Ganging/merging, yes or no?

[Maury Markowitz]

I live outside of Toronto, I have one set of antennas about 25 miles away at 245 degrees, and another 65 miles away at 175 degrees. In spite of using a cheap-o 4-bay and pointing it at about 200 degrees, I do get a reasonable selection of channels from both directions.

But its not a *great* selection. So I'm thinking of "merging" two antennas. Given the 70 degrees between the two sets, it seems like a perfect setup, right?

Reading HDPrimer's page on the topic:

http://www.hdtvprimer.com/ANTENNAS/merging.html

it seems that you cannot easily combine two antennas unless they are pointed in the same direction, otherwise you will suffer 1/2 signal loss.

But...

First off, what does 1/2 signal loss mean? If I have an antenna with 16 db of gain for a particular channel, does that mean the output is 8 db, or 16-3=13 db?

If it is the later, am I correct in thinking that as long as pointing the antenna at the channel gets me more than 3 db improvement, I should do that, right?

And finally, is any of this correct? Someone once told me the loss is only true if the same channel is on both antennas. In my case with 70 degrees, the reception should be close to zero in the "other" antenna for every channel.

But HDPrimer seems to be suggesting this effect is purely in the combiner, so even if you just put in the combiner and no second antenna it would still suffer the loss. This seems more logical to me, but antenna's and logic do not yet mix in my brain...

[Jake V]

I'd first recommend posting your TV Fool Report, the antennas you are using, their location (roof, attic, etc.), any splitters or amps, etc.

It sound like you are somewhere near Ajax. That's pretty flat so Toronto should be easy. Buffalo would likely require something bigger than a "cheap-o 4-bay".

As to combining antennas, sometimes it works and sometimes it does not (usually it does not).

[Maury Markowitz]

Originally Posted by Jake V

As to combining antennas, sometimes it works and sometimes it does not (usually it does not).

Is there a good, theoretically based, answer to the questions?

I want to understand this in general, not in the specific.

[rabbit73]

First off, what does 1/2 signal loss mean? If I have an antenna with 16 db of gain for a particular channel, does that mean the output is 8 db, or 16-3=13 db?

Half the signal loss is 3 dB. But that's not the main problem when using a splitter in reverse as a combiner.

If you combine two identical antennas aimed in the same direction, with identical length feed lines, the gain is 3 dB minus the internal splitter loss of 0.5 dB, for a max possible gain of 2.5 dB.

If you combine two identical antennas not aimed in the same direction, the same signals will interfere with each other when they reach the combining point if they are not in phase. The length of the feed lines is irrelevant in this case.

If you combine two antennas, one for VHF (real channels 2-13) and one for UHF (real channels 14-51), with a UVSJ diplexer, they will not interfere with each other because the UVSJ keeps the two bands separate.

If you use one of the new 8-bay UHF antennas with separate panels that can be aimed in different directions, sometimes it works and sometimes not because of the phase interference mentioned above. It's trial-and-error, with the two panels at 90 degrees apart for your best chances.

And finally, is any of this correct? Someone once told me the loss is only true if the same channel is on both antennas. In my case with 70 degrees, the reception should be close to zero in the "other" antenna for every channel.

Yes, it is, but it is a matter of the strength and phase of interference, not all-or-nothing. What usually happens when there is a problem, some of your channels will be missing after combining.

We have several members here that can do a good job with the antenna theory, but the guys in your area at DHC can do a better job with the specifics of your location. But we will give it a try anyway, if you so desire.

I do get a reasonable selection of channels from both directions.

What are you getting now by callsign and real channel number?

But its not a *great* selection.

What would you consider great, that you aren't getting now?

But first, as Jake V requested, please post a link to an exact address (which will not show) or coordinates (which will be shortened) tvfool report so that we can make an accurate analysis.

How close are my guesses to your exact address report?
http://www.tvfool.com/?option=com_wr...8e03088b382880

http://www.tvfool.com/?option=com_wr...8e03cd784464bd

[Maury Markowitz]

Originally Posted by rabbit73

If you combine two identical antennas not aimed in the same direction, the same signals will interfere with each other when they reach the combining point if they are not in phase.

This is what I thought, and it is not covered on HDTVPrimer - or if it is, it is hidden in the discussion of combiners.

