MR816 noise problem (from "The other side")

Steinberg Hardware
21

Hey, I just got an MR816x and think I may be having a similar problem. My setup includes a late 2008 iMac 2.8 GHz Intel Core 2 Duo, 2GB of RAM, and I believe the Agere/Lucent firewire chipset. My previous interface was a Presonus Firebox, which was somewhat noisy (especially close to when it blew up/popped and started smoking), but lacked the new noise I’m having now.

As soon as I hooked up my new MR816x I noticed what sounded like a microphone outside on a windy day coming through my monitors, which also happen to be early 2000s Tannoy Active Reveals (with matching subwoofer). It is a very low frequency rumble that is almost always present when anything is played through the speakers, Itunes or Cubase. If the main master volume is turned off it usually goes away, but I swear the first time I powered on and the volume was all the way down on the MR816x it still made the noise, but now it doesn’t appear to.

This is what I’ve noticed about it through some testing/debugging attempts. It happens with all outputs, only happens when the firewire cable is connected (directly or by being daisy chained to an external FW hard drive), and has a short decay. That is if I play sound (at low volumes so that it is really noticeable) and hit stop, it takes about a second to fade away. It also seems that the rumble is able to be present when no audio is played back, but this is not the usual case. It is detectable on any channel that too (doesn’t matter if it’s a bass or overhead track in cubase) and shows up on a frequency analyzer around 0-30Hz. If I run a high pass filter and set it all the way up to around 250Hz, the low rumble noise goes away.

That last part really baffles me. I can’t record the low rumble, but I can filter it out live by running a high pass filter on any unmuted channel? For example, I have an overhead track that rumbles, throw on the hi-pass filter (on the track channel or main out channel) and it’s gone, but it’s not possible to work at such a high filter settings. I don’t get it. And the fact that it can be present when no audio is passing through. Weird.

It really doesn’t sound like this is the same problem everyone is talking about, but I have some of the common factors as others. I also get an unreasonable level of hiss/noise (like a preamp with gain turned up high) coming through my speakers that doesn’t change at any volume level, but I can live with that (just bought a TC level pilot that I hope will fix this problem). But when the low frequency, wind like rumble, starts I cannot monitor at low levels anymore. It’s driving me crazy. Also, all noise temporarily disappears when switching clock rates, like everyone else.

To complicate things further, it doesn’t appear to be happening at all right now, which I guess is a good thing. I was beginning to wonder if the convertors are so good and I’m not used to them, that they’re revealing low frequency noise that cheaper convertors missed. But the strange decay, the presence in professional recordings on Itunes, being present when no audio is playing, leads me to believe this is not the case.

I’m guessing it’s some kind of interference with the Imac and external/internal hard drives. It also seems that a lot of Macs have had problems over the last year with firewire audio interfaces. So many variables and I’m not even sure if my problems are the same as others, or if mine is unique. Everything except the Imac is plugged into a Furman power unit, and everything is going to the same socket (maybe that’s a problem). I have a laptop downstairs that I’m going to see if the same behavior occurs on it as well. I’ll post any new findings I come across.

Sorry this post is all over the place. I really want this interface to work properly, it’s got everything I want.

22

Okay, a little update. I think I’m having a multitude of noise issues. I’m not positive, but besides the intermittent outside microphone on a windy day sound, I’m pretty sure I’m having a ground loop hum problem as well. With no audio passing through the unit I can sometimes hear a low hum. Weird thing is, or perhaps it’s expected, if I put a high pass filter on my main output in Cubase and roll the frequency up and down, the low hum disappears. Again this is with no audio passing through and nothing is showing up on the frequency analyzer, but I swear I can get the low hum to go away with the high pass filter. What do you think about this one?

Still not nearly as annoying as the low freq wind rumble, but still something I’d like to eliminate. By the way, the low rumble seems worse when the MR816 is at low volumes (master unit volume). If I increase the volume the rumble seems to drop away, but it can still come back. So damn strange.

23

Mine is the MR816X.

No LF rumbles or wind noise: just classic (but exaggerated) output resistor hiss . . .

24

GOOD NEWS!!! Okay, this is kind of funny, but I’ve found the cause for my LF rumble. Recall how I had been quite stumped on why I would get this rumble with no audio playback, but was able to get rid of it by placing a high pass filter on the master out? Well, some plugins I use have analog output modeling of old units to simulate hum and buzz (50Hz and 60Hz selectable analog noise output). Since this added noise (intentionally added by the plugin mind you) is always on regardless of audio being passed through, this is why I am able to filter it out when everything is stopped. I guess in the past I just wasn’t able to hear it with my old interface. So, I guess this is an example of how accurate/detailed the DAC convertors in the MR816 are.

