Bit Ten Install and Tuning

arun1963

Ok here are the current settings. Will come back and explain them tmrw.

pentoncm

I've been tweaking and its getting much better now. I've went through with pink noise and tried to get the Left and Right within 2-3 db of each other. I originally tried getting them exact, but I found this doesn't work right and I feel that this overcompensates. I've been cutting about an octave past crossover points and this is helping a lot especially with blending the mids with tweeters. Its sounding so much more natural now.

On a side note, I've really starting to like Apocalyptica. Well recorded and a good blend of cello, electric guitar, and powerful drums. Mix of vocalists so its good to tune to.

arun1963

Just looking at your settings.....

Are you running the tweets on ch 1&2? You have these channels cut by 9db. I ran my tweets at about -8db for a long time. While you can dial in good sound, the top end sparkle / edge / shimmer will be tough to dial in. You will also have to cut your mids a bit around the xover point to get the integration with the tweets (like you have). A while back I shared my settings with Mac and he pointed this out. My tweets are now cut by 3db. Maybe something for you to try.

Since you are eq'ing about an octave above and below the xover point, try this exercise. Cut your mids and only play the tweets. Now play the pink noise tracks from about 200hz on. At what frequency do you start hearing your tweets? Is your tweet supposed to play that low? 500hz from your tweet is going to have high levels of distortion, even if the tweet is 20db down, you are still adding that distortion to the sound. Eq an octave below the xover point and then cut everything below say 2khz to -12db L/R. You should hear cleaner sound. You can then go back to the mids and bring up the 500-2khz range a bit.

You've boosted 400hz and 500 is cut deep. I think you're trying to correct the null at 400hz that you mentioned at one point. You can cut a peak but you can't fill in a null. 315 and 400hz are mid bass eaters. You generally need to cut these frequencies a bit to make the mid bass more prominent. Cut 400 to about where you have set 315 and bring 500hz up.

1khz is great for vocal clarity. Bring this up to ~ -2/3db. 1.25khz though adds graininess to the vocals, you would typically set 1.25 a bit lower than 1khz. I'm going out a bit on a limb here as things can vary with the environment and the placements. If you have your mids in the door, the far mid will be much louder in the 2-4khz range due to beaming. While beaming is frequency specific (depending on the diameter of the driver), the on further but more on axis mid will louder a bit below and above the actual beaming frequency. A 6" cone would beam around 2.25khz. So normally 2-4khz would be louder from the far side, with door mounted mids.

Level matching the frequencies to get tonality is the toughest part of tuning, at least for me. If you have access to a 2 ch home setup, spend sometime listen to cd's that you like and are familiar with, then hear the same cd's in the car. This is the real fun part :biggrin:.

pentoncm

I have a couple of questions.

(1) I have 400Hz at 0db on both L and R because that appears where the null is in my car. Right now it sounds right set there and also it makes sense for it to be there with my meter readings. On the left side, with 315 cut by -6db I was getting a reading of 70db, 400 at 0db read at 69db, 500 at -7.5db read about 69db. Also, trying to bring 500 on the near side seems to add mud into the sound especially on guitar and cello. I do need to bring up 500 and 630 on the far side by .5db to 1db for each getting those around the -2 to -3db area.

(2) I agree with you that 1kHz should come up, also it seems like 800Hz needs to go down. There seemed to be still a hollow like sound on the near side, so dropping 800 and raising 1kHz sounded better. I brought 1kHz up to -5db and it is sounding better, I'll keep playing around with it though.

(3) The top end is a little lacking with my tweets set at -9db. Should I cut the mids or the tweeters or both around the crossover points when I raise the tweeter level?

(4) I'm playing around with the sub EQ, I'm trying to get the right combination of cut for 80Hz and 100Hz.

(5) Although 2kHz is cut by 24db with the filter, I still tried cutting everything under 2kHz by -12db and it does clear up the sound. I hadn't thought of this before but it makes sense especially if you are going to be playing the system loud it should definitely help to reduce distortion.

I've been using this frequency graph as a guide, http://www.independentrecording.net/irn/resources/freqchart/ear_sensitivity.htm

arun1963

pentoncm wrote: »

I have a couple of questions.

(1) I have 400Hz at 0db on both L and R because that appears where the null is in my car. Right now it sounds right set there and also it makes sense for it to be there with my meter readings. On the left side, with 315 cut by -6db I was getting a reading of 70db, 400 at 0db read at 69db, 500 at -7.5db read about 69db. Also, trying to bring 500 on the near side seems to add mud into the sound especially on guitar and cello. I do need to bring up 500 and 630 on the far side by .5db to 1db for each getting those around the -2 to -3db area.

