Home Theater Forum and Systems banner

221 - 240 of 887 Posts

·
Registered
Joined
·
19 Posts
I wonder how likely it is that an updated file can be produced. I don't know how the calibration works but surely if the measurements are incorrect the only way to get a valid file is to re calibrate and therefore send the mic back?

Disappointed to say the least, past 3 weeks have seemed to drag, was happy to see the shipped email and find it is now in customs only to stumble onto this. The cynical part of me (which if I'm being honest is a big part!) wonders at the timing of this issue, why all of a sudden change your calibration method / output? I hope it wasn't down to the sudden surge of orders that needed to be fulfilled :/

Anyway I guess there's no point speculating until we hear from them, which is also annoying as everyone is off on holiday for what I can only presume is down to the new year..wish everyone would stick to the Gregorian calendar!..just kidding! :)
 

·
Registered
Joined
·
24 Posts
I worked 9 years in Quality Control at my previous company, gauge verification/calibration was an important part of our responsibilities and some instruments are less sensitive to drift than others, some you have to watch like a hawk and pay attention to your results, checking the results is the most important part.

I would imagine due to the backlog they were in a hurry to get things shipped out and someone either didn't follow procedure or their procedure isn't robust enough to catch a bad result, the readings we were given should have been flagged as out of range from the norm and the problem corrected. To me this looks like a computer interface issue, like they should and didn't do a clean reboot and recal before running the job.

Before I joined this forum I bought a Dayton DATS Audio Test System, some of you are certainly familiar with it, it will measure the TS parameters of drivers and test capacitors etc. I had calibrated it and used it and then let it sit connected for some time, the computer may have gone to sleep, I don't remember but I tried using it again later on to test some capacitors and got some really screwy results, I rebooted and re-calibrated and all was well and this looks like a similar case.

I guess there's a small chance that the readings can be straightened out but I doubt it.

At least that's my 2¢ FWIW......
 

·
Banned
Joined
·
4,838 Posts
I guess there's a small chance that the readings can be straightened out but I doubt it.
It is conceivable that it can all be straightened out on their end, bit I, too, expect that returning the units in question for re-calibration will be necessary.

Mistakes happen. We may never get a complete explanation of how this all came about, but what matters is that
  1. the problem was caught relatively quickly and the user community knows to not be misled by potentially erroneous measurement results
  2. the problem has not cost anyone life or limb or loss of property or business; not minimizing the annoyance factor, just trying to keep it in perspective
  3. the problem is fixable
A good company will jump on a problem and do what it takes to regain customer confidence. I trust we will all be willing to allow MiniDSP the opportunity to make things right and regain OUR confidence, and one day soon we will all be back to posting about the wonderful progress we are making with our various measurement and improvement projects.:sn:
 

·
Registered
Joined
·
8 Posts
dezibel-3: Nice work. Thank you for taking the time. Your charts really tell a story.

In case it is not already obvious to anyone, a UMIK-1 with serial number 7000190 or above is usable WITHOUT the calibration file installed in REW, for the purpose of getting rough measurement results, It would not be usable for critical or highly-accurate measurements at this time. When new calibration files are available for those UMIK-1's - calibration files known to accurately reflect the true frequency response for each unit respectively - then they will be useable for critical measurements with their new calibration files installed in REW.

Edit: Do not try to use one of the 7000190+ units with the calibration file installed. It will cause you to get unusable results.
You are absolutly right. It makes no sense using the "newer" cal-files which are in the range approximately of +6 dB to -8 dB instead of of the "older" files which have a range of about +2.4 to -2.9 dB. The highest deviations are within the frequency <16Hz and >2kHz (> +-1 dB).
I'll post an "AverageCal-File" separatly.
 

·
Registered
Joined
·
8 Posts
Here is the "AverageCalFile" whitch uses the average values from my erlier post.
I could load it into REW and it seems to be ok. If you have any problems with it, let me know.
Good luck :praying:
 

Attachments

·
Registered
Joined
·
9 Posts
Is it verified that old calibration files where correct?
The file "AverageCalFile.cal" have a compensation factor of ~0.1dB at 10KHz.
If the UMIK-1 have the same panasonic capsule as the ECM8000 the new calibration files looks more correct in the 800-20KHz range (if you normalize it).
 

Attachments

·
Premium Member
Joined
·
134 Posts
A simple solution - if only temporary - for those who have the mic already would be to simply replace the lines below 812Hz with a single line at 20 Hz, which has the value recorded for 812 Hz. In fact, just removing all lines below 812 Hz would probably work. If someone that has one of these could do that and run a sweep with a modified file, that would be helpful I think.