I’ve also noticed I can hear low rumble in live room mics and even DI bass tracks that I never noticed before (as well as in professional recordings). Odd, kind of annoying, but at least I know now it is not a defect in the product.

I still believe I may be having my own periodic 50Hz/60Hz hum from my studio setup due to the computer and what not, but that is barely noticeable. I also want to point out that I do have the typical hiss from the MR816, but that’s not too bad and I believe it will be eliminated when I get my TC Electronics Level Pilot and can set the MR’s output volume at a decent level and reduce the volume on the analog side before it hits my active Tannoy Reveal monitors.

All in all, I’m happy today.

25

Hi Mirabela, yep, still visiting, I have the MR816X


Thats great to hear! :smiley: Please let us know how you get on with the TC Electronics Level Pilot

MW, hopefully this level pilot does the trick and we can all be HISS free!! :slight_smile: (and you wont have to get rid of a nice unit! :slight_smile: )

all the best
Sav

26

HI guys !

jlofling, nice to have you in.

Your report is very interesting and well documented. Good work.
Regarding the ground loops, if you can’t get rid of them 100%, describe us the wiring in detail. Probably this will help you chase the loop :smiley:

About the hiss problem, interesting that, all of you guys own the X model, mine is CSX and is noise free.
Could this be a coincidence?

jlofling, have you checked if the snakes are in the headphones too?
Interesting you are using exactly the same monitors as Sav.
Do you have access to other monitors to check if the snakes are still there?

Please note that my CSX is used with HS50M/HS10W and zero snakes here :slight_smile:

Best regards,
M.

27

Just swapped the Adam monitors (A7 and S3X-H) for some MSP7s - same problem. This really is all about the output resistors of the MR816X, surely?

Several similar problems reported by others, but no official solution - bad batch? Bad design? Unknown . . .

28

Well, I just got the TC Level Pilot (a simple volume attenuator) and the hiss is now GONE!!! I just set the master out of the MR816 to 0 and control the volume post DAC with the Pilot. At quiet to normal monitoring volumes it’s dead silent. Pushing it to pretty ridiculous levels (around 1-2 O’clock on the Pilot) there is just a slight hiss coming through. Certainly not like before. And again this is at a level I doubt I would ever monitor at with near fields.

I still think it’s odd that my previous interface didn’t have this hiss, but I’m happy that my setup appears to be working nicely now. Plus the intermittent LF rumble is gone as well (when the source is not the plugins). All I have to do now is put the unit to the test and do some 3 hour live recordings. If I can get 16 channels to record for this duration without ANY hiccups or glitches, I will be ecstatic. Nothing in computer audio has ever worked well for me (starting with the old Yamaha DSP Factory back in the late 90s). I have my fingers crossed and will keep everyone updated on the MR’s performance.

29

The external volume knob is only a “dirty fix” for a problem should not be there in the first place.
Now, using the external volume, you are limiting the 816’s dynamic range, going to the tape era :smiley:

(Now Mr. Chris will tell me to stop writing <<can’t reproduce his obscene language>>).

If you could test your toy a different speaker model, like Yamaha (mine are hiss free on HS50M) we could conclude if there is an incompatibility of some sort with some speakers, or is just a problem with the X model, or is a problem with the QA on Steinberg.

Since Steinberg is keeping quiet about this, I only speculate that they are covering something.
Let’s uncover :smiley:, what about that?

M.

30

Not to be rude, but I’m fairly certain that controlling volume post DAC, on the analog side, is the favored method. When reducing the volume digitally before the D/A conversion you’re reducing your dynamic range and S/N ratio due to bit loss (or something like that). By keeping the digital master out at 0dB (DAW volume also at unity) I get to utilize all of the bit depth. At least that’s what I’ve read (something like that anyhow). I believe this is why most high end digital monitors have volume controls on the speakers themselves and one of the reasons monitoring stations exist. I’ve always thought that my Tannoy’s were too damn loud, they’re permanently stuck at maximum gain.