There are two ways to treat a null in your frequency response. Identify and eliminate the cause of cancellation or bring everything around the null down via your eq, so that you reduce the perception of the FR dip. With the way the rest of the frequencies are set, the null at 400 should not be an issue. A 400hz wave has a wavelength of 33". Measure 400hz at speaker level L&R and then again at ear level. If ear level is down by say 8db then look at the source of cancellation about 30" from the mid. Chances are whatever it is will not be moveable. Like mentioned your eq settings around 400 are already cut so you should be ok.

315-400 is a range that you would look to cut in any case. Bring up 500hz while cutting 80-100hz. If raising 500 is making your MB muddy it's because 80-100 is too hot. Bring 250hz upo a bit too.

pentoncm wrote: »

(2) I agree with you that 1kHz should come up, also it seems like 800Hz needs to go down. There seemed to be still a hollow like sound on the near side, so dropping 800 and raising 1kHz sounded better. I brought 1kHz up to -5db and it is sounding better, I'll keep playing around with it though.

You have already cut 800 fairly deep, not sure if you need to cut more. Reset 400, 500, 80 and 100 and then bring up 1khz. If anything you may need to bring up 800 a bit. Hollowness could be because you have cut 500 and 1khz way more than you should. If you still have hollowness after making these changes try playing only the mid and tweet on each side to confirm which side. Let me know how it goes. If you still have hollowness we will look at some other causes.

pentoncm wrote: »

(3) The top end is a little lacking with my tweets set at -9db. Should I cut the mids or the tweeters or both around the crossover points when I raise the tweeter level?

Cut 5khz and up on your mids and cut 6.3khz and up on your tweets. You will have to find the right balance.

pentoncm wrote: »

(4) I'm playing around with the sub EQ, I'm trying to get the right combination of cut for 80Hz and 100Hz.

If you're running your sub to 60hz then set the eq flat till 60 and cut everything above 60 to -12db. That should do it for now. You can come back and tweak here later if you want.

pentoncm wrote: »

(5) Although 2kHz is cut by 24db with the filter, I still tried cutting everything under 2kHz by -12db and it does clear up the sound. I hadn't thought of this before but it makes sense especially if you are going to be playing the system loud it should definitely help to reduce distortion.

With a 24db slope or even with a passive, the tweet is not in physical danger while playing 500hz but you're just lower the distortion levels. In fact even after you cut everything below 2khz by 12 db, you will still hear the 600hz pink noise from the tweets but at a much lower level than that from a plain 24db slope. Less is more.

arun1963

One other thing about 1-1.25khz. Your ears are most sensitive here and how it sounds can be different from what it measures. The mic and your ears would perceive these frequencies and reflections thereof, differently.

pentoncm

I have attached my updated settings. It has been helpful to switch between my previous setting and the new one.

I dropped the 1.25 and 1.6kHz region on my near woofer. This has reduced the hollow like sound that I was hearing. I brought up 500 and 600 by a bit on both sides as well as 1kHz up to -4db. I cut 400 slightly on both sides, it just sounds better with a little bit of cutting. I dropped 100Hz and 125Hz on both sides a little more which has cleaned up things better and helped with sub integration.

Everything is blending much better now. These settings are way better than my previous posting.

arun1963

Welcome to the rabbit hole. All I can tell you is, it goes deep:eek:

arun1963

Been busy doing stupid things. Like emptying out my attachment folder here, so that I wouldn't be swamped everytime I wanted to attach something. Duh!! it's not a 'history' folder :redface:.

Good news is that the baseline is improving. Thank God for the second Saturday. A two day weekend is welcome. Will post up the current settings.

pentoncm

Do my cuts make more sense now? Does 500 cut by -5.5db sound reasonable? Do cuts of -6.5db for 1.25kHz, and -5.5db for 1.6kHz sound ok as well. These are all referring to the near side.

arun1963

When you set your eq, you are doing two things. First you are balancing for L/R response and next you are level matching across the 10 octaves for tonality. When you balance for L/R think only of that frequency. When level matching, think of 'how it sounds' and 'how you are setting that frequency relative to everything else'. Eg, while you've cut 400hz in absolute terms, in reality you are boosting a ton at this frequency.....relative to how you have set the other frequencies.

