AVS Forum banner

21 - 40 of 6001 Posts

·
Registered
Joined
·
2,538 Posts
True about the VibSensor(s) and I’m sure d00d’s is great!! I’ve been saying this for a while now, I need to build one!! But even if the phone VS may not be the most accurate, and not good below 10hz or even much above 30-50hz (and the lack of multiple locations (yes multiple locations would be awesome!!), it’s still (IMHO) a fantastic tool to have in your tool bag to work with in that very important 10-30hz area. I think its accurate enough in the 10-30 hz area for what most of us need and is easily repeatable, making it easy to track changes in your systems TR with different components that may be added and combined in ones system, showing timing changes etc pretty darn well and most definitely helps to tune with, in conjunction with the ole butt-o-meter. Truly accurate and the be all end all, of course not, but close enough for what most of us need and doesn’t cost anything. Not that I wouldn’t love to mess around with a multipoint ultra-high end VibSensor set up :)

Just my .2 cents
Good point Shelby.....yeah, I think you and Nalleh would benefit a lot from dood's system and being able to measure the TR of each of your devices to see how they're combining. It was certainly an eye opener for me when I started measuring independent TR devices all time aligned on one plot....that was how I discovered platform bending was the biggest enemy of natural feeling single digit TR because the drivers started working against each other. There's no way I could have made that discovery with VS especially with the single digit responses that needed to be measured and the limitation of only measuring one driver at a time.

I think we've all found using REW for optimizing room response has netted additional SPL and highlighted holes in our system response that we didn't know about until we started using REW (and MSO). Same with TR, but even more so from a timing error perspective since TR is much more sensitive to timing.

We really don't know how much TR is being left on the table until we measure each TR component and put them all on one chart similar to how REW does it. That's what dood's system can do.

Not knocking VS, just calling attention to it's limitations as a vibration measuring device in our hobby. It's hard to describe what I'm talking about but once you see vibration of all devices on one plot similar to how REW does it with SPL, then you start to realize how powerful it can be for optimizing all TR devices so they're working together....within milliseconds....even better than the ole buttometer :)

Said another way, I thought my system sounded pretty good years ago. But, then I started using REW and realized I wasn't even close to reaching the full potential of my investment and my room.

Same with multiple TR devices all on one platform....it's hard to know what's being left on the table until each device can be measured and displayed time-aligned. Even the 10-30Hz range has significant cancellations happening that VS won't be able to detect on it's own.

With all the investment in multiple TR devices, another couple hundred dollars for dood's system would be money well spent for anyone wanting to combine those TR devices and ensure they're getting the most from their investment.....just like REW does for us and our room responses today.
 

·
Registered
Joined
·
4,544 Posts
completely agree with @trhought re the need for higher resolution measurements, it's why I wrote that app in the first place :) I remember drawing the analogy with RS SPL meters, back in the day these were the only way to go but then at some point, accurate mics became commonplace and everyone could enjoy better bass! VS is a bit like that imv.

another option which is not that expensive but is more involved is based on https://www.audioxpress.com/assets/upload/files/George NtanavarasAccelerometer.pdf I have one of these also & verified that my rpi device measured accurately by reference to this device (using an ach-01 accelerometer).

if one were to build multiple such pre amps & buy multiple accelerometers (these are not expensive) & an audio device that can accept n input channels then it would be possible to build an extremely high resolution measurement device.

The nice thing about this approach is that it presents as an audio signal so it makes it possible to feed the input into any audio signal analysis software and hence a whole load of options open up without having to roll your own app. The only slightly tricky bit is actually building the preamp, this will be straightforward for some people but trivial for others.

on the rpi front, the thing that I never worked out is how to build an i2c circuit that can handle many devices and longer cables. This is useful because it lets you move the hardware away from the system under test and also makes it easier to support multiple sensors. I know this is possible (and not expensive IIRC) but never got round to working out exactly how to do it (requires an smbus repeater I think).

If anyone reading has the knowledge of how to build such a circuit then that would be a really useful thing to know as then you could really kit out your seating with lots of sensors :) I imagine the rpi4 would help with this also (more data to process in realtime so faster cpu helps).
 

·
Registered
Joined
·
3,582 Posts
Discussion Starter #23
^^^ Great stuff you guys!!! :) :)
 

·
Registered
Joined
·
2,538 Posts
Funny you mentioned the RS SPL meter @3ll3d00d....I was also going to use that as an analogy. Something like using VS for TR measurements is like using REW with an uncorrected RS SPL meter and only measuring 1 speaker in the room.

I'm hopeful there will be some breakthroughs in vibration measurement for our hobby that will allow us to optimize TR in our setups just like REW has done with SPL in our rooms. There has to be an affordable multi-channel solution out there short of borrowing a pro BK rig from work that costs 6 figures.

