Welcome, Guest
Username: Password: Remember me
Here is the place to share your questions/comments about your setup of the UMIK-1 USB measurement microphone


NOTE: This is a "Community" forum. Please be mindful that community members are here to help as part of a community effort. We therefore appreciate your effort in keeping this forum a happy place!

If you have a specific issue (e.g. hardware, failure) and want help from our support team, please use our tech support portal (Support menu - > Contact Us).
Thanks a lot of your help in making a better community.
  • Page:
  • 1
  • 2

TOPIC:

UMIK-2 measuring different from UMIK-1 1 year 1 month ago #51902

  • AustinJerry
  • AustinJerry's Avatar Topic Author
  • Offline
  • Elite Member
  • Elite Member
  • Posts: 162
  • Thank you received: 49
REW measurements:







Why are there significant differences between the measurements?

Also: to reach a 90dB measurement level in REW, I set AVR master volume for the UMIK-1 to -12, but to reach 90dB with the UMIK-2, I need to set MV to -4, and 8db difference. And finally, with the UMIK-1 and using the REW SPL Meter, my listening room's noise floor measures 50dB. With the UMIK-2, it measures 35dB. Why?
Attachments:
The following user(s) said Thank You: [email protected], gratefulnautilus

Please Log in or Create an account to join the conversation.

Last edit: by AustinJerry.

UMIK-2 measuring different from UMIK-1 1 year 1 month ago #51903

  • audiocrazy
  • audiocrazy's Avatar
  • Offline
  • New Member
  • New Member
  • Posts: 5
  • Thank you received: 0
I've a CSL UMIK-1 and UMIK-2 and I second this observation between UMIK-1 and UMIK-2.
I see a 8 dB difference when I do Dirac Live Volume calibration.

Can someone from support please explain this?

Please Log in or Create an account to join the conversation.

UMIK-2 measuring different from UMIK-1 1 year 1 month ago #51904

I can confirm this as well, I have UMIK-1 and UMIK-2 directly calibrated by miniDSP.
Unfortunately Dirac Live (v3) is not working properly with UMIK-2 as the reading s from UMIK-2 will lead to too much bass.

Please Log in or Create an account to join the conversation.

UMIK-2 measuring different from UMIK-1 1 year 1 month ago #51931

  • Ultrasonic
  • Ultrasonic's Avatar
  • Offline
  • Platinum Member
  • Platinum Member
  • Posts: 1706
  • Thank you received: 309

And finally, with the UMIK-1 and using the REW SPL Meter, my listening room's noise floor measures 50dB. With the UMIK-2, it measures 35dB. Why?


The lower noise floor is the one thing that I'd potentially expected, given that the noise floor of the UMIK-1 was rather high and that a lower noise floor of the UMIK-2 is actually one of its claimed improvements.

The differences in the measured frequency response are concerning and I'll be interested to hear what miniDSP have to say about this.

I was tempted to order a UMIK-2 but held off to wait for some reports on their performance. Glad I did on the basis of the posts above!

Please Log in or Create an account to join the conversation.

UMIK-2 measuring different from UMIK-1 1 year 1 month ago #51935

  • AustinJerry
  • AustinJerry's Avatar Topic Author
  • Offline
  • Elite Member
  • Elite Member
  • Posts: 162
  • Thank you received: 49
I have been in contact with Cross Spectrum Labs, and they have received a UMIK-2, which they are in the process of calibrating. I expect we will see a comparison between the factory cal and the CSL custom cal very soon.

Please Log in or Create an account to join the conversation.

UMIK-2 measuring different from UMIK-1 1 year 1 month ago #51952

  • gratefulnautilus
  • gratefulnautilus's Avatar
  • Offline
  • New Member
  • New Member
  • Posts: 1
  • Thank you received: 2
Like OP, I'm also seeing a concerning difference in measured frequency response between the UMIK-1 (using both CSL and MiniDSP calibration files) and the UMIK-2 (using the provided MiniDSP files).

Using a Kef LS50 loudspeaker or a Peerless D26NC56 tweeter for reference, the measured difference between the UMIK-1 using either the CSL and MiniDSP calibration files is minor, almost negligible. But versus the UMIK-2, the difference is huge.





About a 1.2db difference in overall sensitivity and a significant 3-5db difference in the measured upper treble(7-20khz range starting from 3khz).

Also In regards to dynamic range/SNR/noise floor, I'm seeing ~20db noise on average in REW from the UMIK-2 compared to ~29 db of the UMIK-1. Problem is the maximum SPL of the UMIK-2 seems to be comparatively very low at just ~107db versus the ~115db of the UMIK-1. Any louder and it clips badly.

I'm not sure if I set up the microphone incorrectly or I'm interpreting the specifications wrong, but I think with a 120db SNR ADC and a -105dbFS noise floor microphone capsule, I should be seeing a big improvement over the UMIK-1's SNR (-74dBFS IGPA, whatever that means), but with a maximum SPL of 107db and ~20db noise floor, dynamic range/SNR is just 87db. Hardly an improvement over the UMIK-1's 86db DR (115db max SPL and ~29db noise floor).