Or, if someone in AU has one (post-0190) and would like to overnight it to me, I'll do it and compare it against an M30 and a CSL-calibrated EMM-6. If the cal file isn't correct, I'll generate a new one for you as thanks for the loan :)

BTW I compared a UMIK (7000144) against the M30 and the CSL-calibrated EMM-6 and running without the cal file, it's within a dB down to 20 Hz.
 

·
Registered
Joined
·
8 Posts
I requested to have my mic order put on hold/cancelled to this is figured out. The order has just been confirmed so they shouldn't have shipped it yet, etc. Thanks for the heads up everyone.

Jack
 

·
Registered
Joined
·
15 Posts
Ugh, why did I have to find this thread. My serial no. is 7000323 and the cal file is entirely within +/- 3dB and doesn't have that spike at 20hz or 800hz

Maybe its just a database entry error on minidsp's server?

edit: actually no there is a spike at 800hz on mine, but its tiny:

773.187 0.0050
782.812 0.0083
792.557 0.0090
802.424 -0.0025
812.413 0.3771
822.526 0.3898
832.766 0.4034
 

·
Registered
Joined
·
151 Posts
Thanks for posting this info about the calibration files. I just found out about the UMIK-1 and was thinking of ordering one. I'll hold off now until things are cleared up a bit.
 

·
Premium Member
Joined
·
134 Posts
At minidsp forum it's suggested that perhaps they didn't bother doing a calibration below 800hz as most are "within spec" in that range anyway:



http://i.imgur.com/RDXY1NB.png
Um... that's my measurement :) The point of it is that the adjustment to the cal file that I suggested above will probably be a reasonable workaround, not to speculate on what/why miniDSP did ;) I will be trying it with ausvette's mic tomorrow (thanks Dave!).
 

·
Registered
Joined
·
91 Posts
I sent a request to cancel my order. I placed my order on Sunday, right before the posts on the calibration issue. I hope MiniDSP will cancel my order. I was wary of ordering stuff from Hong Kong, I guess I should have followed my instincts! I did read the first 10 pages of this thread first though...

No offense to Chinese companies, but I have little trust in their quality control or that they are using non-counterfeit parts. The fact that they shipped mics out like this tells me they have no quality control whatsoever and are going through a separate manufacturer. Sounds like a dangerous situation to me and a bad way to run a business (if you want to sell stuff in the US).

I ordered a camera charger from China once and it was wired backwards!?! After that experience, I've avoided ordering anything from individual factories found on ebay; that is until this fiasco.

So, if my order is cancelled, is this USB mic ok to use with REW under Windows 7 64-bit?:
http://cross-spectrum.com/measurement/calibrated_umm6.html
 

·
Registered
Joined
·
15 Posts
I am satisfied with their response. So basically it is a matter of scaling of the data due to the 2 different rigs being used for those ranges to improve accuracy. However, that still doesn't explain the anomaly at 20hz.

I guess I got lucky as the "step" at 20hz and 800hz is already quite low on mine:


http://i.imgur.com/lel0wzG.jpg
 

·
Registered
Joined
·
169 Posts
I am satisfied with their response. So basically it is a matter of scaling of the data due to the 2 different rigs being used for those ranges to improve accuracy. However, that still doesn't explain the anomaly at 20hz...
Nor does it answer the question of how low does the calibration go below 20Hz, if at all. Guess we will all have to wait until next week to see what the newly revised calibration files look like. At least it doesn't appear we will have to send it back to Hong Kong for a re-calibration.
 

·
Registered
Joined
·
9 Posts
I am satisfied with their response. So basically it is a matter of scaling of the data due to the 2 different rigs being used for those ranges to improve accuracy. However, that still doesn't explain the anomaly at 20hz.

I guess I got lucky as the "step" at 20hz and 800hz is already quite low on mine:


http://i.imgur.com/lel0wzG.jpg
I got the 0277 serial that was posted earlier in the thread.
The step at 20-30hz on mine is pretty awful, hopefully this is due to the cal file as well.

Posted on the minidsp forum so I hope we get a fix soon.
 

·
Registered
Joined
·
9 Posts
I plotted three graphs using three different calibration files in the rangen 20–800 Hz(ignoring the first sample).
The graphs looked like random noise. I don't think scaling could fix the 20–800 region.
 
221 - 240 of 887 Posts
Top