-4-khz-10-db-sound

arun1963

MacLeod wrote: »

All of this is subjective for the most part and it just gets old with "this sounds better" and "no that sounds better" and on and on.

++ That is the main issue.

Sound is subjective and folks have problems with that. You can't quantify subjective, it's not directly measurable. You can't measure the taste of dark Chuao chocolate, just like you can't measure good sound. You have to experience both. Not just once, but repeatedly, to start catching subtle layers, to start understanding. It takes a while to understand the true extent of difference between the Chuao and a Mars bar, just like it takes a while to understand that in real terms, if your sound is off by 10% its going to sound crappy. The other pitfall with the subjective bit is that everyone would claim to have great SQ sound. "Sounds awesome to my ears", kinda thing.

80% of folks are left brain dominant. This is the analytic side of the brain, the side that is better at processing information. Its the side that needs measurable proof to believe. This half of the brain tries to build the whole from bit up.

The right side of the brain is more intuitive. It works by deconstructing from the whole down to the bit. Each side of the brain controls the opposite side of the body. Experiments have proved that instructions are better received via the right ear so that the left brain can process it, but music appreciation is better via the left ear. A big + for those of us who sit on the right side of the car :cheesygrin:

You're not going to get good sound without some amt of measurement and calibration. But in the end if it measures right but sounds wrong it's wrong and vice versa.