or Connect
AVS › AVS Forum › Audio › DIY Speakers and Subs › HOW TO: Measure using HOLM and ARTA
New Posts  All Forums:Forum Nav:

HOW TO: Measure using HOLM and ARTA - Page 2

post #31 of 155
Quote:
Originally Posted by Myggpower View Post

What was supposed to be a 'cheap' solution for me just keeps getting more and more expensive! Sigh



The internal card in my new HP measures dead flat from 10Hz to 20kHz. The card in my three year old Toshiba needed correction, but after calibration the results read identical to those from my HP. A Behringer mic is like $50, a Blue Icicle USB pre $40. Try it with your internal card to see if it works before being concerned.
post #32 of 155
Quote:
Originally Posted by brandonnash View Post

The realtek should work, but you should run measurements on it to make sure it has a flat response and is full duplex. Just run the line out straight back to the line in and run a sweep. You theoretically should have a perfectly or close to perfect flat response out of it.

My old IBM pc came with a SoundMax card. ( SoundMax could be the software for the card.) Are you saying that I can connect the output to the input and run a sweep like I'm actually measuring a speaker, and then make corrections for the card using Holms?

I have a Eurorack MX 602A mixing board that powers my ECM8000 mic. It has analog connections that I can connect to the sound card.

I suppose I can make a loop out from the sound card to the mic input of the MX 602A and back out from it to the sound card input. Does that sound right? Can I then generate a correction file using Holms? Can I alter the correction file manually?

If something in my setup is not calibrated, and is off above 10k or below 100 hz, would that affect any measurements in the 300hz to 8k range?
post #33 of 155
Thread Starter 
Quote:
Originally Posted by Myggpower View Post

What was supposed to be a 'cheap' solution for me just keeps getting more and more expensive! Sigh

So I need a soundcard with phantom power!

Both the M-audio and Tascam US-144 only has 2 channel sound.. they are around 270$ where I live.

What I don't like about those soundcards, is that they only have 2.0 sound.. It would be nice with a soundcard that has all the features of the M-Audio / Tascam, but with 5.1 or 7.1 - because I watch a lot of movies and series on my PC with 5.1 & 7.1 sound.. Having an upgrade on both stereo and surround would be really nice. Are there any "bang-for-the-buck" cards that have all the features I need ?

Or should I just focus on stereo sound - and use the built-in realtek (like I always have) for movies/series?

Im currently googling, and so far the only cards I've found are very expensive.

I won't hijack this thread further.. maybe someone can shoot me a PM with recommendations, if they have any.

Thanks

You seem to have two goals here.. We are strictly talking about speaker measurements.

If you want muti-channel then why not use digital audio?
post #34 of 155
Thread bump. Also, I'm not seeing the pictures you posted. Only text shows.
post #35 of 155
Thread Starter 
I apologize, I store all my pics on my own webserver but that means when my network goes down so do my pictures
post #36 of 155
Quote:
Originally Posted by penngray View Post

mobilepre loop back calibration

Make sure phatom power is off. (red light on the front)



I use a 1/4" plug, its plugged into 1/L output (on back) and into ch1 inst/line input (on front).


Just my 2 cents, but if your intention is to use the MobilePre's XLR input for your mic. You need to use that connection for your loopback calibration. I don't believe I was involved in your thread at HTGuide, but I have made this same comment a couple of times over there.

Now how much of a difference will it make... who knows. However, if you're going through the trouble of calibrating everything, including your mic. Why not mimic your signal chain exactly.

My 1/4'' to XLR cable cost ~$5. It's worth the minor investment to do it right.
example:http://www.bhphotovideo.com/c/produc..._Phone_to.html

edit: my comments in another thread at HTG
http://www.htguide.com/forum/showpos...52&postcount=2
post #37 of 155
Thread Starter 
Quote:
Originally Posted by smokinghot View Post

Just my 2 cents, but if your intention is to use the MobilePre's XLR input for your mic. You need to use that connection for your loopback calibration. I don't believe I was involved in your thread at HTGuide, but I have made this same comment a couple of times over there.

Now how much of a difference will it make... who knows. However, if you're going through the trouble of calibrating everything, including your mic. Why not mimic your signal chain exactly.

My 1/4'' to XLR cable cost ~$5. It's worth the minor investment to do it right.
example:http://www.bhphotovideo.com/c/produc..._Phone_to.html

edit: my comments in another thread at HTG
http://www.htguide.com/forum/showpos...52&postcount=2

Yeah, I did think about that I actually saw your thread on that and I did buy the XLR to 1/4" adapter to find out.

The difference wasnt meaningful so I kept the pictures. THey are all wired together inside (Input is either XLR or 1/4") but the noise isnt going to change that much.
post #38 of 155
Quote:
Originally Posted by penngray View Post

Yeah, I did think about that I actually saw your thread on that and I did buy the XLR to 1/4" adapter to find out.

The difference wasnt meaningful so I kept the pictures. THey are all wired together inside (Input is either XLR or 1/4") but the noise isnt going to change that much.