Unfortunately, I won't be much of a help when it comes to the circuit needs for multiple accelerometers. I'm an ME with limited circuits experience.....once built a cable TV signal descrambler DIY kit back in the 80's which was pretty neat. The only problem was it started to drift after about an hour of use, but was pretty cool for that first hour of watching MTV unscrambled or the latest boxing PPV match unscrambled :)

Hopefully, there can be some collaboration happening with TR measurement similar to the awesomeness you have unlocked for all of us using your BEQ app. Members with multiple TR devices really don't know what they're missing until their TR is optimized the correct way.....we're certainly caught in the "RS SPL meter" days right now with TR measurement in our hobby.

Fingers crossed!
 

·
Registered
Joined
·
2,538 Posts
completely agree with @trhought
another option which is not that expensive but is more involved is based on https://www.audioxpress.com/assets/upload/files/George NtanavarasAccelerometer.pdf I have one of these also & verified that my rpi device measured accurately by reference to this device (using an ach-01 accelerometer).
On my DIY vibration measurement system wishlist.....my vote would be to use the MPU-6050 rather than the ACH-01 for it's triax capabilities. This presents more challenges from a bandwith perspective with the 3 axes vs. 1 axis, but that would be ideal.

Also, just saw there's a bluetooth version of the 6050 in this link https://www.amazon.com/Accelerometer-Triple-Axis-Acceleration-Transducer-BWT61/dp/B018NL1R0Y/ref=sr_1_4?keywords=6050+vibration&qid=1564785627&s=gateway&sr=8-4

This appears to be a 6 axis accelerometer which would be even more cool if pitch, yaw and roll could be added to the measurement system capability. Makes me wonder if bluetooth would have enough bandwidth for all these channels in the frequencies needed. But, this bluetooth 6050 would eliminate the need for cables, which is a big pain in the neck when connecting 6 or more accelerometers to an analyzer and dressing all those cables so they don't interfere with the measurement.
 

·
Registered
Joined
·
7,585 Posts
Great thread. I look forward to reading what comes out of it. If you don't have TR you don't know what you're missing. It changes the game completely, I don't know what I would do without my Crowsons.
 
  • Like
Reactions: trhought and SBuger

·
Registered
Joined
·
3,582 Posts
Discussion Starter #28
Thanks for starting this thread Shelby! Great idea to help us all collaborate as we continue the chase for that magical TR in our setups.
Great thread. I look forward to reading what comes out of it. If you don't have TR you don't know what you're missing. It changes the game completely, I don't know what I would do without my Crowsons.
Thank guys!! Yeah Hop, with those new iPals of yours AND Crowsons, you definitely got TR!! :D I bet your enjoying the heck out of it all too!!!!
 

·
Registered
Joined
·
4,544 Posts
On my DIY vibration measurement system wishlist.....my vote would be to use the MPU-6050 rather than the ACH-01 for it's triax capabilities. This presents more challenges from a bandwith perspective with the 3 axes vs. 1 axis, but that would be ideal.

Also, just saw there's a bluetooth version of the 6050 in this link https://www.amazon.com/Accelerometer-Triple-Axis-Acceleration-Transducer-BWT61/dp/B018NL1R0Y/ref=sr_1_4?keywords=6050+vibration&qid=1564785627&s=gateway&sr=8-4

This appears to be a 6 axis accelerometer which would be even more cool if pitch, yaw and roll could be added to the measurement system capability. Makes me wonder if bluetooth would have enough bandwidth for all these channels in the frequencies needed. But, this bluetooth 6050 would eliminate the need for cables, which is a big pain in the neck when connecting 6 or more accelerometers to an analyzer and dressing all those cables so they don't interfere with the measurement.
agree about the need for multiple axes, ach-01 makes it easy to run long cables though and takes interface bandwidth out of the equation which is definitely a pain with the mpu6050.

that bluetooth equipped one looks interesting, this is the equivalent of what I was suggested the pi zero w for but smaller, lighter and more plug n play (if it works). All you'd need is a power supply (i.e. a battery and some cables). I was thinking of rewriting the vibe app to simplify it anyway so perhaps that would be a good way to add support for this sort of device at the same time.

https://mbientlab.com/metamotionc/ looks like the non DIY version, $70 per sensor and those sensors are small, light and have built in battery power. It also has an apparently friendly api on top.
 

·
Registered
Joined
·
2,538 Posts
agree about the need for multiple axes, ach-01 makes it easy to run long cables though and takes interface bandwidth out of the equation which is definitely a pain with the mpu6050.

that bluetooth equipped one looks interesting, this is the equivalent of what I was suggested the pi zero w for but smaller, lighter and more plug n play (if it works). All you'd need is a power supply (i.e. a battery and some cables). I was thinking of rewriting the vibe app to simplify it anyway so perhaps that would be a good way to add support for this sort of device at the same time.

https://mbientlab.com/metamotionc/ looks like the non DIY version, $70 per sensor and those sensors are small, light and have built in battery power. It also has an apparently friendly api on top.
metamotionc.....now that looks intriguing! Especially the multiple sensor capability and .csv output with timestamp for alignment :)

The sensor itself looks like a great size, weight and form factor. I use beeswax for fastening to the measurement object. Looks like beeswax would work well with these also with the outer plastic housing....great find!
 