Will post a more detailed response later this evening.

arun1963

You asked if your settings at 500, 1 and 1.6khz looked right. You've cut those frequencies a bunch. 500hz doesn't look right. Let me walk you through my settings and explain why I have things set the way I do and see if that helps.

Starting with the sub, I'm leaving the 20-50 hz range flat on the eq. With the spl meter 31.5 and 40 hz are louder by about 5db but I'm letting it stay that way for now. Ideally I would cut everything above 50hz (xover point) by 12 db, but I haven't cut 63hz fully. That is because I'm crossing the SR mids below their Fs which is ~65hz. At this frequency the mid shows the highest impedence and hence the FR will be much lower, creating a dip in response. By not cutting the sub to -12 at 63, I trying to fill in some of the dip in the mid. I have to be carefull where I set 63hz on the sub, else I will run into localization issues. I can flatten out the response peaks ~31-40hz by changing the box size, but for now it's fine.

With the mids I've cut 20-31hz completely. I'm trying to minimize the frequencies that the mid would struggle with. I'm gradually raising things till 50hz. 50 and 60hz from the mids is about 8db lower than 80-100. This is when I play the mids alone. So the first instinct would be to boost these frequencies by 5db. But I don't do that. Two reasons. First, the null at 50-60 is due to the resistance on the speaker. There's nothing I can do about that, I can't fill that null via the eq. It will only mess up the sound. But when I play my sub and mids together the 50-60 and 80-100 are much better balanced. So I'm filling in the null of the mids with the sub. The good thing here is that our ears can locate source of sound on a lateral plane (front/back/left/right) from about 70hz up. That is also the reason that a lot of recordings will have frequencies below ~70hz in mono.

60hz is where you have the rumble in your bass, so I definately don't want to lose out here. Your mid bass frequencies 70-300hz are more sensitive to arrival times (phase). Hence the speaker nearer you will normally be louder in this range, with the exception of 125hz which seems to buck this trend. 80-100hz can add a boomy sound to your lower end. You want this range to be tight and snappy not boomy. If you look at your settings for 80&100, there's barely a 0.5db difference. I'm not sure that is correct. Try cutting both these frequencies more on your near side so that there is about a 1-1.5db difference between L/R then cut both by about 2db on bot sides.. I typically cut 100 a bit more than 80. You've got that bit covered.

125hz is the lower end of male vocals. It is also the low end for a bass guitar, piano etc. Boosting here (relative to your other mb frequencies) may give the low notes of the bass or piano more, but that would come at the cost of vocals. So keep that balance. Mac is right that 125hz will normally be hotter from the far side. 160hz will again be hotter from the near mid. 200hz is the low end for female vocals. If you use your spl meter you may find that the 250hz - 400hz range is louder than say the 100-200hz range. Keep that in mind when you're level matching. Try and keep your 80-400hz range flatish. That will give you a nice tight MB. I think you may want to bring 250-315 up a bit, while keeping the L/R balance. See how it sounds.

Which brings us to your million dollar question about 400/500hz. I have an issue with how you've set these. I know the placement and the interior of your car will have a major role to play in the FR but cutting 500hz by 5.5db can't sound right. I think if you cut 80-100 like I'm suggesting and then cut 400 while raising 500, you may be in for a surprise. Boosting 400 just messes up the sound for me. Try it and see what you come up with. 500hz is a critical frequency for me. I'm boosting (relative) here as it helps with clarity on vocals and increases the impact on the MB. Now if 80-100 is set too hot to begin with and then you raise 500, you are going to get the feeling of muddy mid bass.

The SR's mids have a peak at ~800hz so I cut the 630-800 range. 1-1.6khz is important for vocal clarity. Too little and your vocals are dull, too much and your vocals are tinny and you get that hollow telephone kind of sound. 1khz is good for clarity, 1.25khz adds texture/graininess to the sound, 1.6khz adds shrillness to the vocals. You need all three qualities, but in the right proportions. Try and play a bit over here. With my mids and tweets crossed at 3khz, I have to start thinking about two sets of drivers when I'm tweaking the the 1.6-5khz range.

2 and 2.5khz look ok but I would definately bring down 3-6khz on the near mid by atleast 2-3dbs. It will help in removing the shrillness and extra bite and you may notice a better lower end. 3 and 4khz is best used in moderation and 6&8 are just adding sibilance.

