1 Topic by integra (edited by integra 2012-01-25 09:04:45)

  • From: Novosibirsk, Russia
  • Registered: 2011-05-16
  • Posts: 86

Topic: Fireface UFX Frequency response limits?

Hi All!

Please see UFX specifications below. RED(text color) are strange??

AD, Line In 1-8, rear
Signal to Noise ratio (SNR): 110 dB RMS unweighted, 113 dBA
Frequency response @ 44.1 kHz, -0.1 dB: 5 Hz - 21.5 kHz
Frequency response @ 96 kHz, -0.5 dB: 5 Hz - 45.5 kHz
Frequency response @ 192 kHz, -1 dB: 5 Hz - 66.5 kHz


DA, Line Out 3-8, rear
Dynamic range (DR): 115 dB RMS unweighted, 118 dBA
Frequency response @ 44.1 kHz, -0.5 dB: 5 Hz - 22 kHz
Frequency response @ 96 kHz, -0.5 dB: 5 Hz - 34 kHz
Frequency response @ 192 kHz, -1 dB: 5 Hz - 50 kHz


In NEW UCX all spec is standart and understand.

Signal to Noise ratio (SNR): 111 dB RMS unweighted, 114 dBA
Frequency response @ 44.1 kHz, -0.1 dB: 10 Hz - 20.6 kHz
Frequency response @ 96 kHz, -0.5 dB: 5 Hz - 45.3 kHz
Frequency response @ 192 kHz, -1 dB: -1 dB: 5 Hz - 90 kHz

DA, Line Out 1-6, rear
Dynamic range (DR): 111 dB, 114 dBA @ 44.1 kHz (unmuted)
Frequency response @ 44.1 kHz, -0.1 dB: 5 Hz - 20.4 kHz
Frequency response @ 96 kHz, -0.5 dB: 5 Hz - 45 kHz
Frequency response @ 192 kHz, -1 dB: 5 Hz - 90 kHz

  • Registered: 2007-06-16
  • Posts: 214

Re: Fireface UFX Frequency response limits?

?? What is the problem??HeadScratch

  • From: Novosibirsk, Russia
  • Registered: 2011-05-16
  • Posts: 86

Re: Fireface UFX Frequency response limits?

Usually uses the full bandwidth(for example Frequency response @ 192 kHz, -1 dB normally is : 5 Hz - 90 kH). In the UFX AD is 5 Hz - 66.5 kHz and DA is 5 Hz - 50 kHz.

In NEW UCX spec all is OK

4 Reply by MC

  • MC
  • MC
  • Administrator
  • Offline
  • From: Germany/Thailand
  • Registered: 2007-04-06
  • Posts: 25,600

Re: Fireface UFX Frequency response limits?

Some converters have a slow slope above the hearing range, wich ruins the usual -0.5 dB values, but they still have the full frequency range up to the aliasing filter. Nothing unusual here.

Regards
Matthias Carstens
RME

MC's Website

5 Reply by panatrope (edited by panatrope 2012-01-25 10:05:59)

  • From: Melbourne, Australia
  • Registered: 2007-05-16
  • Posts: 495

Re: Fireface UFX Frequency response limits?

If you accept Stuart Boothroyd's findings, dead flat (amplitude and phase) to 26kHz is all that is required.   Above that I would prefer a slow phase-linear roll-off to a "flat to the Nyquist frequency" with a huge phase kink - and increased alias components.   (Maybe RME would like to show us the phase response achieved.)

But since we are mentioning specifications, what about noise specs?  Stuart says we need 120dB S/N (equivalent to true 20 bit resolution) - we are not quite there yet, especially on the mic pre-amps.  The currently quoted signal to noise ratio for UFX inputs 9-12 (112db unwtd, 115dBA) is meaningless, because it can only be measured at 0dB pre-amp gain.  What happened to the much more meaningful Equivalent Input Noise (Ein) figures at typical gains that used to be quoted by RME?

De gustibus - et sonus - non est disputandum
  • Registered: 2011-10-30
  • Posts: 63

Re: Fireface UFX Frequency response limits?