In this case, two antennas in different directions, is it not the case that if one of them is 90 degrees off the other, the phasing issue will be eliminated because the output from the second will be zero?

But I still have the combiner loss too, right? And is that loss 3 db?

Originally Posted by rabbit73

What are you getting now by callsign and real channel number?
What would you consider great, that you aren't getting now?

http://www.tvfool.com/?option=com_wr...8e0347474ce022

I'm currently getting lots of channels, including many that the system says are 2edge. On the chart at the link above, I most everything up to WKBW, the exceptions being WUTV and CHEX analog (poor image, no sound, don't care. 9, in high VHF, is in and out.

And that's pretty good, but it will start fading when it gets cooler. By the end of October most of the Buffalo stations will be marginal or gone, and 9 will be barely usable.

So, I'd like to fix the fading, and ideally add ION to my list.

[rickbb]

Originally Posted by Maury Markowitz

In this case, two antennas in different directions, is it not the case that if one of them is 90 degrees off the other, the phasing issue will be eliminated because the output from the second will be zero?

But I still have the combiner loss too, right? And is that loss 3 db?

With the 2 at 90 degrees you "might" get good results by taking advantage of what's called the null direction of the antennas. This works best on directional antennas with narrow beam widths. One antenna is aimed at the others null signal zone, so to speak.

And yes, combiner loss will still be there, usually slightly more than 3db as no combiner is 100% efficient.

I've tried this with 2 DB4's at 90 degrees apart and did not get any better results than a single DB4. These were DYI antennas so I may have other issues in the mix, but I was not impressed enough to continue with the project at this time. I haven't given up on it, but it's seems less likely to be of any improvement to my reception.

Forum member holl_ands did a computer simulation of a vertical DB8 with 90 degree separation, (basically 2 DB4's vertically stacked), and the computer model does show it should be possible to achieve multi direction reception. I used his model in my tests.

[rabbit73]

In this case, two antennas in different directions, is it not the case that if one of them is 90 degrees off the other, the phasing issue will be eliminated because the output from the second will be zero?

No, because it is not likely that the output will be zero. It is true that the pattern null of one antenna at right angles to the other will attenuate the unwanted signals from the other, but it's not zero. Any interference from the other antenna will cause digital errors. The FEC (Forward Error Correction) has a limit to the number of errors that it can correct. Once that limit is exceeded, the signal reaches the "Digital Cliff" where you have pixilation, picture freeze, and finally dropout.

But I still have the combiner loss too, right? And is that loss 3 db?

HDTV primer has led you down the garden path to a false conclusion.

The inherent combiner loss is only about 0.5 dB.

The 3 dB loss that he mentions is the loss that you would have gained if the two antennas were aimed in the same direction.

Don't confuse the 3 dB loss of a splitter used as a splitter, with the 0.5 dB internal loss of a splitter used as a combiner.

Combining two UHF antennas aimed in different directions sometimes works, and sometimes not. There is no way that I can predict or guarantee the outcome. You must use the empirical approach, which is trial-and-error to find out. I certainly don't want to discourage antenna experimentation. I have always learned something from my experiments, especially the ones that didn't work, because I then had to find out why.

I don't think you will be happy until you try it.

[rabbit73]

Thanks for the new tvfool report. I see that you have moved it from Emperor & Pickering Beach to L1S2W9 O'Dell Ct and Turnbull Rd. It doesn't make much difference with LOS signals but it makes a lot of difference with your weak 1 and 2Edge signals.

I'm currently getting lots of channels, including many that the system says are 2edge. On the chart at the link above, I most everything up to WKBW, the exceptions being WUTV and CHEX analog (poor image, no sound, don't care. 9, in high VHF, is in and out.

That sounds pretty good to me.

WUTV is on real channel 14; most 4-bay antennas don't have as much gain at the low end of UHF. And they have even less gain for VHF channel 9.

So, I'd like to fix the fading, and ideally add ION to my list.