It's really late here and I'll come back tmrw night to deal with the tweeters settings. I think I mentioned somewhere that you should get some seat time with even a basic 2ch home setup with cd's you are familiar with this will give you a baseline in terms of the target sound. This tuning bit can get a bit overwhelming to start with. One neat trick I used when I started with the p-800 was, to focus only on one aspect of the sound i.e. vocals. Get the vocals to sound right and you may find that a lot of the instruments fall in line too. Along the way you will also learn what cutting and boosting each frequency does to the overall sound. Stick with it this can take a while but the end results are definately worth it.

pentoncm

I've went and readjusted a bunch of frequencies now. On the near side, 250Hz at -4.5db, 315Hz at -5db, 400Hz at -3db, 500Hz at -4db, 1kHz at -3.5kHz, 1.25kHz at -4.5db, 1.6kHz at -5.5db. I adjusted the far side accordingly to match the changes. This configuration sounds better when I go back and compare the settings from yesterday. This has a much fuller more realistic sound. The midbass has totally come alive now. Also the increase in the 1 and 1.25kHz has provided much more clarity. I brought down 3.15, 5, and 6kHz a bit but I am still tweaking those levels.

pentoncm

New settings, by far the best its ever sounded.

It definitely helped to drop 400Hz and raise up 500Hz. Way cleaner now and much more MB detail on bass lines. I thinkI still need to possibly tweak the 1.6 to 2.5kHz region. Brought 1 kHz way up to -3db and it sounds much better. I think before I just had too many areas messed up. Now it is sounding really good.

Arun thank you for the help. I'm not nearly done but I'm headed in the right direction.

arun1963

A baseline is a setting you come back to the next morning and not have an instant desire to tweak. It's one where you can go through a day listening to different cd's and actually enjoy some of the music. You check it against your last baseline and it's better.

Eventually you'll get down to tweaking that as well cause, 'what's wrong' will catch up with you. I thought the settings I posted last night could have been one, but it isn't. I think I'll give myself about four months, to get to a point where I can lay off the tuning and just enjoy the music. The tweaking just eats up too much of the minds bandwidth (more so when one is limited in the first place)

arun1963

pentoncm wrote: »

Arun thank you for the help. I'm not nearly done but I'm headed in the right direction.

You're sharing your journey just like I am, you don't need to thank me. If you look at my settings below, you'll notice I've set 80-100 closer to what you have. So I'm learning too.

I think I have a baseline. I've run it today back and forth from work and its much better than what I had earlier. I needed something good that would last me a bit so that I can step back from this hobby for a bit. I know I can go mch further, but nows just a bad time for the amount of bandwidth this hobby takes up. I think I'm going to ride with this for a bit. I played a selection of well recorded cd's Floyd, Diana Krall, Donald Fagan, Eagles and I enjoyed the music for a change. Male and Female vocals sound decent.

A seamless transition form your sub to your tweets is critical for good balance. Thats where you get your real impact from. I had been struggling for weeks where parts of the sound were nice but other parts just stood out like a sore thumb. Good vocal clarity, but a bit harsh and in your face. Decent lows, but somehow not enough. So yesterday this is what I did on an impulse.

I started with the sub and mids. I played the pink noise playing only the sub. The last faint starins were at 100hz. Then I played the mids to see where I could hear them faint trickle at 31 (could be my ears playing tricks) faint but definately there at 40hz. Then I played the 40-125hz range. I would play each frequency with sub only, then mids only and then with both. 40hz from the sub and from the mids are two different sounds. I wanted to balance the sub and mid ~60hz and there after have the mids play louder. The mids would focus up about halfway up the windshield. When I play the sub its ok if it focuses up slightly lower, but it should stay in front. If it doesn't you have phase issues you need to addres via TA. I also wanted to have the 60-125 range relatively at the same level. This is what prompted me to raise 80 and 100 from where I ad it earlier.

Next I went and did the same between my mids and tweets. Based on my xover points and slope this meant the 1.25khz to 6.3khz range. Again, at a given frequency play the mids only then the tweets only and then both. Again 3khz from the mids and tweets are two different sounds. While it is much easier to centre the frequencies from the mids, hearing the same frequency from the tweets can make you feel like its pulling towards the tweets. That needs correction. Try this at 1.25-3khz. If at these frequencies the tweets pull to the sides, even after you've cut/boosted either side, come back and play with the delay between your tweets. Typically reduce the delay between the tweets, to get them centred. Above 3-4khz your tweets are more sensitive to axis response so the 1.25-3khz range is best to sort out any phase issues. If you can get your tweets in phase for these frequencies, it will be much easier to dial in L/R balance for the top end.