Anyway, I agree that this noise shouldn’t exist. Something is definitely not working properly since the hiss stays constant whether the MR816’s volume is at -70dB or 0dB (without the TC pilot in the chain). What I’d be interested to know is if your monitors have volume controls on the back, and if so, does turning their gain all the way up and reducing the MR816’s master volume to somewhere between -70dB and -40dB result in a relatively loud hiss. I wouldn’t be that surprised if they do, which I take is what’s happening without running an analog volume attenuator in my setup. Still odd that I don’t recall my previous cheap interface having this problem. I always had to run it’s volume (which I believe was pre D/A convertor) really low as well. Again, I’m fairly content with things now and really enjoying the TC unit. It’s nice to have a tiny volume control right in front of my keyboard, plus it keep my speakers from being damaged when turning my DAW and rack gear on/off.

I might get a chance to borrow a buddy’s M-Audio active monitors. I’ll let you know what transpires when I do. Take care.

31

By the way, the hiss was never present in either headphone out.

32

Hi J,

Thanks for getting back to us on this. Great to hear the TC unit works, I will have to save some pennys and get one.
As we have the same speakers and MR unit, im guessing this has got to work for me too.
Good luck with the recording!
I’ve had my MR for about 2 years now, apart from this silly hiss mine has not let me down! :slight_smile:

Cheers
Sav

33

Hi Jlo and Sav !

First of all, from my knowledge, digital volume is not lowering the bit depth (not more than analog volume).
Imagine you are re-recording an audio stream after your external volume. The amplitude will be lower, so less bit needed. Same with the digital volume (there are some differences regarding the math involved in the amplitude computation, but hey, if we want to step on the audiophile arena, variable resistors are noisy, so… let’s better not :stuck_out_tongue:).

I was not explicit enough, I think. The way I see this fix is: having an external volume knob is actually like using an attenuator in audio chain. You will end up using also digital volume (by meaning of software volume in your media player, mixer faders in you DAW, etc). So, your external volume will be basically an external attenuator and, probably, the go-to knob when accidentally your monitors jump up on the table :smiley:.
However, having already the 816’s knobs at hand (and the nonlinear way they lower the volume in panic situation), this would probably be the closest knob to rich - at least for me).
Now, having an external attenuator all the time, you will loose some amplitude, so, bit depth.
Yes, you are right, with enough space, you can forget abut the 816’s knobs and use exclusively the TC external volume (my space is limited, I sometimes use an external midi control surface, at best).

Now, regarding the monitor test, well, this was my first test when reading Sav’s post: I quickly cranked up my monitors volume knobs, powered down 816 CSX, listen to the noise floow, powered up 816 CSX.
The increase in hiss level was not noticeable.

So, my conclusion, at least on 816 CSX with Yamaha monitors the hiss is not there.
I begin to suspect more and more that only the X model is build cheaper with snakes inside :stuck_out_tongue: but this should ne proved only if one of you guys can barrow another pair of monitors (or an CSX and test it on your monitors).

Either way, the hiss should not be there, in my opinion. Other manufacturers manage to build (cheaper) interfaces that do not need additional external equipment to be bought.
I think Steinberg should be able to do the same.
However, since this is their first interface, we are lab mice :smiley:
Weird the Steinberg/Yamaha customer support is not a first, they should have another quality level, don’t they?

My2c,
M.

34

Regarding the digital vs. analog volume adjustment debate, you should read up on this a bit. You may not have your system calibrated optimally. Here’s a pretty good explanation from GearSlutz:

Originally Posted by the_man361
motu cuemix software is the same as turning the knob on the motu 828, its just controlled by a number of software mixers rather than using the menus on the motu’s front panel.
you can change output levels for all the output pairs, ADAT and spdif from using the software. > :slight_smile: > i think if cuemix crashes, even if you’re ‘not using it’ (because you kinda always are i think), that means your motu interface has bombed out too > :stuck_out_tongue:

This is correct – but the knob isn’t a potentiometer in an analog circuit – it’s just a controller that sets the output level going to the DA stage.

So you are, indeed, losing the least significant bits off the resolution. Which may be no big deal at normal-to-full level listening levels – but it’s something to think about. (Of course, if you’re running powered monitors, you should have your monitor trims set optimally, too, so that your powered monitors aren’t running wide open and you’re turning the digital output level way down to compensate; lately I’ve been reading posts from folks who didn’t seem to know what those little knobs were for and were delighted to find they could reduce the noise of their system by setting them properly.)

But back to losing the LSB’s off the bottom of your digital signal by using a digital output control to set monitor level – my everyday listening rig is a 65 w/side power amp running into some NS10m speakers. Under normal circumstances, I get my normal full volume listening level with the analog volume on the amp at -40 dB.