Does this mean I should change any settings when using the 5 & 6 outputs connected to a headphone amplifier in order to get a wider frequency response? Or is it only a matter of what kHz are fed to the amplifier?

  • From: Germany
  • Registered: 2007-04-10
  • Posts: 9,782

Re: Fireface UFX Frequency response limits?

No changes to any settings will give you a "wider frequency response", esp. if you are playing 44k or 48k audio material, which is limited to just over 20k anyhow.


Regards
Daniel Fuchs
RME

Regards
Daniel Fuchs
RME

RME Support's Website

8 Reply by MC

  • MC
  • MC
  • Administrator
  • Offline
  • From: Germany/Thailand
  • Registered: 2007-04-06
  • Posts: 25,600

Re: Fireface UFX Frequency response limits?

panatrope wrote:

But since we are mentioning specifications, what about noise specs?  Stuart says we need 120dB S/N (equivalent to true 20 bit resolution) - we are not quite there yet, especially on the mic pre-amps.  The currently quoted signal to noise ratio for UFX inputs 9-12 (112db unwtd, 115dBA) is meaningless, because it can only be measured at 0dB pre-amp gain.

Wrong. It can only be measured at +10 dB Gain.

panatrope wrote:

What happened to the much more meaningful Equivalent Input Noise (Ein) figures at typical gains that used to be quoted by RME?

Still there...in the manual...

Regards
Matthias Carstens
RME

MC's Website

  • From: Melbourne, Australia
  • Registered: 2007-05-16
  • Posts: 495

Re: Fireface UFX Frequency response limits?

MC wrote:
panatrope wrote:

But since we are mentioning specifications, what about noise specs?  Stuart says we need 120dB S/N (equivalent to true 20 bit resolution) - we are not quite there yet, especially on the mic pre-amps.  The currently quoted signal to noise ratio for UFX inputs 9-12 (112db unwtd, 115dBA) is meaningless, because it can only be measured at 0dB pre-amp gain.

Wrong. It can only be measured at +10 dB Gain.

Do you mean it is worse at 0dB gain?

MC wrote:
panatrope wrote:

What happened to the much more meaningful Equivalent Input Noise (Ein) figures at typical gains that used to be quoted by RME?

Still there...in the manual...

From the current UFX manual, V1.8 Oct 2011, Technical specifications, p.88:

Microphone In 9-12, front
As AD, but:
• Input: XLR, electronically balanced
• Input impedance: 2 kOhm
• Signal to Noise ratio (SNR): 112 dB RMS unweighted, 115 dBA
• Gain range: 65 dB
• Maximum input level, Gain 0 dB: +12 dBu
• Maximum input level, Gain 65 dB: -53 dBu
• CLIP LED: 0 dBFS
• SIG LED: -46 dBFS

Cannot find any Ein specification here.
Am I looking in the wrong place? V1.9 maybe?

De gustibus - et sonus - non est disputandum

10 Reply by MC

  • MC
  • MC
  • Administrator
  • Offline
  • From: Germany/Thailand
  • Registered: 2007-04-06
  • Posts: 25,600

Re: Fireface UFX Frequency response limits?

Wow - can't find them either...ok, next version.

Regards
Matthias Carstens
RME

MC's Website

11 Reply by spacealf (edited by spacealf 2012-01-26 19:41:28)

  • Registered: 2011-02-08
  • Posts: 261

Re: Fireface UFX Frequency response limits?

Oh, geeze, didn't I read somewhere that it was -151EIN.
Anyway, you can not hear it, so..............................
To me (in my humble opinion), the noise is caused by humans mainly, not equipment.

  • From: Melbourne, Australia
  • Registered: 2007-05-16
  • Posts: 495

Re: Fireface UFX Frequency response limits?

spacealf wrote:

Oh, geeze, didn't I read somewhere that it was -151EIN.
Anyway, you can not hear it, so..............................
To me (in my humble opinion), the noise is caused by humans mainly, not equipment.

Somebody's poor arithmetic.  You can only get that low if the microphone, cables and UFX are at absolute zero degrees.

State of the art is around -130dBV(A) with a 150 ohm input loading.   You can get a good idea by looking at the data sheet for the PGA2500.  (And if it's no longer a PGA2500, it will only be less than a dB better.)