To be totally accurate you need to do the loopback including the power amp used for measuring. Using extreme caution, of course.
post #39 of 155
Thread Starter 
Quote:
Originally Posted by Bill Fitzmaurice View Post

To be totally accurate you need to do the loopback including the power amp used for measuring. Using extreme caution, of course.

I disagree, we are simply looking for the soundcard curve and amp will have no impact on the sound card. The amp is part of the speaker measurement actually.

There are enough threads on doing the soundcard calibrations that this isnt really a discussion that is needed.
post #40 of 155
Quote:
Originally Posted by penngray View Post

I disagree, we are simply looking for the soundcard curve and amp will have no impact on the sound card. The amp is part of the speaker measurement actually.

There are enough threads on doing the soundcard calibrations that this isnt really a discussion that is needed.

My preference is to take the amp into the loop measurement so that the measurement is of the speaker, not the speaker/amp combination. For the most part above 20 Hz most amps are quite linear, but IME below 20 Hz few are.
post #41 of 155
Thread Starter 
Quote:
Originally Posted by Bill Fitzmaurice View Post

My preference is to take the amp into the loop measurement so that the measurement is of the speaker, not the speaker/amp combination. For the most part above 20 Hz most amps are quite linear, but IME below 20 Hz few are.

Good point, definitely in terms of measuring subwoofers.
post #42 of 155
I think the distinction in approaches is that we're designing speakers to be used only by us in our home with the amps we have. We want to measure what we're going to hear (including any fluctuations in the FR of the amp we're using). Bill - you're probably designing speakers to be built by others and being pushed by who knows what receiver/amp in which case you would want to remove the effects of the particular receiver/amp you're using for the measurements from the equation.
post #43 of 155
Quote:
Originally Posted by lennon_68 View Post

Bill - you're probably designing speakers to be built by others and being pushed by who knows what receiver/amp in which case you would want to remove the effects of the particular receiver/amp you're using for the measurements from the equation.

In my case providing the most accurate measurements possible is based on the pissing and moaning I can expect if they aren't. Or even when they are, for that matter.
post #44 of 155
Gating observation.

In my measurements, I found that I needed the gating at about 1.5 ms to 2 ms. After doing a little quick math, this makes sense because my first reflection off the floor would be roughly 2 ms behind the direct signal. This is with the drivers about 30" to 39" above the floor, and measuring from between 3 to 5 feet away.

The gating marker on the spl trace shows up at about 500 hz to 800 hz range. I assume that any frequencies above the marker should be accurate, and any below could be partly contaminated by the reflections.
post #45 of 155
Thread Starter 
Quote:
Originally Posted by lennon_68 View Post

I think the distinction in approaches is that we're designing speakers to be used only by us in our home with the amps we have. We want to measure what we're going to hear (including any fluctuations in the FR of the amp we're using). Bill - you're probably designing speakers to be built by others and being pushed by who knows what receiver/amp in which case you would want to remove the effects of the particular receiver/amp you're using for the measurements from the equation.

but he is also talking about subwoofers and how amps react down around 20Hz.

This thread is about speaker measurements. Subwoofer measurements are well documented using REW.
post #46 of 155
Thread Starter 
Quote:
Originally Posted by kbgl View Post

Gating observation.

In my measurements, I found that I needed the gating at about 1.5 ms to 2 ms. After doing a little quick math, this makes sense because my first reflection off the floor would be roughly 2 ms behind the direct signal. This is with the drivers about 30" to 39" above the floor, and measuring from between 3 to 5 feet away.

The gating marker on the spl trace shows up at about 500 hz to 800 hz range. I assume that any frequencies above the marker should be accurate, and any below could be partly contaminated by the reflections.

Yes, the gated response will depend on the room. You can drag the gated line to see how it changes the response, you will notice the upper frequencies, 1K Hz and above do not really change that much but the lower frequencies do.

My measurements are done in my HT room which has treatements, I also put 4" thick panels on the floor and behind the speaker Im measuring. I can get gating out about 4 ms or more.

You are correct about the marker, everything below includes room interaction and everything above is just the speaker.

HOLM is pretty good with room noise too, someone on DIYaudio.com did a a measurement while playing music the FR response didnt change much at all. The music just should up in the THD as noise. This means we do not need to worry about dead silence in the room.
post #47 of 155
What procedure is recommended for measuring the response below the gating frequency? How about distortion measurements for a woofer?

Is this done by placing the mic close to the woofer?
post #48 of 155
Thread Starter 
Quote:
Originally Posted by kbgl View Post

What procedure is recommended for measuring the response below the gating frequency? How about distortion measurements for a woofer?

Is this done by placing the mic close to the woofer?


Yeah, Im not the expert but I think ground plane measurements are needed and close to the woofer.

Ground plane will remove the floor reflections and moving the mic close will remove other reflections.

You can splice the two responses back together later (accounting for the dB change for the closer mic)
post #49 of 155
Quote:
Originally Posted by penngray View Post

Yeah, Im not the expert but I think ground plane measurements are needed and close to the woofer.