ION is pretty weak, has adjacent channel interference from CHEX analog on 22, and is at 153 degrees magnetic.

http://www.digitalhome.ca/forum/2495601-post2041.html

I'm in Ajax, just east of Toronto. My Toronto stations are at 245 magnetic, and the Buffalo ones at 165. If I point at Toronto I don't get anything from Buffalo, if I point at Buffalo I lose CFTO and the rest get spotty. So I pointed it around 185, splitting the difference.

http://www.digitalhome.ca/forum/2277401-post1100.html

I live in Ajax, just east of Toronto. Using the south-most edge of my roofline as "south" (I don't know what it really is), Buffalo is at about 185 degrees, and the CN Tower is around 255. I have a knock-off four-bay and found I can get a sort of compromise if I point it around 195 degrees. That gets me most of Toronto except 9, and Buffalo kinda works. If I point directly at Toronto or Buffalo I get everything from that location, but of course that's non-optimal.

http://www.digitalhome.ca/forum/2299185-post1974.html

Buffalo is about 160-175 degrees true and the CN Tower is 235. I currently have a four-bay pointed around 190, which sort-a works. I'm debating whether to add another antenna.

My concern is that I have little room to work with, and another 4-bay wouldn't fit. However, if I were to point the 4-bay at the CN Tower, there would be room behind it for something more linear, like a Yagi.

Whatever I do use, I want to be sure that I strongly reject the signals from the Tower. I always thought a Yagi was more directional and would be better at this, but looking over reception charts I can't really see any difference?

Splitting the difference doesn't always work. For best results the antenna should be aimed directly at the transmitter.

Before combining two antennas aimed in different directions, you must be certain that each antenna gets what you want from that direction. Otherwise, there is no point in combining; it's not going to get better.

If splitting the difference doesn't work, and combining doesn't work, you must switch between antennas with an A/B switch, which requires you to rescan after switching unless your TV can add a new channel after scan like a Sony.

An alternative would be to have your primary antenna connected directly to the TV antenna input, and your secondary antenna connected to a separate tuner with its output connected to an aux input of the TV.

You haven't told us much about your antennas, where they are located, and if you are using a preamp.

The way that I optimize aim for an antenna is first aim for max signal strength. Then I adjust aim for best signal quality as defined by SNR and errors. The two are not always at the same azimuth. Signal quality is at least as important as signal strength for digital TV signals.

I first learned about the importance of signal quality when using an Apex DT502 converter box with dual signal bars that measure signal quality and signal strength. I first adjusted aim for max signal strength. I then turned the antenna slightly to the right; the signal strength went down a little, but the signal quality went way up.

[Maury Markowitz]

The antenna is a no-name, and I mean that literally there was no name on the box, at least in English. It is mounted very close to the peak of the roof using a J-mast. The antenna has a clear view over other houses out to about 250 feet.

My current thinking is to use two entirely separate antennas, equip each one with a USB tuner right on the back of the antenna, run their USB's into a Raspberry Pi running MythTV backend, and then run that into the house using Ether (POE to power everything). That way I get live TV everywhere.

The question, then, becomes whether I should use a Yagi for the Buffalo direction. I've asked a couple of places, but no straight answers yet.

[Maury Markowitz]

Originally Posted by rickbb

I've tried this with 2 DB4's at 90 degrees apart and did not get any better results than a single DB4.

Very useful info Rick!

Originally Posted by rickbb

Forum member holl_ands did a computer simulation of a vertical DB8 with 90 degree separation, (basically 2 DB4's vertically stacked), and the computer model does show it should be possible to achieve multi direction reception. I used his model in my tests.

Do you have a pointer to this? Google turns up something unrelated.

[Jake V]

Originally Posted by Maury Markowitz

The antenna is a no-name, and I mean that literally there was no name on the box, at least in English. It is mounted very close to the peak of the roof using a J-mast. The antenna has a clear view over other houses out to about 250 feet.

Can you post a photo of the no-name antenna?

And maybe use a real compass to verify the direction it is pointed in?

[rabbit73]

My current thinking is to use two entirely separate antennas, equip each one with a USB tuner right on the back of the antenna, run their USB's into a Raspberry Pi running MythTV backend, and then run that into the house using Ether (POE to power everything). That way I get live TV everywhere.

So, I'd like to fix the fading, and ideally add ION to my list.

The question, then, becomes whether I should use a Yagi for the Buffalo direction. I've asked a couple of places, but no straight answers yet.

Thanks for telling us a little more about your goals; it helps us answer.