BTW, I think the latest version of REW may have a compatibility issue with the UMIK-2. By default, the input volume is set to ".53"(53 in the windows mixer), which is correct for the 0db point for the UMIK-1, but wrong for the UMIK-2, as the 0db point is "1.00" instead(100 in the windows mixer).
Attachments:
The following user(s) said Thank You: [email protected], Ultrasonic

Please Log in or Create an account to join the conversation.

UMIK-2 measuring different from UMIK-1 1 year 1 month ago #51954

  • AustinJerry
  • AustinJerry's Avatar Topic Author
  • Offline
  • Elite Member
  • Elite Member
  • Posts: 162
  • Thank you received: 49
I have asked MiniDSP Support to comment.

Please Log in or Create an account to join the conversation.

UMIK-2 measuring different from UMIK-1 1 year 1 month ago #51960

I have made some measurements how the UMIK-2 behaves with different Settings, 24bit with 48kHz, 96kHz and 192kHz.
My expectation was that the curves would look -with very minor differences - all the same in REW but they are not.

Additional I have measured the same Speakers with my UMIK-1, no changes in volume of the AVR have been made, just exchanged the UMIK-2 against the UMIK-1 and measured again.





UMIK-2 with different settings only:



Attachments:

Please Log in or Create an account to join the conversation.

Last edit: by [email protected].

UMIK-2 measuring different from UMIK-1 1 year 1 month ago #51975

  • devteam
  • devteam's Avatar
  • Away
  • Administrator
  • Administrator
  • Posts: 9021
  • Thank you received: 1338
Dear All

As our team is looking into this, it's useful that we get a tech support ticket started. That's the official method to get in touch with our support so please if you have some files, share it with us so we can understand the conditions these measurements/comparisons where made with. (0deg/90deg), confirming cal files were indeed loaded in the correct place (potential pitfall with UMIK-2 knowing it's UAC2 and not discovered by default unless you're using ASIO4all as per manual)...

Anyway, we need more info/screenshots and a case with our team so we can follow up. Thanks!

Here is our support portal: support.minidsp.com -> Click New Support Ticket.

DevTeam
miniDSP, building a DSP community one board at a time.

Please Log in or Create an account to join the conversation.

UMIK-2 measuring different from UMIK-1 1 year 1 month ago #51976

  • devteam
  • devteam's Avatar
  • Away
  • Administrator
  • Administrator
  • Posts: 9021
  • Thank you received: 1338

I have made some measurements how the UMIK-2 behaves with different Settings, 24bit with 48kHz, 96kHz and 192kHz.
My expectation was that the curves would look -with very minor differences - all the same in REW but they are not.


@Michael

That low frequency changes based on sample rate is unfortunately a little bug that made its way on the latest firmware build.
There is an HPF on the onboard ADC that needs to be reconfigured based on the sample rate. The higher the sample rate, the higher the cuttoff frequency. We're doing a simple firmware build now to solve that issue.

To get accurate data, please stick to 48kHz for time being. Thanks for your patience.

DevTeam
miniDSP, building a DSP community one board at a time.
The following user(s) said Thank You: [email protected]

Please Log in or Create an account to join the conversation.

UMIK-2 measuring different from UMIK-1 1 year 1 month ago #51977

Thanks for the info, I assume we do NOT need to open a support ticket.

Please Log in or Create an account to join the conversation.

UMIK-2 measuring different from UMIK-1 1 year 1 month ago #52094

  • devteam
  • devteam's Avatar
  • Away
  • Administrator
  • Administrator
  • Posts: 9021
  • Thank you received: 1338
Dear All

Thanks again for all your patience. Here is a long update based on all the investigation we've done in the past week.

UMIK-2 units have now shipped to all customers and hoping you're enjoying your microphone! As it unfortunately can happen with a new product introduction, not everything might go according to plan. We'd like to thank all community members for the great feedback we've received! Here is an update on few topics we'd like to clarify today:

- Low Frequency roll off @ higher sample rate
- Dirac low level message
- High frequency response of UMIK-2 vs UMIK-1
- REW measurements vs Sample rate
- Understanding ASIO vs WDM

Here are some specific answers to each section:

- Low frequency roll off vs Sample Rate: As some of you might have noticed, the UMIK-2 shares the same USB+ADC module than the ambiMIK-1 product. Before the release, we combined the firmware and unfortunately missed to turn off the High Pass Filter for the UMIK-2 scenario. This high pass filter on the ADC needs to be bypassed for a flat response at all sample rates. The new firmware fixing the UMIK-2 HPF bug will be posted in the coming days as we also need to update the manual to include upgrade steps. We'll share the firmware with few beta testers in the mean time.
ACTION ITEM: Upgrade the UMIK-2 firmware once available by mid of this week. Make sure that you don't have any other devices connected to your PC/MAC during upgrade.