Nearfield measuring introduces both line sourcing and zone compression issues. That's why the preferred method for measuring subs is outdoors at 10 meters and 100 watts. But doing so requires a lot of open space and a lack of easily irritated neighbors, and no highways in the vicinity either. Ground conducted LF carries literally for miles. You can't hear it, but it will show up as a 'mystery' source in the 15-25Hz range, with 50-55dB levels not the least bit unusual. Picture your mic as Tonto with his ear to the ground, telling Kemosabe about the approach of unseen riders.
post #50 of 155
As long as THD measurements are included at 100W, to let people know where to set a HPF...

JSS
post #51 of 155
This looks nothing like yours. What happened? Any ideas?
post #52 of 155
Thread Starter 
Quote:
Originally Posted by Bill Fitzmaurice View Post

Nearfield measuring introduces both line sourcing and zone compression issues. That's why the preferred method for measuring subs is outdoors at 10 meters and 100 watts. But doing so requires a lot of open space and a lack of easily irritated neighbors, and no highways in the vicinity either. Ground conducted LF carries literally for miles. You can't hear it, but it will show up as a 'mystery' source in the 15-25Hz range, with 50-55dB levels not the least bit unusual. Picture your mic as Tonto with his ear to the ground, telling Kemosabe about the approach of unseen riders.

For Subwoofer measurements I would have to differ to other experts. I believe Ilkka from HTS did all his great measurements outdoors, close to the ground at 2M.

Anyways, this measurement discussion is about main speaker designs and not subwoofers.

REW has lots of documention and is good enough for subwoofer data.
post #53 of 155
Thread Starter 
Quote:
Originally Posted by kbgl View Post

This looks nothing like yours. What happened? Any ideas?

Im not sure, could you post your settings.

The X axis scale looks strange too.
post #54 of 155
Thread Starter 
btw, I would recommend you post questions here
http://www.diyaudio.com/forums/multi...ml#post2138133

Since other more knowledgeable people can help you also.
post #55 of 155
Quote:
Originally Posted by penngray View Post


Anyways, this measurement discussion is about main speaker designs and not subwoofers.

Mains as well should be measured from at least four meters for an accurate result. It's more difficult with them, as you need to take a groundplane measurement to get a half-space result below the baffle step, and then take an axial measurement with the speaker on its back facing upward and the mic suspended above it to get a half-space result above the baffle step uncorrupted by ground reflections. It's not difficult, just time consuming.
post #56 of 155
Thread Starter 
Quote:
Originally Posted by Bill Fitzmaurice View Post

Mains as well should be measured from at least four meters for an accurate result. It's more difficult with them, as you need to take a groundplane measurement to get a half-space result below the baffle step, and then take an axial measurement with the speaker on its back facing upward and the mic suspended above it to get a half-space result above the baffle step uncorrupted by ground reflections. It's not difficult, just time consuming.


Harman internation, NRC of Canada and just about every expert designer I have read about do not measure from 4 meters. I can conclude there is great accuracy at 1 or 2 meters. Atleast enough to conclude 4 meters would be overkill.
post #57 of 155
Quote:
Originally Posted by penngray View Post

Harman internation, NRC of Canada and just about every expert designer I have read about do not measure from 4 meters.

Depends on how large their room is. Measuring from close range doesn't allow full integration of wavefronts with multiple drivers, nor mitigation of line sourcing and compression zone issues. It all boils down to how accurate a result you want to get. You can take gated nearfield measurements of each driver and cobble them together and get a reasonably accurate result, and if you don't have a large anechoic chamber handy, or a goodly sized backyard, then by all means do so. I've got a large enough yard to allow me to measure at 4 meters, so I do. The results are always different than those I measure at 1 meter.
post #58 of 155
The Realtek HD Audio is actually a very good chip and makes a fair measurement device (i used to measure my speakers with one, till i fried it accidentally, now i use a SB Audigy in a separate computer for this purpose).

For mike phantom power you can simply use an external supply, those are cheaper than a dedicated soundcard.
post #59 of 155
Thread Starter 
Quote:
Originally Posted by Bill Fitzmaurice View Post

Depends on how large their room is. Measuring from close range doesn't allow full integration of wavefronts with multiple drivers, nor mitigation of line sourcing and compression zone issues. It all boils down to how accurate a result you want to get. You can take gated nearfield measurements of each driver and cobble them together and get a reasonably accurate result, and if you don't have a large anechoic chamber handy, or a goodly sized backyard, then by all means do so. I've got a large enough yard to allow me to measure at 4 meters, so I do. The results are always different than those I measure at 1 meter.

I complete agree with you.

it does come down to compromises. Its hard for some not only the space available but moving 100 lbs speakers around is impossible.
post #60 of 155
HOW TO: Measure using HOLM and ARTA

I was reading something on the DIYA forum about this. They said people
are getting erroneous measurements because the minimum requirements
to measure J.HOLMes is a ruler, not a mic.
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: DIY Speakers and Subs
AVS › AVS Forum › Audio › DIY Speakers and Subs › HOW TO: Measure using HOLM and ARTA