I like your idea of two antennas, but I don't think it's a good idea to put the tuners outside in the weather. Maybe this idea by GroundUrMast would work for you if you want to involve computers. I prefer to watch TV on a TV set myself; computers are too fickle and user-unfriendly for me:
An Alternative to Rotators and Antenna Combiners
http://forum.tvfool.com/showthread.php?t=820

ION isn't going to be easy. The curvature of the earth starts to interfere with TV signals at about 70 miles. If you look at the terrain profile by clicking on its callsign in your TVFOOL report, you will see a black arc at the bottom of the profile; that's the earth. And you will also see that only a small portion of the ERP is sent in your direction:



and if you look at the coverage map, you will see that the signal barely makes it to your location:



Antennas Direct makes a popular yagi, the 91XG. Yagis don't cover as wide a frequency range as collinear arrays, like the DB4E, DB8E, CM4221, and CM4228 which are more commonly called 4-bay and 8-bay bowtie antennas. The 91XG gain curve is low at low end of the UHF band, and peaks at the high end. Since you want gain for channel 23, you can see that it doesn't perform as well at that frequency: 524 to 530 MHz, as it does at the high end.
https://en.wikipedia.org/wiki/North_...on_frequencies
https://www.antennasdirect.com/cmss_...y/91XG-TDS.pdf

Collinear arrays have a flatter gain VS frequency curve. This is the tech sheet for the DB8E:
https://www.antennasdirect.com/cmss_files/attachmentlibrary/Technical%20Data%20PDF's/DB8E-TDS.pdf
and the DB4E:
https://www.antennasdirect.com/cmss_...y/DB4E-TDS.pdf

Notice also the tradeoff between gain and beamwidth. The higher gain antenna has a narrower beamwidth, which makes it more critical to aim.

[Maury Markowitz]

First, antennas:

I used to get ION when I first set up the antenna. To do so I had to point the antenna to about 160 degrees, but then I also got all the Buffalo stations just fine. But nothing from Toronto. So I rotated it to catch Toronto and then of course ION disappeared.

I should also note that I can get some of the other NY stations that are just as "deep" into the hinterland as ION right now. So I'm relatively confident I can get ION again, at least in the summer, with an antenna pointed in that direction. If I lose it in the winter, c'est la vie.

Following your logic I'll stick with the array. The only channel that really requires that is Fox, and that one is kinda marginal right now.

Tuners:

Nice link rabbit! That thread is doing just what I propose, so it's good to see I'm not the only one thinking this way! (BTW I did Google for people doing this, but this did not turn up)

I looked at the SiliconDust machine, but as that thread notes, they removed one of the antenna inputs circa 2010, which kind of ruins it for me. I understand why they did this, but my situation demands the former design and they disappeared from the market pronto.

I also looked at the Tablo, which has the same issue. Being from Canada I had a lengthy conversation with them about this issue, but they seem unconvinced. Which is odd, because I believe both of these boxes are nothing more than MythTV back-end running on a lightweight PC - precisely what I'm hoping to do, but replacing the PC with a Raspberry Pi. So what both of them could do is put a USB port on the back and periodically scan to see if someone plugged in a tuner.

The real problem, as you note, is watching it. As to that, I already watch perhaps 90 to 95% of my time using streamed content on my Apple TV. So what we really want is to have all of these products feed into that. Of course the Apple TV famously does NOT allow this...

Well that's about to change. A new one almost certainly comes out on 9 September, and even more likely is a programming kit. If this comes to pass I suspect MythTV front-end will be on it by the end of the month.

So at that point I'll have everything coming in through the Apple TV, including live and recorded TV. It already works for you Roku users, of course.

Roof:

The only reason with thinking about the tuners right on the roof is line losses. But I do intend to experiment with this - I'm going to find a good position for Buffalo and then test the signal strength at the top and bottom of the coax run I already have. Certainly leaving it in the utility room will offer many advantages once the pointing is complete.

Why bother? Because I noticed that the line loss from the antenna to the TV in the basement was *considerable*. I lost well over half the channels! I added a cheap-o RCA distribution amp in the utility room to address this. So the question remains how much I'm losing on the line that's left, maybe lots, maybe next to nothing.

[rabbit73]

Maury:

Thanks for your thoughtful post. I'm glad that I was able to provide a link for the path you have been following.

I don't know much about LANs, but would it be possible to have two HDHRs connected to the network to do what you want?
https://www.silicondust.com/

I understand your concern about signal loss in the coax, which is about 6 dB per 100 ft of RG6 for UHF. The traditional way to offset that loss is to mount a preamp near the antenna to amplify the signals before the coax attenuation. This only works if you don't have any very strong signals that would overload the preamp, which would create spurious signals in the preamp that would wipe out your weakest signals.