I aimed to have my mids call the shots from ~1.25-3khz and then let the tweets take over. Oh and in both sub/mid and mid/tweet this transition should be smooth. So if 1.6-3khz from the tweets is at one level and say 4khz jumps in loudness you will have issues with balance. While integrating the mids and tweets you're looking at making the ascend and decay of each set of drivers as smooth as you can and you're trying to prevent the situation where 4khz from the mids centres up nice and high at your rear view mirror and the when you kick in your tweets it smears and pulls to the side.

I have no clue if any of this made sense, but it worked like a charm.

pentoncm

Just tweaked the settings again. I mainly changed the 5- 8kHz region on my mids.

Now the cymbals don't sound as thin as they did before and vocals sound much more natural. Everything sounds more cohesive and realistic.

I may have an easier time with TA and the crossover region because my tweets are right next to my mids.

pentoncm

I just spent about an hour fine tuning the TA for my tweeters and it has made a big difference in regards to stage height and centering. Cymbals have always pulled to the right and I wasn't happy with the width. I thought it was an EQ problem with the 6.3kHz to 10kHz being to hot on the far side and pulling. Well EQ never fixed the problem, as I cut the far tweeters in that region it just made the cymbals fainter to hear instead of moving their position.

My tweeters are mounted right next to my mids. For example on the left side, the tweeters are directly to the right of the woofer and slightly lower and pointed up and to far side of the opposing headrest. Vice versa on the right side.

When playing my left side only, originally it sounded as though the cymbals were coming from the tweeter location that is to the right of the woofer. I played around with adding delay to the tweeter until it sounded as though the cymbals were hitting in the same vertical plane as the woofer. Also, at this position the cymbals sound clearer.

I repeated this exercise with my right side. I had the same problem there with cymbals sounding like they were closer to me than the right woofer. I added delay and the same goes with getting both on the same vertical plane.

When I listened to both together cymbals were centered directly in front of me. In addition the left side and right side stage height had been improved dramatically. Blending between each tweeter and its corresponding woofer seems to be improved as well.

Does it make sense that the driver that has a more direct angle will need to have extra delay applied to it as compared to the predicted delay based on measured distance? It makes sense to me that tweeters need extra delay if they are aimed partially on/off axis compared to door speakers which are more off axis than on. The driver more on axis should have more direct sound and less reflected sound thereby requiring more delay in order to match up with a woofer that is more off axis and has a larger proportion of reflected sound.

I think this makes sense, any input? Bottom line, it sounds much better.

I have been listening to my 2 channel home system. I have a Sony system that goes back to the early 90s. Its not the best being that it has 8 inch woofers, cone midrange, and cone producing the highs but it helps. Vocals come out of the woofers and the high range cone doesn't sound as crisp as a tweeter. On a positive note I am getting the Grado SR60 headphones for x-mas which should be a good reference. Now I just need a bigger mp3 player cause mine doesn't play lossless so I am stuck with using big old wav files for reference.

arun1963

Will post a reply tomorrow. Hope you're enjoying the journey

arun1963

http://www.independentrecording.net/irn/resources/freqchart/main_display.htm

Not sure if you have this chart, but it gives the fundamental and harmonics range for vocals and some common instruments. So while the attack (fundamental) on a cymbal will be in the 200-800hz range i.e. 2 octaves, 4.5 octaves are devoted to harmonics and decay. If you want to highlight the shimmer (part of decay) you'd boost a bit in the the 6.5-10khz range. Now while this may make the cymbals sound more lifelike, 6.3-8 khz adds sibilance to vocals. With sound its always a trade off. Hence if you're boosting 6-8 you'll do it to the point where it doesn't add sibilance and you'll take the shimmer that this gives.

Same situation down at 125hz. Boosting 125hz can add to the attack on a low piano note or you bass guitar, but it will mess up the male vocals. Initially I had this issue where while fixing something I would pick a frequency higher than what I needed to. I.E If I'm trying to correct the cymbal, I wouldn't think of anything below like 5khz. Ok, back to your post.

With frequencies from 70~400hz our ears are more sensitive to phase (arrival time) differences than frequency response. Hence the closer speaker will be louder. From 400hz~2khz the ears are sensitive to both phase and FR (driver more on axis will sound louder). Above about 2khz its mostly about FR. Hence, frequencies ~3khz and up will be louder from the far side (mid and tweet) and will pull your image to the right. So chances are this is a L/R eq issue.