- Dirac Low level message: Having a much lower noise floor, the UMIK-2 has a larger dynamic range and different sensitivity level compared to a UMIK-1 (about 10dBFS). Unfortunately Dirac Live doesn't translate this dBFS to dBSPL at this stage hence the warning that UMIK-2 is "less sensitive". Indeed the exact same SPL is played and if correctly calibrated, there is enough signal. For the time being, while Dirac is implementing the SPL calibration (similar to REW), we've been advised by Dirac Team to not worry about the lower level reading. Future release of Dirac Live will hopefully solve this warning knowing there is plenty enough signal.
ACTION ITEM: Ignore lower level. Will be solved in future Dirac versions.

- High frequency response of UMIK-2 vs UMIK-1: When building calibration curves for microphone on the factory floor, capsules + microphones are sweeped against a "golden sample" microphone that was calibrated against a B&K reference grade microphone. There are actually 3 process to that calibration (Low Freq/1kHz sensivity factor/High freq). The "raw data" of these 3 measurements is then "post processed" by a script afterward. This post processing of the data is a mathematical process aided by some custom coding to combine the 3 sets of data and generate the cal file format as a txt file. The process is well known and been used for many years for the UMIK-1. Unfortunately when the post processing happened in mass qty for the UMIK-2, the wrong set of reference was used in the code and the calibration file txt output was corrupted even if raw files were right. After investigation of our raw files and coding, we've indeed now found the issue. We'll be issuing the corrected cal files this week that solves this offset of 2dB at the High Frequency. This is obviously less than ideal and we're really sorry for this issue.. When things go into mass manufacturing in large quantity on a factory floor, things can get a bit more complex than usual and has the potential to go wrong... The good news is that we indeed found the culprit and have all the raw files to start with. It's a coding issue, not a hardware so no need for recalibration. We're now reviewing very cal file before posting them. In the mean time, we've removed the old cal files.
ACTION ITEM: Download the new calibration file expected by mid/end of this week.

- REW measurements vs Sample rate: We've noticed this comment from some of you that UMIK-2 was rolling off prematurely when using Java on REW as one increases the sample rate. That behavior is actually expected if one is increasing the sample rate on REW but not on the OS. Java is a virtual machine that actually doesn't have access/control over the device sample rate via the driver. If you change the sample rate on REW to say 192kHz. You also need to change the sample rate from the UMIK-2 via the WDM(windows audio devices) or OSx (Audio & Midi devices) dialog box. A warning is already included in REW manual on this topic but we'll make sure to include it in the manual as a reminder. Manual to be updated this week.
ACTION ITEM: Always match the sample rate on your OS with the Java Sample rate in REW.

- Understanding ASIO vs WDM: There are two types of drivers that one can control under windows.
WDM = The Windows driver that you can see in the bottom right corner of your taskbar. That's the OS driver and controlled by Windows. You indeed can change the sample rate from the control panel as per p8 of the manual. www.minidsp.com/images/documents/miniDSP...-2-User%20Manual.pdf
ASIO = This is the low latency driver controlled by the miniDSP UAC2 Control panel. This driver is actually completely unrelated to WDM.
When you change the sample rate on WDM, it won't affect the ASIO part (i.e. don't look at the miniDSP UAC2 control panel, it's not the same path and unrelated). ASIO on the other hand will overwrite whatever sample rate you have made under WDM.
ACTION ITEM: None.. Just a clarification.

I think that we've talked enough... I'm sure there are other questions.. :-)

As usual, if you have any questions, don't hesitate to contact our support team (support.minidsp.com). Simply create a ticket and we'll take it from there.

DevTeam
miniDSP, building a DSP community one board at a time.
The following user(s) said Thank You: AustinJerry, foundation42, [email protected], [email protected], Ultrasonic, audiocrazy

Please Log in or Create an account to join the conversation.

UMIK-2 measuring different from UMIK-1 1 year 1 month ago #52123

  • audiocrazy
  • audiocrazy's Avatar
  • Offline
  • New Member
  • New Member
  • Posts: 5
  • Thank you received: 0
Thanks for the detail update.
Can you please post here to let us know when the firmware and new cal files are available to downloaded?

Please Log in or Create an account to join the conversation.

UMIK-2 misst sich von UMIK-1 1 year 1 month ago #52173

  • Sascha259
  • Sascha259's Avatar
  • Offline
  • New Member
  • New Member
  • Posts: 2
  • Thank you received: 0
:(
Attachments:

Please Log in or Create an account to join the conversation.

UMIK-2 misst sich von UMIK-1 1 year 1 month ago #52174

  • Ultrasonic
  • Ultrasonic's Avatar
  • Offline
  • Platinum Member
  • Platinum Member
  • Posts: 1706
  • Thank you received: 309

:(


Have you read the post above from @DevTeam that explains the problem and intended solution? Or have you installed new firmware and still found a problem? (My guess is the firmware hasn't been released yet...)

Please Log in or Create an account to join the conversation.

Last edit: by Ultrasonic.
  • Page:
  • 1
  • 2
Moderators: devteam