If that were a digital output control, I’d be reducing my digital s/n from a theoretical 135-140 dB (in 24 bit mode) to a still-respectable 95-100 dB. If I take it lower still, as I might for a low-volume comparison, the reduction in resolution would then likely result in audible degradation of the signal – which wouldn’t happen if the volume reduction was happening on the analog side. (Not that there isn’t potential signal degradation, decreased s/n in some analog circuits at really low levels, but the issues can be somewhat different.)

But that’s only 65 w/speaker. My regular powered monitors are 200 w/speaker.

If I were to reset the trim on those to wide open and then control my monitoring volume solely via the CueMix DSP [or the output level in my DAW or a combination], I’d really have to turn that level way down, effectively throwing out a fairly significant amount of digital signal resolution just to get the monitor level down to a roar.


Now, that example above was for 24 bit. Suppose we’re sending a 16 bit signal out the DA. If we have to reduce the output level by, say, 40 dB, just to not be blasted by our monitoring level, we’re going from potential 96 dB s/n ratio to 56 dB – not even hi fi by 50’s standards. And to extend my example, that would be full listening level with mid-powered speakers (65 w/side). If you had higher powered speakers or were to turn down that (digital) volume control even more to do a low-level comparison, you’d really bet talking about a bit starved signal. Seems to me.

(Again, setting the input trims on your powered monitors is probably the most beneficial and important signal/noise optimization you can do to your rig. At least, it’s the first you should do.)

Also the following section from Digital Volume Control and SNR - SqueezeboxWiki is helpful in explaining it as well.

However, when the DAC is making a quiet signal, you have a little signal and a little noise. If we now consider the noise level in relation to the signal level, the noise is now louder. The noise level hasn’t gone up in absolute terms (eg volts), but relative to the signal it has, so you now have a bad SNR.
Now consider a simple resistor attenuator being fed by a loud (good SNR) signal from the DAC. When the voltage passes through the resistor divider, everything gets attenuated - the signal and noise together. You have the same* SNR coming out of the divider as you had going in, i.e., the DAC’s optimal SNR is preserved.

35

Lool, my system is not calibrated optimally, but I have zero hiss, no matter what :ugeek:
I’m pretty happy with this un-optimally calibration I am using down here :laughing:

I’ll not debate on the analog/digital volume, in your quote some statements are right, some are just twisted, and is normal, since not everyone is using Fourier transforms on daily basics :smiling_imp:

However, back to our sheep, the hiss is there for some people, the hiss is not there for other (running their monitors on full 70W), if you will read all the posts and the long topic on the old forum, you will be up to date on our testings.

I see that your hiss is gone, can you tell us, please (by measurements or by computation) what would be the lowest attenuation applied by your TC Pilot, in order to hill the unpleasant hiss? (Yes, is subjective, but we trust your years)

Also, more important, for this value, what input and what output impedance do you see on your TC Pilot?
If you can’t measure/compute that, at least a DC measurement with you Ohmmeter would be helpful.

Thanks,
M.

36

Sorry to side track things, and also I’m not trying to argue or call you out on anything. I’ve just been reading a lof of things over the years and wanted to share what I heard an thought was the general consensus. Something along the “record at low levels when tracking at 24 bit (-18dBFS)” mantra.

Back to the matter at hand. The TC Level pilot has a range of about 315 degrees (I have it setup in front of me so that 6:00 is 0 and 4:30 is 315 degrees) With the MR816x’s DAW and Master faders set to 0 dBFS (in its virtual mixer), hiss begins to show at around 1:00 (210 degrees of motion, or pi/4 if your familiar with the unit circle :wink:). Similarly, starting with the Pilot all the way up (4:30/315 degrees), hiss is eliminated by 1:00 (105 degree reduction).

From TC’s website;

Quad-core fully balanced XLR connectivity - Level Pilot comes with its own discrete quad-core cabling to ensure perfect stereo imaging and left-right tracking - with no messy cabling to clutter up your workspace.

So low input and output impedance, right? Maybe I can find my old multimeter and see if I can’t figure something out. I’ll let you know when I do, I made need some assistance with that too. Take it easy.

37

Hi Jlo !

Well, don’t trust everything you read :unamused: , this post included :smiley:
Go back to basic physics and math primary class and you’ll understand better for yourself :bulb:
Experimenting is another way to verify your theories (and others, there are so so many “smart” people - this one included :mrgreen: - posting on forums to confuse competition :laughing: or just ignoring common sense and basic physics).