Looking at your TVFOOL report, it seems possible to add a preamp with moderate gain and resistance to overload near the antenna. In spite of the fact that the preamp will add its own noise to the signals, there is another benefit of the preamp which is to improve the total system Noise Figure. This is possible because a good preamp has a NF that is much less than the average ~7 dB NF of the tuner, and by the rules of cascaded NF math, the NF of the preamp primarily determines the total system NF because of its position at the beginning of the chain.

The result of the lower total system NF is greater sensitivity to weak signals. You can see examples of NF improvement here:
http://imageevent.com/holl_ands/files/ota
scroll down to and click on file 10 COMPARE System Noise Figures
With and Without Preamps
http://imageevent.com/holl_ands/file...=0&w=1&s=0&z=4
which was created by majortom of DHC. You can enter your own figures if you have Excel or OO Calc.

I also did an FM fool report based on your estimated location that shows some strong FM signals that could be attenuated by an FM filter to prevent interference with TV reception. See attachment. Roger1818 of DHC told me that the data for Canadian FM stations isn't up to date at FMFOOL.

[Maury Markowitz]

Originally Posted by rabbit73

I don't know much about LANs, but would it be possible to have two HDHRs connected to the network to do what you want?
https://www.silicondust.com/

Yes, but the various products out there aren't designed to work this way - with the exception of the older SD boxes anyway. MythTV, on the other hand, kind of assumes that's the situation - zero or more tuners get folded together into a single list. You can even say that it should pick a particular channel from one receiver vs another if the same signal comes in from more than one source. If I can get this working it might just be a $70 solution to a lot of multi-directional issues. And it's a DVR too.

Originally Posted by rabbit73

This only works if you don't have any very strong signals that would overload the preamp, which would create spurious signals in the preamp that would wipe out your weakest signals.

So I have thought about getting a preamp for the Buffalo station and leaving the Toronto one alone. Channel 19 would otherwise kill me I think.

BTW I've thought over a lot of possible topologies.

One is to put the USB sticks on the antennas and then run a USB cable down to the Raspberry Pi which could be anywhere. But over 5 meters you need a repeater (~$10).

The other is to put all of the gear on the roof, and then run a ether cable up and use POE to power it all. Now you get zero line losses, and the Pi and sticks are small enough to easily weatherproof.

Or if you want to use your existing RG6, you can get an ether adaptor for it and use that instead, but then you can't do POE so you need a power cable of some sort. But you would for a preamp too.

And finally, the easy way, run the RG6 to the box in the basement, accept the 25 feet of signal loss and just run with it.

I think each of these would have upsides and downsides for different people and situations. But I guess the last option is probably the easiest.

Originally Posted by rabbit73

I also did an FM fool report based on your estimated location that shows some strong FM signals that could be attenuated by an FM filter to prevent interference with TV reception.

That's interesting, I never considered that issue. I can get a "channel list" easily enough from the car. Any suggestions on a filter?

[rabbit73]

Any suggestions on a filter?

All FM filters are not equal. If I'm using a preamp with an FM filter, I have to accept its attenuation curve. If I pick a filter for a system without a preamp, or a system that needs even more attenuation than provided by the preamp FM filter, I select a filter that attenuates the frequencies of the strongest FM signals. A HLSJ makes an effective FM filter because it attenuates everything below channel 7 including the FM band. Some filter curves in the attachments.

Attachment 1: Antennas Direct FM trap vs Radio Shack 15-577. Notice that the AD does better at the low end of the FM band than at the high end. The RS attenuates better at the high end, but doesn't at the low 88 MHz end to allow reception of TV CH 6.
Attachment 2: The MCM FM trap does well across the whole FM band.
Attachment 3: A HLSJ used as an FM filter.
Attachment 4: I was concerned that a HLSJ might affect reception of UHF signals. ADTech did this test that shows the attenuation of the Holland HLSJ is of no concern for UHF signals.

If FMFOOL is correct, your two strongest signals are CJKX-FM on 95.9 MHz at -24.1 dBm and CKQT-FM on 94.9 MHz at -25.3 dBm.