While it is usually out of the pre-amps hands with modern condensor mics, dynamics and high-quality ribbons need this performance to avoid noise intruding in good performance halls or naturally quiet spaces.  In these circumstances you can hear it ....

Please don't drag me in the direction of distortion!

De gustibus - et sonus - non est disputandum
  • From: Melbourne, Australia
  • Registered: 2007-05-16
  • Posts: 495

Re: Fireface UFX Frequency response limits?

And while we are on the subject, while everyone likes to quote A-weighting because of the 'big number' syndrome, the "CCITT" program weighting is generally considered by professionals to be a more reliable indicator of perceived noise.  It is not liked in industry because it tends to give a figure on paper that is about 10db worse than an A-wtd measurement.

While it may be a big ask for RME to break with habit and publish CCITT figures for its gear, maybe noise measurement with this weighting could someday be included in DigiCheck?

De gustibus - et sonus - non est disputandum
  • From: Netherlands
  • Registered: 2012-03-25
  • Posts: 10

Re: Fireface UFX Frequency response limits?

Hi, new here, just getting my 1st RME (400 UC) :-),

If I may come back to the frequency response limits, what is actually limiting the low end response?
I've ordered a 400 UC which is specified down to 1Hz.
I have another requirement for instrumentation that requires DC coupling, and I *thought* I read somewhere that the line inputs to the 400 UC were DC coupled, but I can't find it back.
Does anyone here have information about this?

thanks,

Jan Didden

janneman01's Website

15 Reply by MC

  • MC
  • MC
  • Administrator
  • Offline
  • From: Germany/Thailand
  • Registered: 2007-04-06
  • Posts: 25,600

Re: Fireface UFX Frequency response limits?

Only the headphone outputs of the 400/UC are DC coupled (mentioned in this forum). All others have capacitors, especially at the inputs. This is standard with audio devices, you will not find ones that allow DC on input. Additionally most ADCs include DC filters (high-pass) at around 1 Hz to get rid of DC information in the converted audio data.

Regards
Matthias Carstens
RME

MC's Website

  • Registered: 2012-03-26
  • Posts: 1

Re: Fireface UFX Frequency response limits?

Thank you Matthias.

I have some experience with driving audio ADCs from DC through a stage providing 2.5V output common mode voltage; there is no technical problem I am aware of.
Maybe I should modify one channel of my new UC just to try.
After the warranty period of course ;-)

best regards,

jan

  • Registered: 2011-08-07
  • Posts: 3

Re: Fireface UFX Frequency response limits?

Of course it is very disapointing, if the specs drop shortly after product introduction. When I bought my UFX, DA @ 192k (line out 3..8) was specified with 80kHz BW, now with 50 kHz. But because these products get also used for acoustical measurement  -  I am such a user  -  it is of prime interest to know the specs are correct when deciding on what product shall be used for what application.

The low end specs on the UC and UFX for the output are corresponding to my measurements still not correct (-1.5 dB and not -0.5 dB @ 5 Hz); input is o.k.  The high end FR is (mainly) given by the chips (its manufaturer); the low end FR by the device designer (RME) by C-value decisions.  Inputs must be DC decoupled  -  the device must protect itsself from several kind of impacts.

Distortion is a much more complex issue, because the chip suppliers specs are often not complete and the wort case values so poor, that no product supplier will likely not want to base his data on them.  With the UFX I measured outstanding THD data on line input alreday 1 dB below FS:
                                   
Eingang    1    2    3    4    5    6    7    8    Mic 9    TRS 9
K2                                       
- 1 dB    -122    -124    -125    -122    -123    -123    -129    -122       
K3                                       
- 1 dB    -116    -117    -119    -116    -118    -116    -118    -118    -100    -112

But you see here quite some variances from chip to chip  -  thats life.

To my experiences: With UFX you can also get outstanding little distortion also on the line output (0.0001% per harmonics); but you need to reduce the level to -15 dB.  So you can use UFX together with software like ARTA or STEPS for measuring harmonic distortion on high end pre-/ amplifieres, etc.

Best regards