I would definitely look at getting the tweets higher than the mids. Pillars / on dash would be first option, sails would be next. Then angle them right. This will help with stage height and sorting out the axis response a bit. If you draw an imaginary line from pillar to pillar about 12" above your dash, that would define the base of your stage. If you feel something pulling towards the speakers its most likely a TA or L/R eq issue.

Your imaging cues are 2khz and up. That's why the cymbals image up at the tweet when you play left side only. Getting it to the mid level is right. Now add some delay to the mid and see if the mid rises, then delay the tweets to the mid level again. If your tweet was mounted higher than the mids, you'd use the mids to image up at the tweet level. Draw an imaginary line from your rear view to the top of your dash, this is your centre stage. When you play the right side drivers, ideally you should not feel the highs cross this line. If you do, your right tweet needs some cutting.

Does it make sense that the driver that has a more direct angle will need to have extra delay applied to it as compared to the predicted delay based on measured distance? It makes sense to me that tweeters need extra delay if they are aimed partially on/off axis compared to door speakers which are more off axis than on. The driver more on axis should have more direct sound and less reflected sound thereby requiring more delay in order to match up with a woofer that is more off axis and has a larger proportion of reflected sound.

No, delay is related to physical distance. The driver closest to you would have the highest delay. You go beyond the measured distances to partly raise stage height and also to set the relative delay between each set of drivers. Sub arrives a fraction before the two mids, followed by both tweets. Your brain then locks into the highs and pulls everything up. TA has little to do with axis. Axis issues are for your eq.

Arun

pentoncm

With the old settings, I had thought that I needed to increase the levels above 2-5kHz in order to raise the stage. I had played around with this a lot but it didn't sound right in regard to tonality. Instead it sounds better to raise the stage with adding delay to my left tweeter. Isn't it likely that when a mid and tweeter are aligned in time so that they sound as one on the same vertical plane that you reduce cancellation in the crossover region. With the mid and tweeter sounding before like they were in two different vertical planes, the 2-8kHz region sounded like it was coming from two different locations. I would think that it would be impossible in this case for them to cooperate with each other effectively. Adding delay to the left and right tweeters, made the high end much wider and matched where the mids are.

Basically before, the cymbals on the left side were hitting about 1-2 inches to the right of the A pillar. I added delay to match the width of the left mid, now both are coming from the left mirror. The right tweeter had cymbals hitting at the right A pillar while the right mid was at the right mirror. I added delay and this lined those up as well. Now the stage height is nice and even on both side at ear level approximately. Wouldn't it make sense that the image would smear to the right if the left side was blending less than the right in the xover region. With the right tweeter being closer to being aligned with the right woofer there should of been less cancellation than on the left side.

arun1963

Yes you can use TA to raise your stage height. But be careful that the drivers don't fall out of relative phase with each other. When you do it one side at a time you run the risk of your L/R mids and tweets falling out of phase. This will give you a split stage. Lot's of detail on the left and right edges of the stage.

It's simpler to start with all drivers in phase and then add equal delay to each mid and tweet. This will raise your stage height. Do this till the point you feel your mid and sub fall out of phase. Then add a touch of delay on the sub to bring everything back in phase.

pentoncm

I went back and readjusted the TA on the mids. I listened to tracks with a male vocalist speaking to do this. I felt that the settings were a little off. The vocalist wasn't directly in front of me, it was a little left of center in front of me. To fix this I decreased some delay on my far woofer. It went down about .13ms. I think the mids are centered now, the left and right definitely seem more focused if that makes any sense. I disabled the EQ on both when doing this. At the point where I thought it was right it sounded like the vocalist got clearer and jumped out at me a little bit more. I went back and checked the tweeters, basically I started at measured distance and added delay to raise height until the stage raised to the highest point. I feel that the height goes up and then starts to go down, so I set it at the highest point. Also at this point it sounds as though the detail coming through the tweeters is cleaner and clearer. Compared to measured distance, the left tweeter had .5ms extra delay, and the right tweeter had an extra .45ms.

arun1963

Most recordings will have the vocalist at centre stage. I try and place the vocals around the rear view mirror. That gives me about equal width for left and right stage. If you can place your vocals correctly and have them focused in a narrow area (think about the size of a 'mouth'), chances are your phase and L/R eq are in place. You may still need to level match a bit for tonality but as a first step try to get the vocals focused, so that they are not smeared across your stage.

MacLeod

The vocals wont always center right under the mirror because of the time alignment. When adjusting TA, listen carefully at the voice. Its best to use a male voice as itll have the most frequencies represented. Listen closely not just to the voice but the high and low parts of his voice, they wont always move together. You can get his voice to right under the mirror but the highs might be skewed off to the right. Adjust time alignment until the highs and lows of the voice are centered together.