·
Registered
Joined
·
1,825 Posts
Good point Shelby.....yeah, I think you and Nalleh would benefit a lot from dood's system and being able to measure the TR of each of your devices to see how they're combining. It was certainly an eye opener for me when I started measuring independent TR devices all time aligned on one plot....that was how I discovered platform bending was the biggest enemy of natural feeling single digit TR because the drivers started working against each other. There's no way I could have made that discovery with VS especially with the single digit responses that needed to be measured and the limitation of only measuring one driver at a time.

I think we've all found using REW for optimizing room response has netted additional SPL and highlighted holes in our system response that we didn't know about until we started using REW (and MSO). Same with TR, but even more so from a timing error perspective since TR is much more sensitive to timing.
completely agree with @trhought re the need for higher resolution measurements, it's why I wrote that app in the first place :) I remember drawing the analogy with RS SPL meters, back in the day these were the only way to go but then at some point, accurate mics became commonplace and everyone could enjoy better bass! VS is a bit like that imv.

another option which is not that expensive but is more involved is based on https://www.audioxpress.com/assets/upload/files/George NtanavarasAccelerometer.pdf I have one of these also & verified that my rpi device measured accurately by reference to this device (using an ach-01 accelerometer).

If anyone reading has the knowledge of how to build such a circuit then that would be a really useful thing to know as then you could really kit out your seating with lots of sensors :) I imagine the rpi4 would help with this also (more data to process in realtime so faster cpu helps).
Here's another option for a relatively inexpensive accelerometer package from Vernier Software and Technology. It looks like you at least need the following components to get some measurements going. The price for all components is only $249.

When I called Vernier a couple of years ago, they said their products were intended for educational institutions, and wanted assurances that I wasn't buying their equipment for commercial reasons. They indicated they would sell to hobbyists, but I'm not sure how they would react if they started getting a lot of 'hobbyists' orders all at once.

EDIT: I've never made TR measurements before, but instead of measuring each TR device separately to get its relative delay, would it be possible to apply an impulse stimulus to all TR devices simultaneously, and then tune the delay of each device to get the sharpest (narrowest) TR measurement? Maybe you start with the first two devices, get them tuned together, and then add other devices one at a time while tuning its delay to match the previously tuned devices. Just a thought.
 

·
Registered
Joined
·
11,125 Posts
I emailed Vernier and invited them to the party. We'll see if they want to play.
Michael
 

·
Registered
Joined
·
18 Posts
Can you guys provide me some opinions about if the crowson might be a more interesting upgrade vs 2nd sub in my set? I use the PB 4000 right now and by measuring I can see it is still powerful below 20hz at MLP, during movies the tactile feeling overall is great anyway. The problem is the best overall position I can find for that makes a pretty strong dip around 60hz, I am trying to identify if a second sub could solve that but maybe just wastes money or cant even notice a big change from fixing it anyway. So if I can learn to ignore the dip I could buy the crowson instead, but is it very useful? Would I xover on 20hz? Is there any practical job to do at ultra low frequency? Will any movies play that low anyway?

For now I already paired a small buttkicker with headphones for personal enjoyment and setting output on the buttkicker from miniDSP it can mix very well in that situation, really smooth and you won't even expect the buttkicker to be there. But seems like buttkicker+headphones and crowson+pb4000 is really different situations, so what do you guys think about that?
 

·
Registered
Joined
·
4,544 Posts
Re calibration, there was a lot of discussion in one of the older threads (either the VS one or the other one). An impulse (infinite energy at t=0 so contains content from DC to infinity, approximated in audio terms using something like a log swept tone) doesn't really make sense in this context as the device responds over a limited frequency range and can only be measured using bandwidth limited measurement devices. This leads to the use of white noise with a bandwidth to match your measurement capability (to avoid aliasing).

As an aside, I am still not sure why people find the rpi solution so offputting. It is cheap, can provide high resolution measurements (higher than all these other devices) with little effort and leaves you with a useful computer you can use for other purposes (particularly if you get the rpi4, or is extremely cheap if you go the rpi0 route). An rpi is really not a hard thing to setup, easier than windows :) i also plan to turn that into a desktop app with an rta view which should make it a bit easier to pick up and use.
 

·
Registered
Joined
·
11,125 Posts
Because someone like me sees something like this:



and decides it's a project a little above his pay grade.
For the time being, at least. If nothing better comes along, I may reconsider.

I have been known to bite off more than I can chew once or twice before.
Michael
 

·
Registered
Joined
·
11,125 Posts
Okay, you twisted my arm. I'm in.
Since I have an SD card and power supply (actually, several of both) and the cables (if I can find them), investing about $30. I've wasted more on a bad meal. I'm sure I'll enjoy this more.
Thanks for the push.
I think.
Michael
 

·
Premium Member
Joined
·
3,192 Posts
@SBuger what did you prefer about the JBLs in the BOSS compared to the SI 18s cantilevered? I would have thought there would be a significant gain in ULF with the 18s.
 
21 - 40 of 6001 Posts
Top