Welcome, Guest
Username: Password: Remember me

TOPIC: Re: MiniDSP sound quality test

Re:MiniDSP sound quality test 3 years 8 months ago #1282

  • devteam
  • devteam's Avatar
  • OFFLINE
  • Administrator
  • Posts: 2389
  • Thank you received: 228
  • Karma: 65
Alright, my last post because sometimes, trying to argue and explaining basics of measurements is not the best use of our time. :-) The only reason why I'm insisting is because you are providing false information that requires to be corrected by our engineering team.

1) I see that your sound card measures 96dB on a loop back. How did you ever expect to measure over this noise floor (e.g. a device with 100dB SNR) if your limitation is already on the sound card?
We use for some simple quick measurement a EMU1212 which does 120dB in loop back. That's what I call a reference.. :-) Hope you will understand what I just said here is a simpler version: "it is not possible to measure higher specs than your sound card limitation. That's your bottle neck here."

2) Not sure if you had a look at the measurements we pointed you toward, showing some similar specs to what we're advertising, but in case you didn't here are a couple of quick screenshots I did with Right Mark Audio analyzer software. I'm using a PC measurement software to make sure I'm within similar conditions as you are. Board is powered from a USB port, not a DC supply.

See that in an analog loop back (analog in to analog out) through the miniDSP, we are indeed reaching 99dB DR.





Now typical specs in audio (something maybe you're not aware) typically talk of ADC specs, DAC specs. These are NOT done with an analog loopback. They are done with ADC - > I2S (for the ADC specs) and I2S - > DAC for the digital part. Using both ADC and DAC, you would always get lower specs than what the manufacture are rating because noise floor of both conversions + buffer are combined. The truth is with miniDSP, if you were looking at the ADC only or DAC only (with measurement I just talked about), we can rate each of them with a better spec and that's why we're saying > 98dB on the specs. Because we're doing something called "derating" of the specs. A typical practice in audio engineering.

3) Last but not least: We sent our board to the Audio Precision (world leading measurement company) office for 3rd party approval, we got the same specs. We sent our board to Elektor Magazine (one of the most reputable electronic magazine out there), we got the same specs. Finally, others DIYers are also finding the same specs as us (see DIYAUDIO post)

If you get these specs, it's either your measurements (you will NEVER be able to see a measurement over 96dB because your noise floor is already 96db) or maybe your measurement setup. Gain structure matters, ground loops can throw off measurements, sound card specs matters...

Hope this info helps,
Attachments:
MiniDSP, building a DIY DSP community one board at a time.
Last Edit: 3 years 8 months ago by devteam.
The administrator has disabled public write access.

Re:MiniDSP sound quality test 3 years 8 months ago #1284

  • dreite
  • dreite's Avatar
  • OFFLINE
  • Gold Boarder
  • Posts: 276
  • Thank you received: 73
  • Karma: 32
Proger wrote:
The sound card is a E-mu tracker with analog output,
in the first try, I made a loop with :
jack -- rca-- rca -- xlr (so unbalanced)
in the second :
jack -- rca -- minidsp -- rca -- xlr

(I tried to low the volume, but same noise and dynamic)

The minidsp was powered by usb.

The 80Db of noise level is a fact, I always had a little pink noise when I plug the minidsp (and 96sensitivity LS, the 90Db has no noise)
Now I'm using a 5v transformer, and the problem is always here.

Maybe there is no noise with minidigi, but with analog input, I had it.

Thanks,


Proger,

You have a problem with your test setup. The EMU Tracker (I have one also) has very low noise on the inputs, but the line/headphone outputs are quite noisy. So, when configured for loopback or two-channel (differential) measurements it can work well, but (depending upon how you configure it for testing) you can run into pitfalls.

If you have access to a low distortion sine-wave generator use that for your signal source and set up the Tracker preamp for monitoring duty only and set both input knobs all the way counter-clockwise. You'll have to use a different program than Rightmark. Try ARTA instead.

Perform a measurement with nothing connected to your Tracker (so you can see the residual noise,) another with the MiniDSP connected but with no input signal, and another with the MiniDSP input signal at 900mV.

I'm measuring dynamic range numbers even better than quoted by devteam.

Hope that helps.

Cheers,

Dave.
The administrator has disabled public write access.

Re:MiniDSP sound quality test 3 years 8 months ago #1285

  • devteam
  • devteam's Avatar
  • OFFLINE
  • Administrator
  • Posts: 2389
  • Thank you received: 228
  • Karma: 65
Thanks for your very valuable insight Dave on how to configure Arta for a measurement. Looking at the specs of an EMU Tracker, it indeed appears that it should be doing much better than what is quoted here.

DevTeam
MiniDSP, building a DIY DSP community one board at a time.
The administrator has disabled public write access.

Re:MiniDSP sound quality test 3 years 8 months ago #1287

  • Proger
  • Proger's Avatar
  • OFFLINE
  • Fresh Boarder
  • Posts: 4
  • Karma: 0
The Emu tracker does better in 96KHz and 192KHz, here I set 48KHz, like the minidsp.

If i understand, the minidsp add 15-20Db to noise level ? And can do 100Db noise level and dynamic only with 120Db source or more ???

Maybe my test isn't the best, but when I plug the minidsp, I have a pink noise that I haven't without it.
When I set input of minidsp to -10Db, I have less pink noise.
Now that I put the minidsp in my amp, I can't relly test it anymore. I love them :laugh: but the pink noise is a little problem :(

Do you think a power of 12v can change something ?

Devteam -> have you tested the minidsp in a box ? or balanced revA ? an other ?

I am happy to see the "minidsp team" so active :P

Thank,
Sincerely,
The administrator has disabled public write access.

Re:MiniDSP sound quality test 3 years 8 months ago #1289

  • dreite
  • dreite's Avatar
  • OFFLINE
  • Gold Boarder
  • Posts: 276
  • Thank you received: 73
  • Karma: 32
Did you look at the plot I posted on DIYaudio?

www.diyaudio.com/forums/minidsp/175810-a...dsp.html#post2340627

You'll notice those measurements were at 16bit/48khz. (MiniDSP modules powered via USB to the measuring computer.)

Yes, there is a bit of noise on the outputs of the MiniDSP that (probably) wouldn't exist on an analog unit, but it's pretty low. Certainly lower than the measured outputs of my Tracker-Pre and DCX2496. Unless you're listening to Klipschorns using an amplifier with high voltage gain I can't see this being an obtrusive problem.

Cheers,

Dave.
The administrator has disabled public write access.

Re:MiniDSP sound quality test 3 years 8 months ago #1291

  • Proger
  • Proger's Avatar
  • OFFLINE
  • Fresh Boarder
  • Posts: 4
  • Karma: 0
What do you call "low" ?

The DCX has a bad analog input, and there is tweaks for it, (like just remove the input condenser) but it sounds good with digital input.

I have 96Db sensitivity loudspeaker, and I ordered a 107Db sensitivity tweeter.
I am afraid that the little noise I have with my LS will be horrible with it.

My amps are UCD180 and TA2022. The TA2022 make more noise due to better sensitivity. (the ucd looks to work with 2v and unbalanced input)
And the little pink noise is present when I unplug my sound card, but not when I unplug minidsp :S
The administrator has disabled public write access.
Time to create page: 0.370 seconds