In almost every single SQ car Ive listened to, the center image is always a little to the right of the mirror. I think the reason for this (and this is just me guessing out my **** here) is that when you put time alignment on your near side speakers, its not as if youre moving them away from you to the side but rather pushing them forward. So with the time alignment added, the stage is acting not like each speaker is 5' to either side of you but rather the driver side speaker is 5' in front of you and the passenger side is 5' to the side. Not sure if that makes any sense or not but it does in my little mind.

The key to concentrate on is getting the frequencies as equal on each side as possible. This is by far the most important. Having the center image a little to the right is not nearly as important as having all the frequencies where theyre supposed to be and all the images focused properly so you hear ALL of a saxophone in one spot instead of some notes over here and other notes over there.

arun1963

So with the time alignment added, the stage is acting not like each speaker is 5' to either side of you but rather the driver side speaker is 5' in front of you and the passenger side is 5' to the side. Not sure if that makes any sense or not but it does in my little mind.

When I add delay to the L and R side, it feels like I'm moving the stage on each side, a bit towards the centre. So yes, I can force the vocals to be centred at a point, but if over done, I'm actually losing a bit of width. It also seems to affect the tonality as you tend to lose that 'open' sound. In fact part of the reason that I sometimes wind up doing this, is because while 90% of the vocals will be focused, the higher end of the vocals tend to move around a bit. I'm obviously not doing the L/R balacing well enough in the 1.6-3khz range. I do the TA on the mids and tweets seperately.

So, if your vocals are placed just right of the rearview, with you sitting in the left seat, for me it would be just left of the rearview with me sitting on the right. Come to think of it, I've actually had the vocals placed there at some point and then moved them back towards the rearview, cause I thought that was supposedly more accurate :rolleyes:. Ok, I going to try it like you mention.

The key to concentrate on is getting the frequencies as equal on each side as possible. This is by far the most important. Having the center image a little to the right is not nearly as important as having all the frequencies where theyre supposed to be and all the images focused properly so you hear ALL of a saxophone in one spot instead of some notes over here and other notes over there.

You sometimes say really important stuff in a very understated way :razz:. You mentioned that you had a cd where the pink noise would first play for L then the R channel. It seemed like a novel idea and I think I asked a couple of dumb questions, but never really persued that angle. In the interim I got my spl meter and managed to dial in the mids well. But I was never satisfied with the highs. I recalled what you had mentioned and played the 3-16khz range with only the tweets where I would play a frequency and go back and forth between L/R to see which side was louder. It was a shock to discover by how much my near tweets were louder in the 3-8khz range.

Playing only the far tweet, these frequencies imaged up around the rearview but from the near tweet they were loder by 3-5db in some cases and imaged right at the where the tweet was loacted. It's much better now.

The settings now are quite different from what I posted earlier. Will post the revised settings tonite.

Thanks for joining the thread.

MacLeod

What I did was take the filtered pink noise track off the IASCA Setup Disc which is one long track with a narrator telling you each frequency before it plays. I chopped it up to where each frequency was its own track and made one left channel only and the next right channel only. A program like Audacity would be good for this, and really any filtered 1/3 octave pink noise tracks will work.

The best way to do it is to NOT use a SPL meter. Use your own ears. This is why its so important to have left and right sides on separate tracks so you can switch back and forth. If the right side's 4 KHz is bouncing around due to all kinds of reflections, its going to be much louder than the left side but it might not register as louder on the SPL meter.

Listen to each track with your eyes closed (no distractions) and click back and forth between L and R on each track.

Now Im not saying to not use an SPL meter at all because 85% of the time, itll be plenty accurate and itll show you exactly how many db to cut. So use the SPL meter but double check it with your own ears. If the SPL meter says both sides are equal at 800 Hz but the right side blows your ear drums out, then thatll be the correct answer. Your ears are much more sophisticated listening/measuring devices than that $50 Radio Shack SPL meter.

pentoncm

Two things.

(1) My HU had equalization on it this whole time that was set to Natural. Kenwood keeps the eq under the setting system Q. I hadn't checked this before because I didn't think it was the actual EQ and I never engaged it. Luckily it only affected 3 frequencies, 100Hz, 1kHz, and 12.5kHz. 100 and 1kHz were boosted and 12.5 was cut. I had to bring up 100Hz up on my fronts to about the -1.5db to -2.5 db area. If only I used the DeEq function it wouldn't matter.