Your strongest TV signal is CICA-DT at -39.6 dBm, which is 15.5 dB weaker than CJKX-FM.

FM can interfere with TV signals in two ways: fundamental overload, even affecting UHF if the FM signals are very much stronger (not likely in your case), or 2nd harmonic (2x its frequency) of FM signal falling on VHF-High channel frequency.
https://en.wikipedia.org/wiki/North_...on_frequencies

[rickbb]

Originally Posted by Maury Markowitz

Very useful info Rick!



Do you have a pointer to this? Google turns up something unrelated.

Here is the link to the "twisted DB8" model.

http://imageevent.com/holl_ands/omni...d8bayquasiomni

[Maury Markowitz]

Originally Posted by rabbit73

HDTV primer has led you down the garden path to a false conclusion.

The inherent combiner loss is only about 0.5 dB.

Ok, this is the question I was really trying to get answered.

The page in question is confusing (IMHO) and seems to mix a bunch of different issues together. I was trying to understand which bits of that page are inherent to the combiner and which were part of the particular setup he was describing.

[rabbit73]

The page in question is confusing (IMHO) and seems to mix a bunch of different issues together. I was trying to understand which bits of that page are inherent to the combiner and which were part of the particular setup he was describing.

I also find that page confusing, and I have been doing antenna experiments since I was 8 years old when I built my first crystal set radio. I'm now 82, still studying antennas, have been a radio amateur (ham) since the early 50s, having built many ham antennas, and have been making tests and measurements with antennas for digital TV since 2008.

My McAfee security software tells me to stay away from that page, so I look at the Google cache version instead. I previously used the Norton security software that let me look at hdtvprimer pages.

On that page Ken Nist KQ6QV is talking about "How to combine antennas that point in different directions" but starts his explanation with a splitter and then makes a conclusion about a splitter in reverse used as a combiner. He then states there is a loss of 3 dB.

I found it difficult to make that leap to the conclusion that he did because when the two antennas are aimed in different directions, the same signals from each antenna arrive at the combining point with results that vary because of differences in amplitude and phase.

He does go on to state

(It will be explained later that when the two antennas are pointed at the same TV station, the reflected currents subtract to zero, and as if by magic, the 3 dB combiner loss turns into a 3 dB gain.)

I have found, by actual measurements with a signal level meter, that when two identical antennas aimed in the same direction are combined with a splitter in reverse, the gain can be as much as 2.5 dB. When the two antennas are aimed in different directions, the results can not be predicted. Some channels survive the combining, and others are lost.

The discussion is still going on by people that are smarter than I am:

Thread
http://www.digitalhome.ca/forum/186-...rm-fields.html

relevant posts begin here at post #14 between holl_ands and K6STI and continue on page 2 of that thread
http://www.digitalhome.ca/forum/186-...ml#post2629098

Their discussion is about two antennas aimed in the same direction, but the two antennas are not in a uniform signal field because of trees in the signal path, which is a problem similar to combining two antennas aimed in different directions.

holl_ands is a communications engineer who is an expert in computer modeling and K6STI is a ham who designs computer modeling software for antennas. As Jeff Dunham's Peanut says sometimes the discussion goes "woosh" right over my head.

Rather than argue about it endlessly, I prefer to make some measurements to find out for myself, as the link in my signature says.

[Maury Markowitz]

Originally Posted by rabbit73

On that page Ken Nist KQ6QV is talking about "How to combine antennas that point in different directions" but starts his explanation with a splitter and then makes a conclusion about a splitter in reverse used as a combiner. He then states there is a loss of 3 dB.

I found it difficult to make that leap to the conclusion that he did because when the two antennas are aimed in different directions, the same signals from each antenna arrive at the combining point with results that vary because of differences in amplitude and phase.

Right, and I guess that's why there's the whole "antenna trick" page on his site.

I'm sort of getting some of this, but I'm still a little lost about most of it.

For instance, my dad's old VHF/UHF was clearly two different antennas sharing a common structure, one in front of the other. There was only one wire coming out of it, so they had to combine them somehow.

And right now I can look out the window at my neighbour's eight-bay and clear see that it is literally nothing more than two four-bays put side to side. I can even see the separate wires running into the "combiner" in the middle.

So clearly this is not only possible but common.

I do get that having them point in different directions is a whole other can of worms, but I'm not even sure I understand the most basic parts.