Well, actually XLR cable is not having to do with the PILOT’s internal impedance… Usually (actually all the time, to be polite) the input impedance (of amplifiers, active speakers, etc) are high impedance (in theory, infinite impedance would be perfect, but hey, this is Earth over here :blush: ) and output impedance are very low.

Now, the thing is, changing input impedance for your active speaker (remember, now the input impedance for your speaker is not 816, is TC PILOT) will change the frequency response of your speaker amplifier (and the frequency response of your audition).
That is, changing volume on your Pilot will change the frequency response (some people don’t hear it, some do; all people I now hear it after a while).
There are designs for volume knobs that do not change the impedance, but I speculate Pilot is not one of them.

Now, if you want to get an idea about, first of all, please remember that you can’t measure impedance with your multimeter (impedance is changing with frequencies, isn’t that fun? :mrgreen: ) . However you could measure a resistance.
On your multimeter, set the scale on Ohm and, for a good start, 20KOhm. (if you can’t read, try the next up scale).
Check if the multimeter is working by shorting the two leads, you should read zero.
Now keep your leads with your fingers and see the reading. You’ll learn that touching the leads will influence the reading, since is measuring your internal resistance. So, when measuring your Pilot, keep in mind that you should not touch the metallic part of the leads, or cables, or your measuring will be compromise.

Disconnect the Pilot from input and output, both channels.

Good, now you can measure Pilots input resistance (this will be easy, since there are female connectors) between 2 and 3 pins on the XLR female connector. Turn the knob all the way up and down and see if the resistance is changing (my guess is will not).
Ok, now to the real thing: measure the Pilot output impedance (also 2 and 3, but you can play with 2 and 1, respectively 3 and 1). You will probably need a third hand to play with the knob and another one to write down the values, so be prepared :stuck_out_tongue: .
It would be nice to write down the following values: Pilot min volume, max volume, value where the hiss start to be noticed, value where you are usually keeping the Pilot set for a comfortable audition.

All this should take no more than 6 minutes.

Thanks,
M.

38

It appears that I no longer have a multimeter. Perhaps I can borrow one from a friend. Also, I went ahead and ordered another MR816X so that I can run a daisy chained setup. I’ll be sure to let you know if the hiss is present with that unit as well (I’m feeling like it will be). Sometime next month I should also have access to some different monitors. So, all sorts of experiments are about to be carried out on the MR816’s. Hopefully they’re up for the task. Take it easy.

39

Great, it would be interesting if you could have access to a CSX model, just for a quick A/B regarding the snakes …

40

Well, the second MR816x also produces the same hiss when connected straight to the monitors and behaves the same with the TC Level Pilot (ie. thresholds identical). Now, I’m not sure, but it seems like the overall signal-to-hiss ratio has improved slightly. Individually the two units behave the same, but when daisy chained there might be a reduction in the overall hiss level.

In addition, I’m not sure if this is supposed to happen either (as in one of the benefits of running two units), but it also seems that my CoreAudio performance has improved. Is there any way adding a second MR would relieve some of the stress on my CPU? Strange, but I’m not going to question the authenticity of these improvements, placebo effect or not.

EDIT: I also want to say that I’m absolutely loving this new setup and will be putting it to the test this weekend. Already have routing and connection templates setup. Now, I just have to hope my Sony Vaio SZ230p can handle 16 channels of 44.1kHz 24 bit audio for 3+ continuous hours. It has an onboard TI firewire chipset and the DSP Latency Checker Utility results looked pretty good. I’m crossing my fingers. Still not sure which way to record the audio. What would you recommend.

  1. MR816x(2)–>MR816x(1)–>FW 400 HDD–>COMPUTER
  2. FW 400 HDD–>MR816x(2)–>MR816x(1)–>COMPUTER
  3. MR816x(2)–>MR816x(1)–>COMPUTER<–USB HDD
  4. MR816x(2)–>MR816x(1)–>COMPUTER (record to internal hard drive)

I’m leaning towards configuration 1 since it kind of works like that on my Imac and things have been running nicely. At home it’s MR816x(2)–>MR816x(1)–>FW 400 HDD (time machine backup)–>[FW 400 port]COMPUTER [FW 800 port]<–FW 800 HDD (audio recording and sample libraries) <–FW 800 HDD (backup and media library). What do you (or anyone else) think about this?