(2) I've switched over to the Alpine MRX M50. Has anyone ever found that even with time alignment, it sounds better sometimes to invert the phase of the sub? Without the invert it sounds good in the upper sub bass range but the low end impact is not good. Inverting the phase makes the low end come up front and sound clearer.

MacLeod

What is happening is that your sub is out of phase with the mids. Its not that the upper bass sounds better, its that most of the sub is cancelled out in this range so you have a little extra energy coming from the sub but youre getting better detail from the mid's.

If this is the case, you need to play with delay some more because theyre not syncing up. Im a big believer in that everything should be in phase with each other. Its cool to use it if you dont have a lot of TA you can use but if youve got the processing power, time alignment works better than swapping the phase around.

I would suggest keeping the sub's delay at 0 then adding delay equally to the front speakers. 3.5 ms is usually what Ive found to be a good starting point. Thats what I had in my Accord and 4.0 is what Ive got in my Edge. The goal is to hear the sub FIRST, then the front stage. Your brain will cue in on the last thing it hears as the place its coming from. So if you have a kick drum kick in, youll hear the 40-80 Hz energy from the sub, then the punch and slap from the front stage a split second later and since thats the last thing you hear, your brain tells you that the sound is coming from in front of you.

The way youd set this up depends on how your system is set up. If you have mids in the doors and tweets in the pillars, then youd have the most delay on the tweeters so you hear them last and so thats where your brain will tell you most of the sound is coming from. This trick wont work on stuff that has no sound in the upper end. A kick drum has a slap in the upper midrange frequency so thatll help pull the image up. A cello in the lower octaves have nothing in the upper end so youre not going to be able to pull it up to the tweeters if your mids are in the doors.

If you have a 3 way with mids in the kicks, midbass in the doors and tweeters in the pillars, then youd want to hear the sub first, midbass second then youd have to play with the mids and tweets to get them to gel. You really want to hear them at the same time so you dont want to split them up too much. Generally youd want the driver thats the furthest up front to be last but if its in the kicks, you dont want to sacrifice height for depth. Youll have to play with the timing between the mids and tweets and find the best trade off of each.

arun1963

Happy New Year guys, hope you'll have a great year ahead.

The TA on the bit10 is a bit strange as you 'set distance' in inches / cm and then fine tune in ms. The beauty about phase is that its relative and not absolute. So exact measurements are not really important. You can either enter measured distance to the sub or even leave it at 0.

Play some bass heavy tracks and cut all drivers except the sub and the far mid. Start with the far mid at 0 and add delay to it till you hear and feel the sub from the far mid. You want it at a point where you hear the sub from the far mid without losing the clarity in the midbass/mid range. Our ears cannot locate (front/back or left/right) sound below about 70hz. By getting your sub and far mid in phase your brain and ears are cueing the 70hz and below at the same point as 70hz and over i.e. the far mid. Once you have this dialed in cut your sub and play only the mids and now add delay to the near mid till you get the stage focused upfront. Then play only the far mid and tweet and dial in the tweet and then the near mid and tweet. While doing the mids and tweets you're trying to lift the sound higher than the level of the midbass. Now play all drivers together and you will need to do some minor tweaking.

Mac is spot on, you will wind up with your near mid on a delay between 3.5-3.7ms relative to your sub. The delay between the two mids (assuming door mounted) will be ~ 1-1.2ms. If the tweets are mounted higher than the mids i.e. dash level or pillars, the delay between the tweets would be slightly more than that between the mids. Conversely, the tweets are lower than the mids the delay between the tweets would be slightly less than that between the mids. This is because the PLD between drivers reduces the lower they are mounted.

Frequencies that are phase sensitive, tend to be those played by the mids. Hence getting the mids and the sub/mids in phase is vital. However it is equally important to keep the mids and tweets in phase as well. Doing this is actually tougher than just dialing in the sub and mids.

There are two types of phase. Mechanical phase and acoustic phase. Mechanical phase is when the drivers are firing in an exact sequence based purely on distance of each driver from your ears. Mechanical phase is aimed at exact arrival time. Here measured distances are critical. If you were listening to music in a far field environment, free from early reflections with speakers at different distances, then yes you would want to get them in mechanical phase. But a car is a near field environment with tons of early and late reflections. These reflections are causing phase shifts along with summations and cancellations at different frequencies. Hence mechanical phase is of limited use here. The sequence that Mac is describing ie sub followed by mids together and then tweets together etc is acoustic phase.